Task-Based Language Teaching in Foreign Language Contexts

Task-Based Language Teaching:

Issues, Research and Practice (TBLT)
Task-Based Language Teaching (TBLT) is an educational framework for
the theory and practice of teaching second or foreign languages. The TBLT
book series is devoted to the dissemination of TBLT issues and practices,
and to fostering improved understanding and communication across the
various clines of TBLT work.
For an overview of all books published in this series, please see
http://benjamins.com/catalog/tblt

Editors
Martin Bygate John M. Norris Kris Van den Branden
University of Lancaster University of Hawaii at Manoa KU Leuven

Volume 4
Task-Based Language Teaching in Foreign Language Contexts.
Research and implementation
Edited by Ali Shehadeh and Christine A. Coombe
Task-Based Language Teaching
in Foreign Language Contexts
Research and implementation

Edited by

Ali Shehadeh
UAE University

Christine A. Coombe
Dubai Men’s College

John Benjamins Publishing Company

Amsterdam / Philadelphia
 TM
The paper used in this publication meets the minimum requirements of
8

the American National Standard for Information Sciences – Permanence

of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data

Task-based language teaching in foreign language contexts : research and implementation

/ edited by Ali Shehadeh, Christine A. Coombe.
p. cm. (Task-Based Language Teaching, issn 1877-346X ; v. 4)
Includes bibliographical references and index.
11. Language and languages--Study and teaching--Methodology. 2. Second language
acquisition--Methodology. 3. Task analysis in education. I. Shehadeh, Ali.
II. Coombe, Christine A. (Christine Anne), 1962-
P53.82.T356 2012
418.0071--dc23 2012020908
isbn 978 90 272 0723 4 (Hb ; alk. paper)
isbn 978 90 272 0724 1 (Pb ; alk. paper)
isbn 978 90 272 7342 0 (Eb)

© 2012 – John Benjamins B.V.

No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any
other means, without written permission from the publisher.
John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands
John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa
We would like to dedicate this work to the memory of Teresa Pica, a
luminary in the TBLT world, who wrote the foreword to the volume,
but sadly passed away before its publication (Teresa Pica 1945–2011).
 Table of contents

Preface xi

Foreword xv
Teresa Pica

chapter 1
Introduction: Broadening the perspective of task-based language teaching
scholarship: The contribution of research in foreign language contexts 1
Ali Shehadeh

section i.  Variables affecting task-based language learning and performance

chapter 2
Effects of task complexity and pre-task planning on Japanese
EFL learners’ oral production 23
Shoko Sasayama and Shinichi Izumi

chapter 3
Measuring task complexity: Does EFL proficiency matter? 43
Aleksandra Malicka and Mayya Levkina

chapter 4
Effects of strategic planning on the accuracy of oral and written tasks
in the performance of Turkish EFL learners 67
Zubeyde Sinem Genc

chapter 5
Effects of task instructions on text processing and learning
in a Japanese EFL college nursing setting 89
Yukie Horiba and Keiko Fukaya
 Task-Based Language Teaching in Foreign Language Contexts: Research and implementation

chapter 6
Task structure and patterns of interaction: What can we learn
from observing native speakers performing tasks? 109
James Hobbs

section ii.  Implementation of task-based language teaching

chapter 7
Patterns of corrective feedback in a task-based adult
EFL classroom setting in China 137
Noriko Iwashita and Huifang (Lydia) Li

chapter 8
Incidental learner-generated focus on form in a task-based
EFL classroom 163
Paul J. Moore

chapter 9
Qualitative differences in novice teachers’ enactment
of task-based language teaching in Hong Kong primary classrooms 187
Sui Ping (Shirley) Chan

chapter 10
Implementing computer-assisted task-based language teaching
in the Korean secondary EFL context 215
Moonyoung Park

chapter 11
Task-based language teaching through film-oriented activities
in a teacher education program in Venezuela 241
Carmen Teresa Chacón

chapter 12
Task-based language teacher education in an undergraduate
program in Japan 267
Daniel O. Jackson

chapter 13
Incorporating a formative assessment cycle into task-based
language teaching in a university setting in Japan 287
Christopher Weaver
 Table of contents 

chapter 14
Language teachers’ perceptions of a task-based learning programme
in a French University 313
Julie McAllister, Marie-Françoise Narcy-Combes,
and Rebecca Starkey-Perret

epilogue.  What is next for task-based language teaching?

chapter 15
TBLT in EFL settings: Looking back and moving forward 345
David Carless

About the contributors 359

Index 363
 Preface

We are very pleased to welcome Task-based language teaching in foreign language con-
texts: Research and implementation, edited by Ali Shehadeh and Christine Coombe, as
the fourth volume in this series on task-based language teaching. As the volume edi-
tors note in their introduction, the common thread that provides coherence to this
collection of studies is a focus on how tasks are being researched or used in a wide
variety of English as a Foreign Language (EFL) educational settings in order to un-
ravel the complexities and peculiarities of researching and implementing TBLT in
these contexts. As such, the volume adds a range of new studies to the steadily growing
research base on the implementation of TBLT in the language classroom. This the
volume accomplishes not only in quantitative terms, but also in terms of scope: While
historically, many of the available studies into the use of tasks in intact classrooms were
carried out in contexts of second language teaching and learning, this volume focuses
exclusively on the use of tasks in the teaching of English as a foreign language.
The theoretical framework on language learning underpinning the rationale be-
hind TBLT has tended to emphasize that processes of second language (SL) learning
and foreign language (FL) learning have a lot in common. For both SL and FL learn-
ing, social interaction embedded in holistic, goal-directed activity is centrally seen as
the means for deep-level language learning. The importance of comprehensible and
rich input, focus on form, speaking opportunities, and feedback are stressed in both
contexts. As a result, in many publications on task-based language learning, the term
“second language learning” is used by authors as an umbrella term covering both for-
eign and second language learning processes. However, one of the interesting ques-
tions this volume raises is whether SL and FL teaching through tasks have as much in
common. From a TBLT perspective, foreign language teaching may differ from second
language teaching contexts in a number of ways. For example, in a FL context, and
contrary to a SL context, students may lack opportunities and/or pressure to put what
they learnt in the classroom to proper use in the outside world. Secondly, foreign lan-
guage teachers may find it much harder than second language teachers to introduce
authentic material and texts in the classroom, to point out the usefulness of certain
tasks, to motivate their learners to use the target language in the classroom, and to
encourage them to put effort in acquiring and studying it. Thirdly, teaching languages
as a subject (as is typically done in the case of foreign language teaching) may add to
both learners’ and teachers’ view of the target language as an object of study, rather
than as a useful means for functional communication or as something with direct
relevance to learners’ needs. As a result, practitioners in the field of foreign language
 Task-Based Language Teaching in Foreign Language Contexts: Research and implementation

teaching might be expected to be (even) more hesitant than second language teachers
to use tasks in the classroom, and to be more inclined to stick to the use of more tradi-
tional, grammar-based methods of teaching and testing. Fourthly, in most FL contexts,
most or all of the students generally share the same L1, so that in the classroom the
target language is not a lingua franca but an additional and often unnecessary alterna-
tive medium.
Indeed, the various articles in this volume empirically substantiate the wide range
of factors that complicate the introduction and implementation of TBLT in EFL class-
rooms. In their introduction, the volume editors group these factors into three types:
institutional factors (comprising, amongst others, issues like class size, official exam
pressure, available materials, and mixed-proficiency classes), teacher factors (e.g.,
teachers’ beliefs and subjective theories on language teaching, their need to control
what goes on in the classroom, their interactive skills), and student factors (e.g., their
beliefs about effective language learning, their preferences for certain methodological
formats, their level of assertiveness). Many of these factors have been mentioned in
previous publications on the implementation of TBLT as considerably complexifying
the use of tasks in authentic classrooms: the combined presence of large/huge class
sizes, grammar-based exams, forms-focused teacher and student beliefs, lockstep-type
curriculum, and teacher-dominated traditions of teaching have been reported to turn
the introduction of TBLT into a real ‘challenge’ (and we probably should take that as
an understatement). This has been particularly, though not exclusively, the case in
publications on the implementation of TBLT in Asian contexts. In this respect, it is
fascinating to see that this volume reports on studies that were carried out in different
countries around the world: in addition to numerous studies from Asian countries,
European and American settings are also represented, which allows the reader to draw
a wide range of interesting comparisons.
In fact, some of the above-mentioned factors hindering the introduction of tasks
in authentic classrooms do again turn up in a number of chapters in this volume, but
at the same time they do not in other chapters, or they play a surprising and unantici-
pated role. Indeed, some of the chapters in this volume report on success stories with
the use of tasks. Strikingly, the accounts of successful implementation of TBLT do not
appear to be geographically bound: they are reported in Asian, Venezuelan, and French
EFL classrooms. It is also worth highlighting that in most of these success stories the
teacher played a crucial role. This pattern lends support to the insight that the actual
impact of potentially complexifying factors tends to be mediated by the teachers and
the students in the first place, and by other change agents (such as syllabus and exam
developers, and teacher trainers and coaches) in the second place. This corroborates
the basic insight that when it comes to promoting the use of tasks in the foreign lan-
guage classroom, the teacher is a key figure. Structural and institutional measures, like
reducing class sizes and making available new syllabuses, may help to pave the way, but
if teachers lack the skills and the motivation to work with tasks (and the basic belief
that task-based interaction fosters language learning), no real change will take place.
 Preface 

So, in many, perhaps all, foreign language teaching contexts, the implementation of
TBLT will need to include a strategic policy plan on teacher training and support.
Clearly, this volume does not answer all questions on the implementation of TBLT
in EFL contexts. In fact, it raises new questions, as a good book will. As a collection of
a wide variety of studies carried out in different parts of the globe where English lan-
guage learning persists as an important educational target for diverse types of learners,
it illustrates both (a) the possibilities for language educational innovation, in particu-
lar at the classroom, teacher, and learner levels; and (b) the kind of methodologies that
might be used when further pursuing this research path and the kinds of more nu-
anced questions in need of answers. At the same time, the volume provides another
colorful illustration that in different continents, and in different educational settings
(from primary schools to higher education), practitioners are experimenting with
tasks in ever-more sophisticated ways, in the effort to make their language teaching
more functional, more usage-based, and more powerful. Ultimately, it is what happens
in a range of distinctive contexts that provides a measure for the generalizability of the
approach, as well as providing the variety of real world challenges that an approach like
TBLT needs to enable its full development.
 Foreword

Teresa Pica
University of Pennsylvania, USA

Task-based language teaching in foreign language contexts: Research and implementa-

tion – from its title to its table of contents and authors, and across its chapters, this
important volume builds on, and enhances task-based teaching and research in con-
texts where English is not readily available beyond the classroom (e.g., Turkey,
Thailand) or is available in the wider community, but is seldom learned as an L1 there-
in (e.g., Hong Kong). The volume includes contributions from countries such as Japan
and Spain, where publications on task-based language teaching (TBLT) have long been
widely disseminated, and from countries such as Venezuela, whose work on TBLT
might have been unfamiliar to readers prior to this volume. The chapter entries cover
topics and concerns that are characteristic of TBLT across ESL contexts, including the
role of planning in task implementation or task outcomes in fluency and accuracy, as
well as those that are shared between ESL and EFL contexts, such as task-based assess-
ment within a national English curriculum.
“Foreign” languages typically originate in classroom contexts, where they are pur-
sued as an academic endeavor or in compliance with education policy. However, many
foreign language learners go on to use learned languages for academic, business, or
scientific communication beyond their original site of learning. Editors Shehadeh and
Coombe have selected topics and authors that focus on English as an example of such
a foreign language, and on the issues that surround its teaching and research in FL
contexts. In so doing, their collection builds on the substantial body of work by Doughty
and Long on critical “foreign” languages such as Korean, Chinese, and Arabic, and by
Gass and Van Patten on Spanish learning in US university classrooms (see, e.g., Doughty
& Long, 2003; Gass & Alvarez-Torres, 2005; Van Patten, 2002). The success of these
research-based projects, accomplished primarily in controlled settings, has activated a
great deal of thinking about the ways in which tasks can be used to address the scope
and scale of EFL teaching and research in broader contexts, and to do so through a
classroom focus and a longitudinal approach. As it builds on this foundation, the cur-
rent volume pioneers explorations into territories of FL teaching which have so far
been neglected in much of TBLT research.
 Teresa Pica

Over the past several decades, tasks have generated increasing interest among
educators and researchers, and provided them with an opportunity to respond to the
challenges of their fields in complementary and productive ways. The vitality and
versatility of the task as an instructional tool, a research instrument, and a learning
activity becomes especially evident when viewed with respect to the pedagogic, socio-
cultural, linguistic, and cognitive processes that contribute to successful language
learning. Since their formalized inception in the early work of Long and Prabhu
(see, e.g., Long, 1985; Prabhu, 1987), tasks have brought forth a focus on design fea-
tures that have made them suitable as course syllabus units (Willis, 1996), classroom
activities (Nunan, 1989), enhancements to the language curriculum (Crookes & Gass,
1993), and as tools for assessment (e.g., Norris, Brown, Hudson, & Yoshioka, 1998).
These pedagogic features have also made tasks effective treatments for encourag-
ing learners to engage in goal-oriented social interaction during which they exchange
information and negotiate to achieve its comprehensibility. As they do so, they modify
their messages and signal their difficulties. These interaction moves enable learners
(and teachers) to provide each other with modified, comprehensible input and correc-
tive feedback, and to respond to each other with their own modified output. These
linguistic manipulations shed light on cognitive processes such as noticing, attention,
and awareness. Together, the modifications and processes serve as a source of data
for researchers as they study language development and outcomes (Iwashita, 2003;
Mackey, 1999).
The versatility of tasks is further revealed as they are implemented by teachers and
researchers for independent or joint purposes. A problem-solving task, for example,
can originate in a student textbook or professional resource guide. A teacher might
assign the task to a group of students, making sure the information needed to solve the
problem is distributed evenly, so that each student can engage in collaborative plan-
ning and practice. The same task might be adopted by the researcher, and used with
the same student group, in order to study the linguistic features of their planning and
practice and the ways in which they provide modified, comprehensible input and out-
put in their exchange of information. An interview task, used as a classroom activity
for students to access information for writing a report or planning a project, can also
provide data on linguistic forms and manipulations required for question formation
(Mackey, 1999) and on students’ ability to encode and respond to each other’s conver-
sational adjustments (Pica, Kang, & Sauro, 2006). A comparison task applied to deci-
sion-making projects, pictures, and texts can promote turn-taking and classroom
communication, as it provides data for the researcher to study learners’ development
of English morphemes, such as – er and – est (Loschky & Bley-Vroman, 1993).
In addition to their versatility and appeal to teachers and researchers, tasks have
played an important role in responding to theoretical and practical challenges posed
by approaches to learning and instruction, curriculum design, classroom language
study, and assessment of language skills. As such, tasks are helping to address the long-
standing debate over the effectiveness of direct and indirect instructional approaches
 Foreword  

in meeting learners’ linguistic needs. It is widely held that direct approaches are effec-
tive when forms are simple and when criteria for their learning can be transparently
defined in terms of explicit knowledge and assessed by discrete-point testing. For dif-
ficult and complex forms to be internalized and used automatically, however, implicit
and incidental approaches, albeit much slower in their impact on learning, are be-
lieved to be required. Even here, tasks which indirectly engage the strategic use of
target features are seen as offering an essential dimension of learning. As research by
Pica et al. (2006) has revealed, difficult forms can be readily incorporated into text-
based tasks, and both direct and indirect approaches can be applied to their implemen-
tation. Comparison studies are underway to resolve this debate (see, e.g., Ellis, 2003;
Pica, et al., 2006).
Tasks have also been used to address issues that surround the design of the lan-
guage curriculum (see, for example, Pica, 2009). One of the foremost criticisms of
task-based instruction has been that its approach to task specification and selection
does not explicitly address the forms, sequences, and processes of language learning.
Of course, this standard has not been applied to other instructional approaches, per-
haps because their ordered arrangement of linguistic units gives the appearance that
they reflect these characteristics of language and learning. Nonetheless, it is well docu-
mented that language acquisition is a multidimensional process, does not often follow
a stage-wise path, and occurs most efficiently in meaningful contexts, rich in compre-
hensible input, with opportunities for feedback and production of modified output.
These are the very attributes that task-based instruction has to offer, and the chapters
of this volume provide a springboard for further research on the design of an effective,
task-based curriculum.
As Shehadeh (this volume) and others (e.g., Ellis, 2009; Van den Branden, 2006)
point out, it is unfortunate that most of the tasks used in research have been imple-
mented under controlled (i.e., laboratory-like) conditions rather than in authentic
classroom settings. Some studies in this volume as well as in the broader literature have
begun to gather data by implementing tasks in authentic classrooms, but as extra-
curricular activities, added on to the regular classroom agenda. Such researcher-dom-
inated practices have been shown to generate concerns among learners that they are
subjects in a study rather than students in a classroom (see, e.g., Pica, et al., 2006).
These learner perceptions raise questions about the simultaneous validity of tasks as
both attractive classroom activities but also research tools. Whereas considerable
strides have been made in the design of tasks, there remains a need for approaches
where task designs meet the twin demands of being genuine learning activities as well
as effective research tools within authentic classroom settings.
Research to date has revealed the ways in which tasks can be designed to activate
linguistic and cognitive processes that are needed for successful language learning.
However, achieving learning outcomes is ultimately what students, teachers, and re-
searchers want and deserve. Longitudinal studies of sustained task-based coursework
can reveal the extent to which successful outcomes are possible in the foreign language
 Teresa Pica

classroom, and whether tasks can play a defining role. This type of work can also pro-
vide answers to questions that arise after more typical, one or two week studies, as to
why students all too often fail to retain the very features they had been able to use dur-
ing the studies. Was a task-based intervention withdrawn too early? Had it been poor-
ly designed? Were the students simply not ready to internalize the task form or feature?
What was the teacher’s apparent role in implementation and outcomes? Such ques-
tions require close, long-term observation by the researcher and sustained interest
among the researcher, the teachers, and their students. As this volume includes studies
in classrooms where students might remain for lengthy periods, its chapters provide a
promising basis for addressing these questions.
One final concern about tasks relates to education and language policy and the
restrictions it often places on classroom practice. Many teachers and their students are
eager to join researchers in using tasks, but their involvement is limited by a curricu-
lum that is pre-set by policy and tradition (and often by external assessments as well).
As teachers and researchers become more informed, more professionalized, and more
visible and relevant to each other, so too will opportunities arise for their greater
collaboration and dialogue. As long as teachers and researchers find ways to work to-
gether in the classroom, and remain committed to long-term relationships with lan-
guage learners and with each other, tasks provide principled and informed direction
and guidance. In identifying the opportunities for the study of task-based instruction
in EFL contexts, the editors and chapter authors of Task-based language teaching in
foreign language contexts: Research and implementation serve as a source of motivation
for advancing the importance of tasks and their current and potential contributions to
the field of language studies.

References

Crookes, G., & Gass, S. (Eds.). (1993). Tasks and language learning: Integrating theory and prac-
tice. Clevedon, UK: Multilingual Matters.
Doughty, C., & Long, M. (2003). Optimal psycholinguistic environments for distance foreign
language learning. Language Learning & Technology, 23, 35–73.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Ellis, R. (2009). The differential effects of three types of tasks planning in the fluency, complex-
ity and accuracy in L2 oral production. Applied Linguistics, 30(4), 474–509.
Gass, S., & Alvarez Torres, M. (2005). Attention when? An investigation of the ordering effect of
input and interaction. Studies in Second Language Acquisition, 27, 1–31.
Iwashita, N. (2003). Negative feedback and positive evidence in task-based interaction: Differ-
ential effects on L2 development. Studies in Second Language Acquisition, 25,1–36.
Long, M. H. (1985). A role for instruction in second language acquisition: Task-based language
teaching. In K. Hyltenstam, & M. Pienemann (Eds.), Modeling and assessing second lan-
guage development (pp. 77–99). Clevedon, UK: Multilingual Matters.
 Foreword  

Loschky, L., & Bley-Vroman, R. (1993). Grammar and task-based methodology. In G. Crookes
& S. Gass (Eds.), Tasks and language learning, Vol. 1 (pp. 123–167), Clevedon, UK: Multi-
lingual Matters
Mackey, A. (1999). Input, interaction, and second language development. Studies in Second Lan-
guage Acquisition, 21, 557–588.
Norris, J. M., Brown, J. D., Hudson, T., & Yoshioka, J. (1998). Designing second language perfor-
mance assessment. Honolulu, HI: National Foreign Language Resource Center.
Nunan, D. (1989). Designing tasks for the communicative classroom. Cambridge: Cambridge Uni-
versity Press.
Pica, T. (2009). Integrating content-based and task-based approaches for teaching, learning, and
research. In L. Wei & V. Cook (Eds.), Continuum contemporary applied linguistics, Volume 1.
London: Continuum.
Pica, T., Kang, H., & Sauro, S. (2006). Information gap tasks: Their multiple roles and contribu-
tions to interaction research methodology. Studies in Second Language Acquisition, 28,
301–338.
Prabhu, N. S. (1987). Second language pedagogy. Oxford: Oxford University Press.
Van den Branden, K. (Ed.). (2006). Task-based language education: From theory to practice.
Cambridge: Cambridge University Press.
Van Patten, B. (2002). Processing instruction: An update. Language Learning, 52, 755–803.
Willis, J. (1996). A Framework for task-based learning. Harlow, UK: Longman.
 chapter 1

Introduction
Broadening the perspective of task-based
language teaching scholarship: The contribution
of research in foreign language contexts

Ali Shehadeh
UAE University, United Arab Emirates

This chapter introduces the collection of papers and situates it against previous
and current research on task-based language teaching (TBLT). The chapter
first contextualizes the theme of the volume within the field of TBLT, arguing
that there is a need to broaden the perspective of TBLT scholarship to include
the contribution of research in foreign language (FL) contexts. First, it will be
argued, as attested by several contributions to this volume, that the manner in
which TBLT is researched, implemented, and used in FL contexts depends on
conditions and social practices that do not necessarily coincide with those in
second language (SL) contexts. Next, a brief description of what constitutes a
FL, or English as a foreign language (EFL), context will follow as a prelude to the
discussion of the factors and problems that characterize educational settings in
FL contexts. After that, an overview is provided of the chapters that make up this
collection, showing how each chapter contributes to the theme of the volume,
and signaling how it relates to other chapters. The chapter concludes with a few,
but important remarks, on the scope and significance of the volume, calling for
more rigorous research in other EFL settings in the world, which thus far are
underrepresented within TBLT research.

Introduction

In 2009, Kris Van den Branden, Martin Bygate, and John Norris inaugurated a new
book series entitled Task-Based Language Teaching: Issues, Research and Practice with
a volume entitled Task-Based Language Teaching: A Reader. In their introduction to
the volume, Van den Branden, Bygate, and Norris stated:
... there is widespread agreement that tasks, potentially at least, offer a uniquely
powerful resource both for teaching and testing of language. In particular, they
 Ali Shehadeh

provide a locus for bringing together the various dimensions of language, social
context, and the mental processes of individual learners that are key to learning.
There are theoretical grounds, and empirical evidence, for believing that tasks
might be able to offer all the affordances needed for successful instructed language
development, whoever the learners might be, and whatever the context. (p. 11)

The authors based these statements on the extensive and varied literature on task-
based language teaching (TBLT) that has appeared across many journals, edited
volumes, monographs, and special issues in refereed journals like Language Teaching
Research, Language Testing, and International Review of Applied Linguistics, which
speaks of the potential of TBLT as an approach to second/foreign language (L2) learn-
ing and teaching and as a teaching methodology in which classroom tasks constitute
the main focus of instruction and assessment (see Van den Branden et al., 2009, p. 1).
At the same time, research on TBLT has acquired a special value in light of the
shift of focus in research in the last few years from controlled, laboratory conditions to
authentic, naturalistic or classroom contexts (e.g., Ellis, 2009; Van den Branden,
2006b). For instance, Van den Branden (2006b) suggested that “much of the research
concerning TBLT has been conducted in laboratory conditions or in tightly controlled
settings.... Far less empirical research has been carried out where tasks have been used
as the basic units for the organization of educational activities in intact language class-
rooms” (p. 1). However, Van den Branden also points to the expanding research base
that is being developed in intact classrooms or authentic educational contexts, and
indeed combines “a discussion of task-based pedagogical principles with descriptions
of actual applications of task-based language teaching in response to language educa-
tion problems” (p. 13).
This shift illuminates a number of crucial questions such as whether TBLT works
for teachers and learners in the classroom, whether it inspires language teachers when
they prepare their lessons, whether it is compatible with prevailing classroom prac-
tices, how TBLT can be implemented in classes with a wide range of cultural back-
grounds and different levels of proficiency, and how one writes a task-based syllabus
that may need to cover multiple levels of schooling (Van den Branden, 2006b). There
is no wonder, therefore, that Van den Branden et al. (2009) rightly point to the sub-
stantial challenges associated with the belief that “tasks might be able to offer all the
affordances needed for successful instructed language development, whoever the
learners might be, and whatever the context” because “the institutional, cultural, pro-
fessional, and research challenges involved in establishing whether, or how far, this is
the case are substantial” (p. 11).
One of these challenges is providing sufficient empirical knowledge about TBLT
issues, research, and application in varied and authentic foreign language (FL) educa-
tional settings. In 2009, the Asian Journal of English Language Teaching (AJELT) de-
voted a special issue (Volume 19) to the topic titled Task-based Language Teaching in
Asia: Innovation in Research and Practice. Although the AJELT special-issue was a step
 Chapter 1.  Introduction 

in that direction, the issue was limited to the Asian context only, and mainly to the
challenges involved in implementing TBLT. For instance, the issue’s guest editors,
Jonathan Newton and Rebecca Adams (2009), stated that “The purpose of this special
issue is to present recent research which further investigates what challenges are in-
volved in implementing TBLT in Asia, how these challenges are experienced by
teachers and students, and how the challenges are responded to” (p. 11).
Some edited collections have focused on relevant phenomena, of course. For ex-
ample, the volume by Leaver and Willis (2004) focused on TBLT implementation in FL
contexts; and the volume by Van den Branden, Van Gorp, and Verhelst (2007), based
on papers presented during the first international TBLT conference, was specifically
devoted to studies into the use of tasks in the classroom. Other edited volumes have
also focused on the use of tasks in the language classroom as well as on issues like tasks
for language testing and assessment, and how tasks can be used for specific classroom
activities or whole courses (e.g., Bygate, Skehan, & Swain, 2001; Edwards & Willis,
2005; Garcia Mayo, 2007; Shehadeh & Coombe, 2010; Van den Branden, 2006a).
Other authored books have focused on other TBLT-related issues like the relation-
ship between research, teaching, and tasks (Ellis, 2003), the effect of planning on task
performance (Ellis, 2005), or tasks within a broad educational and social perspective
(Samuda & Bygate, 2008). On the other hand, a lot of the early work on TBLT was car-
ried out in second language (SL) contexts like Ellis (1987), Lyster and Ranta (1997),
Pica (1987), Pica and Doughty (1985a, 1985b), Van den Branden (1997), and many of
the studies in Bygate, Skehan, and Swain (2001). However, no volume or journal spe-
cial-issue was specifically devoted to providing empirical accounts of TBLT research
and practice in varied and authentic FL, particularly English as a Foreign Language
(EFL), contexts (see also Carless, this volume, for other differences between this vol-
ume and others like Edwards and Willis, 2005, and Shehadeh and Coombe, 2010,
which have also in part explored TBLT in EFL settings).
The concern in the current research on TBLT is that knowledge about TBLT in SL
contexts gets naturalized inadvertently as being about TBLT in general, with the impli-
cation that it is universally valid and easily generalizable across other contexts, includ-
ing FL contexts (see also Manchón, 2009a, 2009b; Ortega, 2009, for similar arguments
in relation to L2 writing in FL vs. SL contexts). For instance, in her introductory
chapter to a recent volume (Manchón, 2009a) that is devoted to theory, research, and
pedagogy on L2 writing in FL contexts, Manchón (2009b) argues that “...mainstream
pedagogical discussions have rarely debated whether or not instructional recommen-
dations for SL contexts apply to FL settings” (p. 2). She also pointed out that “the SL
bias of scholarly work in the field [...] means that the bounds of claims of official dis-
course have not been sufficiently tested across diverse contexts (much less across wide-
ly varying EFL contexts)” (pp. 16–17).
Along the same lines, it is possible to make similar arguments about TBLT re-
search. As will be discussed below, and attested by several contributions to this vol-
ume, the manner in which TBLT is researched, implemented, and practiced in FL
 Ali Shehadeh

contexts depends on a whole set of conditions and social practices that do not neces-
sarily coincide with those in SL contexts. In the following section, I will provide a brief
description of what constitutes a FL context (features, characteristics) as a prelude to
the discussion of the factors and problems that typically constitute educational settings
in FL contexts.

What is a foreign language context?

A distinction is often made between a foreign language (FL) context and a second
language (SL) context. An FL context describes a setting in which the teaching of a
language other than the native language usually occurs in the student’s own country
and as school subject only. An SL context, on the other hand, describes a setting in
which a target language other than the learners’ native language is the medium of in-
struction. It also describes a native language in a place as learnt by people living there
who have another first language. For instance, English in the UK would be called the
SL of many immigrants.
Richards and Schmidt (2010) describe the difference between a foreign language
and a second language more extensively as follows: A foreign language is “a language
which is not the native language of large numbers of people in a particular country or
region, is not used as a medium of instruction in schools, and is not widely used as a
medium of communication in government, media, etc. Foreign languages are typically
taught as school subjects for the purpose of communicating with foreigners or for
reading printed materials in the language” (pp. 224–225). For example, English is de-
scribed as a foreign language in France, Japan, China, Venezuela, and other regions. A
second language, on the other hand, refers to “a language that plays a major role in a
particular country or region though it may not be the first language of many people
who use it” (ibid, p. 514). A second language is a language that is widely used as a me-
dium of communication (e.g., in education and in government) and which is usually
used alongside another language or languages. For example, English is described as a
second language in countries such as Singapore, Nigeria, India, and the Philippines.
Strictly speaking, the distinction between a FL and a SL applies to any language. In
practice, however, the distinction is most often applied to English language teaching,
because to date English is the most frequently taught second or foreign language in the
world and because there exists significant research about English-as-second-or-for-
eign language in the world too. For instance, Polio and Williams (2009), describing the
state of writing in English-as-a-foreign language, claim that “English is the dominant
foreign language in most settings outside of North America and is certainly the only
one in which there exists significant research on writing” (p. 494). It comes as no sur-
prise, therefore, that most of the scholarship on TBLT, this volume included, comes
from English as a second and/or foreign language contexts. Thus, an English-as-a-for-
eign language (EFL) context, or English non-dominant region, describes settings in
 Chapter 1.  Introduction 

which English is taught as a school subject in the students’ home location, which is the
case with the collection of studies comprising this volume. Henceforth, therefore, in
this introduction FL settings will be referred to as EFL settings. (See Howatt and
Widdowson, 2004, for fuller discussion of English as a second language and English as
a foreign language; Fujii and Mackey, 2009, and Moore, this volume, for important
distinctions between foreign and second language contexts in task-based learner-
learner interaction; and Sato, 2011, for a discussion of the use of the terms foreign and
second language contexts in SLA research.)1

Factors and problems associated with educational settings in EFL contexts

In spite of all the potential value of TBLT noted above, and in spite of the policies of a
number of governments and educational authorities worldwide that support curricu-
lar innovations in favor of TBLT, traditional, language-centered, and teacher-centered
instruction still persists in many EFL settings. Adams and Newton (2009), citing evi-
dence from a large body of classroom-based research on current English teaching in
Asia, state that “[r]esearch conducted across East Asian contexts has overwhelmingly
suggested that curricular policies have had limited overall impact on English language
teaching, which remains traditional with an explicit grammar-teaching focus” (p. 2).
Such resistance to curricular changes applies to the adoption of TBLT in particular. For
instance, Adamson and Davison (2003) and Luk (2009) found very minimal adoption
of the task-based Hong Kong Target Oriented Curriculum by teachers and schools
(see also Chan, this volume). Similarly, Zhang (2007), describing the situation in
mainland China, speaks of the “limited, sporadic, unsystematic, and sometimes con-
tradictory dissemination of TBLT by various disseminators, including educational au-
thorities, teacher trainers, university scholars, and textbook writers” (p. 76). Similar
observations have been noted in other EFL contexts such as Taiwan (Chao & Wu, 2008),

1. Some authors, e.g., Harmer (2007, pp. 19–21), suggest that the traditional distinction be-
tween EFL and ESL has become difficult to sustain in recent years. First, English has become a
language for international communication, especially on the internet; as a consequence the so-
called foreign language (English) invades learners’ lives more than the so-called second lan-
guage they are learning (when for instance they are living in a region dominated by a language
other than their mother tongue). Second, in many places around the world, communities have
become truly multilingual making it harder to distinguish second languages from foreign lan-
guages. In Flanders, for example, English has become such a dominant foreign language
(on television, on the internet, in the entertainment industry, in business, etc.) that it can hardly
be called a foreign language anymore. Harmer suggests that ESOL (English to Speakers of Other
Languages) is gaining ground as an umbrella term. However, Harmer admits that although
ESOL reflects a more multilingual global reality, it “ignore(s) the context in which language-
learning takes place” (p. 20).
 Ali Shehadeh

Korea (Shim & Baik, 2000), Japan (Jackson, this volume; Gorush, 2000), and Venezuela
(Chacón, this volume).
A number of factors may be identified as challenges for adopting TBLT in many
EFL settings, including administrative constraints, exam pressures, cultural pressures
and expectations, time pressures, and available materials (e.g., Adams & Newton, 2009;
Iwashita & Li, this volume). Following Adams and Newton (2009), these challenges
may be grouped under three main factors: institutional factors, teacher factors, and
student factors. These are considered and discussed separately below.

Institutional factors

A number of institutional factors are shown to hinder the adoption of TBLT in many
EFL settings like focused exams and assessments, large class sizes, and mixed-profi-
ciency classes. First, studies in several EFL settings including China (Hu, 2002; Zhang,
2007), Hong Kong (Chow & Mok-Cheung, 2004), Korea (Shim & Baik, 2000), and
Japan (Gorush, 2000) have shown that the measurement of success in L2 teaching and
learning is frequently sought through norm-referenced, summative, knowledge-based,
vocabulary-and grammar focused exams. These likely hinder successful adoption of
TBLT in the classroom. For instance, Hu (2002) and Zhang (2007) show that grammar
and vocabulary knowledge-based high stakes national examinations are the main bar-
riers preventing the adoption of TBLT in China, because they do not take into account
or reflect communicative curricular objectives. They rely rather too heavily on multi-
ple-choice testing formats, leading administrators and teachers to revert to traditional,
explicit, and rote-learning approaches for teaching to the test (Carless, 2007; Li, 1998;
Littlewood, 2004).
Similarly, large class sizes and mixed-proficiency classes are also frequently cited
as factors that limit the adoption of TBLT courses in many EFL settings, such as Taiwan
(Chao & Wu, 2008), Korea (e.g., Jeon, 2006; Li, 1998), Hong Kong (Carless, 2002),
China (Zhang, 2007), and Venezuela (Chacón, this volume). For instance, Li (1998)
and Littlewood (2007) argue that it is difficult to implement TBLT in large classes be-
cause it requires a range of participatory structures including whole class work, group
work, and pair work which interfere with local notions of good classroom manage-
ment, which is usually defined in these contexts as students working individually and
quietly and not causing any disruption (Cortazzi & Jin, 1996). Likewise, Chao and Wu
(2008) pointed out that many teachers find it difficult to select appropriate tasks for use
in public schools in Taiwan because their students are often streamed by age and not
by level of proficiency. Along the same lines, Chacón (this volume) states that “EFL in
Venezuela [...] is taught in formal classroom situations. In these settings, learners do
not have sufficient opportunities to interact in English in the classroom in part due to
large class sizes” (p. 255).
 Chapter 1.  Introduction 

Teacher factors

A number of teacher factors have been found to challenge the adoption of TBLT in
several EFL contexts. First, many teachers struggle to make sense of new approaches to
language learning and teaching because of uncertainty about the nature of tasks,
doubts about the effectiveness of TBLT itself for their traditional notions of what con-
stitutes language learning, beliefs that tasks are no more than another face of the tradi-
tional exercises and drills, or else they simply do not know how to implement TBLT in
their teaching practices (e.g., Adamson & Davison, 2003; Carless, 2009; Chacón, this
volume; Chan, this volume; Jeon, 2006; Zhang, 2007). Second, many teachers feel more
secure and in control in traditional, teacher-fronted, teacher-centred instruction.
These teachers feel uncomfortable with the shifts in teaching styles and classroom dy-
namics required by TBLT because they feel that these reduce their ‘authority’ from the
role of instructor to that of a facilitator or counsellor (e.g., Carless, 2004, 2007;
McAllister, Narcy-Combes, & Starkey-Perret, this volume). Third, teachers in many
EFL settings consider TBLT an alien concept not applicable to their specific teaching
context or educational setting because it is incompatible with their own experiences of
language learning and teaching. For instance, Cortazzi and Jin (1996) pointed out that
many TBLT principles are incompatible with the Confucian heritage for language edu-
cation in China that advocates the authoritative role of the teacher, memorization, and
rote learning. Many of these teachers consider traditional methods of teaching more
appropriate than TBLT (see also Iwashita & Li, this volume).2

Student factors

Research exploring EFL students’ views and beliefs has also suggested that many stu-
dents express doubts about the effectiveness of TBLT, which both matches their teach-
ers’ views and echoes their conservative parental beliefs about education. In particular,
many students in these settings have expressed a preference for traditional methods of
teaching that promote accuracy over fluency, individual or independent work over
pair- and group work, and reliance on the teacher as an authority figure over taking
risks through speaking, as favored by TBLT courses (e.g., Adamson & Davison, 2003
in Hong Kong; Lee, 2005; Zhang, 2007 in China; Li, 1998 in Korea; Eguchi & Eguchi,
2006 in Japan). For instance, many students in the studies by Eguchi and Eguchi (2006)
and Lee (2005) were reluctant to take risks or make mistakes in English in order to save
face, which undercuts the value of interactive and production tasks necessary for lan-
guage development.

2. It has to be acknowledged that pretty much the same applies in Europe – traditional meth-
ods die-hards can still be found in places like the UK, France, Spain, Italy, Germany and Hungary,
amongst others.
 Ali Shehadeh

Overall, and in spite of perceived value of TBLT and the increasing interest in
TBLT research that is being built up in EFL contexts (such as Ellis, Tanaka, & Yamazaki,
1995; Nobuyoshi & Ellis, 1993; Skehan & Foster, 1997), and by a number of studies in
recent edited collections (e.g., Edwards & Willis, 2005; Garcia Mayo, 2007; Leaver &
Willis, 2004; Shehadeh & Coombe, 2010; Van den Branden, 2006a; Van den Branden
et al., 2007), TBLT research and implementation in many EFL settings still face a
number of challenges. It must be pointed out, however, that these challenges are NOT
problems for/of TBLT itself. Rather they are problems for ANY successful language
teaching in these settings, due to the fact that many of the schooling-related policies
and practices are NOT aligned with notions of effective language pedagogy. For ex-
ample, very large numbers of Confucian-educated international students attempting
to enter the US higher education systems incapable of functional use of English would
suggest that there are real problems with the ‘appropriate’ traditional approaches.
On the other hand, while acknowledging that the tasks of researching and doing
TBLT in some EFL settings are challenging as shown above, it must be mentioned that
there are numerous cases, experiences, and studies which demonstrate success stories
of TBLT research and implementation in these same settings, including McDonough’s
research in Thailand (e.g., McDonough, 2004; McDonough & Chaikitmongkol,
2007, 2010), Robinson’s work in Japan (e.g., Robinson, 2007; Robinson, Cadierno, &
Shirai 2009; Cadierno & Robinson, 2009), and Gilabert’s work on EFL in Spain
(e.g., Gilabert, 2007; Gilabert, Baron, & Llanes, 2009). Against this backdrop, and in
pursuit of encouraging effective language learning, this volume puts together a series
of new studies that explores a range of TBLT research and implementation possibili-
ties, and strengthens and widens the range of EFL contexts represented in research on
TBLT. The volume presents recent research that reports on how TBLT ideas are being
investigated by researchers and language professionals and being successfully put into
practice by teachers and learners in these educational settings.

An overview of the book

It is worth restating that the common thread that provides coherence to this collection
of studies is how tasks are being researched or used in authentic and a wide variety of
EFL settings in order to unravel the complexities of researching and implementing
TBLT in these contexts. All contributions to the book: (a) are focused on task-based
language teaching or task-supported language teaching (Samuda & Bygate, 2008);
(b) are conducted in educational contexts, motivated by local relevance, or have local
implications for TBLT research and practice; (c) are based on actual research studies
that are grounded in the relevant literature, research, and theory; and (d) are moti-
vated by evidence-driven practice. Yet, the book emphasizes the rich range of possi-
bilities, avenues, and issues that this collection of studies opens up in undertaking such
an activity. It describes some of the challenges and illustrates some of the successes in
 Chapter 1.  Introduction 

researching, implementing, and utilizing TBLT in EFL settings, with an aim of making
a fresh and enriching contribution to our knowledge about TBLT research and imple-
mentation in these settings.
The thirteen studies that make up this volume fall into two main sections. Section I
consists of five studies that have investigated how different variables affect learners’
interaction and performance within a TBLT framework, serving as a lead-in to the
study of a number of contextual adaptations of TBLT for implementation, which con-
stitute the focus of Section II. Some variables explored in Section I include the effects
of task complexity and pre-task planning on EFL learners’ oral production, whether
language proficiency mediates the perception of task difficulty, whether intended task
complexity differences are reflected in the language production (i.e., fluency, accuracy,
and complexity) of learners of different proficiency levels, effects of strategic planning
on the accuracy of learners’ performance on spoken and written narrative tasks, effects
of task instructions on text processing and L2 learning, how task structure influences
native speaker interaction, and how EFL learners might benefit from exposure to such
interaction.
Following this introduction, Sasayama and Izumi (Chapter 2) test Skehan’s
(1996, 1998) limited capacity hypothesis and Robinson’s (1995, 2003) cognition hy-
pothesis by investigating the effects of task complexity and pre-task planning on EFL
learners’ oral production in a Japanese educational setting. They collected data from a
population underrepresented in previous studies despite its obvious importance for
both research and teaching, namely, EFL high school students who have limited oral
L2 proficiency. Sasayama and Izumi found that increased task complexity positively
affected the specific measure of syntactic complexity, but negatively affected global ac-
curacy and fluency; whereas planning time positively affected global syntactic com-
plexity, but like task complexity, negatively affected fluency. These findings partially
support and partially disconfirm both Robinson’s and Skehan’s hypotheses, posing
questions about the value of making blanket universal generalizations on the linguistic
consequences of task manipulation. Findings of the Sasayama and Izumi study suggest
the importance of employing task-discourse sensitive measures in investigating the
effect of task complexity on EFL learners’ language use.
Also testing Robinson’s cognition hypothesis and Skehan’s trade-off hypothesis,
Malicka and Levkina (Chapter 3) investigate whether language proficiency mediates
the perception of task difficulty and whether intended task complexity differences are
reflected in the language production (i.e., fluency, accuracy, and complexity) of learn-
ers of different proficiency levels in English. The researchers collected data from
37 participants (20 advanced and 17 pre-intermediate learners of English) who were
undergraduate students at a Spanish university. The participants were provided with
two tasks differing in cognitive complexity levels and manipulated along ± few ele-
ments and ± reasoning dimensions. Two kinds of instruments were used to measure
participants’ perceptions of task difficulty: self-reported difficulty ratings and time es-
timation of task completion. Complexity, accuracy, and fluency measures were used to
 Ali Shehadeh

analyze participants’ speech production on the two tasks. With respect to the partici-
pants’ perception of complexity, the researchers found no significant differences be-
tween the high and low proficiency groups. Regarding performance measures, the
complex task triggered greater lexical and structural complexity and accuracy in the
high proficiency group, but adversely affected fluency. In the low proficiency group,
fluency was boosted on the complex task, while the other areas remained intact irre-
spective of cognitive task complexity. As the researchers state, the main conclusion to
be drawn from this study is that the proficiency level plays a key role in attention al-
location, whereby trade-off effects between fluency, accuracy, and complexity are more
pertinent to lower L2 proficiency learners, which become less marked, if not disappear
altogether, as proficiency increases.
In part like Chapters 2 and 3, Genc (Chapter 4), focusing on the issue of planning,
investigates the effects of strategic planning on the accuracy of learners’ performance
on spoken and written narrative tasks in a Turkish EFL university setting. Most of
Genc’s findings provide further support for research conducted in other, primarily
ESL, settings. Most notably, she found that in unplanned conditions, EFL students
were more accurate on the written tasks than on oral tasks; in contrast, in planned
conditions the effect of modality (oral, written) on accuracy was not influential; stra-
tegic planning did not have a significant effect on accuracy for the oral task while it
appeared to have had an adverse effect on accuracy for the written task; and learners
spent more attention and time on vocabulary and organization than grammar during
the planning time before and during the task performance phases. The main contribu-
tion of Genc’s study is that it extends the study of a number of task-design and perfor-
mance issues, including planning time, accuracy, and oral and written modalities, into
an EFL setting.
Remaining within the territory of task conditions and how they might influence
students’ processing capacity and subsequent learning, Horiba and Fukaya (Chapter 5),
report on a study that investigated the effects of task instructions on text processing
and L2 learning. They collected data from 70 limited L2 proficiency college nursing
students in a Japanese EFL context. The participants processed texts about a health-
care case under three different task conditions. Some students were informed of a re-
call task in one language and later recalled in the same language (the L1-only and the
L2-only conditions), while others recalled in a different language (the L1-L2 condi-
tion). After recall, the participants took a vocabulary test on unfamiliar words contained
in the text. The investigators found that the L1-only condition facilitated content recall
whereas the L2-only condition led to increased incidental vocabulary acquisition.
Horiba and Fukaya argue that these findings imply that there may be a trade-off be-
tween content learning and language learning for limited L2 proficiency students
(i.e., those often the students in EFL contexts). The main contribution of Horiba and
Fukaya’s study to this volume is that it addresses two essential questions on TBLT re-
search in EFL settings: Can limited L2 proficiency students learn both content and
 Chapter 1.  Introduction 

language simultaneously in the L2? and How can task conditions influence students’
text processing capacity and subsequent learning?
In the final contribution to Section I (Chapter 6), Hobbs expands the TBLT re-
search base into a largely unexplored area, namely, how task structure influences na-
tive speaker (NS) interaction, and how EFL learners might benefit from exposure to
such interaction. Hobbs examined NS task performance over a range of nine tasks,
focusing on the ways in which patterns of interaction are influenced by task design,
and showing how recordings of NS task interaction can be used to better equip learn-
ers to perform similar tasks without recourse to L1. The study sought to identify fea-
tures that might offer a basis for predicting performance by other NSs on similar tasks,
and that learners could be expected to benefit from exposure to. For instance, even
when students are accustomed to a particular task type, activities using NS transcripts
can encourage them to experiment with alternative expressions. Findings of Hobbs’
study show that valuable insights can be obtained by analyzing NS task performance,
enabling L2 researchers to gain a deeper understanding of how task design and task
selection can influence interaction. And this, consequently, enables them to make rea-
soned pedagogical recommendations for the implementation and practice of TBLT in
the classroom, including task selection and design, sequencing of tasks, task complex-
ity, and task demands.
Section II consists of eight studies that focus on TBLT contextual adaptation, im-
plementation, or related issues in EFL educational contexts. The studies explore themes
such as patterns of teacher – student and student-student interaction in a TBLT class-
room, effectiveness of learner-generated focus on form in oral task-based interaction,
differences in teachers’ enactment of TBLT, designing and implementing computer-
assisted TBLT lessons, enactment of TBLT in teacher education programs, task-based
language assessment, and teacher perceptions of task-based language teaching.
Following on the issue of interaction raised in Chapter 6 by Hobbs, but this time
from a purely classroom interaction perspective, Iwashita and Li (Chapter 7) investi-
gate patterns of teacher-student and student-student interaction in a task-based oral
EFL classroom in a university setting in China. Eight hours of classroom interaction
data were analyzed for various types of feedback and uptake. The study was conducted
against the backdrop of classroom-based TBLT research in Asian counties in general,
and in China in particular, that has focused to date mainly on the factors that ostensi-
bly hinder the implementation of communicative, TBLT methodologies in these con-
texts. By contrast, these investigators found strong and active student participation in
the classroom, and extensive teacher – student and student-student interaction, de-
spite the large class size and the students’ unfamiliarity with a teaching methodology
that was very different from the traditional Chinese way of learning and teaching. The
main implication and contribution of Iwashita and Li’s findings is that the apparent
successful implementation of TBLT in this study was not only due to the classroom
teacher’s familiarity with TBLT, her strong belief in it, and her good relationship with
her students, but also to the willingness of these students to accept new methodologies
 Ali Shehadeh

and modes of learning which are so different from their past learning experiences,
their old beliefs about learning, and from the traditional methodologies they were ac-
customed to.
In a longitudinal study, Moore (Chapter 8) investigates oral task-based interac-
tion in an undergraduate EFL classroom in a Japanese university context. He analyzed
data obtained through language-related episodes (LREs) from four focal learners and
their partners (N = 8) in two oral presentation tasks for the effectiveness of learner-
generated focus on form (FONF). He also conducted a qualitative microanalysis of
one learner’s interaction with partners of similar proficiency on two similar tasks,
separated by a period of seven months, to investigate the influence of context. Provid-
ing further support for previous studies, Moore found that there was little focus on
form in interaction and there was much variability across dyads. Qualitative analysis
revealed that the effectiveness of FONF in interaction and performance may have
been influenced by the learners’ shared background (including L1 use), individual dif-
ferences in terms of engagement in LREs, learners’ perceptions of each other’s lan-
guage proficiency, and other interpersonally negotiated features of the interaction.
Pedagogically, Moore’s study highlights the potential of attempting to refine learners’
FONF to encourage more on-task negotiation, including forms which may be essen-
tial to task performance.
Chan (Chapter 9) also considers patterns of interaction (who spoke about what, in
response to what, and with what effect) as part of her investigation of how TBLT is
enacted in primary ESL classrooms in Hong Kong. Framed within a qualitative re-
search framework, Chan’s study collected data from 20 lessons taught by four teachers
on the same topic (the weather) from individual lesson plans, teaching materials, in-
terviews with these teachers, and tasks completed by the students. The researcher
found that teachers differed in enacting TBLT in their classrooms along six dimen-
sions: (1) strategic use of visual support to manage task demands; (2) contextualizing
input to make connections between old and new knowledge; (3) simultaneous atten-
tion to task demands for progression in complexity; (4) provision of scaffolding
through task sequencing and adjustment of task variables; (5) creating conditions in
noticing form and salient features; and (6) creating conditions for restructuring to oc-
cur. Chan argues that these findings imply that what is most important in shaping
learning in an authentic TBLT classroom is not the task per se, but rather the inter-
weaving of pedagogic strategies (e.g., scaffolding) at various levels of complexity as
teachers respond to students’ needs in the immediacy of the classroom environment.
An additional, but important, contribution of Chan’s investigation is that it focuses on
pupils (aged between 7–9 years) at lower primary levels, an age group not often studied
in research into TBLT practice or in TBLT theory.
It should be noted here that the decision to include a contribution from Hong
Kong, in which the status of English language teaching might be considered to be ESL
rather than EFL, was based on an important consideration; namely, the educational
setting in Hong Kong includes most of the challenges observed in other EFL settings
 Chapter 1.  Introduction 

for adopting and implementing TBLT principles or framework (see pp. 6 ff above;
Carless, this volume). Indeed, Chan enumerates some of these challenges in the intro-
duction to her chapter and also in her Conclusions and Implications section. For in-
stance, in her concluding section, Chan states: “Local studies and my personal interac-
tion with in-service teachers [in Hong Kong] seem to suggest that teachers find the
concept of TBLT difficult to grasp” (p. 206).
In a research area that has received comparatively little attention, Park (Chapter 10)
illustrates how computer-assisted TBLT lessons can be designed and implemented in a
conventional English classroom in a Korean school setting. He also investigates stu-
dents’ L2 development in writing as well as their perceptions of TBLT. A total of 61
Grade 7 students at a Korean middle school participated in the study. The participants
were divided into an experimental group (N = 30) that was taught with the TBLT les-
son plans, and a control group (N = 31) that was taught in a conventional teacher-
centered and forms-focused approach. For each unit, two task-based writing tests
(pre/post-test) and a conventional unit test on grammar and reading comprehension
were administered. A paired sample t-test of the two groups revealed that the mean
scores of the experimental group were significantly higher than those of the control
group. The experimental group also exceeded the control group in the conventional
unit tests. The main finding of the study is that TBLT can be effective in improving
students’ communicative competence while not hindering form-focused L2 learning.
On the other hand, students and teachers both found the TBLT lessons effective and
motivating. Park discusses a number of implications of the study for EFL teachers as
well as for administrators, curriculum designers, and materials writers.
Like Chan’s and Park’s studies, Chacón’s contribution (Chapter 11), framed within
a qualitative research framework, reports on the successful enactment of TBLT in a
teacher education program in Venezuela. The investigator explored ways of enhancing
EFL students’ oral skills using TBLT through film-oriented activities. She collected
data from 50 third year students enrolled in the program over a ten-week period. Her
data sources included student diaries, student recordings, and focus group interviews.
The investigator found that implementing TBLT through a cooperative learning proj-
ect using films was successful and beneficial for L2 learning in multiple ways, includ-
ing improvements in the students’ fluency and intelligibility in L2, their listening
comprehension, and their vocabulary building skills. Most of the participants in the
study expressed positive attitudes towards the TBLT-based cooperative learning proj-
ect they took part in. At the same time, the project was successful because it fostered
collaboration between learners and facilitated language learning through the use of
authentic input, interaction, and communication tasks in the classroom. In line with
the recent EFL teaching curriculum reforms in Venezuela, which recommend the
adoption of TBLT in teacher education programs in the country, the main implication
of Chacón’s study is that TBLT enabled these would-be teachers to develop their Eng-
lish competence as learners, and at the same time gave them firsthand knowledge
about TBLT methodology and practice in order to utilize it successfully in their future
 Ali Shehadeh

teaching situations. Chacón argues that these findings and implications are quite ma-
jor for the EFL setting of Venezuela -and perhaps for other EFL settings like the ones
explored in this volume- where English is usually taught following a traditional, linear
curriculum in formal classrooms, where there is hardly any time or opportunity for
interaction or negotiation of meaning, and where teachers face numerous constraints
such as lack of materials and school facilities, tight schedules, and large classes
(as documented in Chacón, 2005).
Also focusing on teacher education programs, Jackson (Chapter 12) investigated
aspects of novice teacher cognition among 15 participants in a one-semester, task-
based, undergraduate seminar on language teaching methods in a university setting in
Japan. Jackson collected data from retrospective comments, classroom discourse, and
survey results on the effectiveness of this task-based teacher education approach. He
found that the participants both gained and shared knowledge related to teaching prac-
tice through classroom tasks. In addition, the participants expressed very positive at-
titudes towards the TBLT instructional practices (see also the contribution by Chacón,
this volume). The chapter discusses the potential of task-based teacher training to sup-
port curricular innovation initiatives, like TBLT, which are designed to enhance lan-
guage education in Japan. Jackson concludes that “task-based second language teacher
education offers opportunities for novice teachers to explore roles and responsibilities,
develop a collaborative culture around language teaching, and plan and implement
communicative teaching practices designed for school-based learners” (p. 282).
The study by Weaver (Chapter 13) focuses on assessment and testing in EFL con-
texts utilizing TBLT principles. It is worth noting that within the framework of TBLT
methodology, the main goal of task-based language assessment (TBLA) is measured
against the extent to which assessment can successfully achieve a close link between
the testee’s performance during the test and his/her performance in the real world
(e.g., Ellis, 2003; Long & Norris, 2000; Shehadeh, 2012; Weaver, this volume). For in-
stance, Weaver reminds us at the outset of his chapter that:
At its core, task-based language assessment (TBLA) involves evaluating the degree
to which language learners can use their L2 to accomplish given tasks. A well-
designed and implemented assessment can also provide teachers and language
learners with a detailed account of task performance that can inform future task-
based instruction and L2 development. (p. 287)

Along these lines, Weaver proposes a formative assessment cycle that can inform the
development, implementation and evaluation of assessment tasks in a task-based syl-
labus. He reports the results of a study in which he examined 46 Japanese university
business students’ competence to deliver a PowerPoint presentation in English. He il-
lustrates how a many-faceted analysis of student ratings of their PowerPoint presenta-
tions combined with a discourse analysis of the presentations revealed a significant gap
between the students’ PowerPoint and their English presentation skills, which then
provides focal points for informing future learning and teaching. He also shows how a
 Chapter 1.  Introduction 

significant gap existed between these students’ descriptive and explanatory skills, and
how this too can be a focal point for future learning and teaching. By way of exemplify-
ing this successful formative assessment cycle that is informative, refined, and reliable,
Weaver provides a detailed and compelling account of one participating student’s
competence to deliver a PowerPoint presentation in relation to the different task re-
quirements and multiple assessment criteria. The main implication of Weaver’s study,
in his own words, is that “A formative assessment cycle can help teachers establish a
framework for systematically implementing TBLT in their classrooms” (p. 307).
In the final contribution to Section II, Julie McAllister and her team (Chapter 14)
examine, as part of a large-scale research project, teachers’ self-perceptions and atti-
tudes to learning and teaching after shifting from face-to-face teacher-centered ap-
proaches to computer-mediated and task-based teaching. Their findings are based on
interviews with 14 teachers involved in a task-based blended learning program for
first-year Business English undergraduate university students in Northwest France.
The researchers found that most of the teachers accept and are adapting to the new,
multifaceted role of the teacher implied in the TBLT program. However, some teachers
voiced some concerns and reservations about successful implementation of TBLT, in-
cluding the increased workload associated with the provision of more personalized
support for students, the shift away from a transmission-based approach to teaching,
and some institutional and cultural constraints. The main contribution of the chapter
by McAllister and her team is that it focuses on one of the key factors for successfully
implementing TBLT in EFL settings, namely, the teacher’s role (see Factors and Prob-
lems Associated with Educational Settings in EFL Contexts above). Indeed, the study
is of particular value given that teachers play a key role in the successful implementa-
tion of any TBLT program.
An important theme that runs explicitly or implicitly through all eight contribu-
tions to Section II, in particular those by Chacón (Chapter 11), Jackson (Chapter 12),
and McAllister et al. (Chapter 14) is that successful task-based teacher training/educa-
tion programs are necessary to support curricular innovation initiatives like TBLT,
which are designed to enhance language education in these EFL settings. Such pro-
grams have strong implications for the teacher trainees’ own language learning, for
their self-perceptions and their attitudes to TBLT, and for their language teaching ca-
pabilities, as many of them will become EFL teachers in the future (see Carless, this
volume, for a detailed discussion of the issue).
In the closing chapter, Carless (Chapter 15) contextualizes the volume’s contribu-
tion to TBLT in EFL contexts in particular, and in the wider contexts of TBLT and re-
search in general. Specifically, he identifies and discusses five main themes arising in
the collection: research methodologies used, contextual adaptations to TBLT, TBLT in
Chinese contexts, assessment and TBLT, and teacher education and TBLT. He con-
cludes his chapter and the volume by outlining some issues in task-based language
teaching in need of further exploration.
 Ali Shehadeh

Concluding remarks

It is worth concluding this introductory chapter with a number of remarks: First, as

one might have noticed, the geographical spread of the contributions to this collection
is worth some comment (see also Carless’s chapter, this volume). As can be seen from
the Table of Contents, the majority of the chapters come from Asian settings, with the
remaining including one from Latin America (Venezuela), one from Eastern Europe
(Turkey), and two from Western Europe (Spain and France). Six of the Asian contribu-
tions come from Japan alone. As Carless says in his chapter, this spread may partly
reflect editorial orientation; yet, arguably it could also reflect developmental trends
across Asia in general and Japan in particular. That is, it might simply reflect the likeli-
hood that any worldwide dissemination of TBLT or any other trend is likely to proceed
asymmetrically, spread occurring more easily and quickly in some areas than others
for contextual and chance historical reasons – and that this applies both in terms of
education and research. With the increase in global communications, however, such
dissemination is likely to become more even in the future. Publications such as this
one are intended to promote further spread and dissemination of TBLT. Viewed from
this perspective, this volume of studies adds to the large amount of work addressing
TBLT in EFL learning in Europe in places like Spain, France, Holland, and Germany;
and at the same time encourages exploration of how TBLT is being researched and
practiced in other – largely underrepresented – EFL parts of the world like the Middle
East and the Gulf, Africa, and other Latin American and Eastern European counties.
Second, in spirit with the mission statement of the International Consortium on
Task-Based Language Teaching (ICTBLT), formed in 2005, which states that “The ba-
sic mission of the ICTBLT is to foster excellence in TBLT work – from theoretical,
empirical, and practical perspectives – across the diverse contexts of language educa-
tion worldwide,” this collection of studies reflects the increasing amount of scholarship
underway in exploring and using TBLT in EFL contexts. With its focus on how TBLT
ideas are investigated and being put into practice beyond the contexts represented in
the majority of the literature, the book extends research on TBLT to new boundaries
and accesses new research findings. With this remark in mind, the ultimate aim of this
collection of studies was to explore the potential of TBLT for language learning in
general, and in EFL contexts in particular, and to promote further studies. The volume
provides strong indications that the research and implementation of TBLT in EFL set-
tings is of a different kind because it takes into account different contextual variables
from those found in ESL settings. It is thus a step towards acknowledging the contribu-
tions of TBLT research in EFL settings and the right of such research to question or
resist pedagogies developed for ESL settings. The book puts TBLT research in EFL
contexts on the scene.
Finally, as stated earlier, the collection provides firsthand knowledge about how
TBLT is being researched, implemented, and practiced in EFL contexts. It thus adds to
the ongoing debate on the subject given the “considerable diversity in the theoretical
 Chapter 1.  Introduction 

scope, research, and applied practice that corresponds with the TBLT name” (Van den
Branden et al., 2009, p. x). I therefore believe that the collection will appeal to SLA
researchers and research students in applied linguistics. Likewise, because a number of
the contributions explore the various ways in which TBLT principles may be incorpo-
rated in the curriculum or utilized by the classroom teacher, I believe that the book
will be of value too to course designers and language teachers who come from a broad
range of formal and informal educational settings encompassing a wide range of ages
and types of language learners.

References

Adams, R., & Newton, J. (2009). TBLT in Asia: Constraints and opportunities. Asian Journal of
English Language teaching, 19, 1–17.
Adamson, B., & Davison, C. (2003). Innovation in English language teaching in Hong Kong
primary schools: One step forward, two steps sideways? Prospect, 18(1), 27–41.
Bygate, M., Skehan, P., & Swain, M. (Eds.). (2001). Researching pedagogic tasks: Second language
learning, teaching, and testing. Harlow, UK: Longman.
Cadierno, T., & Robinson, P. (2009). Language typology, task complexity and the development
of L2 lexicalization patterns for describing motion events. Annual Review of Cognitive Lin-
guistics, 6, 245–277.
Carless, D. (2002). Implementing task-based learning with young learners. English Language
Teaching Journal, 56(4), 389–396.
Carless, D. (2004). Issues in teachers’ reinterpretation of a task-based innovation in primary
schools. TESOL Quarterly, 38, 639–662.
Carless, D. (2007). The suitability of task-based approaches for secondary schools: Perspectives
from Hong Kong. System, 35, 595–608.
Carless, D. (2009). Revisiting the TBLT versus P-P-P debate: Voices from Hong Kong. Asian
Journal of English Language teaching, 19, 49–66.
Chacón, C. (2005). Teachers’ perceived efficacy among English as a foreign language teachers in
middle schools in Venezuela. Teaching and Teacher Education, 21, 257–272.
Chao, J. C., & Wu, A. M. (2008). A study of task-based learning and teaching in a large EFL class.
Saarbrücken: Verlag Dr. Muller.
Chow, A. K. W., & Mok-Cheung, A. H. M. (2004). English language teaching in Hong Kong
SAR: Tradition, transition, and transformation. In H. W. Kam & R. Y. L. Wong (Eds.). Lan-
guage policies and language education: The impact in East Asian countries in the next decade
(pp. 150–177). Singapore: Times Academic Press.
Cortazzi, M., & Jin, L. (1996). English teaching and learning in China. Language Teaching 29,
61–80.
Edwards, C., & Willis, J. (Eds.). (2005). Teachers exploring tasks in English language teaching.
London: Palgrave Macmillan.
Eguchi, M., & Eguchi, K. (2006). The limited effect of PBL on EFL learners: A case study of Eng-
lish magazine projects. Asian EFL Journal, 8(3), 207–225.
Ellis, R. (1987). Interlanguage variability in narrative discourse: Style shifting in the use of the
past tense. Studies in Second Language Acquisition, 9, 1–19.
 Ali Shehadeh

Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Ellis, R. (Ed.). (2005). Planning and task performance in a second language. Amsterdam: John
Benjamins.
Ellis, R. (2009). The differential effects of three types of tasks planning in the fluency, complex-
ity and accuracy in L2 oral production. Applied Linguistics, 30(4), 474–509.
Ellis, R., Tanaka, Y., & Yamazaki, A. (1995). Classroom interaction, comprehension, and the
acquisition of L2 word meanings. Language Learning, 44, 449–491.
Fujii, A., & Mackey, A. (2009). Interactional feedback in learner-learner interactions in a task-
based EFL classroom. International Review of Applied Linguistics, 47, 267–301.
Garcia Mayo, M. (Ed.). (2007). Investigating tasks in formal language learning. Clevedon, UK:
Multilingual Matters.
Gilabert, R. (2007). The simultaneous manipulation of task complexity along planning time and
[± Here-and-Now]: Effects on L2 oral production. In M.P. Garcia Mayo (Ed.). Investigating
tasks in formal language learning (pp. 44–68). Clevedon, UK: Multilingual Matters.
Gilabert, R., Baron, J., & Llanes, M. (2009). Manipulating task complexity across task types and
its influence on learners’ interaction during oral performance. International Review of Ap-
plied Linguistics, 47, 367–395.
Gorush, G. J. (2000). EFL educational policies and educational cultures: Influences on teachers’
approval of communicative activities. TESOL Quarterly, 34, 675–710.
Harmer, G. (2007). The Practice of English Language Teaching (4th ed.). Harlow, UK: Pearson
Longman.
Howatt, A.P.R., & Widdowson, H.G. (2004). A history of English language teaching (2nd ed).
Oxford: Oxford University Press.
Hu, G. (2002). Recent important developments in secondary English-language teaching in the
People’s Republic of China. Language, Culture and Curriculum 15(1), 30–49.
International Consortium on Task-Based Language Teaching (ICTBLT). Link: http://cteno.be/
downloads/ictblt_support_letter.pdf (accessed on 13 June, 2011).
Jeon, I. J. (2006). EFL teachers’ perception of task-based language teaching: With a focus on
Korean secondary classroom practice. Asian EFL Journal, 8, 192–206.
Leaver, B. L., & Willis, J. R. (Eds.). (2004). Task-based instruction in foreign language education:
Practices and programs. Washington, DC: Georgetown University Press.
Lee, S. M. (2005). The pros and cons of task-based instruction in elementary English classes.
English Teaching, 60, 185–205.
Li, D. F. (1998). “It’s always more difficult than you plan and imagine”: Teachers’ perceived dif-
ficulties in introducing the communicative approach in South Korea. TESOL Quarterly, 32,
677–703.
Littlewood, W. (2004). The task-based approach: Some questions and suggestions. English Lan-
guage Teaching Journal, 58(4), 319–326.
Littlewood, W. T. (2007).Communicative and task-based language teaching in East Asian class-
rooms. Language Teaching 40, 243–249.
Long, M. H., & Norris, J. M. (2000). Task-based language teaching and assessment. In M. Byram
(Ed.). Encyclopedia of language teaching (pp. 597–603). London: Routledge.
Luk, J. (2009). Preparing EFL students for communicative task performance: The nature and
role of language knowledge. Asian Journal of English Language teaching, 19, 67–90.
Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake: Negotiation of form in
communicative classrooms. Studies in Second language Acquisition, 19, 37–66.
 Chapter 1.  Introduction 

Manchón, R. (Ed.). (2009a). Writing in foreign language contexts. Clevedon, UK: Multilingual
Matters.
Manchón, R. (2009b). Introduction: Broadening the perspective of L2 writing scholarship: The
contribution of research on foreign language writing. In R. Manchón (Ed.) (2009a),
pp. 1–22. Clevedon, UK: Multilingual Matters.
McDonough, K. (2004). Learner-learner interaction during pair and small group activities in a
Thai EFL context. System, 32, 207–224.
McDonough, K., & Chaikitmongkol, W. (2007). Teachers’ and learners’ reactions to a task-based
EFL course in Thailand. TESOL Quarterly, 41(1), 107–132.
McDonough, K., & Chaikitmongkol, W. (2010). Collaborative syntactic priming activities and
EFL learners’ production of wh-questions. The Canadian Modern Language Review, 66(6),
817–842.
Newton, J., & Adams, R. (Eds.). (2009). Task-based language teaching in Asia: Innovation in
research and practice (special issue). Asian Journal of English Language teaching, 19.
Nobuyoshi, J., & Ellis, R. (1993). Focused communication tasks and second language acquisi-
tion. ELT Journal, 47, 203–110.
Ortega, L. (2009). Studying writing across EFL contexts: Looking back and moving forward. In
R. Manchón (Ed.) (2009a), pp. 209–231. Clevedon, UK: Multilingual Matters.
Pica, T. (1987). Second-language acquisition, social interaction, and the classroom. Applied Lin-
guistics, 8, 3–21.
Pica. T., & Doughty, C. (1985a). Input and interaction in the communicative language class-
room: A comparison of teacher-fronted and group activities. In S. Gass & C. Madden (Eds.).
Input in second language acquisition (pp. 115–132). Rowley, MA: Newbury House.
Pica. T., & Doughty, C. (1985b). The role of group work in classroom second language acquisi-
tion. Studies in Second Language Acquisition, 7, 233–248.
Polio, C., & Williams, J. (2009). Teaching and testing writing. In M. Long, & C. Doughty, (Eds.).
The handbook of language teaching (pp. 486– 517). Malden, MA: Blackwell.
Richards, J., & Schmidt, R. (2010). Longman dictionary of language teaching and applied linguis-
tics (4th ed.) London: Longman.
Robinson, P. (1995). Task complexity and second language narrative discourse. Language Learn-
ing, 45, 99–140.
Robinson, P. (2003). The cognitive hypothesis of adult, task-based language learning. Second
Language Studies, 21, 45–107.
Robinson, P. (2007). Task complexity, theory of mind, and intentional reasoning: Effects on L2
speech production, interaction, uptake and perceptions of task difficulty. International Re-
view of Applied Linguistics, 45, 193–214.
Robinson, P., Cadierno, T., & Shirai, Y. (2009). Time and motion: Measuring the effects of
the conceptual demands of tasks on second language production. Applied Linguistics, 28,
533–554.
Samuda, V., & Bygate, M. (2008). Tasks in second language learning. Basingstoke, UK: Palgrave
Macmillan.
Sato, M. (2011). Constitution of form-orientation: Contributions of context and explicit knowl-
edge to learning from recasts. The Canadian Journal of Applied Linguistics, 14, 1–28.
Shehadeh, A. (2012). Task-based language assessment: Components, development, and imple-
mentation. In C. Coombe, P. Davidson, B. O’Sullivan, & S. Stoynoff (Eds.). The Cambridge
guide to second language assessment (pp. 156–163). Cambridge: Cambridge University
Press.
 Ali Shehadeh

Shehadeh, A., & Coombe, C. (Eds.). (2010). Applications of task-based learning in TESOL.
Alexandria, VA: TESOL.
Shim, R., & Baik, M. (2000). South and North Korea. In H. W. Kam & R. Y. L. Wong (Eds.).
Language policies and language education: The impact in East Asian countries in the next
decade (pp. 246–261). Singapore: Times Academic Press.
Skehan, P. (1996). A framework for the implementation of task-based instruction. Applied Lin-
guistics, 17 (1), 38–62.
Skehan, P. (1998). A cognitive approach to language learning. Oxford: Oxford University Press.
Skehan, P., & Foster, P. (1997). Task type and task processing conditions as influences on foreign
language performance. Language Teaching Research, 1, 185–211.
Van den Branden, K. (1997). Effects of negotiation on language learners’ output. Language
Learning, 47, 589–636.
Van den Branden, K. (Ed.). (2006a). Task-based language education: From theory to practice.
Cambridge: Cambridge University Press.
Van den Branden, K. (2006b). Introduction: Task-based language teaching in a nutshell. In K.
Van den Branden (Ed.). (2006a), pp. 1–16. Cambridge: Cambridge University Press.
Van den Branden, K., Bygate, M., & Norris, J. (Eds.). (2009). Task-based language teaching: A
reader. Amsterdam: John Benjamins.
Van den Branden, K., Van Gorp, K., & Verhelst, M. (Eds.). (2007). Tasks in action: Task-based
language education from a classroom-based perspective. Newcastle: Cambridge Scholars
Press.
Zhang, E. Y. (2007). TBLT innovation in primary school English language teaching in mainland
China. In K. Van den Branden, K. Van Gorp, & M. Verhelst (Eds.). Tasks in action: Task-
based language education from a classroom-based perspective (pp. 68–91). Newcastle:
Cambridge Scholars Press.
section i

Variables affecting task-based language

learning and performance
 chapter 2

Effects of task complexity and pre-task

planning on Japanese EFL learners’ oral
production

Shoko Sasayama and Shinichi Izumi

University of Hawai‘i at Manoa, USA and Sophia University, Japan

This study sought to test Skehan’s (1996, 1998) limited capacity hypothesis and
Robinson’s (1995, 2003) cognition hypothesis by investigating the effects of task
complexity and pre-task planning on EFL learners’ oral production (see also
Genc, this volume). Twenty three Japanese-L1 high school students were given
two sets of picture-based narrative tasks: a simple task with fewer characters and
a complex task with more characters appearing in cartoon-based stories. Ten of
these participants were given pre-task planning time, whereas thirteen were not.
The results indicate that (a) the increased task complexity positively affects the
specific measure of syntactic complexity, but negatively affects global accuracy
and fluency; and (b) planning time positively affects global syntactic complexity,
but negatively affects fluency. These findings partially support and partially
disconfirm both Robinson’s and Skehan’s hypotheses, posing questions about
making blanket predictions on the linguistic consequences of task manipulation.
The findings also show the importance of employing task-discourse sensitive
measures in investigating the effect of task complexity on learners’ language use.
The main value of the findings of the study comes from its focus on Japanese
EFL high school students who have limited oral L2 proficiency, a population
underrepresented in previous studies, despite their obvious importance for both
research and teaching.

Introduction

The use of tasks in second language (L2) classrooms has recently been gaining enor-
mous popularity. Unlike traditional exercises which focus exclusively on the manipu-
lation of language forms, tasks are designed to encourage learners to pay primary
attention to meaning and simultaneously attend to the form that is necessary to convey
meaning. Tasks, therefore, are believed to be useful tools to promote the development
of the form-meaning connections that are crucial for L2 learning. However, it is still an
 Shoko Sasayama and Shinichi Izumi

open question as to what kinds of tasks can effectively trigger such learning processes
and under which conditions. Many SLA researchers (e.g., Bygate, Skehan, & Swain,
2001; Ellis, 2003, 2005; Ortega, 1999, 2007; Robinson, 2001a, 2001b, 2005; Skehan,
1998) acknowledge that various factors mediate the learning processes. One such fac-
tor that has received considerable attention in SLA research is the role of pre-task plan-
ning. The question of interest is whether and how giving learners time to plan their
utterances before engaging in the task affects their language use in terms of complex-
ity, accuracy, and fluency. Another factor that is hypothesized to mediate the effects of
task lies within the task itself, that is, the question of how the complexity inherent to
the task itself affects the learners’ language use. The current study focuses on these two
factors in an effort to further our understanding of the processes involved in task-
based L2 use and learning. Specifically, it manipulates two independent variables, ±few
elements and ±planning time, in order to examine their respective, as well as their
combined, effects on the production of oral narratives by Japanese EFL learners.

Task complexity and L2 task performance

Task performance and attentional capacity

In the current SLA literature, two major competing claims exist regarding how task-
related variables affect learners’ performance. On the one hand, Skehan (1996, 1998)
assumes a single-resource model of attention and claims that learners are not capable
of paying simultaneous attention to the three main aspects of language use: complex-
ity, accuracy, and fluency. He argues that attention to one aspect is likely to compro-
mise attention to the others. Thus, attention to complexity, for instance, likely results
in decreased accuracy, and vice versa. This is known as the ‘limited capacity’ hypoth-
esis (also referred to as the ‘trade-off ’ hypothesis). Robinson (1995, 2001a, 2001b,
2003, 2005, 2007), on the other hand, assumes a multiple-resource model of attention
and argues that learners are capable of attending to different aspects of language per-
formance as needs arise in the task. He views structural complexity and accuracy as
arising from functional complexity in discourse and, hence, increased functional de-
mands imposed by the task should have detectable linguistic consequences. In this
model, known as the cognition hypothesis, concurrent attention to different aspects of
L2 use is considered not just possible, but natural.
These two opposing positions held by Robinson and Skehan derive from the differ-
ent theoretical frameworks adopted in the conception of task complexity. In his Tri-
adic Componential Framework for task design, Robinson (2005) makes distinctions
between three categories of factors: task complexity, which concerns inherent task
characteristics relating to cognitive demands posed to the participants; task difficulty,
which relates to learner factors as reflected in learners’ perceptions of how difficult
tasks are; and task conditions, which concerns interactive demands of the task. Task
 Chapter 2.  Effects of task complexity and planning on oral production 

complexity is further divided into two categories: resource-directing and resource-­

depleting dimensions. Resource-directing dimensions are those in which particular
features of the task direct learners’ attention to specific aspects of the language; for ex-
ample, whether the event to be described focuses on here-and-now or there-and-then
([±here-and-now]), and whether the task requires elaborating on the reasons for deci-
sion/choice made or not ([±no reasoning demands]).1 Increasing task complexity
along resource-directing dimensions is expected to lead to interlanguage development,
as learners are pushed to produce linguistically more complex and more accurate
speech to meet increased functional demands. In the case of resource-depleting di-
mensions, increasing task complexity along these dimensions disperses, not directs,
learners’ attention over many non-specific areas of the L2 performance; for example,
whether planning time is allowed prior to the engagement in the main task ([±planning]),
and whether the learner has relevant background knowledge to perform the task or not
([±prior knowledge]). Although these factors do not by themselves push learners to
direct their attention to any particular linguistic forms, a decrease in complexity in
these dimensions is hypothesized to bolster the attention-directing effect of the in-
crease in the complexity of the resource-directing dimensions.
Skehan (1996, 1998), on the other hand, argues for analysing the complexity of
tasks based on three principal areas: language, cognition, and performance conditions.
Language is concerned with code complexity, that is, how complex the language is re-
quired to be for completion of the task. Cognition is concerned with cognitive com-
plexity, referring to the intrinsic cognitive demands posed to the participants and to
the amount of task-related knowledge or information the learners possess. Perfor-
mance conditions are concerned with the communicative stress the task places on the
participants during task performance. Skehan contends that all these factors affect
how learners allocate their attention during a task, as well as what L2 performance
results. However, given that L2 learners’ knowledge of the language is still limited and
is under conscious control, which takes up considerable attentional resources, cogni-
tively demanding tasks are likely to draw their attention away from language forms
and encourage them to rely on already developed automatized L2 knowledge. This
may have the consequence of learners achieving the task outcome by sacrificing com-
plexity along with any accompanying potential for IL development (“the safety first
approach” – Skehan & Foster, 2001, p. 189). Conversely, learners may attempt to stretch
their IL capacity to its limit, with the consequence of increasing complexity at the ex-
pense of accuracy (“the accuracy last approach” – Skehan & Foster, 2001, p. 189). To
alleviate these problems, Skehan views pre-task planning as a particularly useful exter-
nally-manipulable performance condition that can regulate cognitive load. The op-
portunity to plan ahead of the task is believed to allow the learners extra processing

1. In all these cases, plus (+) denotes lower task complexity and the minus (-) denotes higher
task complexity.
 Shoko Sasayama and Shinichi Izumi

space so that they can allocate their limited attentional resources strategically to focus
on formal aspects during task performance.
Based on these different models, Robinson and Skehan make contrasting predic-
tions about the effects of cognitive task complexity. Robinson hypothesizes that
increasing task complexity along resource-directing dimensions has the effect of si-
multaneously improving complexity and accuracy, while fluency may be negatively
affected. Where resource-depleting dimensions are concerned, an increase in task
complexity is hypothesized to negatively influence all three aspects of L2 performance.
In contrast, Skehan argues that when cognitive task complexity is high, accuracy and
complexity enter into severe competition for attentional resources and, therefore,
these two aspects cannot be promoted simultaneously. He argues, however, that by
providing pre-task planning time, which can be viewed as decreasing task complexity
along a resource-depleting dimension in Robinson’s terms, learners can allocate their
attention with greater ease, with the consequence of improving some, but not neces-
sarily all, aspects of task performance. When cognitive task complexity is increased by
depriving learners of pre-task planning time, Skehan’s argument coincides with Rob-
inson’s, in that all three aspects of task performance are likely to be negatively
affected.

Previous studies on the effects of planning on L2 performance

A number of studies have examined how planning affects learners’ L2 production

(e.g., Ellis, 1987; 2005; Ellis & Yuan, 2004; Foster & Skehan, 1996; Mehnert, 1998;
Ortega, 1999; Yuan & Ellis, 2003). To take a couple of examples here for illustration,
Foster and Skehan (1996) used different tasks to examine the effects of two planning
conditions – guided and unguided planning conditions – on the learners’ L2 use. The
results showed that planning had a greater effect on accuracy than on complexity or
fluency for the personal information exchange task, whereas it promoted complexity
and fluency in the case of the narrative task. The results of the narrative task obtained
here are also corroborated by most other studies utilizing this type of task. Moreover,
guided planners were found to be more fluent than unguided planners in the narrative
task, but no obvious differences emerged for the other tasks. From these results, Foster
and Skehan argued for the competence-expanding role of pre-task planning, while
admitting that simultaneous improvement of complexity, accuracy and fluency is hard
to attain. Another study was conducted by Ortega (1999), who examined the effects of
pre-task planning on L2 Spanish learners’ linguistic performance in a story-retelling
task. She found that syntactic complexity and fluency were positively influenced by
planning, but lexical complexity was not. The results for accuracy were mixed, with
significant effects found for noun-modifier agreement, but not for articles. Ortega also
investigated the process of planning by administering retrospective interviews with
the participants and found quite a bit of individual variability in the extent to which
learners attended to form or meaning during pre-task planning. It appears that learners’
 Chapter 2.  Effects of task complexity and planning on oral production 

interpretations of the task requirements, as well as their general orientation to the task
and their L2 proficiency, mediate what they do during planning time.
In general, three robust findings emerge from the results of previous studies on
planning: (a) pre-task planning typically promotes fluency and syntactic complexity of
the learners’ L2 performance; (b) results for accuracy are inconsistent, as some studies
show greater accuracy in the planning condition for some forms or measures, but not
others, and other studies show no detectable effects of planning on accuracy; and
(c) all three aspects of complexity, accuracy, and fluency may not be improved simul-
taneously by pre-task planning. In spite of these findings, there is still uncertainty re-
garding what inherent task characteristics interact with pre-task planning to affect the
learners’ L2 performance (see also, Genc, this volume).2

Previous studies on the effects of inherent task characteristics on L2 performance

Other researchers have focused on the effects of inherent task characteristics

(i.e., resource-directing dimensions of task complexity) on learners’ performance
(e.g., Gilabert, 2007; Ishikawa, 2007; Kuiken & Vedder, 2007; Robinson, 1995, 2001b,
2007). Robinson (1995), for example, used monologic narrative tasks and manipulated
task complexity along the ±here-and-now dimension. The results indicated that the
complex task significantly promoted lexical complexity, with a trend for greater accu-
racy and greater dysfluency as well. No statistically significant differences were ob-
served for syntactic complexity, however. In a more recent study, Robinson (2007)
used an interactive narrative task and manipulated task complexity along the ±reason-
ing demands dimension. The results indicated no statistically significant differences
among the different tasks in accuracy, syntactic complexity or fluency, while signifi-
cant effects on lexical complexity were found for simpler tasks. These results are in
contradiction to the cognition hypothesis. Robinson attributed these results to his fail-
ure to control for the resource-depleting dimensions adequately. In addition to using
general measures, Robinson also used in this study theoretically-motivated specific
measures of lexical and syntactic complexity. Based on findings in L1 acquisition re-
search, Robinson hypothesised that increasing conceptual demands of reasoning
would lead learners to use more cognitive state terms (e.g., “think,” “know”) and com-
plex syntactic complementations (e.g., “he thinks that he is ...”). The analysis using
these specific measures indeed showed this to be the case.
Another recent study was conducted by Ishikawa (2007), who used a story-writing
task and examined the effects of manipulating task complexity along a resource-­
directing dimension of ±here-and-now. All participants were given five minutes to

2. Learners’ task performance may likely be affected by manipulations of other variables as

well. For instance, task repetition studies (e.g., Bygate, 1996, 2001; Bygate & Samuda, 2005) raise
the possibility that different aspects of complexity, accuracy and fluency could be affected on
different iterations of task performance.
 Shoko Sasayama and Shinichi Izumi

plan and thirty minutes to write a story. The results revealed that syntactic complexity
and accuracy were significantly higher in the complex task condition, with fluency be-
ing negatively affected. Gilabert (2007) used monologic, oral narrative tasks and ma-
nipulated task complexity along the resource-directing dimension of ±here-and-now
and the resource-depleting dimension of ±planning. The results revealed that plan-
ning significantly promoted greater fluency and lexical complexity, but not syntactic
complexity or accuracy. As for the resource-directing dimension, the here-and-now
conditions elicited significantly more fluent speech, whereas the there-and-then con-
ditions promoted significantly greater accuracy. There was also some indication that
the learners’ performance in the there-and-then task was enhanced when given plan-
ning time, albeit non-significantly.
In sum, these previous studies lend only partial support to Robinson’s cognition
hypothesis. It appears that although lexical complexity and accuracy may improve si-
multaneously, syntactic complexity and accuracy may not. The only exception is
Ishikawa (2007), who found a simultaneous increase in syntactic complexity, lexical
complexity and accuracy in the complex task. This may be due to the use of a written
task in his study, in which the processing load may have been eased for learners more
than would have been the case in a speaking task. In continued investigation of the
cognition hypothesis, particularly needed are studies that examine the effects of ma-
nipulation in the resource-directing dimensions and resource-depleting dimensions
concurrently, so that the validity of both Skehan’s and Robinson’s claims can be exam-
ined in a comparable manner. The study reported below constitutes an attempt to do
this by investigating the effects of both pre-task planning and ±few elements in the
same study.

Research aims and hypotheses

Three unique features of the current study should be noted. First, as mentioned earlier,
this study is one of the first attempts to examine the possibly synergistic (mutually
potentiating) effects of manipulating task complexity along both resource-directing
and resource-depleting dimensions on L2 learners’ task performance. Second, in addi-
tion to using global measures, the current study employed theoretically-motivated
specific measures in analysing L2 task performance. Robinson’s cognition hypothesis
predicts that tasks made complex along resource-directing dimensions prompt learn-
ers to pay attention to “task relevant, communicatively non-redundant language”
(Robinson, 2001b, p. 35). If so, it is reasonable to expect that specific measures focus-
ing on task-relevant forms can capture the effects of such manipulation better than can
global measures (see the thematic issue of Applied Linguistics, 2009, 30(4), for exten-
sive discussion on this and other related issues). In our study, we looked at the use of
noun modifiers, as these are the forms most likely to be affected by the manipulation
of ±few elements. Third, in terms of our study participants, we focused on Japanese
EFL high school students who had limited oral L2 proficiency – a population thus far
 Chapter 2.  Effects of task complexity and planning on oral production 

not highlighted enough in previous studies, despite their obvious importance for both
research and teaching.
Two main research questions (RQ) guided the study:
Research Question 1. What effects does the manipulation of task complexity along a
resource-directing dimension ([±few elements]) and a resource-depleting dimension
([±planning]) have on L2 learners’ task performance in terms of complexity, accuracy,
and fluency?
Research Question 2. How does analysis using specific selected language measures dif-
fer from analysis using general measures of L2 task performance for complexity and
accuracy?
Based on the predictions made by Robinson and Skehan in their respective theoretical
frameworks as outlined above, the following hypotheses were formulated:
Hypothesis 1: A complex task will elicit more complex and more accurate, though less
fluent, speech than a simple task (RQ1 focusing on the impact of a resource-directing
factor).
Hypothesis 2: Specific measures of complexity and accuracy will provide more eviden-
tial support to Robinson’s cognition hypothesis than will general measures of L2 task
performance (RQ2).
Hypothesis 3: Planners will produce more complex and more fluent speech than will
non-planners, while the effects on accuracy will be restricted (RQ1 focusing on the
impact of a resource-depleting factor).
Hypothesis 4: A decrease in complexity in the resource-depleting dimension
(i.e., [+planning]) will bolster the attention-directing effect of the increase in the re-
source-directing dimension (i.e., [–few elements]) (RQ1 focusing on the interaction
between both resource-directing and resource-depleting factors).

Methodology

Participants

Twenty-three Japanese EFL high school students participated in the study on a volun-
tary basis. They were students in year 11, ranging in age from 16 to 17. There were
seven male and 16 female students. The regular English classes these students took in
the school were streamed, and all participants were enrolled in the highest level class.
However, the class streaming was done based on the students’ knowledge of written
English, as is typically the case with EFL in Japan. Therefore, it turned out that as a
group their oral proficiency was variable and generally limited.
 Shoko Sasayama and Shinichi Izumi

Tasks and procedures

Participants engaged in two monologic narrative tasks of different levels of complexity.

Task complexity was varied along the resource-directing dimension of ±few elements.
A simple task required learners to narrate a sequenced set of pictures with two charac-
ters, which was adapted from the Picture Arrangement subset of the Wechsler Adult
Intelligence Scale-Revised, Japanese version at level 3 (Shinagawa, Kobayashi, Fujita, &
Maekawa, 1990). A complex task required learners to narrate a sequenced set of pic-
tures adopted from Elder and Iwashita (2005). These pictures contained four male
characters, one female, one policeman, two ambulance attendants, and a dog. The par-
ticipants, therefore, had to differentiate between these characters, which made this
task cognitively more complex.
The experiment was conducted on two separate days with six other collaborators
in a one-on-one setting.3 Participants in the experiment were asked to narrate a story
based on the given set of pictures, so that a listener – a collaborator or one of the re-
searchers – could correctly sequence the same set of pictures based on their narration.
The listener, however, did not interact with the speaker except to occasionally encour-
age the speaker to keep on speaking. All participants engaged in both the simple and
complex tasks. The order of the tasks was counterbalanced. At the same time, for both
tasks, half of the participants were assigned to the planning condition (n = 10), and the
other half to the non-planning condition (n = 13). Those in the planning condition
were given five minutes of pre-task planning time for both tasks, and those in the non-
planning condition were given only one minute to familiarize themselves with the
picture stories.4 The instructions given for the planners were as follows (original in
Japanese): “Consider in detail ‘what’ you would like to say, and ‘how’ and ‘in what or-
der’ you would like to say it. You can take a memo, but you will not be allowed to look
at your memo during the task.” No questions or access to a dictionary were allowed
during planning. No time limit was set for the completion of the task.5

3. These collaborators were graduate students majoring in TESOL or applied linguistics. All of
them participated in a thirty-minute training session prior to the data collection.
4. Although Mehnert (1998) found in her monologic task that complexity was promoted only
when participants were given ten-minute planning time, our pilot study showed that ten minutes
was too long especially for a simple task. Since the experimental design of our study necessitated
that the length of the planning time be equivalent for the two tasks, we made a compromise
decision of five minutes as adequate time to be given to both tasks.
5. Planners spent 84.30 seconds on average for the simple task (SD = 31.61) and 155.70 sec-
onds for the complex task (SD = 65.79). Non-planners, on the other hand, spent 50.85 seconds
for the simple task (SD = 16.02) and 110.46 seconds for the complex task (SD = 28.09).
 Chapter 2.  Effects of task complexity and planning on oral production 

Data analysis

Participants’ utterances were transcribed and divided into T-units and dependent claus-
es. The T-unit was defined as “a main clause plus all subordinate clauses and non-clausal
structures attached to or embedded in it” (Hunt, 1970, p. 189). In this study, the T-unit
was chosen over the C-unit or AS-unit, as our one-way monologic tasks elicited few el-
liptical utterances. The dependent clause was defined to include coordination,6 subordi-
nation, and embedding (Celce-Murcia & Larsen-Freeman, 1999). To address our
research hypotheses, data were analysed using two-way repeated-measures analyses of
variance (ANOVA), using one between-subject (i.e., with or without planning time) and
one within-subject (i.e., two levels of task complexity) factorial design. The dependent
variables in the current study were analyzed by seven measures as discussed below.
Syntactic complexity was analyzed by clauses per T-unit, which was calculated fol-
lowing the formula of the total number of clauses divided by the total number of
T-units. The same measure has been used by other researchers such as Foster and
Skehan (1996) and Robinson (2001b, 2007) for L2 spoken corpora and Ishikawa (2007)
for L2 writing corpora. Clauses included both dependent clauses and independent
clauses. In addition to this global measure, syntactic complexity was also analyzed by a
specific measure, that is, by examining the use of noun modifiers. We decided to focus
on these particular forms because these were the task-relevant forms that may be most
likely to be affected by the manipulation of ±few elements. As Robinson (2007) argues,
complex tasks with many elements require distinguishing similar elements in the task
and this requirement can be met by the use of noun modifiers. For the purpose of our
analysis, the following forms attached to nouns were counted as noun modifiers: rela-
tive clauses, prepositional phrases, participle phrases, compound nouns, and adjec-
tives. Determiners and number words were excluded from the analysis, as these were
not used in our data to distinguish one noun from another in the learners’ narration.
Lexical variety was measured by calculating the mean segmental type/token ratio
(MSTTR). By dividing the text or utterance into certain segments and taking the aver-
age of the type/token ratio for each segment, the MSTTR minimises the influence of
text or utterance length (Ellis & Barkhuizen, 2005; Ellis & Yuan, 2004). To calculate the
MSTTR, the transcript of each participant for each task was divided into 25-word seg-
ments, and the average number of different words or types included in each segment
was calculated.
Accuracy was analyzed by calculating the percentage of error-free clauses. Several
researchers, such as Foster and Skehan (1996) and Mehnert (1998), have used the same
accuracy measure for the analysis of task performance. Errors were defined at phrasal

6. Coordination here means “the process of combining two constituents [or sentences]”
(Celce-Murcia & Larsen-Freeman, 1999, p. 461). An example would be “The man was arrested
and put into prison” or “She tried to open the door but couldn’t.” In the present study, when a
subject is inserted in the second sentence (e.g., “The man was arrested and he was put into
prison”), it was not counted as a coordination but as two separate sentences.
 Shoko Sasayama and Shinichi Izumi

and sentential levels, rather than at discourse or pragmatic levels. For this reason, it
was not counted as an error when different verb tenses were used within a single task
performance and when common nouns substituted pronouns. To calculate the per-
centage of error-free clauses, the number of error-free clauses was divided by the total
number of clauses and multiplied by 100. As with syntactic complexity, accuracy was
also analyzed by using a specific measure, that is, by calculating the percentage of er-
ror-free use of noun modifiers. This analysis was done to examine whether increased
task demands along the resource-directing dimension would promote not only the
frequency of use of the task-relevant form, but its accuracy as well.
Fluency was analyzed by calculating the percentage of repeated words or phrases
per total number of words. Repetitions were divided into two types: verbatim and sub-
stitutive repetitions. Verbatim repetitions often occur “when hesitating, creating time
to find an appropriate word” (Bygate, 1996, p. 141). Substitutive repetitions, on the
other hand, are used “when correcting a word or grammatical feature” (Bygate, 1996,
p. 141) upon identification of any errors in speech. These two types of repetitions were
used jointly and then separately in the analysis in order to examine both the frequency
of repetitions generally and the nature of those repetitions more specifically. For both
types of repetition, the number of repeated words or phrases was counted, then divided
by the total number of words and multiplied by 100. Fluency was also analyzed by the
number of pruned syllables per total number of seconds taken to complete a task.7

Results

In this section, the results of the study will be presented to answer the two research
questions that guided this study. Table 1 shows the descriptive statistics for complexity,
accuracy, and fluency of the learners’ performance in different conditions. First, the
results of the effects of task complexity and planning as analyzed by global measures
will be presented. Then, the results from the analysis using the specific measures will
be reported.8

Analysis by global measures

First, when the data were analyzed for global syntactic complexity, it appears that the
complex task elicited approximately the same proportion of clauses per T-unit as did

7. An independent rater coded a subset of the data (52%). The simple percentage agreement
between the raters was more than 90% in all cases, except for the global accuracy measure,
which reached 83.33%. All disagreements were subsequently discussed and resolved between
the raters.
8. Interpretations of statistical analysis reported below should be taken with caution due to
multiple uses of ANOVAs in this study.
 Chapter 2.  Effects of task complexity and planning on oral production 

Table 1.  Descriptive statistics for complexity, accuracy, and fluency in different conditions

Syntactic Syntactic Lexical Accuracy Accuracy Fluency 1 Fluency 2

complex- complex- complex- (global) (specific)
ity ity ity
(global) (specific)

Clauses Use of MSTTR % of % of % of Syllables

per T-unit noun error-free accurate repeated per total
modifiers clauses use of words/ number of
noun phrases seconds
modifiers*

[+few elements] 1.33 1.00 0.64 50.06 75.00 9.55 0.87

[+planning] (0.22) (1.25) (0.10) (24.47) (43.30) (7.27) (0.39)
[+few elements] 1.04 0.08 0.61 42.03 0.00 5.74 1.14
[–planning] (0.08) (0.28) (0.08) (25.22) (NA) (3.31) (0.50)
[–few elements] 1.28 4.80 0.64 35.27 38.71 14.24 0.73
[+planning] (0.34) (4.61) (0.10) (22.13) (32.54) (8.56) (0.40)
[–few elements] 1.13 3.23 0.64 22.67 54.11 8.16 0.74
[–planning] (0.15) (2.49) (0.10) (19.25) (30.22) (4.31) (0.38)
Note. Standard deviations are shown in parentheses.
*The data of only those who produced noun modifiers were subject to analysis for the specific accuracy. The
number of the samples is five subjects for the [+few elements +planning] condition, one for the [+few ele-
ments -planning] condition, eight for the [–few elements +planning] condition, and 11 for the [–few ele-
ments –planning] condition.

the simple task, while planners produced a greater number of clauses per T-unit than
did non-planners (see Figure 1). A two-way repeated-measures ANOVA revealed no
main effects for task complexity (F = 0.13, p = 0.72), a significant main effect for plan-
ning (F = 10.12, p < 0.01), and no significant interaction effect (F = 1.91, p = 0.18).
Therefore, participants with pre-task planning time produced significantly more com-
plex speech than did those without planning time, whereas differences in ±few ele-
ments did not result in any significant differences in the learners’ linguistic complexity.
For lexical complexity, approximately the same scores were obtained across all
conditions (see Figure 2). A two-way repeated-measures ANOVA revealed no signifi-
cant main effect for task complexity (F = 0.83, p = 0.37), nor for planning (F = 0.37,
p = 0.55), and no interaction effects (F = 0.94, p = 0.34). Therefore, neither task com-
plexity nor planning conditions affected lexical variety to any significant degree.
As for global accuracy, participants appear to have produced more accurate speech
when engaging in the simple task. Participants also seem to have produced more ac-
curate speech under the planning condition (see Figure 3). A two-way repeated-mea-
sures ANOVA revealed a significant main effect for task complexity (F = 21.00,
p < 0.01); however, no significant main effect for planning (F = 1.56, p = 0.23), nor any
 Shoko Sasayama and Shinichi Izumi

Global syntactic complexity

1.40
1.20
1.00
0.80 Simple task
Mean

0.60 Complex task

0.40
0.20
0.00
+ Planning – Planning

Figure 1.  Mean scores on the global syntactic complexity measure by task complexity
and planning

Lexical complexity
0.70
0.60
0.50
0.40 Simple task
Mean

0.30 Complex task

0.20
0.10
0.00
+ Planning – Planning

Figure 2.  Mean scores on the lexical complexity measure by task complexity
and planning

Global accuracy
60.00

50.00

40.00
Simple task
Mean

30.00
Complex task
20.00

10.00

0.00
+ Planning – Planning

Figure 3.  Mean scores on the global accuracy measure by task complexity and planning
 Chapter 2.  Effects of task complexity and planning on oral production 

interaction effects (F = 0.22, p = 0.64) were obtained. Therefore, the simple task elicited
significantly more accurate speech than did the complex task, while planning did not
significantly affect the degree of learners’ linguistic accuracy.
As for fluency, participants produced more repetitions when engaging in the com-
plex task than in the simple task, and under the planning condition than under the
non-planning condition (see Figure 4). A two-way repeated-measures ANOVA re-
vealed a main effect for task complexity (F = 6.12, p < 0.05), a main effect for planning
(F = 5.81, p < 0.05), and no interaction effects (F = 0.62, p = 0.44). In other words, par-
ticipants produced significantly less fluent speech for the complex task than for the
simple task and under the planning condition than under the non-planning condition.
Separate analyses of the two types of repetitions – verbatim and substitutive rep-
etitions – were conducted. These analyses revealed that the complex task elicited sig-
nificantly more verbatim repetitions than did the simple task (M = 4.18, SD = 4.88 for
the simple task; M = 6.98, SD = 6.59 for the complex task) (F = 4.66, p < 0.05). How-
ever, no significant differences were found for substitutive repetitions (M = 3.22, SD =
2.55 for the simple task; M = 3.82, SD = 2.54 for the complex task) (F = 1.18, p = 0.29).
The analyses also revealed that planners produced significantly more verbatim repeti-
tions than did non-planners (M = 8.19, SD = 7.41 for planners; M = 3.57, SD = 3.37 for
non-planners) (F = 6.12, p < 0.05). However, no significant differences were found for
substitutive repetitions (M = 3.70, SD = 3.01 for planners; M = 3.38, SD = 2.15 for non-
planners) (F = 0.14, p = 0.71). In other words, the significant results for fluency here
were attributable to the use of verbatim, but not substitutive, repetitions.
With regard to another fluency measure, the number of pruned syllables per total
number of seconds taken to complete a task, participants produced more syllables
when engaging in the simple task than in the complex task, especially under the non-
planning condition (see Figure 5). A two-way repeated-measures ANOVA revealed a
significant main effect for task complexity (F = 31.06, p < 0.01), no significant main

Fluency 1
16.00
14.00
12.00
10.00
Simple task
Mean

8.00
Complex task
6.00
4.00
2.00
0.00
+ Planning – Planning

Figure 4.  Mean scores on the fluency measure (percentage of repeated words/phrases) by
task complexity and planning
 Shoko Sasayama and Shinichi Izumi

Fluency 2
1.20

1.00

0.80
Simple task
Mean

0.60
Complex task
0.40

0.20

0.00
+ Planning – Planning

Figure 5.  Mean scores on the fluency measure (number of syllables) by task complexity
and planning

effect for planning (F = 0.63, p = 0.44), and a significant interaction effect between task
complexity and planning (F = 7.24, p < 0.05). In other words, fluency was best promoted
when participants engaged in the simple task under the non-planning condition.

Analysis by specific measures

With regard to the specific measure of complexity, the complex task seems to have trig-
gered much greater use of noun modifiers than did the simple task (see Figure 6). A
two-way repeated-measures ANOVA revealed a main effect for task complexity
(F = 30.65, p < 0.01), no main effect for planning (F = 1.97, p = 0.18), and no significant
interaction effect (F = 0.27, p = 0.61). Thus, the complex task elicited significantly more
use of noun modifiers than did the simple task, but planning did not affect the use of
these forms to a significant degree.

Specific complexity
6.00

5.00

4.00
Simple task
Mean

3.00
Complex task
2.00

1.00

0.00
+ Planning – Planning

Figure 6.  Mean scores on the specific complexity measure by task complexity
and planning
 Chapter 2.  Effects of task complexity and planning on oral production 

Table 2.  Summary of comparisons among task conditions*

Complexity Accuracy Fluency

Syntax Lexis

Global Specific Global Specific** # of # of

repetitions syllables***

Planning planning planning planning planning planning planning planning

time
∨ = = = = ∨ =
no no no no no no no
planning planning planning planning planning planning planning
Task complex complex Complex complex complex complex complex
complexity
= ∨ = ∧ = ∨ ∧
simple simple simple simple simple simple simple
*The symbol “∨” in the table indicates that the condition above it outperformed the condition below it,
and the reverse pattern holds for “∧.” The symbol “=” indicates no significant difference between the two
conditions.
**The results shown here for specific accuracy is only tentative due to the limited sample size available for
this analysis.
***The significant interaction effect obtained for this variable indicates that fluency was best promoted
when participants engaged in the simple task under the non-planning condition.

As for the specific measure of accuracy, the mean percentage of error-free use of noun
modifiers was calculated. It has to be noted first that the number of participants who
produced noun modifiers varied among the four groups. In particular, only one par-
ticipant in the +few elements -planning condition produced any at all, and that was
only a single noun modifier. Therefore, excluding this group, the data from the other
three groups were submitted to a one-way ANOVA. The ANOVA revealed no signifi-
cant differences among the three groups (F = 1.78, p = 0.19). Therefore, despite some
differences observed in the descriptive statistics, accuracy did not differ among the
three groups on this specific measure.
Summary of the overall results of the study is shown in Table 2.

Discussion

Hypothesis 1 posited that a complex (i.e., [–few elements]) task would elicit more com-
plex and accurate though less fluent speech than would a simple (i.e., [+few elements])
task. This hypothesis was confirmed by fluency measures, where the complex task elic-
ited significantly less fluent speech, a result in line with other studies on the effects of
task complexity (e.g., Gilabert, 2007; Ishikawa, 2007; Robinson, 2001b). However,
 Shoko Sasayama and Shinichi Izumi

Hypothesis 1 was not confirmed by the global syntactic complexity, lexical variety, and
global accuracy measures. In fact, the global accuracy measure indicates results op-
posite to Hypothesis 1: the simple task elicited significantly greater accuracy than did
the complex task. It is possible that the complex task, by requiring the participants to
generate and organize their thoughts to convey a complex storyline, may have pre-
vented them from paying due attention to accuracy, whereas the simple task may have
been simple enough for most participants to complete without posing excessive cogni-
tive demands, which allowed them to produce more accurate speech. This explanation
may find some support in the number of words produced by the participants for each
task, where significantly more words were elicited in the complex task (M = 69.04, SD
= 41.77) than in the simple task (M = 51.43, SD = 25.86) (F = 21.89, p < 0.01).
Hypothesis 2 predicted that specific measures would provide more evident sup-
port to Robinson’s cognition hypothesis than would global measures. This hypothesis
was partially confirmed in that, whereas the global complexity measures did not cap-
ture the significant positive effects of task complexity, the specific measure showed a
significantly greater use of noun modifiers in the complex task than in the simple task.
The cognition hypothesis predicts that higher functional demands of a task will be ac-
companied by greater linguistic demands, with observable linguistic consequences in
complexity and accuracy. In our study, the complex task was made cognitively more
complex by increasing the number of characters appearing in the task, which should
pose higher functional demands for distinguishing among similar characters through
the use of noun modifiers. The results obtained in this study are consistent with this
prediction, but importantly this was captured only by the specific, and not by the glob-
al complexity measures. Although global measures have been widely used in the inves-
tigation of task complexity, these results suggest that future studies will benefit from
the use of specific measures, which will allow us to examine more closely the possible
consequence of increased functional demands on learners’ L2 performance. This, we
believe, would be more consistent with the notion of “resource-direction” as proposed
by Robinson in his cognition hypothesis. As for the specific accuracy measure, no firm
conclusion can be obtained in this study due to the limited sampling of the relevant
form. Future research, therefore, will need to elicit data from a greater number of sub-
jects on the use of the target form to allow for adequate analysis.
Hypothesis 3 predicted that planners would produce more complex and fluent
speech than would non-planners. This hypothesis was confirmed by the global syntac-
tic complexity measure; participants under the planning condition produced signifi-
cantly more complex speech than did those under the non-planning condition, a result
consistent with the findings of other planning studies (e.g., Foster & Skehan, 1996;
Ortega, 1999). Following Skehan (1996), pre-task planning appears to have helped
decrease cognitive demands of the tasks, which freed up learners’ attentional capacity
to complexify their language. Hypothesis 3, however, was not supported by the fluency
measures. When fluency was analyzed by the percentage of repeated words, planners
 Chapter 2.  Effects of task complexity and planning on oral production 

produced significantly less fluent speech than did non-planners. This result is contrary
to other studies on pre-task planning (e.g., Foster & Skehan, 1996; Ortega, 1999).
A possible reason for this may be found by looking at the significantly greater
number of words produced by the participants for the complex task. As learners en-
gaging in the complex task had to manage more ideas and express them with limited
interlanguage resources, they may have needed to create time to cope with the greater
cognitive and linguistic processing involved. This they appear to have managed by the
use of verbatim repetitions. It is possible that this situation was exacerbated by the
planning time, as planning time allowed learners to elaborate on their ideas and add
even more things to say. In contrast, when engaging in the simple task, learners had a
clear and simple story to tell, and thus they could presumably perform the task at one
go without worrying about processing time. Especially when deprived of planning
time, learners may have had no choice but to focus on the bare minimum of the story,
which further simplified their narration. The fact that significant differences were
observed only for verbatim repetitions and not for substitutive repetitions for both
complexity and planning variables lends some support to this interpretation because
verbatim repetitions are often used to buy time to allow for greater online processing,
whereas substitutive repetitions are typically used for self-correction to achieve greater
accuracy. For accuracy, no significant effect of planning was observed for either global
or specific measures, a finding which is in line with other studies on pre-task planning
(e.g., Ellis & Yuan, 2004; Foster & Skehan, 1996; Yuan & Ellis, 2003).
Finally, Hypothesis 4 predicted that a decrease in complexity in the resource-de-
pleting dimension would bolster the attention-directing effect of the increase in the
resource-directing dimension. This would mean that participants engaging in the
complex task under the planning condition would outperform the other groups in
complexity and accuracy. No interaction effects, however, were found for any of the
measures, except for one of the fluency measures, the number of pruned syllables per
total number of seconds taken to complete a task. This measure showed that learners
were most fluent when engaged in the simple task under the non-planning condition.
This, we argue, is due to the reasons outlined in the previous paragraph. Although
Hypothesis 4 was not supported in this study, it should be noted that there was some
suggestive evidence from the specific complexity measure that the -few elements
+planning condition elicited the highest instances of noun modifiers of all groups, al-
beit non-significantly so (see Table 1). The possible synergistic effect of resource-di-
recting and resource-depleting dimensions of task complexity, therefore, remains an
interesting area that deserves further investigation.
In sum, results of the current study, together with those of a number of previous
studies, seem to suggest that L2 learners with limited language proficiency have diffi-
culty in paying attention to aspects of complexity, accuracy, and fluency simultane-
ously (see also Genc, this volume). At the same time, it seems that learners’ attentional
allocation can be altered to some extent by manipulating inherent task characteristics
(e.g., [± few elements]) and task implementation conditions (e.g., [± planning]).
 Shoko Sasayama and Shinichi Izumi

Although the research results are somewhat mixed, some common tendencies can
nevertheless be noted: (1) cognitively complex tasks can elicit task-relevant, complex
language structures; (2) pre-task planning can encourage learners to challenge the
limit of their interlanguage and to produce complex language structures in general;
and (3) simple tasks can be effective for fluency development.
Pedagogically, this suggests that teachers are well advised to select tasks and their
implementation conditions according to their purposes. If development of a particular
form is desired, for example, teachers could utilize a complex task by considering care-
fully what kind of form is necessary or useful to complete the given task. If they wish
to focus on learners’ development of linguistic complexity in general, teachers could
provide students with opportunities for pre-task planning. It remains to be seen in
future research, however, how balanced language development in terms of complexity,
accuracy, and fluency can really be achieved in the long term through the effective use
of different tasks and different implementation conditions.

Conclusion

This study set out to investigate the effects of task complexity and pre-task planning on
L2 learners’ task performance. The main findings of the study were: (1) increased task
complexity along ±few elements positively affects the specific measure of syntactic
complexity, but negatively affects global accuracy and fluency; and (2) planning time
positively affects global syntactic complexity, but negatively affects fluency. These find-
ings partially support and partially disconfirm both Robinson’s and Skehan’s hypoth-
eses on task complexity and pre-task planning. The results suggest that increasing
cognitive complexity of tasks indeed poses higher linguistic demands on learners, with
observable consequences on their use of specific forms, but not necessarily of any form
in general. It also seems that L2 learners with limited L2 proficiency, as was the case
with the high school EFL students who participated in the present study, face intrac-
table cognitive difficulty in performing complex tasks (see also Jimarkon & McGrath,
this volume). While EFL learners’ performance can be enhanced in part by the avail-
ability of pre-task planning time, the difficulty seems severe enough to make simulta-
neous improvement in complexity, accuracy, and fluency hard to attain.
Given these results, we are led to conclude that Robinson’s and Skehan’s claims
should be construed as complementary rather than contradictory. Robinson’s cogni-
tion hypothesis has strength in capturing the close connection between functional
demands and linguistic demands posed by the task, while Skehan’s limited capacity
hypothesis has the advantage in capturing the nature of L2 learners’ language use
which is often effortful and attention-consuming. Both positions, we feel, are neces-
sary to account for the complex nature of L2 learners’ language use and learning. The
remaining task is not to decide which position to take, but to show exactly how their
seemingly contradictory claims can be reconciled. The current study suggests, in this
 Chapter 2.  Effects of task complexity and planning on oral production 

vein, that future research will likely benefit from examining the two positions concur-
rently and, in so doing, employing task-discourse sensitive measures of complexity
and accuracy to examine how specific functional demands of the task are related to
specific linguistic demands affecting learners’ L2 use under different conditions.

References

Bygate, M. (1996). Effects of task repetition: Appraising learners’ performances on tasks. In J.

Willis & D. Willis (Eds.). Challenge and change in language teaching (pp. 136–146). London:
Heinemann.
Bygate, M. (2001). Effects of task repetition on the structure and control of oral language. In M.
Bygate, P. Skehan, & M. Swain (Eds.). Researching pedagogic tasks: Second language learning,
teaching, and testing (pp. 23–48). Harlow, UK: Pearson Education.
Bygate, M., & Samuda, V. (2005). Integrative planning through the use of task-repetition. In R.
Ellis (Ed.). Planning and task performance in a second language (pp. 37–74). Amsterdam:
John Benjamins.
Bygate, M., Skehan, P., & Swain, M. (Eds.). (2001). Researching pedagogic tasks: Second language
learning, teaching, and testing. Harlow, UK: Pearson Education.
Celce-Murcia, M., & Larsen-Freeman, D. (1999). The grammar book: An ESL/EFL teacher’s
course. Boston, MA: Heinle & Heinle.
Elder, C., & Iwashita, N. (2005). Planning for test performance: Does it make a difference? In R.
Ellis (Ed.). Planning and task performance in a second language (pp. 219–237).  Amsterdam:
John Benjamins.
Ellis, R. (1987). Interlanguage variability in narrative discourse: Style shifting in the use of the
past tense. Studies in Second Language Acquisition, 9, 1–20.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Ellis, R. (Ed.). (2005). Planning and task performance in a second language. Amsterdam: John
Benjamins.
Ellis, R., & Barkhuizen, G. (2005). Analysing learner language. Oxford: Oxford University Press.
Ellis, R., & Yuan, F. (2004). The effects of planning on fluency, complexity and accuracy in sec-
ond language narrative writing. Studies in Second Language Acquisition, 26, 59–84.
Foster, P., & Skehan, P. (1996). The influence of planning and task type on second language per-
formance. Studies in Second Language Acquisition, 18, 299–323.
Gilabert, P. (2007). The simultaneous manipulation of task complexity along planning time and
[± Here-and-Now]: Effects on L2 oral production. In M.P. Garcia Mayo (Ed.). Investigating
tasks in formal language learning (pp. 44–68). Clevedon, UK: Multilingual Matters.
Hunt, K. (1970). Syntactic maturity in school children and adults. Monograph of the Society for
Research in Child Development, 35, 1–67.
Ishiwaka, T. (2007). The effect of manipulating task complexity along the [± Here-and-Now]
dimension on L2 written narrative discourse. In M.P. Garcia Mayo (Ed.). Investigating tasks
in formal language learning (pp. 136–156). Clevedon, UK: Multilingual Matters.
Kuiken, F., & Vedder, I. (2007). Cognitive task complexity and linguistic performance in French
L2 writing. In M.P. Garcia Mayo (Ed.). Investigating tasks in formal language learning
(pp. 117–135). Clevedon, UK: Multilingual Matters.
 Shoko Sasayama and Shinichi Izumi

Mehnert, U. (1998). The effects of different lengths of time for planning on second language
performance. Studies in Second Language Acquisition, 20, 83–108.
Ortega, L. (1999). Planning and focus on form in L2 oral performance. Studies in Second Lan-
guage Acquisition, 21, 109–148.
Ortega, L. (2007). Meaningful L2 practice in foreign language classrooms: A cognitive-interac-
tionist SLA perspective. In R. DeKeyser (Ed.). Practice in a second language: Perspectives
from applied linguistics and cognitive psychology (pp. 180–207). Cambridge: Cambridge
University Press.
Robinson, P. (1995). Task complexity and second language narrative discourse. Language Learn-
ing, 45, 99–140.
Robinson, P. (2001a). Task complexity, cognitive resources, and syllabus design: A triadic frame-
work for examining task influences on SLA. In P. Robinson (Ed.). Cognition and second
language instruction (pp. 285–317). Cambridge: Cambridge University Press.
Robinson, P. (2001b). Task complexity, task difficulty, and task production: Exploring interac-
tions in a componential framework. Applied Linguistics, 22, 27–57.
Robinson, P. (2003). The cognitive hypothesis of adult, task-based language learning. Second
Language Studies, 21, 45–107.
Robinson, P. (2005). Cognitive complexity and task sequencing: Studies in a Componential
Framework for second language task design. International Review of Applied Linguistics in
Language Teaching, 43, 1–33.
Robinson, P. (2007). Task complexity, theory of mind, and intentional reasoning: Effects on L2
speech production, interaction, uptake and perceptions of task difficulty. International Re-
view of Applied Linguistics in Language Teaching, 45, 193–213.
Shinagawa, F., Kobayashi, S., Fujita, K., & Maekawa, H. (1990). Manual of the Japanese Wechsler
Adult Intelligence Scale – Revised. Tokyo, Japan: Psychological Corporation.
Skehan, P. (1996). A framework for the implementation of task-based instruction. Applied Lin-
guistics, 17 (1), 38–62.
Skehan, P. (1998). A cognitive approach to language learning. Oxford: Oxford University Press.
Skehan, P., & Foster, P. (2001). Cognition and task. In P. Robinson (Ed.). Cognition and second
language instruction (pp. 183–205). Cambridge: Cambridge University Press.
Yuan, F., & Ellis, R. (2003). The effects of pre-task planning and on-line planning on fluency,
complexity and accuracy in L2 monologic oral production. Applied Linguistics, 24, 1–27.
 chapter 3

Measuring task complexity

Does EFL proficiency matter?

Aleksandra Malicka and Mayya Levkina

University of Barcelona

The current study explored whether language proficiency mediates the

perception of task difficulty and whether intended task complexity differences
are reflected in the language production (i.e., fluency, accuracy, and complexity)
of learners of different proficiency levels in English. 37 participants (20 advanced
and 17 pre-intermediate learners of English) took part in the experiment, all
of them undergraduate students at a Spanish university, aged between 18 and
25. They were provided with two tasks of hypothetically differing cognitive
complexity levels manipulated along ± few elements and ± spatial reasoning
dimensions. Two kinds of instruments were used to measure participants’
perceptions of task difficulty: self-reported difficulty ratings and time estimation
of task completion. Complexity, accuracy, and fluency measures were used
to analyze participants’ speech production on the two tasks. As far as the
perception of complexity is concerned, no significant differences between the
high and low proficiency groups were found. Regarding performance measures,
in the high proficiency group the complex task triggered greater lexical and
structural complexity and accuracy, to the detriment of fluency. In the low
proficiency group, on the complex task, fluency was boosted while the other
areas remained intact irrespective of cognitive task complexity. These results are
discussed in terms of Robinson’s Cognition Hypothesis and Skehan’s Trade-off
Hypothesis.

Introduction

Since the early 1990s, there has been an increasing interest in using tasks as a potential
instructional mechanism through which language development and acquisition can
be fostered. One area which has received particular focus is that of grading task com-
plexity according to hypothesized cognitive demands on L2 users’ attention involved
in performing different kinds of tasks. As a result, two competing but complementary
psycholinguistic frameworks have been developed which give different accounts of
 Aleksandra Malicka and Mayya Levkina

attention allocation – and associated cognitive complexity – during task performance:

Robinson’s Cognition Hypothesis (Robinson, 2001, 2003, 2005) and Skehan’s Trade-off
Hypothesis (Skehan, 1998).
The current study sought to find out whether the perception of task difficulty is
mediated by language proficiency and whether intended task complexity differences
are reflected in language production (i.e., fluency, accuracy, and complexity) of learn-
ers of different proficiency levels in L2 English. The role of language proficiency in
task-based instructional settings is an under-researched area, and this study aimed to
shed some light on its potential role by measuring the performance of high and low
proficiency groups in an EFL context in Spain.
In Robinson’s theory, when L2 learners engage in task performance, attention to
aspects of the task draws on different resource pools, potentially enabling the learner
to enhance performance on all three areas of production, that is, fluency, accuracy, and
complexity. Of particular interest, for certain kinds of tasks, increasing complexity
may simultaneously foster attention to (and development of) linguistic complexity
and accuracy, though with a possible detriment to fluency. More specifically, the
Cognition Hypothesis makes a series of claims about the effects of increasing cognitive
complexity of tasks along certain dimensions (Robinson, 2005). The increased concep-
tual and cognitive demands in the case of complex tasks ostensibly generate greater
accuracy and complexity in performance because increased demands boost interac-
tion and require higher attention to and memory for input, as well as possibly long
term retention of input.
Robinson’s Cognition Hypothesis stands in contrast to Skehan’s Trade-off Hypoth-
esis. According to Skehan (1998), attentional resources are limited, and therefore the
three performative areas of fluency, accuracy, and complexity are in competition for
the available attentional resources. Hence, learners have to prioritize between the di-
mensions of performance in terms of where they allocate attention. As a result, “tasks
which are cognitively demanding in their content are likely to draw attentional re-
sources away from language forms” (Skehan & Foster, 2001, p. 189). Engaging in a
cognitively complex task thus leads to trade-off effects between linguistic complexity
and accuracy: when the output is linguistically complex, it ostensibly happens to the
detriment of accuracy, and vice versa.
Although these theories present conflicting accounts regarding how attention is
manipulated during task performance, one basic construct which is common to both
frameworks is that of task complexity. This concept has been of interest to researchers
since the late 1980s, but it was Skehan and Robinson who “most clearly articulated and
researched the concept of cognitive complexity” (Nunan, 2004, p. 86). In his frame-
work, Skehan (1998) proposed a three-way distinction between code complexity
(language required), cognitive complexity (thinking required), and communicative
stress (performance conditions). However, the primary theoretical framework of the
study at hand is Robinson’s Cognition Hypothesis.
 Chapter 3.  Measuring task complexity 

For Robinson (2001), task complexity is the “result of the attentional, memory,
reasoning, and other information-processing demands imposed by the structure of the
task on the language learner” (p. 28). Task complexity is one of the major components
in Robinson’s Cognition Hypothesis, which distinguishes between different kinds of
factors, each making its contribution to overall task performance. These factors are
cognitive, interactive, and learner factors, referring respectively to task complexity,
task condition, and task difficulty. Together they interact to bring about qualitative
changes in performance, and possibly also to contribute to language development.
However, of these factors only task complexity is the controllable one in terms of cog-
nitive demands; it is considered a design feature of tasks that can be manipulated in-
tentionally by placing increased or reduced cognitive demands on the learner, and
thereby supposedly eliciting specific linguistic behavior. Ultimately, according to
Robinson and Gilabert (2007), increasing the cognitive demands of tasks will “push
learners to greater accuracy and complexity of L2 production” (p. 162).
Within the concept of task complexity, Robinson’s model distinguishes impor-
tantly between resource-directing and resource-dispersing variables. The manipulation
of the former “has the potential to direct learners’ attentional and memory resources
to the way the L2 structures and codes concepts, so leading to interlanguage develop-
ment” (Robinson, 2005, p. 4). Recent years have seen a growing body of research into
these dimensions of task complexity, which have been investigated in multiple con-
texts, with multiple learner populations and under various research paradigms.
Pertinent to the current study are the resource-directing dimensions of task complex-
ity, and more specifically, ± elements and ± reasoning demands, a brief review of which
is presented in the following section.

Investigating task complexity and proficiency

Previous research on ± elements and ± reasoning demands

This section reviews studies which have investigated ± elements and ± reasoning de-
mands. Two criteria employed for including these particular studies are their focus
on oral tasks (as opposed to written ones) and on production outcomes (as opposed
to development or acquisition). The following review inquires particularly into the
different interpretations of the findings obtained and aims to see whether language
proficiency, one of the major variables in the current study, is present in these
interpretations.
In a study by Michel, Kuiken, and Vedder (2007), 44 intermediate learners of
Dutch engaged in simple and complex versions of a decision-making task, manipu-
lated along ± elements and ± dialogic dimensions. The complex task was found to
boost accuracy to the detriment of fluency, but hardly any effects were detected
for linguistic complexity. The findings gave only partial support to the Cognition
 Aleksandra Malicka and Mayya Levkina

Hypothesis. The results obtained for fluency and accuracy were attributed to the deci-
sions regarding the measures; it was suggested that the accuracy measures of repair
behavior and two other fluency measures, rather than being related to accuracy or flu-
ency, were related to the construct “repair” (Michel et al., 2007).
In two studies by Michel (2011a, 2011b), 64 intermediate-level learners of L2
Dutch engaged in performing a decision-making task manipulated along ± number of
elements. The first study (Michel, 2011a) found that only lexical complexity (as mea-
sured by Guiraud’s index) was fostered as a result of engaging in the complex task,
which gives little support for the Cognition Hypothesis. Among the possible explana-
tions were task design (insufficient difference in cognitive load between the two task
versions) and insufficiently sensitive measures. The second study (Michel, 2011b), in
which the same materials and subjects were involved as in the first study, measured
performance by means of a specific measure (the use of conjunctions), as opposed to
the first study, which only used global measures. The complex task was found to boost
the use of only one conjunction. Except for that one conjunction, however, the find-
ings go in the opposite direction to that predicted by the Cognition Hypothesis: it was
the simple task that fostered the use of the conjunctions. Possible reasons for these
findings included, among others, task design (not salient enough difference between
the simple and the complex task), and the fact that the variable ± elements on its own
may not bring about qualitative, but only quantitative changes in performance.
In a study by Robinson (2007), L1 Japanese learners of English performed three
narrative tasks manipulated along ± reasoning demands. The results indicated that the
complex tasks resulted in more complex speech on specific, but not general measures,
and they therefore only partially supported the Cognition Hypothesis. This result was
interpreted through task design: the complexity of all the tasks was increased simulta-
neously along resource-directing (± intentional reasoning) and resource-dispersing
(± planning time and ± dual task) dimensions, whereas the Cognition Hypothesis
makes different predictions for increasing resource-dispersing and resource-directing
dimensions. It states that increasing the former likely results in less accurate, complex,
and fluent production, and increasing the latter leads to greater accuracy and complex-
ity, to a possible detriment of fluency.
Gilabert (2007a) explored the effect of task complexity and self-repairs on oral
production with lower-intermediate and advanced Spanish learners of L2 English. The
proficiency of the participants was measured by means of X-lex and Y-lex (vocabulary-
based language ability tests). Out of the three tasks used, two are pertinent to the cur-
rent study: an instruction-giving task and a decision-making task, manipulated,
respectively, along ± number of elements and ± reasoning demands.
In the former task, the findings were in line with the Cognition Hypothesis in
that both higher rate and amount of self-repair were triggered by the complex task.
In the case of the decision-making task, the simple and complex versions of the two
tasks triggered comparable amounts of errors and repairs. Results on two measures
(frequency and the amount of self-repair) were found to be in line with the Cognition
 Chapter 3.  Measuring task complexity 

Hypothesis. Overall, the findings for the latter task did not provide as strong results
in favor of the Cognition Hypothesis as the former one. A psycholinguistic explana-
tion might suggest that the learners paid a substantial amount of attention to justify-
ing their choices, focusing on structural complexity, to the detriment of speech
monitoring.
Regarding the role of proficiency in Gilabert (2007a), except for different error
patterns (as expected, low-proficiency participants made more errors than high profi-
ciency ones), no significant differences between the low and high proficiency groups
were found regarding self-repair behavior. However, the complex version of the narra-
tive task triggered more self-repair behavior in the low-proficiency speakers (as mea-
sured by the number of errors per AS-unit and the ratio of error to words). One reason
that may account for this finding is that high proficiency speakers, given their ad-
vanced level of the L2, did not have difficulties with verb morphology, unlike their
lower-intermediate counterparts.
On the basis of the above review, despite a growing interest in researching ± ele-
ments and ± reasoning demands in the Cognition Hypothesis, language proficiency,
one of the variables under investigation in the present study, has not been studied in a
very systematic way. In the studies reported here, little or no information is provided
regarding the participants’ proficiency in the L2, with only one research design (Gilabert,
2007a) deliberately dividing the participants according to proficiency criteria.
Consequently, where mixed results were obtained in studies, this has been attrib-
uted most frequently to either task design (combining resource-directing and
resource-dispersing dimensions or insufficient differences between the complexity
levels) or the measures employed (global versus specific). While such interpretations
may be valid, none of the studies reported here contemplated the potential role of the
participants’ proficiency as a, or perhaps the, factor explaining the results obtained.
This gap is noteworthy in that the findings of studies reported here frequently give a
rather ambiguous picture of the effects of task complexity on performance, and, as a
result, they fail to give full support to either of the two hypotheses presented at the
beginning of this paper. Hence, the contribution of the current study is to initiate an
exploration into whether L2 proficiency should be considered a factor when drawing
conclusions about the impact of task complexity on L2 learners’ production. Before
turning to our study of the impact of L2 proficiency on oral production, it is important
to consider how we might scientifically determine the cognitive complexity of a par-
ticular task.

Measuring task complexity

When dealing with different constructs of cognitive task complexity, they should be
measured robustly in order to ensure their adequate operationalization. However, to
date researchers have mostly relied on the design of the tasks rather than on
 Aleksandra Malicka and Mayya Levkina

independent measures of cognitive complexity. In some previous studies, when mea-

sured, the most common instrument used to determine the cognitive load of de-
signed tasks was self-reported data based on Likert-scale questionnaires (Robinson,
2001; Gilabert, 2005; Gilabert, 2007b). As research in cognition has shown, humans
do seem to have the capacity to assess the mental burden involved in completing
tasks with different cognitive loads. However, this kind of self-report questionnaire
is by no means the only way of measuring the cognitive complexity realized in di-
verse communication tasks.
Along these lines, Norris and Ortega (2003, 2009) have called for employing a
range of measures previously used in cognitive psychology, such as subjective time
estimations (once the tasks are completed), reaction time data for dual tasks, or physi-
ological data, including heart rate variation, skin conductivity, papillary response, and
even brain activation and blood flow patterns. The last two of the mentioned tech-
niques, as they require special equipment, are quite difficult to use when dealing with
linguistic research that involves the collection of spoken L2 performances. However,
the first of the above-mentioned instruments is easily administrable both in laboratory
and classroom contexts.
Regarding subjective time estimation judgments, these have been extensively
used in psychological literature. However, the tasks used were not language tasks but
psychological ones, such as card sorting, word classification, and anagram manipu-
lation. In these kinds of tasks, certain key features have been detected that may po-
tentially affect subjects’ perception of task complexity (Block, Hancock, & Zakay,
2010). Among them, cognitive load, attentional demands, processing demands, or
judging prospectively versus retrospectively may make a difference in judging the
time on task. For instance, Block et al. (2010) found that, where judging tasks retro-
spectively is involved, the subjects estimate the most difficult task as the one that it
took them longer to complete and also time judgment is likely to be less accurate
when dealing with a more complex task. Although this technique is quite new in the
field of applied linguistics, it has proven to be a good predictor of people’s perception
of cognitive load in psychology where attention and processing had been implied.
Consequently, several studies have started using this tool for measuring cognitive
task complexity independently (e.g., Baralt, 2009; Recio, 2011). It is clear that, on the
one hand, more studies in the field are needed which use this kind of independent
measure. On the other hand, their use is crucial in task-based studies and in studies
dealing with the Cognition Hypothesis, as it will make it possible to scientifically
demonstrate the cognitive differences between tasks with different cognitive loads
and attentional and processing demands, instead of estimating task complexity sole-
ly on the basis of task design.
Whereas methods and instruments of measuring task complexity are beginning
to be researched systematically, the question as to whether L2 proficiency plays a role
in task performance where cognitive complexity is manipulated is still open. The
 Chapter 3.  Measuring task complexity 

motivation for the current study was to initiate investigations into the potential that
cognitive task complexity may interact with language proficiency in determining
variable learner responses to tasks and, in particular, variability in language produc-
tion on oral tasks.

Research questions

The current study investigated the following research questions:

(1) Is the perception of task complexity on tasks designed to represent cognitive
complexity differences actually different for more versus less proficient speak-
ers of English as a foreign language?
(2) Are there observable differences in how designed cognitive task complexity
affects more versus less proficient L2 English speakers’ linguistic accuracy,
complexity, and fluency?
The questions were motivated by the need to explore whether it is language proficiency
that mediates the perception of task difficulty and whether task complexity is reflected
in language production differences (i.e., in terms of fluency, accuracy, and complexity)
for learners of different proficiency levels in English as an L2. In the case of both re-
search questions the null hypothesis is adopted since L2 proficiency is a variable in the
Cognition Hypothesis which has received almost no attention thus far, and neither the
Cognition Hypothesis nor the Trade-off Hypothesis make statements about the role of
proficiency in task-based performance.

Methodology

Participants

The participants in this study were 37 (20 high-proficiency and 17 low-proficiency)

Spanish and Catalan students of EFL from two higher education institutions: Interna-
tional College of Tourism and the University of Barcelona, both located in Barcelona,
Spain. They were students in their first, second, or third years of English language edu-
cation. Participation in the project was voluntary and the participants received course
credit for it in order to incentivize their participation. The university groups were as-
signed to either a high or a low proficiency group in the study on the basis of the
Oxford Placement Test and X-lex (Meara & Milton, 2005) and Y-lex test (Meara &
Miralpeix, 2006).
 Aleksandra Malicka and Mayya Levkina

Materials

Measuring proficiency
To determine participants’ level of proficiency, two different tests were applied: X-lex
and Y-lex measures, and the Oxford Placement Test1. Both tests are considered to be
good predictors of global proficiency levels, and they have been extensively used by the
Language Acquisition Research Group of the University of Barcelona, as well as in
many other studies in EFL contexts. For the current study those students whose scores
were within either the lower-intermediate or advanced categories were taken into ac-
count. Out of the original 60 participants 37 were chosen on the basis of their profi-
ciency test results. These proficiency levels (lower-intermediate and advanced) were
chosen as it was thought that if any differences in either perception of complexity or
performance were to be found among proficiency levels, these would be most notice-
able if the levels represented two extremes of proficiency. In order to gain comprehen-
sive insights into the dynamics between proficiency, task difficulty, and cognitive
complexity, it would of course be useful to see performances from learners across the
full spectrum of proficiencies. However, in the case of the tasks used in this study, a
certain baseline level of proficiency in the L2 was considered essential for task comple-
tion, thus the tasks used might have proven linguistically too challenging for levels
lower than low-intermediate.

Measuring the perception of task difficulty

Two psychological tests were employed to measure the participants’ perceptions
of task difficulty: an affective variable questionnaire and a time judgment task
(see App­endix B). In the Affective Variable Questionnaire a six-point Likert scale was
used. The questionnaire consisted of six statements enquiring about task difficulty,
time pressure, participants’ confidence, interest, and motivation. The participants were
asked to evaluate these statements by circling the number which best reflected their
perception. In the time judgment task the participants were asked to identify the task
that, according to their perception, took them longer to perform, in order to subse-
quently compare their perceptions with the real time it took them to complete each of
the tasks. Whereas questionnaires have been used somewhat in empirical task-based
studies (Robinson, 2001; Gilabert, 2005; Baralt, 2010), time judgment is a new psycho-
logical tool to measure task difficulty (and estimate cognitive complexity by exten-
sion), and to our knowledge in the field of applied linguistics it appeared for the first
time in the doctoral study designed by Baralt (2010).

1. The Oxford Placement Test is a multiple choice test with 60 items, and it is targeted mainly at
lexis and syntax. It distinguishes the following proficiency levels: 0–17 beginner, 18–29 elementary,
30–39 lower intermediate, 40–47 upper intermediate, 48–54 advanced, and 55–60 very advanced.
 Chapter 3.  Measuring task complexity 

Tasks

Finally, the participants were asked to complete two oral instruction-giving tasks2.
Two versions (designed to be cognitively simple or complex) of an instruction-giving
task were used, in which ± reasoning demands and ± elements were manipulated
(see Appendix A). In both versions the participants were asked to explain as accu-
rately as possible where they would place furniture items in different areas of an
apartment in London they had just moved into. The distinction between the simple
and the complex task was established by means of the number of elements the par-
ticipants were asked to deal with and also by the level of spatial reasoning demands
the task required from them.
Regarding ± elements, in the simple version the participants were given a list of six
furniture items, and their task was to explain where they wanted the objects to be
placed on the map of the apartment. In the complex task, the participants were pro-
vided with a list of fifteen objects, from which they had to choose five, taking into
account several conditions (i.e., budget, price, and color). As for ± spatial reasoning
demands, in the simple version the participants were given two points of reference
(two furniture items were already placed in the room), which in cognitive psychology
(Becker & Carroll, 1997) has been shown to be helpful and less demanding in terms of
spatial reasoning. In the complex task, no points of reference were provided.

Procedure

Once two proficiency groupings were established, the two tasks were administered to
the participants, with all task performances audio recorded. The recording took place
in an extracurricular setting, in 5 sessions with 7 or 8 participants per session perform-
ing the task one-by-one with the researcher in the role of listener. The length of the
individual task performance ranged from approximately 1 to 5 minutes. The partici-
pants were first given the instructions, then the layout of the apartment, and after one
minute of planning time they were instructed to start performing the first task. Clari-
fications of instructions were provided when necessary. After completing the first task,
an affective variable questionnaire was administered, after which the whole procedure
was repeated with the second task, with the time judgment questionnaire adminis-
tered after performing the second task (see Appendix B). The time judgment question-
naire asked the participants to circle the task which they thought it had taken them the
longest to complete after performing the two tasks. They were not asked to estimate the
time of each task in minutes or seconds. To avoid possible carry-over effects, approxi-
mately half of the participants were administered the simple-complex order of the
tasks and half of them performed the tasks in reverse order.

2. Both tasks had been previously piloted with 20 university students in their L1 to ensure the
difference in their cognitive complexity.
 Aleksandra Malicka and Mayya Levkina

Production CAF measures

The CLAN mode of CHILDES (MacWhinney, 2000) was used to transcribe and codify
the collected performance data for indices of fluency, linguistic complexity, and accu-
racy. Briefly, coding involved dividing transcriptions into syllables for fluency, tagging
syntactic types of sentences and dividing them into clauses and AS-units for structural
complexity, and tagging different kinds of errors for accuracy, in order to compute the
CAF measures explained below. Interrater reliability of 93% simple agreement was es-
tablished during dual coding of an initial sample of 10% of randomly selected data. The
rest of the data was coded by one or the other of the researchers independently.
Two measures were used to analyze fluency: Unpruned Speech Rate A and Pruned
Speech Rate B (Mehnert, 1998; Yuan & Ellis, 2003). The first measure is the average
number of syllables produced per minute of unpruned speech, including repetitions,
false starts, self-repairs, and so on. In Rate B these performance features were excluded
from the counting. These measures are considered robust, since they include two of
the three sub-components of fluency, distinguished by Skehan (2003): pausing and
speed. Therefore, it has been extensively employed in the literature (Mehnert, 1998;
Ortega, 1999; Gilabert, 2007b; Michel et al., 2007).
Lexical complexity was computed by means of the Guiraud’s index of lexical rich-
ness (Guiraud, 1954), which is calculated by dividing the number of word types in a
speech sample by the square root of the number of word tokens produced. The square
root was introduced to reduce the effect of differences in text length (Vermeer, 2000).
Two other measures used in the present study (tokens/types and number of types)
have also been shown to correlate significantly with the Guiraud’s index in early stud-
ies on vocabulary and, consequently, they are considered to be a good predictor of
lexical diversity.
Structural complexity was measured by dividing words per AS-unit (Foster,
Tonkyn, & Wigglesworth, 2000), words per clause, and clauses per AS-unit. These
measures were chosen following previous studies on language production analysis for
oral tasks (e.g., Gilabert, Baron, & Levkina, 2011).
Accuracy was calculated by means of two general measures (errors per AS-units
and errors per clauses) and two specific measures (errors in spatial expressions per
AS-units and errors in spatial expressions per clause). The choice of the two specific
measures is justified by the nature of the tasks, in which a wide use of spatial expres-
sions was expected (see Appendix A).

Statistical analysis

Three different kinds of statistical analysis were used in the study: (a) descriptive sta-
tistics to retrieve information about means and standard deviations; (b) Wilcoxon
signed-rank comparison tests to provide pairwise inferential comparisons, which
determine where potential significant differences are yielded; and (c) Pearson
 Chapter 3.  Measuring task complexity 

correlations to analyze the results of the newly employed psychological measure of

time estimation. Non-parametric tests were used due to non-normal data distribu-
tions. Alpha decision levels were set at p < .05.

Results

Research Question 1: Perception of task difficulty

Two measures were employed to analyze learners’ task difficulty perceptions. As can be
observed in Table 1, non-parametric pairwise comparisons of the results obtained
from the Affective Variables Questionnaire did not display any statistically significant
differences for any of the variables in either the high or low proficiency groups.
However, the results of descriptive analysis showed the tendency in both profi-
ciency groups to estimate the complex task as more difficult to complete (see Table 2).

Table 1.  Pairwise Comparison of Affective Variables: Wilcoxon signed-rank test

High proficiency students Low proficiency students

Dependent
Simple Task vs. Complex Task Simple Task vs. Complex Task
Variable
Z p Z p

Difficulty   –.79   .43 –1.25 .21

Enough Time   –.85   .39 –1.27 .20
Confidence –1.47   .14   –.37 .71
Interest    .00 1.00 –1.03 .31
Did it easily   –.37   .72 –1.13 .26

Table 2.  Descriptive statistics of affective variables questionnaire: Means

and standard deviations

Dependent High proficiency students Low proficiency students

Variable
Simple Task Complex Task Simple Task Complex Task

M SD M SD M SD M SD

Difficulty* 4.22 1.35 4.06 .98 3.94 1.03 3.65 1.06

Enough Time 4.11 1.41 4.44 1.04 3.94 1.52 3.41 1.46
Confidence 4.00 1.24 4.39 1.04 3.41 1.46 3.29 1.49
Interest 4.33 1.28 4.44 1.19 3.53 1.13 3.76 1.20
Did it easily 4.22 1.17 4.17 .92 3.24 1.35 3.41 1.38
Note: *In the 6-point Likert scale offered to the participants “6” means “easy task” and “1” means “difficult
task”, respectively.
 Aleksandra Malicka and Mayya Levkina

Specifically, the results for perception of difficulty show, on the one hand, greater
differences between simple and complex tasks (4.22 vs 4.06 in the high proficiency
group and 3.94 vs 3.65 in the low proficiency group), where the simple task is marked
as the more difficult one and, on the other hand, the different ways in which low and
high proficiency participants are estimating the tasks (4.22 vs 3.94 for simple tasks;
4.06 vs 3.65 for complex tasks) where low proficiency participants consider the com-
plex task to be more difficult compared to the high proficiency group’s perception
(see Figure 1).
In the Estimation Time Judgment Task, the Pearson Correlation between partici-
pants’ estimated time and real time showed different patterns of time perception in the
two groups. High proficiency participants estimated the duration of their performanc-
es rather inaccurately in comparison with the actual time on task; the correlation be-
tween their estimated time and real time was not significant (r = .30; p = .20). The low
proficiency participants were somewhat more precise in judging the time they needed
to perform each of the tasks; however, the strength of relationship here is still quite
marginal (r = .51; p = .04).
When analyzing the data on Estimation Time Judgment Task qualitatively, the fol-
lowing pattern was observed. As explained before, previous research showed that a
more cognitively demanding task tends to be estimated as the one that takes longer to
complete. In the present study two tendencies emerged. In the high proficiency group
half of the participants considered the simple task as the one that took the longest to
complete, whereas the perception of the other half was the opposite. However, in the
low proficiency group 12 out of 17 participants estimated the complex task as the one
that took the longest to do. When comparing estimated time with the real time the
participants needed to do each task, 60% (high proficiency group) and 65% (low pro-
ficiency group) correctly estimated the time of task completion.

6
5,5
5 4,22 4,06
3,94
4,5 3,65
4 High
3,5 Low
3
2,5
2
1,5
1
Simple task Complex task

Figure 1.  Task difficulty

 Chapter 3.  Measuring task complexity 

Research Question 2: Impact of task complexity on L2 production

Table 3 shows that for high proficiency participants the complex task elicited more flu-
ent speech, which was also linguistically more complex and more accurate. In the low
proficiency group the same pattern is observed for fluency and lexical complexity, but
not for the other measures of structural complexity or accuracy, where ASU per
S-nodes, spatial expressions, and errors per clauses show the opposite patterns.
In what follows, the results of Wilcoxon pairwise comparisons are presented to
detect statistically significant differences for fluency, lexical and structural complexity,
and accuracy for both high and low proficiency groups (see Table 4).
With respect to fluency, there was a statistically significant impact of task com-
plexity on students’ performance in the case of the low proficiency group for both
measures Rate A (Z = –2.63; p = .01) and Rate B (Z = –2.27; p = .02). However, no
significant difference in fluency between the two tasks was found for high proficiency
students.
Concerning the results of linguistic complexity, the complex task generated lexi-
cally richer speech in the high proficiency group, as detected by Guiraud’s Index
(Z = –2.84; p = 0.01) and the number of tokens (Z = –3.08; p = .00). Their speech was
also found to be syntactically more complex, as measured by words per ASU (Z = –2.46;
p = .00) and words per clauses (Z = –2.25; p = –.02). However, the low proficiency
group was not found to be significantly affected by task complexity in terms of lexical
or structural complexity.
Finally, regarding accuracy, high proficiency students displayed more accurate
speech in the more complex task, as reflected by two specific measures, errors in spa-
tial expressions per ASU (Z = –2.56, p = .01) and errors in spatial expressions per
clauses (Z = 3.10, p = .00), whereas low proficiency participants performed less accu-
rately on the more complex task, as yielded by errors in expressions per clauses
(Z = –2.25; p = .02).

Summary of results

In this study, descriptive statistics showed that both groups perceived the complex task
as being more difficult, as measured by Affective Variable Questionnaire. However,
this difference in the perception of tasks was not captured in statistical analysis
(Wilcoxon signed-rank test), which will be discussed in the following section. The
time estimation task showed two patterns. Firstly, the low proficiency group was more
sensitive to task complexity, as most of them identified the complex task as the one it
had taken them the longest to perform. Secondly, both groups obtained similar results
when comparing real time vs estimated time, with 60% (high proficiency group) ver-
sus 65% (low proficiency group) correctly estimating the time.
 
Table 3.  Descriptive statistics of tasks: Means and standard deviations.

Dependent Variable High proficiency students Low proficiency students

Simple Task Complex Task Simple Task Complex Task

M SD M SD M SD M SD

Unpruned Speech Rate A 125.55 28.95 127.99 27.13 95.53 20.24 104.35 22.22
Fluency
Pruned Speech Rate B 116.38 25.89 119.61 24.81 85.01 17.99   94.19 21.63
Lexical Complexity Guiraud’s Index    4.96    .54    5.51    .72   4.51    .55    4.69    .60
Tokens/Types    2.63    .39    2.70    .37   3.06    .56    3.20   1.00
Nº of types   65.80 18.35   82.75 23.38 63.00 18.49   70.24 23.49
Aleksandra Malicka and Mayya Levkina

Structural Complexity Words/ASU   20.59   6.71   16.85   3.60 17.99   3.57   17.15   4.09
Words/clauses    6.40    .88    5.80    .92   6.46   1.75    5.98   1.47
ASU/SN     .53    .23     .55    .15    .59    .22     .56    .15
Accuracy Errors/ASU     .31    .27     .19    .17    .50    .38     .40    .31
Errors/Clause     .10    .07     .08    .08    .19    .17     .15    .11
Spatial Expressions Errors/ASU     .21    .19     .08    .09    .29    .33     .14    .17
Spatial Expressions Errors/Clause   4.32    .10    2.52    .03   2.37    .28    4.44    .06
 Chapter 3.  Measuring task complexity 

Table 4.  Impact of Task Complexity on learners’ performance: Wilcoxon signed-rank test

Dependent High proficiency students Low proficiency students

Variable
Simple Task vs. Complex Task Simple Task vs. Complex Task

Z p Z p

Fluency Unpruned   –.82 .41 –2.63    .01*

Speech Rate A
Pruned Speech   –.86 .39   –2.27*   0.02*
Rate B
Lexical Guiraud’s –2.84 0.01* –1.40   .16
Complexity Index
Tokens/Types   –.78 .43   –.59   .55
Nº of types –3.08   .01* –1.71 0.88
Structural Words/ASU –2.46 0.01*   –.98   .33
Complexity Words/clauses –2.25 0.02*   –.64   .52
ASU/SN   –.68 .49   –.36   .72
Accuracy Errors/ASU –1.83 .07   –.73   .46
Errors/Clause   –.78 .43   –.52   .60
Spatial –2.56   .01* –1.50    .134
Expressions
Errors/ASU
Spatial –3.10   .00* –2.25    .02*
Expressions
Errors/Clause

Task complexity significantly influenced accuracy, and lexical and structural complex-
ity, which all improved in the complex task performance in the high proficiency group.
The complex task also triggered significantly positive effects on fluency in the low pro-
ficiency group, but with a simultaneous decrease in accuracy.

Discussion

The aim of the present study was to investigate whether students’ L2 proficiency may
have an effect on their perception of task difficulty and on aspects of their oral produc-
tion on tasks with different levels of designed cognitive task complexity (± spatial rea-
soning and ± few elements). In this discussion, we consider possible interpretations of
the findings for each of these questions in turn.
 Aleksandra Malicka and Mayya Levkina

Research Question 1: Effects of L2 proficiency on the perception of task difficulty

Research Question 1 addressed the effects that L2 proficiency may have on the percep-
tion of task difficulty. Two different instruments were used to answer this question. The
Affective Variable Questionnaire did not display any statistically significant differences
in perception of task difficulty for any of the groups. However, the overall tendency, as
observed in descriptive statistics, was to identify the more complex task as more difficult
to perform, which corresponds to the expected results based on previous studies. More-
over, the high proficiency group found the complex task less difficult than the lower
proficiency group, which may be associated with their proficiency level interfering in
task difficulty estimation (or with the actual reduction in difficulty of both tasks as a re-
sult of their proficiency). The fact that the low proficiency group found the complex task
more difficult than the simple one could be directly related to their proficiency level.
Hence language proficiency could be the factor determining task difficulty perception.
The second instrument employed in the study was a Time Estimation Task, in
which the participants were asked which task took them longer to complete. Accord-
ing to some previous studies in cognitive psychology (Block et al, 2010), people tend
to fail in guessing the time they take to complete a cognitively more complex task and
also they are likely to consider a cognitively more complex task as the one that requires
more time to complete. That finding is explained by the involvement of a high cogni-
tive load which affects participants’ attention, creates a more stressful situation, and
therefore makes them misestimate and/or exaggerate the real time of task completion.
The present measure is a new tool in applied linguistics and TBLT studies, first used
only recently by Baralt (2010). In the present study, a Pearson correlation along with
qualitative analysis was used to compare the relationships between estimated and ac-
tual time in each group, and it showed that, on the one hand, in the high proficiency
group the perception of task difficulty (i.e., time required) did not correlate with task
complexity to the same degree as in the lower proficiency group. This finding suggests
that the higher proficiency group had greater difficulties in perceiving task complexity.
At the same time, qualitative analysis suggests that most low proficiency participants
correctly pointed to which task required more time to be completed, whereas higher
proficiency participants were less accurate in their judgment. This finding is partially
in line with Baralt’s study (2010) and also with a body of psychological research on the
topic for higher proficiency participants, where a significant impact of task complexity
on time estimation was observed. Simultaneously, no similar results were detected in
the lower proficiency group, which suggests that learners’ proficiency plays a certain
role in task complexity perception.

Research Question 2: The impact of L2 proficiency on performance

The second research question asked if there are differences in how task complexity af-
fects high versus low proficiency speakers’ speech. The results of our study suggest that
 Chapter 3.  Measuring task complexity 

designed cognitive task complexity may have affected high and low proficiency speak-
ers in different ways. In the following paragraphs this result will be interpreted in terms
of the theoretical underpinnings of the study. In particular, we discuss how the find-
ings provide insights into the issues of attention allocation and trade-off effects, and
the role of language proficiency in light of the two hypotheses will be discussed.
The results of this study lend partial support to both hypotheses described at the
beginning of the paper. In the high proficiency group on the complex task, lexical and
structural complexity and accuracy were boosted, to the detriment of fluency. This
finding is in line with the predictions of Robinson’s Cognition Hypothesis, which states
that as tasks are made more complex, learners stretch their interlanguage (in terms of
lexis, syntax, and accuracy) to meet the cognitive demands of the task, with fluency
being potentially negatively affected. In turn, in the case of the low proficiency partici-
pants, the cognitive complexity of the task enhanced this group’s performance in the
area of fluency, the other areas remaining intact irrespective of cognitive task complex-
ity (except for one specific accuracy measure, on which these participants performed
better on the simple task than on its complex counterpart). This finding is partially in
line with Skehan’s Trade-off Hypothesis, which states that complex tasks trigger better
performance in one area of production to the detriment of others. In our study, how-
ever, the other areas (all except fluency) were not negatively affected; rather, they were
found not to be influenced by cognitive task complexity. For this reason the Trade-off
Hypothesis is only partially supported in the low proficiency group.
We believe that the results obtained for the two groups of speakers may be due to
different patterns of attention allocation during task performance, triggered at least in
part by different L2 proficiency levels in the two groups. The more advanced English
speakers, due to their quite rich lexical and syntactic repertoires, were able to simulta-
neously attend to multiple performance areas, it seems. In other words, in the case of
this group, attention drew on various resource pools due to their advanced L2 level,
resulting in overall enhanced performance on the complex task. The low proficiency
speakers on the other hand, given their limited linguistic resources in the L2, when
faced with the more complex task, had to prioritize one area of performance over the
others (a finding consistent with the predictions of the Trade-off Hypothesis). In terms
of attention allocation, these participants’ attentional capacity was insufficient for them
to be able to simultaneously pay attention to more than one aspect of performance. A
proficient L2 speaker has attentional control over multiple areas of performance, while
for low-proficiency speakers attention may need to be more selective, permitting only
limited control of the language. Interpreted in this way, the findings might indicate
that the capacity of human attention is a dynamic process rather than a stable system:
as proficiency in the second language develops, attentional capacity stretches to meet
the cognitive demands of the more challenging complex task.
Given that the study provides some evidence that proficiency plays a role in atten-
tion allocation, what we observed is that trade-off effects are perhaps more pertinent
to lower L2 proficiency learners, and they become less marked, if not disappear, as
 Aleksandra Malicka and Mayya Levkina

proficiency increases. This observation, again, lends partial support to both hypothe-
ses. In light of these findings, we may speculate that the Trade-off Hypothesis holds
true for speakers of limited proficiency, while the Cognition Hypothesis explains phe-
nomena occurring when more advanced L2 speakers engage in task-based perfor-
mance. This pattern indicates that the different psycholinguistic accounts depicted by
the two hypotheses may hold true for speakers at different points in their interlan-
guage development. We might argue that, in the same way as there exists a continuum
of proficiency levels, we might see the two hypotheses not as a dichotomy, but rather a
continuum of shifting attentional capacities and interactions with other phenomena
like task design.

Limitations and conclusion

This study explored the role of only two proficiency levels, and those levels were not
two extremes of proficiency, but rather two groups of relatively different proficiency
levels (low-intermediate vs advanced learners). A different pattern of task performance
could be found if the same tasks were administered to a number of proficiency levels,
ranging from beginner to native-like levels. Additionally, a comparison of L1 baseline
data, contrasted with a range of proficiency levels of L2 speakers, would probably pro-
vide further valuable insights into the issue of attention allocation and the nature of
trade-off effects.
The findings reported in the current study come from an EFL setting. It has to be
born in mind that the major variable in the study – language proficiency – was EFL,
and not ESL proficiency. Different attention allocation, perception, and performance
patterns could have been found had the study been carried out in an ESL context. To
make any definitive statements about the role of proficiency, both EFL and ESL profi-
ciencies would need to be investigated in relation to task-based performance.
The role of L2 proficiency in task-based performance is an under-researched area,
but the small-scale study reported here raises the issue of whether L2 proficiency
should be considered as one of the crucial variables in the Cognition Hypothesis and
similar frameworks. Further research, involving more participants, more proficiency
levels, and different task types is needed to further explore the role of proficiency in
various settings and under a variety of research paradigms.

References

Baralt, M. (2010). Task complexity, the cognition hypothesis, and interaction in CMC and FTF
environments. Unpublished doctoral dissertation, Georgetown University.
Becker, A., & Carroll, M. (1997). The Acquisition of spatial relations in a second language.
Amsterdam: John Benjamins.
 Chapter 3.  Measuring task complexity 

Block, R. A., Hancock, P. A., & Zakay, D. (2010). How cognitive load affects duration judgments:
A meta-analytic review. Acta Psychologica, 134, 330–43.
Foster, P., Tonkyn, A., & Wigglesworth, G. (2000). Measuring spoken language: A unit for all
reasons. Applied Lingustics 21 (3), 354–375.
Gilabert, R. (2005). Task complexity and L2 narrative oral production. Unpublished doctoral dis-
sertation, University of Barcelona.
Gilabert, R. (2007a). Effects of manipulating task complexity on self-repairs during L2 oral pro-
duction. International Review of Applied Linguistics 45(3), 215–240.
Gilabert, R. (2007b). The simultaneous manipulation of task complexity along planning time
and (±here-and-now): effects on L2 oral production. In M. P. García-Mayo (Ed.), Investigat-
ing Tasks in Formal Language Learning (pp. 44–68). Clevedon, UK: Multilingual Matters.
Gilabert, R., Baron, J., & Levkina, M. (2011). Manipulated task complexity across task types and
modes. In P. Robinson (Ed.), Second language task complexity (pp. 105–138). Amsterdam:
John Benjamins.
Guiraud, Pi. (1954). Les caractères statistiques du vocabulaire. Paris: Presses Universitaires de
France.
MacWhinney, B. (2000). The CHILDES project: Tools for analyzing talk. Mahwah, NJ: Lawrence
Erlbaum Associates.
Meara, P., & Milton, J. (2005). X-Lex: The Swansea Levels Test [CD-ROM]. Newbury, UK: Express
Publishing.
Meara, P., & Miralpeix, I. (2006). Y-Lex: the Swansea Advanced Vocabulary Levels Test. v2.05.
Swansea: Lognostics.
Mehnert, U. (1998). The effects of different lengths of time for planning on second language
performance. Studies in Second Language Acquisition, 20, 52–83.
Michel, M. C., F. Kuiken, & Vedder, I. (2007). The influence of complexity in monologic versus
dialogic tasks in Dutch L2. International Review of Applied Linguistics, 45(3), 241–259.
Michel, M. (2011). Cognitive and interactive aspects of task-based performance in Dutch as a
second language. ‘s-Hertogenbosch: Uitgeverij BOXPress.
Norris, J. M., & Ortega, L. (2003). Defining and measuring SLA. In C. Doughty & M. H. Long,
(Eds.), Handbook of second language acquisition (pp. 716–761). London: Blackwell.
Norris, J. M., & Ortega, L. (2009). Towards an organic approach to investigating CAF in In-
structed SLA: The case of complexity. Applied Linguistics, 30, 555–578.
Nunan, D. (2004). Task-Based Language Teaching. Cambridge: Cambridge University Press.
Ortega, L. (1999). Planning and focus on form in L2 oral performance. Studies in Second Lan-
guage Acquisition, 21, 109–148.
Recio, M. (2011). The effects of task complexity on L2 oral production as mediated by differences
in working memory capacity. Unpublished Master dissertation, University of Barcelona.
Robinson, P. (2001). Task complexity, task difficulty and task production: Exploring interactions
in a componential framework. Applied Linguistics, 22(1), 27–57.
Robinson, P. (2003). Attention and memory during SLA. In C. Doughty & M. H. Long (Eds.),
The handbook of second language acquisition (pp. 631–678). Malden, MA: Blackwell.
Robinson, P. (2005). Cognitive complexity and task sequencing: Studies in a componential
framework for second language task design. International Review of Applied Linguistics,
43(1), 1–32.
Robinson, P. (2007). Task complexity, theory of mind, and intentional reasoning: Effects on L2
speech production, interaction, uptake and perceptions of task difficulty. International Re-
view of Applied Linguistics, 45(3), 193–213.
 Aleksandra Malicka and Mayya Levkina

Robinson, P., & Gilabert, R. (2007). Task complexity, the Cognition Hypothesis and second lan-
guage learning and performance. International Review of Applied Linguistics, 45(3),
161–176.
Skehan, P. (1998). A cognitive approach to language learning. Oxford: Oxford University Press.
Skehan, P., & Foster, P. (2001). Cognition and tasks. In P. Robinson (Ed.), Cognition and Second
Language Learning (pp. 183–205). Cambridge: Cambridge University Press.
Tavakoli, P., & Skehan, P. (2005). Strategic planning, task structure, and performance testing. In
R. Ellis (Ed.), Planning and task performance in a second language (pp. 239 – 277). Amster-
dam: John Benjamins.
Tavakoli, P., & Foster, P. (2008). Task design and L2 performance. Language Learning, 58(2),
429–473.
Vermeer, A. (2000). Coming to grips with lexical richness in spontaneous speech data. Language
Testing, 17(1), 65–83.
Yuan, F., & Ellis, R. (2003). The effects of pre-task planning and on-line planning on fluency,
complexity and accuracy in L2 monologic oral production. Applied Linguistics, 24 (1),
1–27.
 Chapter 3.  Measuring task complexity 

Appendix A

Simple task
“COCINA”
Acabas de mudarte a un nuevo piso en Londres. Ya tiene algunas habitaciones amue-
bladas, pero todavía faltan la cocina y la habitación individual por amueblar. Hagamos,
primero, la cocina.
Abajo tienes una lista de objetos que te gustaría colocar allí:

Una nevera Cuatro sillas

Un fregadero Armarios

Una campana Un horno

¿Qué es lo que tienes que hacer?

Imagínate en tu nuevo piso a punto de colocar los nuevos muebles y utensilios en

su sitio. Por favor, lo más detalladamente posible explica al repartidor de muebles
(I) dónde quieres que él ponga cada uno de los objetos en tu cocina y (II) justifica
tu decisión.
A la hora de explicar, puedes hacer referencia a dos objetos que ya están en la cocina.
Tienes 1 minuto de preparación antes de empezar.
 Aleksandra Malicka and Mayya Levkina

Complex task
“Habitación Individual”
Ahora vamos a amueblar la habitación individual. Esta vez todavía no tienes decidido
lo que vas a colocar allí.
Abajo tienes una lista de objetos que podrían quedar bien en la habitación:

400 € 300 € (por UNidad) 400 €

1500 € 1500 € 900 €

1200 € 150 € 800 €

200 € 600 € 1700 €

2400 € 800 € 600 €

 Chapter 3.  Measuring task complexity 

Primero, escoge 5 objetos de los 15 ofrecidos. Ten en cuenta que tienes (I) 4000 €
de presupuesto, (II) te gusta mucho el azul y (III) te gustan cosas modernas. Se-
gundo, explica al repartidor de muebles, qué objetos quieres comprar y dónde pien-
sas colocarlos y justifica tu elección.
Tienes 1 minuto de preparación antes de empezar.

¿Qué es lo que tienes que hacer?

Plan of the apartment

Plano de tu piso•

Cocina• Habitación individual•

Baño•

Sala de estar•
Habitación doble• Baño•
 Aleksandra Malicka and Mayya Levkina

Appendix B

Affective variable questionnaire and time judgment task

 Name:  ________________________________
Anxiety and Perceived Difficulty Questionnaire
Please tell me how much you agree or disagree with the following statements by cir-
cling a number from 1 to 6:

Strongly Disagree Slightly Slightly Agree Strongly

disagree disagree agree agree
1 2 3 4 5 6

“Cocina”
1. This task was easy. 123456
2. I had enough time to prepare myself to do the task. 123456
3. I was relaxed and comfortable completing the task. 123456
4. I enjoyed doing the task. 123456
5. I could complete the task easily. 123456
“Habitación individual”
1. This task was easy. 123456
2. I had enough time to prepare myself to do the task. 123456
3. I was relaxed and comfortable completing the task. 123456
4. I enjoyed doing the task. 123456
5. I could complete the task easily. 123456
Time judgement questionnaire
Please circle the task you think it took you longer
to complete:
“Cocina” “Habitación individual”
 chapter 4

Effects of strategic planning on the accuracy

of oral and written tasks in the performance
of Turkish EFL learners

Zubeyde Sinem Genc

Uludag University, Turkey

This study investigates the effects of strategic planning on the accuracy of EFL
learners’ performance on oral and written narrative tasks. The study addresses
the complex relationship between the planning process and accuracy in task
performance. Data collected from 60 learners in a Turkish EFL educational
setting showed that strategic planning did not significantly affect accuracy on the
oral task, and that it had an adverse effect on the written task. Modality (oral or
written) made a significant difference under unplanned conditions. On the other
hand, a post-task questionnaire revealed that we need to examine what learners
attend to during the strategic planning phase to understand how they actually
use the planning time and whether they change their focus of attention during
the task performance phase. Through its focus on these aspects of task-based
language teaching (TBLT) and learning, including planning time, accuracy, and
oral and written modalities, this study brings to the fore these important aspects
of how TBLT works in EFL settings.

Introduction

One of the main areas of research in task-based language learning and teaching is the
relationship between task variables and language production, with a growing body of
research focusing on language production in terms of different aspects of L2 perfor-
mance. Skehan (1996) presents three aspects of language production: fluency, accura-
cy, and complexity, which have been used to examine the effects of various task
variables. These variables can be classified under two main categories: task design vari-
ables and task implementation factors. A task implementation factor that has attracted
a lot of interest among researchers is planning. As Ellis (2005) points out, this may take
the form of pre-task or within-task (i.e., online) planning. Pre-task planning, in turn,
includes rehearsal and strategic planning. The subtypes that Ellis identifies for pre-task
 Zubeyde Sinem Genc

planning provide a basis for the present study as it aims to investigate the effects of
strategic planning on an important aspect of language production, namely, accuracy.
The effect of strategic planning has been investigated by other researchers, but
results have been mixed. For instance, in their study of the effects of strategic planning
on language production, Foster and Skehan (1996) used three different task-types and
found that task-type was a main factor such that learners were more fluent and more
accurate on personal information and decision-making tasks than on a narrative task.
In other words, the effects of strategic planning varied depending on the task-type. In
addition, the difference between guided and unguided strategic planning was signifi-
cant for fluency but not accuracy. However, the results in Skehan and Foster’s (1997)
study were not consistent with these findings in their 1996 study because they found
that planning affected accuracy positively on the narrative task, but not on the other
two tasks. Interestingly, in a more recent study, Skehan and Foster (2005) found that
the effect of planning was not uniform even on the same task, namely, the decision-
making task. Gilabert (2007), on the other hand, found no interaction between task-
type and planning where accuracy was concerned, in contrast to Foster and Skehan’s
(1996) study. More critically, Wendel’s (1997) study produced no evidence of whether
planning had any effect on the accuracy of learners’ production.
More mixed findings were provided by other studies too. Wigglesworth (1997)
found that planning time improved the complexity and accuracy of highly proficient
learners’ production, especially on tasks with a high cognitive load. However, low-pro-
ficiency learners did not benefit from the opportunity to plan the tasks. The finding that
advanced-level learners can better benefit from planning was supported by Ortega’s
(1999) study as well. Subsequent studies have also found the proficiency level of learners
to be an important factor determining the effects of strategic planning (e.g., Kawauchi,
2005; Tavakoli & Skehan, 2005). However, unlike Wigglesworth’s and Ortega’s findings,
Kawauchi found that low-level learners did benefit from planning more in terms of ac-
curacy, which was in contrast with advanced-level students who did not gain any sig-
nificant increase in accuracy and complexity under the planned condition.
Learner orientation too has been found to affect the interaction between planning
and language performance. Learners have been found to produce more fluent, more
complex, and more accurate language if they have positive attitudes towards planning.
For example, Tajima (2003) showed that planning time resulted in more fluency, more
accuracy, and more lexical complexity because learners were positively oriented
towards the planning time. Learner orientation has also been studied in terms of learn-
ers’ focus of attention. Sangarun (2005) found that the focus of guided planning af-
fected learner performance because “learners do best when they engage in strategic
planning that is focused jointly on meaning and form” (p. 131). Similar conclusions
were made by Ortega (2005), who reported that learners did not necessarily separate
attention to form versus meaning during pre-task planning. On the other hand,
Mochizuki and Ortega (2008) found that pre-task planning affected accuracy posi-
tively when it was guided planning, that is, when learners’ attention was drawn to form
 Chapter 4.  Performance of Turkish EFL learners 

during the phase of strategic planning. In sum, previous research findings provide
mixed support for the effect of strategic planning on the accuracy of L2 learners’ per-
formance, which calls for more research.

Modality of language production and accuracy

In the last 10–15 years, there has been growing interest in written tasks, and more re-
cent research shows that written tasks allow for a significant reduction in errors and
the production of lexically more varied texts, especially when task complexity is ma-
nipulated (e.g., Kuiken & Vedder, 2007). The interest in written as well as oral tasks can
also be observed in the work of Ellis and Yuan (2005), who investigated the effects of
on-line planning on written and oral tasks and found that L2 learners were more ac-
curate in written than in oral production under both careful on-line planning and
no-planning conditions. They reported that these differences were greater across mo-
dality than in planning conditions. In other words, the type of modality affected learn-
ers’ production of accurate language more than planning did. Another study on the
role of modality was conducted by Adams (2005) who examined whether learners’
orientation to form would vary between oral and written subtasks and whether writing
would make learners more likely to focus on form. The results of her study revealed
that writing increased learners’ orientation to form and that the links between engage-
ment in tasks and development of forms was most obvious when a writing component
was included.
Part of the importance of these research findings is that one of the major objec-
tives of TBLT is to promote focus on form (i.e., accuracy) in the context of meaningful
communication. However, learners may or may not be oriented towards form during
the communicative tasks they are involved in. Hence the two variables of planning
condition and modality appear to be two important factors capable of mediating learn-
ers’ attention to form and affecting accuracy in learners’ performance. Following the
empirical study of the effects of on-line planning in Ellis and Yuan (2005), the present
study seeks to explore the effects on form of strategic planning in tasks of different
modalities and to examine whether learners’ production is similar or different in terms
of accuracy with respect to these modalities. Indeed, the mixed findings regarding the
effects of strategic planning on accuracy (see above) make it even more compelling to
examine learners’ performance on tasks of different modalities in an EFL setting, an
underrepresented territory in the majority of TBLT research. We would therefore want
to see whether the modality of tasks makes any difference on the impact of strategic
planning on accuracy in such a context.
A related issue is how L2 learners actually use the planning time they are given.
Previous studies have pointed to the importance of the role of learners’ attention in
task-based language learning (Ellis, 2005; Kuiken & Vedder, 2007; Ortega, 1999, 2005;
Skehan & Foster, 2001). How learners’ attentional resources are allocated during
 Zubeyde Sinem Genc

planning time and task completion is considered to have crucial effects on task perfor-
mance. For instance, Ellis (2005) states that “Even when asked to engage in form-fo-
cused planning they [learners] may not do so, preferring to use the time given [to]
them to sequence ideas and to work out the semantic linkages among propositions”
(pp. 23–24). Ortega (2005) agrees that it is the learner who determines the focus of
attention and how the planning time is actually used. Based on the findings of her
study, Ortega (2005) states that “[i]n spite of holding a meaning-oriented interpreta-
tion of the task, learners paid attention to form during planning without any specific
instructions to do so” (p. 106).
Bygate and Samuda (2005) indicate that learners may not be able to carry over the
forms they plan into their performance of tasks. On the other hand, Skehan’s trade-off
effects suggest that learners have to choose what aspect of production to focus on dur-
ing planning because of their limited processing capacity. He states that learners may
focus on fluency and complexity at the expense of accuracy or vice-versa. It is obvious
that we need to observe what learners actually attend to in order to understand their
focus of attention during pre-task planning and during the completion of tasks of dif-
ferent modalities, which will be the focus of this study.

Research questions

Based on the various considerations outlined above, this study aims to investigate the
effects of strategic planning on accuracy in written and oral performance of learners
in a Turkish EFL setting. The study also tries to tap into the focus of learners’ atten-
tion during the pre-task planning time and during their performance of tasks. Within
this framework, the following research questions were formulated for the current
investigation:
Research Question 1: Do EFL learners produce more accurate oral language when
they have time for unguided strategic planning?
Research Question 2: Do EFL learners produce more accurate written language
when they have time for unguided strategic planning?
Research Question 3: Does the modality of EFL learners’ productions (oral vs.
written) have any effect on the accuracy of their language output when they have
time for unguided strategic planning?
Research Question 4: Does the modality of EFL learners’ productions (oral vs.
written) have any effect on the accuracy of their language output when they do not
have time for any planning?
Research Question 5: What do EFL learners actually attend to during pre-task
planning and task completion phases?
 Chapter 4.  Performance of Turkish EFL learners 

The first research question examines learners’ performance on oral tasks under planned
and unplanned conditions. The second question examines learners’ performance on
written tasks under planned and unplanned conditions. The third question compares
learners’ performance on written and oral tasks under planned conditions. The fourth
question compares learners’ performance on written and oral tasks under unplanned
conditions. The fifth question examines what learners attend to most during the pre-
task planning time and during their performance of the tasks.

Method

Context and participants

This study was conducted in a classroom setting at a one-year intensive EFL program
at a university in Turkey. A total of 60 EFL learners participated in the study, 36 (60%)
of these were female and 24 (40%) were male. Students enrolled in the program have
to improve their proficiency in English before being allowed to start their undergradu-
ate studies. The 60 students had been placed in the low-intermediate level through an
institutional placement exam administered by the university. At the time of the study,
these students were receiving 25 hours of English instruction each week taught by na-
tive Turkish-speaking instructors: 7 hours for grammar, 4 for reading, 2 for writing,
and 12 hours for integrated skills that included listening, speaking, and vocabulary.
Prior to enrolling in the program, these students had studied English as a foreign lan-
guage for 9 years. The age range of the students who participated in the study was be-
tween 18 to 22 years, with an average of 19 years. The 60 students were divided into
four equal groups of 15 students each and randomly assigned to one of the four groups
shown in Table 1 below.
As Table 1 displays, the study employed a between-subjects design with two levels
of pre-task planning conditions and two levels of modality. In order to make the text
reader-friendly, the groups shown in Table 1 will be referred to as follows:
– Group 1: S + P (i.e., oral task with planning)
– Group 2: S – P (i.e., oral task without planning)
– Group 3: W + P (i.e., written task with planning)
– Group 4: W – P (i.e., written task without planning)

Table 1.  Design of the study

Planning No planning

Oral task Group 1 (N = 15) Group 2 (N = 15)

Written task Group 3 (N = 15) Group 4 (N = 15)
 Zubeyde Sinem Genc

Tasks

As stated earlier, this study sought to investigate the written and oral performance of
60 students in an EFL setting. A narrative monologue task was considered appropriate
for both types of modalities. A monologue task was selected in order to avoid the risk
that had been observed in previous studies (e.g., Foster & Skehan, 1996; Skehan &
Foster, 1999) using interactive tasks whereby some students’ performance was found
to be less active as a result of their interlocutor’s patterns of talk, such as their use of
comprehension checks, confirmations, and clarification requests. Ellis and Yuan (2005)
warn that using two tasks with two different sets of pictures to elicit oral and written
narratives needs to be considered a serious limitation because “it is ... possible that the
differences attributed to modality were in part, at least, the result of task differences”
(p. 191). In order to cater for this limitation, the same set of six pictures was used to
elicit stories from the learners in all four groups, which ensured consistency in terms
of task complexity and encoding difficulty. This, of course, required that the separate
groups perform under different planning conditions: one for the oral tasks, and the
other for the written tasks.

Data collection procedures

The participants performed the tasks as an activity in their regular classroom situation
with their regular instructor guiding the process. The activity was designed to fit with
their regular program of study. The researcher was not present during data collection
so as not to distract the learners and influence their performance. The classroom in-
structor was briefed about the study and the processes involved. Before using the tasks
with the target groups, she piloted the tasks with a different group of students. This
also ensured her familiarity with the procedure. The data were collected on four differ-
ent days, each day with a different group. The instructions were given in the students’
first language, Turkish (see Appendix A for task instructions). The participants were
given a set of six pictures (Appendix B). Following Yuan and Ellis (2003), the students
were encouraged to produce at least 4 sentences for each of the six pictures, and asked
to be prepared to tell or write the story in the pictures with or without planning. Under
the no-planning condition, students were not given any time before task performance
as they were asked to start telling or writing the story right within a minute of seeing
the pictures. Under the strategic planning condition, they were given 10 minutes for
planning for both oral and written tasks. During task performance, the time given for
completing the written task was adjusted to suit the written mode. In other words,
students had 8 minutes for their performance of the oral task and 15 minutes for the
written task. The amount of time allowed for each task was determined based on the
performance of the students in the pilot study on the tasks. The students’ performance
was recorded for transcription and analysis.
 Chapter 4.  Performance of Turkish EFL learners 

In addition to that, and in order to answer the fifth research question of the study,
each student filled out a questionnaire immediately after completing the task
(Appendix C). The post-task questionnaire, adapted from Yuan and Ellis (2003), solic-
ited information about how students spent their time before and during the tasks. In
other words, it aimed to obtain self-report data from the students about how they al-
located their attention during pre-task planning and during task completion. Students
were asked to give a priority order for grammar, vocabulary, and organization. They
were asked to put these three component areas in an order of priority according to the
time and attention they paid for each one during the strategic planning and task perfor-
mance phases. Data collected through the questionnaire only focused on the students’
reports: in other words, actual allocation of students’ attention was not measured.

Accuracy measures and coding of data

The measures used most widely in the literature to examine accuracy include percent-
age of error-free clauses, percentage of correct verb forms, target-like usage of plurals,
target-like usage of articles, number of self-corrections or repetitions, target-like usage
of verb tenses, lexical errors (or target-like usage of vocabulary), errors per T-unit, er-
rors per 100 words, and target-like usage of negation. Previous studies differ in their
choice of measure. For instance, Skehan and Foster (1999) define accuracy as “[t]he
ability to avoid error in performance, possibly reflecting higher levels of control in the
language as well as a conservative orientation, that is, avoidance of challenging struc-
tures that might provoke error” (p. 96). Skehan and Foster used the global measure of
percentage of error-free clauses to assess accuracy, arguing that prior studies had found
global measures of accuracy to be sensitive for detecting differences between experi-
mental conditions. Another commonly accepted global measure of accuracy is the
percentage of correct verb forms. These measures were used in the studies by Ellis and
Yuan (2005), Foster and Skehan (1996), Sangarun (2005), Tajima (2003), Yuan and
Ellis (2003), and Wendel (1997).
Following these studies, the current investigation measured accuracy in terms of
error-free clauses and correct verb forms. Error-free clauses refer to those clauses that do
not contain any errors of syntax, morphology and/or lexical choice. Correct verb forms
refer to all accurately used finite verbs in terms of tense, aspect, modality, and subject-
verb agreement (Ellis & Yuan, 2005). To check for interrater reliability of measures, two
raters, the researcher and a native English teacher, independently coded a sample of 40%
of the data (i.e., six learners in each group). There was a 93% agreement between the two
raters. The rest of the data were then analyzed and coded by the researcher.

Analysis of data

The normal distribution of the data collected was tested by the Shapiro-Wilk test, a test
sensitive to both skewedness and kurtosis. The test verifies whether or not the data
 Zubeyde Sinem Genc

collected came from a normally distributed population and is accepted as one of the
most reliable tests for normality for small to medium-sized samples (Conover,1999;
Royston, 1995). After checking that the data met the criteria for normality, parametric
and non-parametric tests were then applied. A series of t-tests on the accuracy data
and Mann-Whitney U tests on the questionnaire data were used for the comparisons
between the groups to address the research questions under investigation. The alpha
level for statistical significance was set at p < .05. When the parametric tests (i.e., t-tests)
were used, descriptive statistics were given as mean scores and standard error of mean
(SEM). Descriptive statistics included median scores, minimum, and maximum values
with non-parametric tests (i.e., Mann-Whitney U tests) were used.
It is necessary to note that the results of this multiple t-testing need to be read with
caution because it is usually not acceptable to run a series of t-tests without first run-
ning an ANOVA or MANOVA test. The reason why a series of t-tests and Mann-
Whitney tests were run in this study is because the data analysis did not include
comparisons between all four groups investigated. In other words, Groups S – P and
W + P (i.e., unplanned group on the oral task versus planning group on the written
task) were not compared. Besides, Groups S + P and W – P (i.e., planning group on the
oral task versus unplanned group on the written task) were not compared either be-
cause these comparisons would not be related to any of the research questions that
guided this study.

Results

Results are reported in light of the five research questions that guided this study. Spe-
cifically, the results will be reported with respect to the construct of accuracy (research
questions 1–4), and with respect to the post-task questionnaire (research question 5),
respectively.

Accuracy

Table 2 presents a summary of the descriptive results for accuracy. It illustrates the
percentages of error-free clauses (EFC) and the percentages of correct verb forms
(CVF) for all groups, and Table 3 reports the inferential results showing whether there
were any significant differences between the groups. The effects of planning condition
and modality on the accuracy of the students’ performance are discussed separately.
Table 2 shows that the planning group on the oral task produced more accurate
clauses and verb forms than the no-planning group. The planning group produced
36.07% error-free clauses whereas the no-planning group produced 28.12% accurate
clauses. In terms of correct verb forms, the planning group produced 58.74% correct
verb forms while the no-planning group produced 42.86%.
 Chapter 4.  Performance of Turkish EFL learners 

Table 2.  Descriptive results for accuracy on oral and written tasks
with and without planning

Planning No Planning

Oral Task EFC% CVF% EFC% CVF%

  Mean 36.07% 58.74% 28.12% 42.86%
  SEM 0.02 0.05 0.03 0.05
Written Task
  Mean 37.78% 48.38% 39.52% 62.29%
  SEM 0.04 0.05 0.03 0.02

Planning, however, had adverse effects on accuracy on the written task. Table 2 shows
that the no-planning group produced more accurate clauses and verb forms than the
planning group on the written narrative task. The planning group on the written task
produced 37.78% error-free clauses whereas the no-planning group produced 39.52%
error-free clauses. In the case of verb forms, the planning group achieved 48.38% of
accuracy while the no-planning group’s accuracy was 62.29%.
Table 3 presents the inferential results for the accuracy data and shows whether
there were any differences between the groups on oral and written tasks under the
planning and no-planning conditions.
As shown in Table 3, the differences between the planning and no-planning groups
on the oral task were not significant (that is p = 0.066 for the error-free clauses (EFC)
and p = 0.053 for the correct verb forms (CVF)). Thus, although the differences be-
tween the two conditions came very close to a level of significance, pre-task planning
did not significantly affect students’ accuracy on the oral task.
Table 3 also shows that modality did not affect accuracy significantly in the planned
groups. There was no significant difference between the planned oral group (S + P) and
the planned written group (W + P) in terms of error-free clauses (p = 0.728) or correct
verb forms (p = 0.212). However, effects of modality on accuracy were observed on

Table 3.  Differences between groups for accuracy on oral and written tasks
with and without planning

Groups compared EFC CVF

p-value p-value
S + P vs. S – P 0.066 0.053
S + P vs. W + P 0.728 0.212
S – P vs. W – P 0.026* 0.004*
W + P vs. W – P 0.754 0.046*
*p < .05
 Zubeyde Sinem Genc

tasks under the unplanned condition, with the learners producing more accurate lan-
guage in writing than in speaking under the unplanned condition. The difference was
statistically significant for both error-free clauses (p = 0.026) and correct verb forms
(p = 0.004). Turning to the difference between the planning and no-planning groups
on the written task, this was statistically significant only in the case of correct verb
forms (p = 0.046), but not that of error-free clauses (p = 0.754).
For the sake of convenience, a summary of these results is provided below:
– Modality made a significant difference under the unplanned condition. The stu-
dents were more accurate on the unplanned written tasks than the unplanned oral
tasks.
– Modality did not affect accuracy significantly under planned conditions. The dif-
ference between the two planned groups was not statistically significant.
– Strategic planning did not have any significantly positive effects on students’ ac-
curacy for the oral task. However, the differences were very close to the level of
significance. The planned group produced more accurate language than the un-
planned group on the speaking task.
– Strategic planning had an adverse effect on accuracy for the written task. The un-
planned group produced more accurate language than the planned group on the
writing task.

Students’ priority ordering of grammar, vocabulary, and organization

Table 4 presents the descriptive results for students’ priority order for grammar, vo-
cabulary, and organization before and during task phases as reported by the partici-
pants in the post- task questionnaire. Median score 1 indicates that the item was the
first in the order, meaning that the students reported giving it the greatest amount of
time and attention. Median score 2 indicates that the item was in the second place in
the order, and Median score 3 indicates that the students gave the least amount of time
and attention to the given item. The unplanned groups did not have median scores for
the Before Task column because they were not given any strategic planning time. Thus,
these groups could only report orders for their use of time and attention in the During
Task column only. The planned groups were able to report orders for both the Before
Task and During Task columns.
It is important to note that we are discussing self-report data here because we did
not have direct evidence of how much time the participants actually used for grammar,
vocabulary, or organization before or during these tasks. We relied on the priority or-
der as reported by the students themselves in the post-task questionnaire.
As Table 4 shows, the participants on the planned oral group reported that before
the task they spent most time and attention on organization, followed by vocabulary,
and finally grammar. During the task, grammar and vocabulary received equal
amounts of time and attention from the participants in the planned oral group, and
 Chapter 4.  Performance of Turkish EFL learners 

Table 4.  Results of students’ priority ordering of grammar, vocabulary and organization
before and during task performance

Planning No Planning

Oral Task Before Task During Task During Task

G V O G V O G V O

Median 3 2 1 2 2 3 3 1 2
Minimum 2 1 1 1 1 1 1 1 1
Maximum 3 2 3 3 2 3 3 2 3

Written Task Before Task During Task During Task

G V O G V O G V O

Median 2 2 1 3 2 1 3 2 1
Minimum 2 1 1 1 1 1 1 1 1
Maximum 3 3 3 3 3 3 3 3 3
Key: G = grammar, V = vocabulary, O = organization

organization was lowest for the amount of time allocated. The participants in the un-
planned oral group reported giving the greatest amount of time and attention to vo-
cabulary, followed by organization and finally grammar during the task. The order of
the items in terms of amount of time allocated for the groups performing the written
task under planned and unplanned conditions was almost the same: Organization was
in first place, followed by vocabulary, and finally grammar. The only difference was that
the planned written group reported paying an equal amount of attention to vocabulary
and grammar before the task, but still put organization at the top of their list.
Table 5 presents the inferential results for the data regarding the priority ordering
of grammar, vocabulary, and organization. It reports the results of a Mann-Whitney U
analysis of the differences between the groups, in terms of the priority of attention
which the participants reported allocating to grammar, vocabulary, and organization
before the task (i.e., during the strategic planning time) and during the completion of
the tasks phase.
Table 5 shows that there was a statistically significant difference only between the
unplanned oral and unplanned written groups in terms of attention to vocabulary
(p = 0.021). In other words, under the unplanned condition, the participants reported
giving significantly more time to vocabulary when speaking than when writing. There
were no significant differences between the other groups. A summary of the priority
ordering as reported by all groups is given in Table 6 below:
 Zubeyde Sinem Genc

Table 5.  Differences between groups for priority ordering of grammar, vocabulary,
and organization before and during task

Groups compared Before Task (p-value) During Task (p-value)

G V O G V O

S + P vs. S – P 0.098 0.367 0.539

S + P vs. W + P 0.217 0.539 0.713 0.148 0.683 0.217
S – P vs. W – P 0.653 0.021* 0.217
W + P vs. W – P 0.806 0.325 0.713
*p < .05

Table 6.  Overall summary of students’ priority ordering of grammar, vocabulary,

and organization

Group Before Task During Task

S+P Organization Grammar = Vocabulary

Vocabulary Organization
Grammar
S–P Vocabulary
Organization
Grammar
W+P Organization Organization
Vocabulary = Grammar Vocabulary
Grammar
W–P Organization
Vocabulary
Grammar

Discussion

A summary of the results of the study is first provided in Table 7 in relation to the four
research questions that concern the effect of pre-task planning and modality on stu-
dents’ accuracy.
The first research question asked whether learners would produce more accurate
oral language when they have time for strategic planning. The results showed that the
difference between the planned and unplanned groups on the oral task was not statisti-
cally significant, though coming quite close to the level of significance. This result sup-
ports findings obtained in previous studies including Gilabert (2007), Rutherford
 Chapter 4.  Performance of Turkish EFL learners 

Table 7.  Overall effect of pre-task planning and modality on students’ accuracy

Research question Groups compared Explanation

1 S + P vs. S – P No significant differences observed

2 W + P vs. W – P*
3 S + P vs. W + P
4 S – P vs. W – P*
W – P was significantly more accurate in CVF
(p = 0.046), but not in EFC (p = 0.754)
No significant differences observed
W – P was significantly more accurate in both
EFC (p = 0.026) and CVF (p = 0.004)

(2001), Wendel (1997), and Yuan and Ellis (2003) which also found no significant dif-
ferences for the effect of strategic planning on accuracy measures in oral performances
of learners. Foster and Skehan (1996) too found that planning promoted accuracy on
personal information and decision-making tasks, but not on the oral narrative task
used in their study.
The second research question asked whether EFL students would produce more
accurate written language when they have time for strategic planning. The results re-
vealed that strategic planning had an adverse effect on accuracy on written narrative
tasks. Students produced more accurate written language when they did not have time
for planning. This finding supports those studies, which are few in number, that have
investigated the effects of planning on written language. For instance, Wiggleworth’s
(2001) study found adverse effects of planning on the students’ oral performances on
both structured and unstructured tasks. In contrast, Ellis and Yuan (2005) examined
the effects of careful within-task planning on oral and written performance. They
found that careful on-line planning promoted accuracy and stated that “the careful
group produced more correct clauses and verbs than the pressured group, differences
that were statistically significant and reflected in large effect sizes for both modalities”
(p. 188). Their results are not consistent with the findings of this study, maybe in part
because Ellis and Yuan investigated the effects of within-task planning, which is differ-
ent from the present study which examined the effect of pre-task planning on oral and
written tasks.
The third research question sought to find out whether the modality of the stu-
dents’ production (oral vs. written) has an effect on the accuracy of their language
output when they are given time for strategic planning. The results found no signifi-
cant differences between the planned oral and planned written task performances of
the students in terms of accuracy. In other words, modality did not lead learners to
produce more accurate language under strategic planning conditions. This finding is
similar to the results obtained by Ellis and Yuan (2005) who also found that the differ-
ence between the two careful groups was not statistically significant.
The fourth research question asked whether the modality of learners’ production
(oral vs. written) has an effect on the accuracy of their language output when they do
not have planning time. The results indicate that the difference between the groups of
 Zubeyde Sinem Genc

unplanned oral and unplanned written tasks was significant in terms of both correct
clauses (p = 0.026) and correct verb forms (p = 0.004). Students produced more accu-
rate written language than oral language under unplanned conditions. This finding is
consistent with the results of Ellis and Yuan (2005) who found that “the participants’
language was more accurate when writing than when speaking” (p. 186).
The fifth question sought to shed light on what the learners actually attend to dur-
ing the planning and task completion phases. Results of the post-task questionnaire
showed that the students reported putting more emphasis on the organization of the
story during planning time, and more emphasis on vocabulary and organization dur-
ing the task performance phase. Grammar seemed to draw the least attention from the
learners. This might explain the differential effects of planning and modality on accu-
racy observed in this study. These learners were more concerned with vocabulary and
organization than with the grammatical accuracy of their output.
The findings of this study are in line with the results in Wendel (1997), who inter-
viewed the learners to understand what they actually did when they planned. He found
that all learners focused on sequencing the narrative events, with only three of them
reporting attention to grammar. Similarly, students in this study said they preferred to
use their planning time for organizing their production, but that they also attended
to vocabulary and organization during task performance. They paid the least attention
to grammar. This could be seen as a strategic response on the part of the learner to the
conditions created by the teacher/materials writer. It is also consistent with Skehan’s
(1998) account of trade-off effects, suggesting that learner choice may be to some ex-
tent determined by the external conditions of the task. Skehan’s limited capacity hy-
pothesis proposes that demanding tasks “consume more attentional resources ... with
the result that less attention is available for focus on form” (p. 97). In other words, the
students may have traded off attention to one aspect of production, that is, vocabulary
and organization, to the detriment of the other, that is, grammar, due to limitations in
attentional resources.

Pedagogical implications

Perhaps the most important implication of this study is that it extends research on
important aspects of TBLT and learning research (planning time, accuracy, and oral
and written tasks) to EFL settings, settings not well represented in much of the existing
literature on TBLT (see also Sasayama & Izumi, this volume). The study has shown
that strategic planning does not necessarily lead students in EFL settings to produce
more accurate language on oral narrative tasks. A related implication is that these stu-
dents may not always need pre-task planning time in order to produce more accurate
language on written tasks; this study revealed that learners actually were more accurate
on written tasks under unplanned conditions and significantly less accurate when they
were given time for planning. A related implication is that learners may choose to
 Chapter 4.  Performance of Turkish EFL learners 

spend their planning time focusing on vocabulary and organizational matters. In this
respect, it is recommended to hold guided planning sessions with the students in
which they are specifically asked to pay attention to the accuracy of the target struc-
tures. However, it does not guarantee more accuracy because, as this study has shown,
the modality of the task may also influence whether or not planning time has an im-
pact. This might be again due to the attention and time students spend on organization
and vocabulary rather than on grammar during strategic planning. Thus, teachers may
find guided planning time to be quite beneficial when their aim is to foster accuracy,
but only under certain circumstances. The positive effects of guided planning on learn-
ers’ accuracy have been documented by previous research as well. For example,
Mochizuki and Ortega (2008) found that guided planners produced more accurate
relative clauses than unguided planners.
The results of this study point to the likely importance of the students’ orientation
towards the task and the planning time. Although written modality naturally allows
for more opportunities to focus on form, students’ orientation towards the organiza-
tion of the story during the written task might obscure a more accurate performance
on written modality. In this sense, teachers might need to guide their students to focus
on form during planning. However, the students’ own focus will also be significant in
determining the outcome.

Limitations of the study and directions for further research

This study is not without its limitations. First, the sample size investigated was rela-
tively small. When single phase data collection is conducted in studies, greater num-
bers of participants need to be secured in order to obtain more reliable findings and be
able to make stronger generalizations. Besides examining larger groups, another op-
tion for obtaining more reliable results would be to conduct a longitudinal study where
the individual learner’s performance is examined over a lengthy period of time, which
may reveal different and more comprehensive findings. It is worth noting that the re-
sults obtained for the groups investigated here may not reflect the performance of
individual learners. This is a well-known limitation for averaging group data. For in-
stance, Larsen-Freeman (2009) states that “the average length of error-free T-units
work[s] well at the group level, but not necessarily for individuals. Averaging group
data has its limitations” (p. 585). This is an important issue for research on TBLT in
particular where individual learner variables affect learner performance.
Another limitation is that the study only examined the accuracy of learner perfor-
mance and did not look at other dimensions of language production (such as fluency
and complexity). This focus was chosen because accuracy, unlike fluency and com-
plexity, was the dimension of language production which has shown the least consis-
tent results in previous studies. However, complexity, accuracy, and fluency interact
and exist in an organic relationship with one another (see Norris & Ortega, 2009;
 Zubeyde Sinem Genc

Sasayama & Izumi, this volume), so it may be that factors unexamined in the current
study were exerting a causal influence on the language production of the learners.
Similarly, only low-intermediate learners were investigated in the current study, and it
may well be that learners at other levels of proficiency would respond differently to the
opportunity for pre-task planning.
One way to cater for these limitations is to conduct longitudinal, multivariate
studies in the future in order to better explore learner performance on tasks under dif-
ferent conditions over lengthy periods of time. EFL settings would indeed be ideal
contexts for such research because students usually remain for a lengthy period of time
in classroom instruction. Such research would address a gap in current TBLT research
and application. For instance, Ellis (2009) states that one major limitation in the ma-
jority of existing research “is the lack of a longitudinal study of the effects of strategic
planning. This is obviously desirable to establish whether any benefit gained from
planning a task at one time carries over to a later time” (p. 505). Such research would
also enable us to examine the effect of individual learner variables on language pro-
duction over time. It is obvious therefore that further research is needed in these im-
portant areas of TBLT and learning.
Despite these limitations, this study has helped to extend TBLT research into new
territory, showing that EFL students were more accurate on unplanned written tasks
than unplanned oral tasks. In other respects, the study found no differences in accu-
racy levels – when seen in terms of the effects of modality under planned conditions
and in terms of the impact of strategic planning for the oral task. Strategic planning
had an adverse effect on accuracy for the written task. In terms of their reports on their
focus of attention, learners apparently spent more attention and time on vocabulary
and organization than grammar during the planning time before and during the task
performance phases. Confirming or contesting these findings in other authentic FL
educational contexts awaits further research.

References

Adams, R. (2005). L2 tasks and focus on form: A role for modality? Paper presented at the Inter-
national Conference on Task-Based Language Teaching, September 21–23, 2005, Leuven,
Belgium.
Bygate, M., & Samuda, V. (2005). Integrative planning through the use task-repetition. In R.
Ellis (Ed.) Planning and task performance in a second language (pp. 37–74). Amsterdam:
John Benjamins.
Conover, W. J. (1999). Practical nonparametric statistics (3rd ed.). New York, NY: John Wiley &
Sons.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Ellis, R. (Ed.). (2005). Planning and task performance in a second language. Amsterdam: John
Benjamins.
 Chapter 4.  Performance of Turkish EFL learners 

Ellis, R. (2009). The differential effects of three types of task planning on the fluency, complexity,
and accuracy in L2 oral production. Applied Linguistics, 30, 474–509.
Ellis, R., & Yuan, F. (2005). The effects of careful within-task planning on oral and written task
performance. In R. Ellis (Ed.) Planning and task performance in a second language
(pp. 167–192). Amsterdam: John Benjamins.
Foster, P., & Skehan, P. (1996). The influence of planning on performance in task-based learning.
Studies in Second Language Acquisition, 18, 299–324.
Gilabert, R. (2007). The simultaneous manipulation of task complexity along planning time and
(± Here-and-Now): Effects of L2 oral production. In Garcia-Mayo (Ed.). Investigating tasks
in formal language learning (pp. 44–68). Clevedon, UK: Multilingual Matters.
Kawauchi, C. (2005). The effects of strategic planning on the oral narratives of learners with low
and high intermediate proficiency. In R. Ellis (Ed.). Planning and task performance in a
second language (pp. 143–164). Amsterdam: John Benjamins.
Kuiken, F., & Vedder, I. (2007). Task complexity and measures of linguistic performance in L2
writing. IRAL, 45, 261–284.
Larsen-Freeman, D. (2009). Adjusting expectations: The study of complexity, accuracy and flu-
ency in second language acquisition. Applied Linguistics, 30, 579–589.
Mochizuki, N., & Ortega, L. (2008). Balancing communication and grammar in beginning-level
foreign language classrooms: A study of guided planning and relativization. Language
Teaching Research, 12, 11–37.
Norris, J. M., & Ortega, L. (2009). Towards an organic approach to investigating CAF in SLA:
The case of complexity. Applied Linguistics, 30, 555–578.
Ortega, L. (1999). Planning and focus on form in L2 oral performance. Studies in Second Lan-
guage Acquisition, 21, 109–148.
Ortega, L. (2005). What do learners plan? Learner-driven attention to form during pre-task
planning. In R. Ellis (Ed.). Planning and task performance in a second language (pp. 77–109).
Amsterdam: John Benjamins.
Royston, J. P. (1995). A remark to algorithm AS 181: The W test for normality. Applied Statistics,
44, 547–551.
Rutherford, K. (2001). An investigation of the effects of planning on oral production in a second
language. Unpublished MA Thesis. University of Auckland, New Zealand.
Sangarun, J. (2005). The effects of focusing on meaning and form in strategic planning. In R.
Ellis (Ed.). Planning and task performance in a second language (pp. 111–141). Amsterdam:
John Benjamins.
Skehan, P. (1996). A framework for the implementation of task-based instruction. Applied Lin-
guistics, 17, 38–62.
Skehan, P. (1998). A cognitive approach to language learning. Oxford: Oxford University Press.
Skehan, P., & Foster, P. (1997). Task-type and task processing conditions as influences on foreign
language performance. Language Teaching Research, 1, 185–211.
Skehan, P., & Foster, P. (1999). The influence of task structure and processing conditions on nar-
rative retellings. Language Learning, 49, 93–120.
Skehan, P., & Foster, P. (2001). Cognition and tasks. In P. Robinson (Ed.) Cognition and second
language instruction (pp. 183–205). Cambridge: Cambridge University Press.
Skehan, P., & Foster, P. (2005). Strategic and on-line planning: The influence of surprise informa-
tion and task time on second language performance. In R. Ellis (Ed.). Planning and task
performance in a second language (pp. 193–216). Amsterdam: John Benjamins.
 Zubeyde Sinem Genc

Tajima, M. (2003). The effects of planning on oral performance of Japanese as a foreign language.
Unpublished doctoral dissertation. Purdue University, USA.
Tavakoli, P., & Skehan, P. (2005). Strategic planning, task structure and performance testing. In
R. Ellis (Ed.). Planning and task performance in a second language (pp. 239–273). Amsterdam:
John Benjamins.
Wendel, J. (1997). Planning and second language narrative production. Unpublished doctoral
dissertation. Temple University, Japan.
Wigglesworth, G. (1997). An investigation of planning time and proficiency level on oral test
discourse. Language Testing, 14, 21–44.
Wigglesworth, G. (2001). Influences on performance in task based oral assessments. In M.
Bygate, P. Skehan, & M. Swain (Eds.). Researching pedagogic tasks, second language learning,
teaching and testing (pp. 186–209). London: Pearson.
Yuan, F., & Ellis, R. (2003). The effects of pre-task and on-line planning on fluency, complexity
and accuracy in L2 monologic oral production. Applied Linguistics, 24, 1–27.
 Chapter 4.  Performance of Turkish EFL learners 

Appendix A

Instructions for each task

Oral Task – With Strategic Planning:
“You have seen a set of pictures. These pictures tell us a story. In a short while, I would
like you to tell this story in English. Before you tell the story, you have 10 minutes to
plan what you are going to say.
Tell the story in as much detail as you can. You can take some notes on the paper
I gave you. But don’t write a complete sentence in Turkish or English. When you begin
to tell the story, I will take the paper and the pictures away. You have 8 minutes to tell
the story. You must produce at least 4 sentences for each of the 6 pictures. If you like,
you can produce more than 4 sentences for each picture.
Please prepare now.
(After 10 minutes) Now, it is time for you to begin telling the story.
Oral Task – No Planning:
“You have seen a set of pictures. These pictures tell us a story. Now, I would like you to
tell this story in English. Please tell the story in as detailed as you can. You have 8 min-
utes to tell the story. You must produce at least 4 sentences for each of the 6 pictures. If
you like, you can produce more than 4 sentences for each picture.
Please begin telling the story.
Written Task – With Strategic Planning:
“You have seen a set of pictures. These pictures tell us a story. In a short while, I would
like you to write this story in English. Before you write the story, you have 10 minutes
to plan what you are going to write.
Write the story in as much as detail you can. You can take some notes on the paper
I gave you. But don’t write a complete sentence in Turkish or English. When you begin
to write the story, I will take the paper and the pictures away. You have 15 minutes to
write the story. You must produce at least 4 sentences for each of the 6 pictures. If you
like, you can produce more than 4 sentences for each picture.
Please prepare now.
(After 10 minutes) Now, it is time for you to begin writing the story.
Written Task – No Planning:
“You have seen a set of pictures. These pictures tell us a story. Now, I would like you to
write this story in English. Please write the story in as detailed as you can. You have
15 minutes to write the story. You must produce at least 4 sentences for each of the 6
pictures. If you like, you can produce more than 4 sentences for each picture.
Please begin writing the story.
 Zubeyde Sinem Genc

Appendix B

Set of pictures used in tasks

 Chapter 4.  Performance of Turkish EFL learners 

Appendix C

Post-task questionnaires
Oral Task – With Strategic Planning:
Name:
Proficiency level in English:
Age: Major:
I. During the 10-minute planning period, how did you plan? Did you think more
about grammar, vocabulary, or the best way to organize your story?
Put these in an order of priority according to time and attention you spent for each
during the10-minute planning period before you told the story.
II. While you were telling the story, did you think about grammar, vocabulary, or the
best way to organize your story?
Put these in an order of priority according to time and attention you spent for each
while you were telling the story.
Oral Task – No Planning:
Name:
Proficiency level in English:
Age: Major:
While you were telling the story, did you think about grammar, vocabulary, or the best
way to organize your story?
Put these in an order of priority according to time and attention you spent for each
while you were telling the story.
Written Task – With Strategic Planning:
Name:
Proficiency level in English:
Age: Major:
I. During the 10-minute planning period, how did you plan? Did you think more
about grammar, vocabulary, or the best way to organize your story?
Put these in an order of priority according to time and attention you spent for each
during the10-minute planning period before you wrote the story.
II. While you were writing the story, did you think about grammar, vocabulary, or the
best way to organize your story?
Put these in an order of priority according to time and attention you spent for each
while you were writing the story.
 Zubeyde Sinem Genc

Written Task – No Planning:

Name:
Proficiency level in English:
Age: Major:
While you were writing the story, did you think about grammar, vocabulary, or the
best way to organize your story?
Put these in an order of priority according to time and attention you spent for each
while you were writing the story.
 chapter 5

Effects of task instructions on text

processing and learning in a Japanese
EFL college nursing setting

Yukie Horiba and Keiko Fukaya

Kanda University of International Studies, Japan and Tsuda College, Japan

This study investigated the effect of task instructions on text processing and
learning. Seventy limited L2 proficiency college nursing majors in a Japanese EFL
context read a text in English and recalled the content of the text. Some students
were told in advance to expect a recall task in a particular language which they
later did in the same language (the L1-only and the L2-only conditions). Others
were told to expect an L1 recall task which subsequently they were actually
asked to do in the L2 (the L1-L2 condition). All students took an unplanned
vocabulary test on unfamiliar words contained in the text. The results suggest
that the L1-only condition enhanced content recall while the L2-only condition
was more conducive to vocabulary acquisition, with the L1-L2 condition being
ineffective on both measures. Results of the study are discussed in terms of
resource allocation and levels of representations. An important contribution of
this study is that it addresses major questions like ‘whether limited L2 proficiency
learners can learn both content and language simultaneously in the L2’, and ‘how
task conditions can influence students’ text processing capacity and subsequent
learning’ in EFL within the principles of TBLT.

Introduction

Processing and comprehending texts (or discourse) for professional purposes is an

important skill for individuals, whether the text is presented in the first language (L1)
or the second/foreign language (L2). People acquire knowledge about the content of
the texts and later use that knowledge when they perform various tasks that require
information about it, such as recall/retelling, answering questions, making decisions,
and carrying out professional activities. Students in EFL classrooms are often given
similar tasks for practice. They are asked to process a text written in the L2 for compre-
hension so that they learn what the text is all about (e.g., a topic about the target cul-
ture/society, a subject matter, or professional domain). Oftentimes students are also
 Yukie Horiba and Keiko Fukaya

expected to acquire knowledge of the target language in which the text is written.
Given the limitations of human beings’ cognitive capacity, to what extent can students
learn both content and language through processing a text that is presented in the L2?
How does task condition influence students’ text processing and subsequent learning?
The present study was designed to provide some answers to these questions in the
context of an English as a foreign language (EFL) program for college nursing majors
in Japan.

Theoretical framework

Text comprehension is an active process in which the reader engages in both bottom-
up and top-down processing in order to construct a coherent representation of the text
in memory (Graesser, Gernsbacher, & Goldman, 2003). It is generally agreed among
researchers that the resulting representation of the text in memory consists of multiple
levels. Major levels are the surface code, the propositional textbase, and the situation
model (Kintsch, 1998; van Oostendorp & Goldman, 1999). The surface code level rep-
resents the memory of the surface linguistic code of the text such as word order and
spelling, and this memory decays fast. The meaning of the text is represented as text-
base and situation model. The textbase represents the propositional content of the text
(i.e., basic ideas explicitly asserted in the text); it consists of a microstructure (i.e., local
propositions) and a macrostructure (i.e., gist of the text). The macrostructure is some-
times directly signaled in a text, but often it must be inferred by the reader. The situa-
tion model level (sometimes called the mental model) concerns the memory for the
referential content (i.e., people and objects, events, ideas and opinions) or micro-world
that the text is describing (Zwaan & Radvansky, 1998). The result of comprehension is
the textbase only in rare cases.
Usually in order to understand a text, comprehenders supplement the textual in-
formation with various information (i.e., inferences) from their knowledge and expe-
rience (long-term memory), so as to achieve a personal interpretation of the text that
is related to other information held in long-term memory. Thus, the mental represen-
tation of a text a reader constructs includes text-derived propositions (the textbase, not
necessarily complete) plus varying amounts of knowledge elaboration and knowledge-
based interpretations of the text – the situation model. The representation of the text
will be retrieved and used in later tasks, such as answering questions and recall/retell-
ing that requires the use of information about the text. Hence, the construction of the
textbase and the situation model is considered particularly important for learning
from texts (Kintsch, 1998; McNamara, 2007).
Because of the limitations of human beings’ cognitive capacities, allocating cogni-
tive resources during text processing is important for successful comprehension. Allo-
cation of cognitive resources during text processing depends in part on the reader’s goal
and strategy use in a given situation (Alexander, Jetton, & Kulikowich, 1996; Kintsch,
 Chapter 5.  Text processing and learning 

1998; McNamara, 2007). Reader goal and strategy use are often influenced by task re-
quirements. Research has shown that L1 readers generate different patterns of infer-
ences depending on the task instructions (e.g., whether reading is to explain, predict, or
understand) (Magliano, Trabasso, & Graesser, 1999). Even with the same text, readers
process and represent the text differently when they are told to read a news story as op-
posed to a literary text (Zwaan, 1994) or when they are told to read for entertainment
rather than for study (van den Broek, Lorch, Linderholm, & Gustafson, 2001).
The general model of text processing and representation described above has been
increasingly accepted among L2 reading researchers. Research has shown that L2 read-
ers generate inferences and construct text bases and situation models of texts (Barry &
Lazarte, 1998; Horiba, 1996). L2 readers allocate cognitive resources differently and
construct different patterns of text representation depending on task type (i.e., to read
freely vs. for coherence) (Horiba, 2000). However, L2 text processing tends to be con-
strained by level of language proficiency (Lee & Schallert, 1997), in particular vocabu-
lary knowledge (Koda, 2005), as well as the text’s linguistic features (Barry & Lazarte,
1995; Sasayama & Izumi, this volume; Stevensen, Schoonen, & de Glopper, 2003;
Zwaan & Brown, 1996). In this connection, research has shown that the use of the L2
in recall, as opposed to the use of the L1, may reduce the amount of content informa-
tion recalled (Donin & Silva, 1993; Lee, 1986), which may be related to language pro-
ficiency and L1-L2 distance (Chen & Donin, 1997). However, it is not yet clear wheth-
er the disadvantage of recall in L2 is caused by the difficulty of producing sentences in
L2 recall (i.e., the problem of output) or whether it is also caused by the difficulty of
encoding content information into memory, even when students are given advance
notice of an L2 recall task before reading. Furthermore, it has not been made fully clear
how the effect of a recall task may interact with the effect of language proficiency and
vocabulary knowledge in L2 text comprehension.
As for language learning through text processing, research has shown that L2 stu-
dents may learn new words incidentally through reading (Hulstijn, Hollander, &
Greidanus, 1996). Incidental vocabulary acquisition is influenced by various factors
such as the reader’s vocabulary knowledge and topic familiarity (Pulido, 2009), item
characteristics (Hulstijn et al., 1996), as well as task type (Hulstijn & Laufer, 2001; Kim,
2006). Regarding the task effect, Laufer and Hulstijn (2001) propose that incidental
vocabulary learning may be largely explained by the degree of cognitive and motiva-
tional involvement while processing the word in a given situation. Ellis (1997) claims
that the mapping of form to meaning for a new word is strongly affected by how much
the learners elaborate, integrate, and engage with the new semantic and conceptual
knowledge.
In SLA research it is repeatedly argued that attention must be allocated to the pro-
cessing of form as well as meaning of the language, in order for an item to be learned
(Long & Robinson, 1998; Swain & Lapkin, 2001; VanPatten, 1996; among many
others). Due to the limitations of cognitive resources or attention, there may be a
trade-off in performance between different aspects of L2 processing or production
 Yukie Horiba and Keiko Fukaya

(e.g., complexity, accuracy, and fluency) (Skehan, 1998; see also Sasayama & Izumi,
this volume). Recently, some researchers (Bygate, 2001; Izumi, 2003; Skehan, 2009;
Tavakoli & Foster, 2008) have started to use ideas from Levelt’s (1989) psycholinguistic
model of L1 speaking in order to describe the effect of task variables (e.g., planning
time, repetition, task complexity) on L2 production as a function of resource alloca-
tion during processing. However, historically speaking the mainstream SLA research
has been largely uninformed by cognitive or psycholinguistic research on text compre-
hension, and there is little research so far on the connections between text processing
and learning in L2 context – how L2 text (as input) processing results in memory rep-
resentations (as changes in long-term memory), from which information is retrieved
for use in retelling/recall (as output).

The study: Setting and task

In the present study, a task is conceptualized as a goal-oriented activity, which involves

a meaningful, real-world process of language use, and which can engage any or all of
the four language skills as well as cognitive processes (Ellis, 2003; Skehan, 1998). Ac-
cording to Nunan (2004), tasks can be best analyzed in terms of three elements: goals,
input, and procedures. The type of task investigated in this study is a task in which an
individual reads a text (the input) for comprehension and later recalls the content of
the text as if to inform someone who has not read it (the goals). The task consists of a
sequencing of two (sub)tasks: first reading a text for comprehension, and second re-
calling its content. College students majoring in nursing processed passages about a
health-care case selected from a professional journal of nursing written in the L2. It is
reasonable to claim that the task is a type of authentic classroom activity which re-
sembles a real-life communicative task in the target profession.
During the read-and-recall procedure, students received task instructions regard-
ing the language of their recall on two occasions. Before reading the text, students were
given a first set of instructions: they were told to read the text and were informed of a
later recall task and which language (L1 or L2) they should use for the recall task. After
reading, the students were given a further set of instructions: they were told to recall
what they understood of the content of the text (without referring back to the text) and
again, which language they should use for their recall. However this time, one of the
three groups was told to carry out the recall in a different language from the one they
had been initially primed to use. Upon receiving the encoding instructions, students
were expected to set their plans and start executing their strategies in order to process
the text for a later recall task in the language they had been told to expect to use. Upon
receiving the retrieval/output instructions, the researchers expected the students ei-
ther to proceed by executing their plans and strategies to recall in the originally primed
language (i.e., the L1-only condition and the L2-only condition) or to have to reset
 Chapter 5.  Text processing and learning 

their plans and strategies in order to produce a recall in the different language (i.e., the
L1-L2 condition).
In this study, content recall and vocabulary acquisition were examined as a func-
tion of task condition. Students receiving a different set of task instructions were ex-
pected to allocate cognitive resources differently to various levels of text processing
which further influence the memory representation of the text. By examining and
comparing students’ performance in content recall and vocabulary acquisition under
these task conditions, this study attempts to shed light on how strategic text processing
induced by task instructions is related to resource allocation and subsequent learning
for these EFL learners.
Based on the objectives of the study and the considerations above, the specific re-
search questions posed for the present study were as follows:
Research Question 1: Is there a significant relationship between task condition and the
amount of content information recalled from the target text?
Research Question 2: Is there a significant relationship between task condition and the
number of new words incidentally acquired from the text?
Research Question 3: Is each given task condition equally effective for content learning
and incidental vocabulary acquisition?

Method

Participants

As stated above, the study was conducted in the context of the learning of English as a
foreign language. The target population was college undergraduates who were native
speakers of Japanese and were enrolled in a professional college of nursing in their
home country Japan. As with many other EFL contexts in the world, the importance of
the English language has been increasingly recognized for educational, social, eco-
nomical, and political considerations, both in society in general, and in some profes-
sions like health care, in particular.
Seventy Japanese-speaking EFL students who were college nursing majors partici-
pated in the study. They consisted of 68 females and 2 males. Of these 70 students, 31
were freshmen, 14 sophomores, 16 juniors, and 9 seniors. The students had six years of
English training in high schools. They were enrolled in a small prestigious private col-
lege in Japan which offers undergraduate and graduate programs in nursing. All cours-
es are taught in Japanese, except for English language courses, which are taught by
three instructors (two native speakers of Japanese and one native speaker of English).
Students are required to take four English courses (about three hours per semester)
 Yukie Horiba and Keiko Fukaya

each during the first two years and none during the remaining two years. The partici-
pants were paid volunteers.

Task conditions

Each participant read a text written in English and later recalled the content of the text.
Three different task conditions were manipulated through a combination of instruc-
tions given before reading, and instructions given after reading at the time of recall
regarding the language (L1 or L2). Before reading, some students were informed that
they would later be asked to recall the content of a text using the L1, while others were
informed that their recall task would be in the L2. After reading, the students first
solved some arithmetic problems (given as a distracter to erase the contents of their
short-term memory) and then did the recall task. At the time of recall, the students
who were given advanced notice of the recall-in-L2 task were asked to recall in the L2
(i.e., the L2-only condition; N = 23). Half of the students who were earlier informed of
the recall-in-L1 task were asked to write their recall in the L1 (i.e., the L1-only condi-
tion; N = 24), whereas the remaining students were – unexpectedly – asked to produce
their recall in the L2 (i.e., the L1-L2 condition; N = 23). Participants were encouraged
to write down as much as possible of what they remembered from the content of the
text as if to inform someone who had never read the text before.

Materials

The L2 proficiency and vocabulary knowledge measures. Because there is ample evidence
in the literature that the level of L2 proficiency and of L2 vocabulary knowledge can
affect text comprehension and learning (Bernhardt, 1991; Grabe, 2009), the proficiency
level of these participants and their level of L2 vocabulary knowledge were measured
using a TOEFL-ITP test and the Vocabulary Levels Test (VLT) (Nation 2001), respec-
tively. The scores on these tests were used to check whether the participants were dis-
tributed equally in terms of L2 proficiency across cells. The TOEFL-ITP test consisted
of reading and structure sections. The VLT test consisted of words from the 2000, 3000,
5000 levels and academic vocabulary. (In this study, words from the 10,000 level were
not used.) There were 10 questions for each level and 12 for the academic vocabulary.
Reading materials. Two short English-language narrative texts (about 430 words
in length), the Decision text and the Hope text, were used in the study (see Appendix).
Each text describes a medical case of a patient and his relationship to his family and
the medical professionals. The texts were originally selected from the American Jour-
nal of Nursing and adjusted so that the two texts would be similar in length and linguis-
tic difficulty.
Each text contained 25 words which were identified to be potentially unknown to
the participants on the basis of the results of pilot tests. Each of these 25 words was un-
derlined with the L1 translation provided underneath. This type of glossing is commonly
 Chapter 5.  Text processing and learning 

used in L2 reading materials and therefore was considered appropriate for the present
study. Among the 25 underlined words, 15 words were selected as targets; all the target
words were at the 5000-word level or above, listed in Mosby’s Medical and Nursing and/
or Steadman’s Medical Dictionary and considered of educational importance. Among
the remaining 10 underlined words five were related to health care and five were not.
The text adaptation was made through discussion with a group of applied lin-
guists, EFL instructors, and a native English professor of nursing. The target words and
the other unknown words were selected through pilot tests with other groups of EFL
learners (including a group of 56 students) whose linguistic competence and educa-
tional background were similar to the participants in the present study.
The recall test. In the recall test, participants were asked to write down everything
that they remembered of the content of the text that they had just read as if to inform
someone who had not read the text. In the L1-only condition, they wrote their recall
in the L1 (Japanese). In the L1-L2 condition and the L2-only condition, they wrote
their recall in the L2 (English).
Vocabulary acquisition test. After the recall task, the participants took an unplanned
vocabulary test, the purpose of which was to examine their incidental learning of the
target words contained in the text. The test consisted of two production and two recep-
tion subtests. In the production-in-isolation subtest, participants were asked to write
the target word for each L1 translation (i.e., translate from L1 to L2). In the production-
in-context subtest, they were asked to complete the target word form given the first
syllable of the word provided in its original sentence context. In the form-recognition
subtest, they were asked to judge whether or not they thought each of the given words
had appeared in the passage they had just read. In the meaning-recognition subtest,
they were asked to write in L1 the meaning of each of the given words (i.e., translate
from L2 to L1). In addition, they were asked to report whether or not they had known
each of the given words prior to the reading. For the production-in-isolation and the
two recognition subtests, the 15 target words plus 20 fillers (10 other words from the
same passage and 10 words from outside) were presented in a random order.

Procedure

After the general purpose and experimental procedures were explained in their L1, the
participants signed a consent form and filled out a basic learner-profile questionnaire.
First, the participants took the TOEFL (80 min.). Second, they were randomly as-
signed to one of the three task conditions described earlier and one of the two pas-
sages each by receiving an envelope that contained the reading materials. They read a
text (12 min.), solved some arithmetic problems (given as a distracter to erase the
contents of their short-term memory), and wrote their recall (30 min.). After the recall
task, they took a vocabulary acquisition test (20 min.) and then the VLT (25 min.). A
total of five meetings were scheduled to accommodate as many students as possible
within a two-month period.
 Yukie Horiba and Keiko Fukaya

Analysis of data and scoring

L2 proficiency and vocabulary knowledge. Responses on the TOEFL-ITP test were

scored by the TOEFL office. Responses on the VLT test were scored independently by
two judges using predetermined answer keys. Interrater reliability for scoring the VLT
responses was 1.00.
Recall. Each text was first parsed into statements or events (each statement repre-
senting an action, a state, or an event) (Trabasso, Secco, & van den Broek, 1984;
Trabasso, van den Broek, & Suh, 1989); 59 and 48 events were identified for the Hope
text and the Decision text, respectively. Using this list of events as a template, each re-
call protocol was scored independently by two judges. Interrater reliability for the
scoring of recall was .93; all the discrepancies were resolved through discussion.
Vocabulary acquisition. First, “genuine” new words for each participant were iden-
tified based on his/her prior-knowledge report as part of the meaning-recognition
subtest. Then the number of target words was adjusted for each participant, ranging
between 10 and 15 (M = 12.8, SD = 1.6). Individuals had similar numbers of target
words regardless of task condition, F(2,69) = .138, p = .87, and to text topic, F(1,69) =
1.918, p = .17.
Responses on the two production subtests and the meaning-recognition subtest
were scored by awarding one point for a correctly produced word and a half point for
a partially correct word. Responses on the form-recognition subtest were scored by
subtracting the number of falsely recognized dummy words which were not in the pas-
sage from the number of correctly recognized target words. The maximum score for
each subtest ranged between 10 and 15 points. All the responses on the vocabulary
acquisition test were scored independently by several judges and inter-rater reliability
of over .90 was achieved. All the disagreements were resolved by another scoring by
one of the raters.

Results

Results of the study will be presented for level of language proficiency, recall, and vo-
cabulary acquisition. For both recall and vocabulary acquisition, the main results will
be reported first and then the results of an analysis of variance (ANOVA) with task and
text as between-subjects factors will be reported.

Overall language proficiency and vocabulary knowledge

First of all, the mean and the standard deviation of the percentage scores on the TOE-
FL were M = 65.7 and S.D. = 7.6. In order to see if the participants were equally distrib-
uted across cells, a two-way ANOVA with task and text as between factors was con-
ducted. There were no significant main effects of task, F(2,69) = .449, p = .64, and text,
 Chapter 5.  Text processing and learning 

F(1,69) = .187, p = .67. The interaction between task and text was not significant,
F(2,69) = 2.351, p = .10. In other words, the students’ levels of language proficiency
were not significantly different across the three groupings for task condition and across
the two texts. As for the VLT, the mean and the standard deviation were M = 67.5 and
S.D. = 14.9. A two-way ANOVA showed similar null results. There were no significant
main effects of task, F(2,69) = .601, p = .55, and text, F(1,69) = .032, p = .86. The inter-
action between text and task was not significant, F(2,69) = 1.311 p = .28. Thus, the
participants were equally distributed across task conditions and between texts in terms
of overall language proficiency and of vocabulary knowledge. There was a relatively
high correlation between TOEFL and VLT, r = .689, p = .0001.

Recall

Table 1 shows a summary of the percentage for events recalled for each text by task
condition. Overall, those in the L1-only condition performed best, followed by those
in the L1-L2 condition, and those in the L2-only condition did poorest. The content of
the Decision text was recalled better than that of the Hope text.
A two-way ANOVA revealed that task, F(2,69) = 23.710, p = .0001, and text,
F(1,69) = 28.438, p = .0001, had a significant effect on recall, and that the effect of the
interaction between task and text did not reach a significant level, F(2,69) = 2.594,
p = .08. Tukey-Kramer HSD tests (α = .05) conducted for comparison between task
conditions indicated that the L1-only condition produced significantly more content
information than the L1-L2 condition and the L2-only condition. The difference be-
tween the L1-L2 condition and the L2-only condition was not significant.

Vocabulary acquisition

Table 2 shows a summary of the total scores on the vocabulary acquisition test by text
and task condition. On average, those in the L2-only condition outperformed those in
the L1-only condition and the L1-L2 condition. The target words in the Hope text were
scored higher than those in the Decision text. A two-way ANOVA revealed that the

Table 1.  Percentage of events recalled by text and task condition

Task the Decision text the Hope text Total

n M (SD) n M (SD) n M (SD)

L1-only 11 62.7 (22.9) 13 37.0 (12.5) 24 48.4 (22.0)

L1-L2 12 38.1 (14.7) 11 16.5 (8.2) 23 27.8 (16.1)
L2-only 10 27.0 (15.5) 13 19.5 (7.3) 23 22.8 (11.9)
All 33 42.9 (23.0) 37 24.8 (13.2) 70 33.3 (20.5)
 Yukie Horiba and Keiko Fukaya

Table 2.  Vocabulary acquisition (total scores) by text and task condition

Task the Decision text the Hope text Total

n M (SD) n M (SD) n M (SD)

L1-only 11 5.1 (2.9) 13 5.9 (4.3) 24 5.5 (3.7)

L1-L2 12 3.6 (2.1) 11 6.4 (6.5) 23 5.0 (4.8)
L2-only 10 6.3 (6.3) 13 10.0 (6.9) 23 8.4 (6.8)
All 33 4.9 (4.1) 37 7.5 (6.1) 70 6.3 (5.4)

Table 3.  Vocabulary acquisition (subtest scores) by task condition

Task Production- Production- Form- Meaning-

in-isolation in-context recognition recognition

n M (SD) M (SD) M (SD) M (SD)

L1-only 24 .23 (.47) .44 (.54) 3.04 (2.16) 1.81 (1.67)

L1L2 23 .26 (.47) .43 (.75) 2.83 (2.57) 1.43 (1.91)
L2-only 23   .74 (1.32)   .87 (1.28) 4.04 (1.94) 2.76 (3.11)
All 70 .41 (.87) .58 (.92) 3.30 (2.27) 2.00 (2.34)

effect of task, F(2,69) = 2.481, p = .09, and of text, F(1,69) = 2.880, p = .05, approached
significance, and the effect of the interaction between task and text was not significant,
F(2,69) = 0.496, p = .61.
The vocabulary acquisition test consisted of four subtests. A summary of the sub-
test scores is shown in Table 3. The patterns of the subtest scores were similar across
task conditions. Recognition scores were higher than production scores. Regardless of
the type of subtest, those in the L2-only condition performed better than those in the
L1-only and the L1-L2 conditions. Separate ANOVAs conducted for each subtest
showed that there were no significant effects for task.

Discussion

The results reported above will be discussed in light of the research questions that
guided the study. The first research question asked if there is a significant relationship
between task condition and amount of content information recalled of the text. The
recall results showed that those in the L1-only condition recalled significantly
more content information of their text than those in the L1-L2 and the L2-only condi-
tions. In fact, this effect of task condition on recall was found even when the effects of
overall language proficiency and vocabulary knowledge were statistically controlled as
 Chapter 5.  Text processing and learning 

covariate. Therefore, it is reasonable to think that these L2 students processed, repre-

sented, and recalled their text differently depending on the kind of task instructions
they received.
During the read-and-recall task, they were given task instructions on two occa-
sions. The first set of instructions was given before reading a text, and the second at the
time of recall. How did the two sets of instructions affect their cognitive processes and
strategies during the course of encoding textual information into memory and retriev-
ing it for output? Presumably, when students received the first set of task instructions,
they set a goal and plans for the read-and-recall procedure. Those who were informed
of a later L1-recall task may have focused on the conceptualization of the content of a
text, actively utilizing background knowledge and generating knowledge-based infer-
ences, while progressing through the text. Emphasis on conceptual processing would
result in stronger textbase and stronger situation model representations. When they
received an L1-recall task instruction at the time of recall (i.e., the L1-only condition),
students presumably continued executing their plans, retrieving content information
of the text from long-term memory. This would explain their superior content recall.
In contrast, those in the L2-only condition, who were informed of an L2 recall
task, were expected to set a different goal and employ different strategies, compared to
those who were informed of a forthcoming L1 recall task. Presumably students in the
L2-only condition would have allocated a large portion of their cognitive resources to
linguistic processing and hence engaged less actively in conceptual processing while
progressing through the text. Their resulting representation of the text would likely be
less coherent at the textbase and situation model levels. At the time of recall, they re-
ceived an instruction to carry out the recall in the L2 and therefore continued follow-
ing their plans, retrieving text memory for output. This L2 orientation likely was the
reasons for poor content recall.
As suggested in previous studies (e.g., Donin & Silva, 1993; Lee, 1986), the superi-
ority of the L1-only condition over the L2-only condition in recall might also be ex-
plained, at least in part, by the fact that L2 students with limited language proficiency
could express what they understood of the text by using L1 much better than by using
L2. However, the comparison between the L1-L2 condition and the L2-only condition
and between the L1-only condition and the L1-L2 condition in this study suggests that
task instructions may have influenced the encoding and construction of representa-
tion of the text in memory.
Those in the L1-L2 condition, who were told to expect an L1 recall task before
reading, likely processed and represented their text in memory in a similar manner to
those in the L1-only condition. When they were given the second task instructions
(i.e., recall in L2) at the time of recall, they needed to change their plans and strategies
to perform the recall task using L2 instead of L1. Assuming that their text memory
contained little surface code information, since they had been working towards L1
recall, they therefore likely had to reconstruct L2 forms mostly on the basis of their
content information in memory. This reconstruction of the L2 linguistic code would
 Yukie Horiba and Keiko Fukaya

be relatively easy where a stable, coherent (textbase and situation model) representa-
tion of the text exists in memory, but it would be more difficult where the representa-
tion of the text is underdeveloped.
The two narrative texts used in the study differed in terms of comprehension dif-
ficulty. Those in the L1-only condition recalled 62.7% of the content information of the
Decision text while those in the same condition recalled only 37% of the content of the
Hope text. Therefore, for the easier Decision text, those in the L1-L2 condition were
able to produce a greater amount of content information from an L1 processing base,
even though using L2 in the recall phase, compared to those in the L2-only condition.
For the more difficult Hope text, those in the L1-L2 condition could not produce much
content information.
The second research question asked if there is a significant relationship between
task condition and number of new words incidentally acquired from the text. The re-
sults of the vocabulary acquisition test showed, descriptively speaking, that those in the
L2-only condition performed better than those in the L1-L2 and the L1-only conditions.
The effect of task condition on vocabulary acquisition was very close to significance.
Although strictly speaking non-significant, this result is worth discussing. First, it
appears that those who processed their text with the intention of recalling in L2 and
wrote their recall using L2 (i.e., the L2-only condition) may turn out to be in a better
position to learn target words than those in other conditions. If this turns out to be the
case, it might be because those in the L2-only condition constructed a relatively stron-
ger surface code of the text including the target words. When they produced their
recall using L2 words and sentences, the surface code information of the text was reac-
tivated and the opportunities to (re)process the formal information of the words in the
text increased. This continued attention to the surface code of the text and increased
opportunities to process linguistic (L2) information throughout the read-and-recall
procedure may prove to help incidental learning of form and form-meaning relations
for the target words. In contrast, students in the L1-only condition and the L1-L2 con-
dition may be emphasizing conceptual processing and allocating fewer resources to
linguistic processing during reading. Their resulting representations of the text would
then contain little surface code information. Those who recalled in their L1 (i.e., the
L1-only condition) did not have the opportunity to (re)process the formal information
of the words during recall. Those who were given the opportunity for output or recall
in L2 (i.e., the L1-L2 condition) likely found it harder to recover much of the surface
linguistic information. This may cause students in the L1-only condition and the L1-
L2 condition to do poorly on the vocabulary acquisition test. As suggested in prior
research, the connection between form and meaning for the new target words may be
affected by the amount of elaboration and integration (Barcroft, 2004; Ellis, 1997). The
finding in this study that those in the L1-L2 condition did not perform better on vo-
cabulary acquisition than those in the L1-only condition may point to the importance
of the encoding stage (over the retrieval/output stage) of text (input) processing
for learning.
 Chapter 5.  Text processing and learning 

Second, it is possible that the target words were perceived as important by these L2
students regardless of task condition. The target words were all related to health care
and presented with the L1 glossing. In each text, there were 25 unfamiliar words with
the L1 glossing provided, among which 20 words were related to health care (including
15 target words). It may be that these nursing majors found these health care related
words relevant and tried to encode the information of the words in memory, even
though they were not told in advance of a vocabulary acquisition test. This would re-
duce the effect of the task condition.
The third research question asked if the same set of task instructions was equally
effective for content learning and incidental vocabulary acquisition. The results of re-
call and vocabulary acquisition tests together suggest that a task condition that facili-
tates the learning of the content information of a text may turn out to be less effective
for incidental acquisition of new words contained in the text. Likewise, a task condi-
tion that is more conducive to vocabulary acquisition may not be as effective for the
learning of the content of the text. However without a significant result on the vocabu-
lary acquisition test, responses to this question cannot be conclusive.
Based on these findings, it is reasonable to consider that there may be some kind of
trade-off between conceptual processing and linguistic processing during text process-
ing for these students whose L2 proficiency is limited. As already mentioned, students
in the L1-only condition seem to have strategically emphasized the conceptual process-
ing aspect of the text and may (although this has not been shown) have allocated fewer
cognitive resources to the linguistic processing aspect during the course of reading-
and-recalling, resulting in superior content learning and little incidental vocabulary
acquisition. In contrast, students in the L2-only condition may have allocated more of
their cognitive resources to the linguistic processing aspect of the text during reading
and recalling, resulting in poor content learning but performing relatively better on
incidental vocabulary acquisition. Thus, different sets of task instructions may have
induced different sets of strategic processing, resulting in different learning outcomes
between the various conditions. The idea of a trade-off in resource allocation leading to
different learning outcomes in the context of this study is in line with cognitive research
on text processing and learning, and seems to corroborate the limited-attention model
proposed by Skehan (1998, 2009) and task repetition research (Bygate & Samuda, 2005;
see also Genc, this volume; Sasayama & Izumi, this volume).
How was level of L2 proficiency related to text processing and learning in a par-
ticular task condition? Additional analysis indicated that there were reliable, moderate
to relatively high correlations between the VLT and recall in all three conditions
(r = .42–.61) and between the VLT and vocabulary acquisition scores in the L1-only
and the L2-only conditions (r = .52–.53; c.f., the L1-L2 condition, r = .18). Further-
more, there were reliable moderate correlations between recall and vocabulary acqui-
sition in the L2-only condition (r = .43; c.f., the L1-only and the L1-L2 conditions,
r = .10, –.18). These correlation results indicate that the effects of L2 proficiency on text
processing and learning were not the same across the three conditions. The correlations
 Yukie Horiba and Keiko Fukaya

between recall and vocabulary acquisition for the L2-only condition suggest that those
who were more successful in conceptualizing the content of a text, through analysis of
the linguistic information presented as text, and in recalling or reconstructing the text
(no matter how fragmented it might be) using the L2, had a better chance to develop
in their L2 knowledge some new connections for the unfamiliar words contained in
the text. On the other hand, the lack of correlations between recall and vocabulary
acquisition in the L1-only and the L1-L2 condition, and the lack of correlations be-
tween the VLT and vocabulary acquisition in the L1-L2 condition, seemed to suggest
that language proficiency is particularly important at the encoding stage of text pro-
cessing. Further, they may indicate that vocabulary knowledge used when processing
a text for comprehension, which should involve knowledge of many words and word
relations, may not correspond to the kind of vocabulary knowledge used when per-
forming on the vocabulary acquisition test, the focus of which was on a small number
of words contained in the text. Thus, level of language proficiency, in particular vo-
cabulary knowledge, intervened in the effect of task condition on L2 students’ content
recall and vocabulary acquisition from the text. The way language proficiency influ-
ences L2 students’ text processing and learning seemed to be partly explained by how
cognitive resources were allocated under specific task conditions.

Conclusions and implications of the study

Based on the results obtained in this study, the following conclusions can be made.
First, the cognitive processes and strategies EFL students employ under a particular
task condition can affect the recall of content of a text and may affect incidental learn-
ing of unfamiliar words contained in the text. Second, a task condition which empha-
sizes conceptual processing can facilitate content learning, but may not be effective for
incidental vocabulary learning. In contrast, a task condition which involves heavy lin-
guistic processing may turn out to be more effective for incidental vocabulary learning
than a task condition emphasizing conceptual processing, but may not be beneficial
for content learning. Third, the level of language proficiency, in particular vocabulary
knowledge, affects students’ processing and learning from the text. Fourth, due to the
limitation of the cognitive capacity, it is unlikely that simultaneous learning of content
and language from text takes place automatically when limited-L2 proficiency stu-
dents engage in a read-and-recall task.

Pedagogical implications

A number of pedagogical contributions and implications may be drawn based on the

findings of the study. The most important contribution of this study is that it sought to
explore whether limited L2 proficiency learners can learn both content and language
 Chapter 5.  Text processing and learning 

simultaneously in the L2, and how task conditions influence students’ text processing
capacity and subsequent learning in an EFL setting within the framework of TBLT. The
results of this study suggest that the type of learning from an authentic, professional
text written in the L2 is influenced by task instructions and that content learning and
language learning may not occur simultaneously in a single task. Task design and im-
plementation, including task instructions, needs to be carefully conceived in terms of
goals, input and procedures by taking into account the limitation of cognitive capacity
as well as level of L2 proficiency.
For content learning, the implication is that teachers could explore ways in which
EFL students can be encouraged to conceptualize what is described in a text. For stu-
dents with limited L2 proficiency, it looks as though the use of their L1 in recalling or
retelling of the content of a text can be beneficial. At the same time, vocabulary learn-
ing through reading a text is not automatic (Hulstijn et al., 1996). Given the limitation
of cognitive capacity, students might benefit from recycling or a cyclical use of reading
tasks (c.f., Bygate & Samuda, 2005; Lynch & Maclean, 2000). Through repetition of the
read-and-recall activity, students may be better able to identify more ideas and rela-
tions in a text, constructing more coherent textbase and situation model representa-
tions (Horiba, 1996). Once they have understood what the text is all about, students
may be better able to identify specific words used in the text or analyze how ideas or
concepts are linguistically presented. Task instructions can be used to guide students’
cognitive processes and strategies to achieve a particular learning goal.
Another implication concerns the design of material. It seems that the use of L1
glossing can facilitate EFL students’ text processing and learning (c.f., Hulstijn et al.,
1996). In this study, target words in a text – infrequent healthcare related words in-
cluding technical terms – were underlined with L1 translations provided underneath.
This kind of presentation can free L2 students from cognitive overload of word recog-
nition, and make it easier to process the words together with other parts of the sen-
tences containing them and to build representations of the content of the text. Al-
though the number of new words learned was low, as shown in the vocabulary
acquisition test scores in this study, there was some evidence of incidental vocabulary
learning. If the target words had not been presented without L1 glossing, these stu-
dents would have been not only unsuccessful in guessing the meaning of the words
from context, but also their comprehension and learning from the text would have
suffered greatly.

Implications for future research

This study raises a number of questions that need to be addressed in future research.
First, because students’ cognitive processes and strategies are often influenced by their
prior experiences with processing and communicating in L2, it is important to extend
the findings obtained here to other EFL or ESL settings, with different levels of lan-
guage proficiency, and with greater exposure to L2 input and interaction. This is
 Yukie Horiba and Keiko Fukaya

important because this enables us to know whether and in what way the specific set-
ting (EFL or ESL), the proficiency level of the learners, and the amount of exposure to
L2 impact the way learners process content and language in L2, and what task condi-
tions can influence students’ processing capacity and subsequent learning.
Second, this study examined and compared the effects of three task conditions in
which recall task instructions were manipulated. A fourth condition, the L2-L1 format
in which individuals are informed of an L2 recall task and later produce their recall in
L1, was not used. Inclusion of this fourth condition may help elucidate the effect of
task instructions on resource allocation involved in text (or input) processing and re-
trieval/recall (or output). This awaits further investigation.
Finally, the relation between a text (as input) and what is learned through working
with the text can be scrutinized by more focused research. For instance, qualitative
analysis of the relation between text and recall (or output), the vocabulary learned, and
the examination of online processing during the read-and-recall procedure, might
shed light on the connection between input, output, and learning.

Acknowledgement

The study reported in this chapter was supported by a grant-in-aid for scientific re-
search (B2) No.15320071 from Japan Society for the Promotion of Science.

References

Alexander, P. A., Jetton, T. L., & Kulikowich, J. M. (1996). Interrelationships of knowledge, inter-
est, and recall: Assessing a model of domain learning. Journal of Educational Psychology, 87,
559–575.
Barcroft, J. (2004). Theoretical and methodological issues in research on semantic and struc-
tural elaboration in lexical acquisition. In B. VanPatten, J. Williams, & S. Rott (Eds.), Form-
meaning connections in second language acquisition (pp. 219–234). Mahwah, NJ: Lawrence
Erlbaum Associates.
Barry, S., & Lazarte, A. A. (1995). Embedded clause effects on recall: Does high prior knowledge
of content domain overcome syntactic complexity in students of Spanish? Modern Lan-
guage Journal, 79, 491–504.
Barry, S., & Lazarte, A. A. (1998). Evidence for mental models: How do prior knowledge, syn-
tactic complexity, and reading topic affect inference generation in a recall task for nonna-
tive readers of Spanish? Modern Language Journal, 82, 176–193.
Bernhardt, E. B. (1991). Reading development in a second language. Norwood, NJ: Ablex.
Bygate, M. (2001). Effects of task repetition on the structure and control of oral language. In
Bygate, M., P. Skehan, & M. Swain (Eds.). Researching pedagogical tasks: Second language
learning, teaching and testing (pp. 23–48). Harlow, UK: Pearson Education.
 Chapter 5.  Text processing and learning 

Bygate, M., & Samuda, V. (2005). Integrative planning through the use of task-repetition. In R.
Ellis (Ed.), Planning and task performance in a second language (pp. 37–74). Amsterdam:
John Benjamins.
Chen, Q., & Donin, J. (1997). Discourse processing of first and second language biology texts:
Effects of language proficiency and domain-specific knowledge. Modern Language Journal,
81, 209–227.
Donin, J., & Silva, M. (1993). The relationship between first- and second-language reading com-
prehension of occupation-specific texts. Language Learning, 43, 372–401.
Ellis, N. (1997). Vocabulary acquisition: Word structure, collocation, word-class, and meaning.
In N. Schmitt & M. McCarthy (Eds.). Vocabulary: Description, acquisition and pedagogy
(pp. 122–139). Cambridge: Cambridge University Press.
Ellis, R. (2003). Task-based language teaching and learning. Oxford: Oxford University Press.
Grabe, W. (2009). Reading in a second language: Moving from theory to practice. Cambridge:
Cambridge University Press.
Greasser, A. C., Gernsbacher, A., & Goldman, S. R. (2003). Handbook of discourse processes.
Mahwah, NJ: Lawrence Erlbaum Associates.
Horiba, Y. (1996). Comprehension processes in L2 reading: Language competence, textual co-
herence, and inferences. Studies in Second Language Acquisition, 18, 433–473.
Horiba, Y. (2000). Reader control in reading: Effects of language competence, text type, and task.
Discourse Processes, 29, 223–267.
Hulstijn, J., Hollander, M., & Greidanus, T. (1996). Incidental vocabulary learning by advanced
foreign language students: The influence of marginal glosses, dictionary use, and reoccur-
rence of unknown words. Modern Language Journal, 80, 327–339.
Hulstijn, J., & Laufer, B. (2001). Some empirical evidence for the involvement load hypothesis in
vocabulary acquisition. Language Learning, 51, 539–558.
Izumi, S. (2003). Comprehension and production processes in second language learning: In
search of the psycholinguistic rationale of the output hypothesis. Applied Linguistics, 24,
168–196.
Kim, Y. (2006). Effects of input elaboration on vocabulary acquisition through reading by Ko-
rean learners of English as a foreign language. TESOL Quarterly, 40, 341–373.
Kintsch, W. (1998). Comprehension: A paradigm for cognition. Cambridge: Cambridge Univer-
sity Press.
Koda, K. (2005). Insights into second language reading: A cross-linguistic approach. Cambridge:
Cambridge University Press.
Laufer, B., & Hulstijn, J. (2001). Incidental vocabulary acquisition in a second language: The
construct of task-induced involvement. Applied Linguistics, 22, 1–26.
Lee, J. F. (1986). On the use of recall tasks to measure L2 reading comprehension. Studies in
Second Language Acquisition, 8, 201–212.
Lee, J. W., & Schallert, D. L. (1997). The relative contribution of L2 language proficiency and L1
reading ability to L2 reading performance: A test of the threshold hypothesis in an EFL
context. TESOL Quarterly, 31, 713–739.
Levelt, W. J. M. (1989). Speaking: From intention to articulation. Cambridge, MA: The MIT
Press.
Long, M., & Robinson, P. (1998). Focus on form: Theory, research, and practice. In C. Doughty
& J. Williams (Eds.). Focus on form in classroom second language acquisition (pp. 15–41).
Cambridge: Cambridge University Press.
 Yukie Horiba and Keiko Fukaya

Lynch, T., & Maclean, J. (2000). Exploring the benefits of task repetition and recycling for class-
room language learning. Language Teaching Research, 4, 221–250.
Magliano, J. P., Trabasso, T., & Graesser, A. C. (1999). Strategic processing during comprehen-
sion. Journal of Educational Psychology, 91, 615–629.
McNamara, D. S. (2007). Reading comprehension strategies: Theories, interventions, and technolo-
gies. New York, NY: Lawrence Erlbaum Associates.
Nation, I. S. P. (2001). Learning vocabulary in another language. Cambridge: Cambridge Univer-
sity Press.
Nunan, D. (2004). Task-based language teaching. Cambridge: Cambridge University Press.
Pulido, D. (2009). How involved are American L2 learners of Spanish in lexical input processing
tasks during reading? Studies in Second Language Acquisition, 31, 31–58.
Skehan, P. (1998). A cognitive approach to language learning. Oxford: Oxford University Press.
Skehan, P. (2009). Modeling second language performance: Integrating complexity, accuracy,
fluency, and lexis. Applied Linguistics, 30, 474–532.
Stevenson, M., Schoonen, R., & de Glopper, K. (2003). Inhibition or compensation? A multi-
dimensional comparison of reading processes in Dutch and English. Language Learning,
53, 765–815.
Swain, M., & Lapkin, S. (2001). Focus on form through collaborative dialogue: Exploring task
effects. In M. Bygate, P. Skehan, & M. Swain (Eds.). Researching pedagogic tasks: Second
language learning, teaching, and testing (pp. 99–118). New York, NY: Longman.
Tavakoli, P., & Foster, P. (2008). Task design and second language performance: The effect of
narrative type on learner output. Language Learning, 58, 439–473.
Trabasso, T., Secco, T., & van den Broek, P. W. (1984). Causal cohesion and story coherence. In
H. Mandl, N. L. Stein, & T. Trabasso (Eds.). Learning and comprehension of text (pp. 83–111).
Hillsdale, NJ: Lawrence Erlbaum Associates.
Trabasso, T., van den Broek, P., & Suh, S. (1989). Logical necessity and transitivity of causal rela-
tions in the representation of stories. Discourse Processes, 12, 1–25.
Van den Broek, P., Lorch, R. F., Jr., Linderholm, T., & Gustafson, M. (2001). The effects of readers’
goals on inference generation and memory for texts. Memory & Cognition, 29, 1081–1087.
Van Oostendorp, H., & Goldman, S. R. (Eds.) (1999). The construction of mental representations
during reading. Mahwah, NJ: Lawrence Erlbaum.
VanPatten, B. (1996). Input processing and grammar instruction in second language acquisition.
Norwood, NJ: Ablex.
Zwaan, R. A. (1994). Effects of genre expectations on text comprehension. Journal of Experimen-
tal Psychology: Learning, Memory, and Cognition, 20, 920–933.
Zwaan, R. A., & Brown, C. M. (1996). The influence of language proficiency and comprehension
skill on situation-model construction. Discourse Processes, 21, 289–327.
Zwaan, R. A., & Radvansky, G. A. (1998). Situation models in language comprehension and
memory. Psychological Bulletin, 123, 162–185.
 Chapter 5.  Text processing and learning 

Appendix

The Decision text is shown below. In the test passages, each target word (*____) or
non-target word (____) was underlined with its L1 translation provided underneath.
Who Decides the Treatment?
Michael Cantos, a 15-year-old who has recurrent *metastatic Ewing *sarcoma, has
been hospitalized with fever and *neutropenia, common complications of his recent
chemotherapy. Michael lives with his parents, two younger siblings, and his paternal
grandmother. His parents and grandmother were born in the Philippines and emi-
grated to the United States about 30 years ago; all three of the Cantos children were
born in this country. When Michael was first diagnosed, he was told that this type of
cancer was aggressive and had already spread from the primary site in his *pelvis to his
*bronchi and parenchyma. Treatment consisted of surgical *resection, a year of che-
motherapy, and six weeks of *radiation. During the past year, whenever Michael asked
if the cancer was fatal and what was the mortality rate, the *palliative team members
have responded both with veracity and reassurance, declaring, “Some patients die, but
we’re all fighting very hard to cure you.” Michael hasn’t forgotten a word.
On this admission, a routine chest X-ray reveals a large *lesion in his right lung.
Additional X-rays reveal multiple smaller lung lesions and a large pelvic *neoplasm.
Just two months earlier his routine *surveillance scans were normal. Now, on learning
the results of the chest X-ray, Michael asks if his mother can stay overnight in the hos-
pital with him. Michael also tells his parents that he wants to hear his scan results and
treatment options at the same time they do. His parents are shocked. They’d prefer
withholding such “dismal news” from Michael, but they agree to honor his wishes.
This decision deeply upsets Michael’s grandmother, however, and on their next visit
Michael says, “Why does Grandmother always have to pray the rosary over me –
doesn’t she know it doesn’t work?”
In the team conference, a new registered nurse expresses frustration with the
grandmother’s “constant interference”; she says that by praying the rosary over him,
the grandmother may be upsetting Michael further. The nurse says she can’t support
the grandmother and the boy at the same time, and asks the team for help. The team
decides to meet every two weeks, or more often if necessary.
The team also schedules a meeting to discuss Michael’s *prognosis and communi-
cation within the family. As usual, the patient and his parents are encouraged to bring
anyone they want. Michael immediately says that he wants only his parents present,
and his parents agree. At the meeting the *pediatric *oncologist initially presents in-
formation, with the palliative care nurse practitioner summarizing or restating impor-
tant points to ensure that everyone understands. Decisions are made by consensus.
 chapter 6

Task structure and patterns of interaction

What can we learn from observing
native speakers performing tasks?

James Hobbs
Iwate Medical University, Japan

Most EFL learners in Japan have few opportunities to observe native speakers
(NSs) performing in task-like scenarios, and consequently struggle to access
the lexical phrases and formulaic chunks that facilitate fluent communication
in goal-oriented conversation tasks. This often results in the use of the L1
during tasks. This chapter reports the findings of a study of native-speaker task
interaction over a range of task types, focusing on the ways in which patterns
of interaction are influenced by task design, and showing how recordings of NS
task interaction can be used to better equip learners to perform similar tasks
without recourse to L1. Findings of the study imply that valuable insights can be
obtained by analyzing NS task performance, enabling L2 researchers to obtain
a deeper understanding of how task design and task selection can influence
interaction. As a consequence, this will enable researchers to make principled
recommendations to materials writers, syllabus designers, and classroom
teachers for a reasoned TBLT implementation and practice in the classroom
regarding issues such as task selection and design, sequencing of tasks, control of
task complexity, and identification of task demands.

Introduction

For almost three decades an unbroken thread linking varied and sometimes conflicting
perspectives on the role of tasks has fed the belief that meaning-focused interaction is
a driving force of language acquisition, not simply an end product facilitated by lan-
guage learning (e.g., Ellis, 2003; Long, 1985; Nunan, 1989, 2004; Prabhu, 1987; Samuda
& Bygate, 2008; Willis, 1996). This argument is supported by the evidence that sponta-
neous, fluent, nativelike discourse consists mostly of chains of multi-word chunks of
meaningful language acquired mainly through natural exposure and stored in subcon-
scious memory (Becker, 1975; Lewis, 1993; Nattinger & DeCarrico, 1992; Pawley &
 James Hobbs

Syder, 1983). If we want learners to produce more language in this way, and become
less dependent on paying conscious attention to language form, it seems logical to en-
courage them to experiment freely with language while performing tasks. Thus, while
a focus on specific forms may be embedded in a meaningful task, it is vital that learn-
ers’ conscious attention be focused primarily on meaning while performing tasks.
Many teachers new to task-based language teaching (TBLT) are unconvinced by
this argument, and wonder why it is not better simply to place tasks at the end of a
presentation-practice-production (PPP) cycle. If we want a task to elicit a specific
structure, they argue, surely it is better to make this explicit and encourage learners to
focus conscious attention on using this structure while performing the task. However,
to believe this is to miss a fundamental point in the case for TBLT: learners can indeed
be frustratingly adept at sidestepping the language forms to which tasks are designed
to direct them, but if teachers respond by imposing lexical shackles on them (Say it this
way, please!), then they may be diluting the focus on meaning that drives language
development, and denying learners valuable opportunities to experiment with alterna-
tive language forms. Instead, teachers committed to TBLT should respond by seeking
a deeper understanding of how task design influences interaction, and by developing
ways to identify the appropriate task language for form-focused instruction. This lan-
guage should emerge from tasks, not be imposed on them.
An approach accessible to individual teachers in local contexts is to observe how
native speakers (or advanced nonnative speakers) perform tasks that resemble those to
be performed by learners. Hobbs (2005) used this approach to discover that much of
the language that supports interaction in opinion-exchange interview tasks consists of
interactive lexical phrases: Learners perform better if they can access appropriate fixed
and semi-fixed L2 phrases to discuss task procedure (I’ll go first; Shall we start by ...
ing...?), agree and disagree (Yeah, right; I’m not sure), give feedback (OK; Really?), and
so on. This chapter focuses on a larger study of native speaker (NS) performance on a
range of structurally diverse tasks, and addresses three questions that should interest
TBLT researchers and practitioners of TBLT:
– How do patterns of NS interaction vary between different kinds of tasks?
– To what extent can NS use of interactive lexical phrases be predicted on the basis
of task design?
– How can insights obtained by observing NS interaction be presented to learners to
help them improve their own performance on tasks?

Research into task interaction

The issue of how task structure influences NS interaction appears to be a largely unex-
plored area, and analyses of NS task performance are often small-scale investigations
by lone teachers in EFL contexts (e.g., Baigent, 2005; Cox, 2005; Hobbs, 2005). This is
 Chapter 6.  Task structure and patterns of interaction 

surprising, for if NS norms are to form the basis of what is taught in class, there appear
to be significant risks involved in deciding the target language and behaviours for giv-
en task scenarios without actually looking at how NSs perform in such situations.
Indeed, Cox (2005) found that in the case of open tasks (without a fixed outcome/solu-
tion), relatively inexperienced teachers trained in PPP methodology made largely
inaccurate predictions about the language used by NSs in a set of tasks.
Of course, this is not to deny the equal or perhaps greater importance of under-
standing the many variables that influence learner output. The heavy focus on learner
output in the research literature is understandable, and has generated valuable insights
into how learner output is influenced by factors such as task repetition (e.g., Bygate,
2001), planning time (e.g., Foster & Skehan, 1996; Gilabert, 2007), corrective recasts
(e.g., Nicholas, Lightbown, & Spada, 2001), and corrective precasts that anticipate
learner difficulties before they occur (Ellis, Basturkmen, & Loewen, 2001). Research
has also shown that we can influence learner performance in predictable ways by alter-
ing task design. For example, Loschky and Bley-Vroman (1993) show that some tasks
naturally elicit particular structures, and describe how changes in task design can
make particular language more or less likely to appear. Mackey (1999) describes how
a series of tasks successfully elicited targeted question forms. Tavakoli and Foster
(2008) and Skehan and Foster (1999) show that narrative tasks elicit more fluent and
accurate language when incorporating a tighter structure with clear time sequences.
Interestingly, related research also shows that this variation in fluency is not seen when
native speakers perform such tasks (Foster & Tavakoli, 2009). However, such com-
parisons of learner output with the performance of native speakers in similar situa-
tions are still comparatively rare, and much remains to be done to help researchers and
teachers understand not only how task structure and lesson procedures are likely to
influence learner output, but also how NS task performance can inform the identifica-
tion of target language associated with a particular task, and how NS recordings can
best be used in class.
Practical guides for implementing TBLT encourage teachers to make room in the
lesson cycle for examples of native or advanced speakers performing tasks that mirror
those that learners will subsequently perform (Nunan, 1989, 2004; Willis, 1996; Willis
& Willis, 2007). However, few reports are available of what has actually been discov-
ered by observing NS task performance, and fewer still of how valuable such data can
be for teachers in EFL settings. Carter (1998) notes that the language that learners
encounter in the classroom often consists largely of contrived dialogues lacking the
very features that identify discourse as nativelike. This is a particular concern if learn-
ers have few opportunities outside the classroom to compare their own output with
that of NSs. However, given evidence that learners notice salient features of transcribed
talk even when their attention is not formally directed to them (Takahashi, 2005), it
would seem important for teachers to expose learners to examples of NS task perfor-
mance, not only through pre-task listening, but also in materials for post-task form-
focused study.
 James Hobbs

Methodology

Context

Task interaction between NSs can generate predictable discourse patterns featuring
recurring lexical chunks. Previous research (e.g., Hobbs, 2005) showed that question-
based opinion-exchange tasks typically included (1) an opening move (Are you ready?),
(2) question markers (Now I’m going to ask you about...), (3) feedback (That’s interest-
ing), (4) a move to return the question (How about you?), (5) pause fillers and vague
language (Let me think..., ...or something), and (6) an ending move (Let’s stop there).
However, when such tasks were used with low-intermediate Japanese university stu-
dents, many of these moves were either performed in the L1 or simply omitted. A
question about summer holiday plans, for example, was seen by most not as a spring-
board for discussion, but as a prompt to simply take turns to compose and deliver a
grammatically accurate sentence (e.g., I want to go to the beach). Kumaravadivelu
(2003) notes how such perceptual mismatches can cause learners to fail to match
teachers’ expectations, but perhaps another reason why learners frequently revert to
the L1, or avoid particular interactive moves, is that they simply lack the lexical re-
sources needed to perform these moves smoothly in the L2.
Hobbs (2005) goes on to describe how this problem was addressed, for these par-
ticular opinion-exchange tasks, by using NS recordings in class within the task-based
learning (TBL) framework proposed by Willis (1996). In the pre-task, learners an-
swered meaning-focused questions based on an NS recording. After the task, they
listened again and studied the transcript, focusing on the use of interactive lexical
phrases. After only a few lessons, significant improvements were noted; not only did
the learners show more fluent use of lexical phrases, but also the near-elimination of
code-switching, and topic development beyond the previously ubiquitous one-sen-
tence answers.
While Hobbs (2005) investigated only question-based pair interview tasks, the
encouraging results suggested that NS performance on other types of task might also
reveal distinct patterns of interaction and lexical choice that could form a basis for
classroom instruction. Moreover, these results invite speculation that common fea-
tures of otherwise distinct tasks might exert a predictable influence on interaction. The
current study, larger in scope and scale, sought to investigate these issues using tran-
scripts of NS performance on nine different tasks, seeking connections between task
structure, interaction patterns, and lexical phrases associated with specific interactive
moves, as well as considering ways to present findings to learners.

Participants

Three pairs of NSs of English participated in the present study. Two speakers were
resident in Japan. Two others were friends of the researcher in the UK and the USA,
 Chapter 6.  Task structure and patterns of interaction 

Table 1.  English native speaker participants

Nationality Age Language-teaching Experience

(Gender) experience living in Japan

Pair 1 Speaker A American (F) 30s Yes Yes

(USA) Speaker B American (M) 30s No No
Pair 2 Speaker A British (M) 30s No No
(UK) Speaker B Irish (F) 30s No No
Pair 3 Speaker A American (F) 30s Yes Yes
(Japan) Speaker B American (M) 30s Yes Yes

respectively, and each solicited the help of a friend. Finding six volunteers among col-
leagues in Japan would have been easier, but I wanted to include speakers who were
neither language teachers nor were familiar with characteristics of Japanese learners.
Participant details are summarized in Table 1.
With only six speakers it was clear that it would be difficult to draw firm conclu-
sions about native speaker language, and the sample was certainly too small to
investigate variations between different standard varieties of English. It may have been
unrealistic to expect the study to do more than simply open up avenues for future re-
search on the issues of how patterns of NS interaction vary between tasks, and the
extent to which the use of interactive lexical phrases can be predicted based on task
design. However, with regard to the third research question – how insights gained
from observing NS task performance can help learners perform better on tasks – the
potential was considerably greater. With regard to this third question in particular, the
research described here is offered as an example of what is achievable by lone research-
ers working with limited resources in parts of the world where access to NS volunteers
may be limited, and where teaching duties may place significant restrictions on the
amount of time that can be devoted to research.

Tasks

The three pairs of NSs each performed nine tasks fitting the description of “an activity
that requires learners to use language, with emphasis on meaning, to attain an objec-
tive” (Bygate, Skehan, & Swain, 2001, p. 11). Care was taken not only to include a range
of structurally distinct tasks but also to pose a genuine cognitive challenge for NSs, and
so to avoid simple tasks that might tempt speakers to consciously act out the roles of
imaginary students. As one goal was to compare interaction between different task
types, I chose nine tasks based roughly on Prabhu’s (1987) three task categories of
reasoning-gap, information-gap, and opinion-gap.
 James Hobbs

1. Reasoning-gap activities (tasks 1–3): These were “ordering and sorting tasks”
(Willis, 1996, p. 26), which involve organizing given information. Task 1 required
speakers to arrange historical events in chronological order:

TASK 1: The events below occurred during 1980–89, one for each year. Arrange them
in chronological order.
A nuclear disaster occurs in Chernobyl, USSR
John Lennon is assassinated by Mark David Chapman
Michael Jackson releases the album “Thriller”
Margaret Thatcher’s Conservatives win a third consecutive UK general election
The Berlin Wall falls
The first compact discs are released
Prince Charles marries Diana Spencer
Ben Johnson is stripped of Olympic gold after testing positive for steroids
Live Aid takes place simultaneously in London and Philadelphia
The AIDS virus is discovered
(Task created by researcher)

Task 2 (from Jones & Kimbrough, 1987, p. 54) gave the statement I don’t understand.
Do you speak English? translated into 12 different foreign languages, together with a list
of the 12 languages, and asked speakers to match them. Task 3 involved ordering pic-
tures to make a story (Appendix 1).
2. Information-gap activities (tasks 4–6) required speakers to exchange given infor-
mation. In task 4 speakers took turns defining words:

TASK 4: The words below are all connected with health and fitness. Explain each word
on your paper to your partner without saying the word. For the spaces in your list
listen to your partner’s explanation and guess the word.
Student A
START: doctor →______→diet→______fever→______→AIDS→______
FINISH←←hangover←←exercise←______←smoking
(Student B words: nurse-gym-hospital-vitamins-dentist-aerobics-cigarettes-drugs)
(Adapted from Bunday & Randell, 1996, p. 93)

In task 5, one speaker saw a 100-word summary of the life of Abraham Lincoln, the
other a similar text about John F. Kennedy, and they were asked to identify coinci-
dences and similarities through discussion only, without looking at each other’s text
(from Jones & Kimbrough, 1987, p. 7). Task 6 was a picture-based spot the differences
 Chapter 6.  Task structure and patterns of interaction 

task (Appendix 2). For all three tasks in this category the term information gap signals
a need for the exchange of information between speakers, but it should be noted that
Prabhu (1987) uses the term slightly differently to identify tasks in which information
is changed from one form to another.
3. Opinion-gap activities, which were all decision-making tasks. These required the
exchange of opinions, and had no correct solution. In task 7 speakers discussed
the desirability of hypothetical laws:

TASK 7: Imagine you have the power to enact the following laws, but only if you both
agree. Which laws will/won’t you enact?
1.  Doubling the tax on cigarettes in order to discourage people from smoking.
2.  Doubling the tax on alcohol in order to discourage people from drinking.
3.  Requiring all school students to wear a school uniform.
4.  Raising the driving age to 21 in order to reduce traffic congestion and road accidents.
5.  Legalising the use of marijuana.
(Task created by researcher)

Tasks 8 and 9 were also created by the researcher. Task 8 asked speakers to discuss the
relative merits of printed reference books as compared with reference materials on
CD-ROM, and agree on three advantages of each; in task 9 each speaker was asked to
suggest three interesting places to visit in his/her country of birth, and the listener in
each case then decided which of the suggested places he/she would most like to visit.
The similarities between the tasks in each group provided a basis for analysis, but
the variety was felt to be sufficient that any common patterns of interaction observed
would be significant. For example, task 4 differs in many ways from tasks 5 and 6, so it
would be interesting if patterns of interaction reflected the common need to exchange
given information. However, it was necessary to record several pairs performing each
task in order to confirm that findings did not reflect only the conversational style of a
particular speaker. Overall, the project had a 3x3x3 structure: three pairs of speakers,
three task categories, three tasks per category.

Data collection procedures

The researcher did not attend any recording sessions. Each pair received printed in-
structions and task papers. They knew that the recordings would be analysed carefully,
but knew nothing of the research goals. Each pair recorded their own unrehearsed
performance of all nine tasks, in one or more sessions. Although no order was speci-
fied, all pairs performed the tasks in order 1–9, each pair taking roughly 60 minutes in
total. The recording was then sent to the researcher.
 James Hobbs

Coding and analysis of data

The recordings were transcribed by the researcher, yielding a research corpus of some
16,000 words, including all the interaction except for a few utterances that could not
be clearly discerned.
Analysis was also performed by the researcher, and began with simply browsing
transcripts for recurring features in three areas: interactive structure (e.g., initiation-
response-feedback), specific interactive moves (e.g., seeking agreement; commenting
on task progress), and lexical choices (e.g., I think...). In some cases, potential areas of
interest were decided in advance by intuition and investigated by analyzing transcripts
(e.g., in tasks without a fixed order/procedure, how do speakers go about deciding
this? In tasks requiring agreement, how is agreement/disagreement expressed? etc.).
At other times the researcher simply noticed a word or phrase encountered elsewhere
in the transcripts, or recognized a particular interactive move or pattern seen earlier.
Where such features were identified, comparisons were made with other pairs and
other tasks to determine the degree to which the feature correlated with the task cate-
gory, as opposed to being associated only with a particular task and/or speaker.
Frequency counts (e.g., of specific lexical items) were performed by computer where
possible, but analysis of interactive structure was necessarily carried out manually us-
ing printouts of transcripts. It did not seem prudent to fix an arbitrary value for a
quantitative definition of recurrence, as the items being considered were often quite
disparate in nature, with frequency not necessarily the best indicator of significance.
For example, a particular interactive move appearing only once in a transcript might
nevertheless be significant if also appearing once in several other transcripts for tasks
of a similar type, while conversely a specific lexical item encountered 20 or more times
might be less interesting if used by only one speaker, and hence clearly not something
that could easily be predicted. In short, the aim was not simply to validate or disprove
specific predictions, but rather to keep a sharp eye open for any way in which task
design appeared to have influenced the interaction, to identify features that offer a
basis for predicting performance by other NS on similar tasks, and to consider ways
that learners could be exposed to recordings and transcripts of such interaction in
order to perform better on their own classroom tasks.

Results

Findings of the study are presented by task category, focusing on the most striking
features common to tasks within each category, as well as on notable differences that
appear related to specific features of individual tasks.
 Chapter 6.  Task structure and patterns of interaction 

Reasoning-gap activities (Tasks 1–3)

At first glance, transcripts of unplanned NS interaction often appear as a confusing

jumble of half sentences, restarts, interruptions and overlaps, and vague language (you
know; whatever). For reasons discussed later, this is especially true in the case of deci-
sion-making activities (tasks 7–9). For example:
Extract 1 (pair 1, task 8)
A: you just find what your/what your word is whereas a book you can find. you know.
other things.
B: it makes you more cur/it makes me more curious.
A: yeah. how can we explain that?
B: well picture curiosity.
A: ok.
B: ...if you don’t have a computer.
A: it’s.. yeah.
B: I mean books will do fine.
A: books are/books are/books don’t require. uh. you know. power. or expensive upgrades
or whatever. you know. all of that.
However, a focus on the overall structure of the interaction often reveals useful in-
sights surprisingly close at hand. This is very evident in the case of the reasoning-gap
activities. For example, on beginning task 1 two pairs made strikingly similar prag-
matic and linguistic choices:
Extract 2 (pair 2, task 1)
A: task one..............[comment to tape?/initiating interaction] ok.. shall we put down the
ones we know. and then try and take a guess at the others? [suggesting procedure]
B: yes. [agreeing]
A: um. [pause filling/stalling] well John Lennon was assassinated in nineteen eighty.
[suggesting answer] I remember that [reasoning] .......uh [pause filling]. the Berlin
Wall fell in nineteen eighty-nine....[suggesting answer] Charles and Di married in.
eighty-one? [suggesting answer/seeking agreement]
B: yeah. eighty-one. [agreeing]
Extract 3 (pair 3, task 1)
B: ok. here we go. [initiating interaction]
A: here we go. [confirming]
B: how do we get started? [asking about task procedure]
A: so let’s start with the ones we know for sure. [suggesting procedure]
B: ok. [agreeing]
A: then we can/the other ones we’ll...guess. [suggesting procedure]
B: ok. [agreeing]
 James Hobbs

A: so we know that John Lennon was/that was nineteen...eighty. [suggesting answer]

B: ok. nineteen eighty. [agreeing]
A: right.[closing discussion of that item]
A: right. [confirming] and then the Berlin wall that’ll be eighty-nine. [suggesting
answer]
Besides recurring lexical items such as the ones we know and guess, we find here we go
meaning let’s start; rising intonation used to signal uncertainty and seek agreement
(eighty-one?); yeah and ok signalling agreement; ok and right used as discourse mark-
ers (ok. here we go; right. and then...); and so on. However, while even the shortest and
apparently simplest of extracts can offer useful insights, we must resist the urge to im-
mediately reveal to our students the way to approach a particular task, or the correct
phrases to use. As noted earlier, the essence of TBLT lies in not imposing such de-
mands, and moreover a look at how the third NS pair begins this task reveals a quite
different approach:
Extract 4 (pair 1, task 1)
A: in chronological order...oh God!...Live Aid was nineteen eighty-one...I think. ‘cause I
remember fully.
B: this is eighty-four eighty-five. so Live Aid had to be after that. I think the AIDS virus
is first.
While even a mere two lines here show pertinent examples of I think used to hedge
commitment to an answer, and (be)cause and so signalling reasoning, clearly this ex-
tract shows a path of interaction quite different from that followed by pairs 2 and 3.
Nevertheless, the full transcripts for this task do reveal a range of interactive moves
related to task structure, all of which could be highlighted in post-task language-focus
activities:
1. Task organizers, addressing progress and task procedure: So there’s one for each
year; What have we got left? etc.
2. Suggesting an answer: (I think) that’s eighty-eight; That was nineteen eighty; That’ll
be eighty-four, etc.
3. Seeking agreement: It had to have been eighty-one, right?... That was eighty-six...do
you reckon? etc.
4. Agreeing: ok (35 instances), yes/yeah (21), right (5).
Significantly, though, the high frequency of one lexical chunk is often due to one
speaker or pair. Sometimes this may simply reflect speaker characteristics; for exam-
ple, there is no obvious reason why 23 of 35 occurrences of ok should have come from
pair 3. In other cases, though, task characteristics appear to be a crucial factor. It is
probably no coincidence that the use of I think to hedge commitment to answers was
most evident with pair 1 (19 occurrences), who found the task more difficult than did
other pairs, and achieved the lowest number of correct answers. This is precisely the
 Chapter 6.  Task structure and patterns of interaction 

sort of connection between tasks and interaction that a busy teacher selecting a task
might fail to anticipate – the harder the task, the more learners will want to hedge their
commitment to answers.
Some similar patterns appear in task 2, reflecting the shared focus on ordering
given information. First, task organizers:
so what are we up to? ([Pair]1[Speaker]B]
so we’ve got Danish Greek Finnish and Swahili (2B)
we’re left with Swahili Finnish and Greek (2B)
where’s the other one then? (2B)
we’ll just do the ones we know first (3A)
so we’ve got Finnish Greek and Polish. and Swahili (3A)
that leaves us with Swahili Finnish and Polish (3A)
which leaves...Finnish (3B)
etc.
The chunk the ones we know is familiar from task 1, occurring here again in various
combinations that can be summarized as follows:
(shall we/we’ll just/let’s) (do/put down/start with) the ones we know
Similarly, we’ve got (left) often appears in comments on progress:
we’ve got (all the ones in the middle/eighty-one/Finnish/Danish/an empty space)
what have we got left?
We can also note variations of the verb leave:
that leaves us with Swahili Finnish and Polish
which leaves...Finnish
we’re left with Swahili Finnish and Greek
what have we got left?
Drawing attention to all these possible variations might not be an urgent priority with
lower level learners, but more advanced learners could easily expand their own lexical
repertoire by analysing such features in transcripts.
Turning to instances of suggesting an answer, task 2 shows considerably less varia-
tion in the form of the verb be as compared with task 1, and the significant factor
seems to be that task 1 focuses on the past. In task 2, 31 of 53 suggested answers are
variations of the pattern
(I think) (this/this one/that/German, etc.) is (maybe/probably)....
Resembling task 1 are the many examples of seeking agreement, although expressions
range from what do you reckon? to ’that seem right? The transcripts also mirror task 1
in the frequency of yeah/yes, right, and ok used to signal agreement, with agreement
often signalled even if not actively sought:
 James Hobbs

Extract 5 (pair 1, task 2)

B: this is probably Greek.
A: yeah. I think you’re right. it is Greek....Finnish is this one.
B: ok.
Task 3, with no correct solution, also features task organizers, again seemingly con-
nected with the ordering and sorting requirements of the task:
so we can arrange the pictures first. (1A)
you start off with number one and then you can do odd numbers. (2B)
shall we do numbering? and then we’ll just ad lib our story as we go along. (3A)
However, other striking features relate to the storytelling element of this task. During
the discussion phase all pairs use short, often elliptical, narrative phrases in the present
tense, for example:
well I think. they take off. the world explodes. falls to a new planet. meets the guy.
they do a/....... (3A)
When speakers present their story at the end, they use complete sentences, but fre-
quent pauses suggest that the cognitive challenge of linking a series of events, and the
associated simultaneous focus on meaning and language form, poses a significant
challenge even for a native speaker:
and having landed on Saturn...it came across a...two-eyed. four-handed. two-
legged...thing called Bob. who wasn’t very happy to see the rocket from Stoke-on-
Trent. because it had landed in his flower bed. (2A)
Other storytelling language that might deserve post-task attention includes openings
(it all started with..., once upon a time), temporal adjuncts (then, as soon as), and com-
ment adjuncts (fortunately, unfortunately). Clearly, then, interaction reflects the order-
ing and sorting element in the tasks, but in task 3 the elements related to storytelling
are more striking.

Information-gap tasks (Tasks 4–6)

Tasks 4, 5, and 6 require speakers to exchange given information, but show variety in
other respects. Tasks 5 and 6 focus on similarities and differences, unlike task 4; task 6
uses visual stimuli, the other two tasks written information; task 4 has a fixed proce-
dure, while the other two are flexible. Task 4, unsurprisingly, features three-part ex-
changes – clue, answer, feedback – resembling the initiation-response-feedback (IRF)
pattern common in teacher-learner interaction:
Extract 6 (pair 3, task 4)
A: ...............something that you smoke. [clue]
 Chapter 6.  Task structure and patterns of interaction 

B: cigarette? [answer]
A: yes. [feedback]
B: ....................what you have the day after you’ve been drinking too much. [clue]
A: hangover? [answer]
B: yes. [feedback]
While recurring lexical phrases typically reflect the choice of one particular speaker,
some patterns recur in task 4 in ways that could be highlighted in post-task activities,
as in the following examples:
someone who cures your ailments. (2A)
something you get....which makes you come down in a sweat. (2A)
something that’s not very good for your health. (2B)
Task 5 reveals a similar IRF interaction pattern:
Extract 7 (pair 3, task 5)
B: what day of the week was Lincoln killed?
A: Friday.
B: it was a Friday. that’s the same.
However, this could be reduced to a series of two-part statement + feedback exchanges:
Extract 8 (pair 2, task 5)
B: they were both presidents.
A: yes.
B: of the United States.
A: that’s very true.
The IRF pattern is again dominant in task 6:
Extract 9 (pair 3, task 6)
A: and next to that is a flower shop?
B: no. it goes café. travel agency. newsstand.
A: ok. I have a flower shop.
The need to describe a scene naturally results in the pattern there is/are being the most
noticeable lexical chunk, appearing some 50 times in the transcripts. Even here,
though, there are clear alternatives:
do you have a guy playing a violin? (1B)
does your market have the fruit stand outside? (1A)
I’ve got a lamp post (1B)
The influence of the need to exchange information is thus clearly evident, but predict-
ing specific lexical chunks is again difficult, and striking features in the transcripts also
 James Hobbs

reflect more specific aspects of each particular task: the need to define words (task 4),
compare written facts (task 5), or describe a scene (task 6).

Opinion-gap tasks (Tasks 7–9)

The opinion-gap tasks also share features while remaining structurally distinct. Task 7
(proposed legislation) and task 8 (CD-ROMs vs. books) both require agreement, but
the latter is less likely to involve disagreement. In task 9 (places to visit) agreement is
unnecessary. Unsurprisingly, instances of expressing an opinion and agreeing are com-
mon, while vague language and pause fillers to maintain fluency are far more common
than in earlier tasks. In opinion-gap tasks we expect to find variations of I think. How-
ever, interestingly these appear more in task 7 than in tasks 8 and 9, with 52 instances
of I (don’t) think in the task 7 transcripts, but only 8 in task 9 (also compare think used
to mark opinions with [especially in task 1] think used to hedge commitment to an-
swers). We also find a rich array of expressions to state opinions in task 7, including I
do believe that, I have a feeling that, and I definitely have a negative reaction to, plus
discourse organizers that weaken commitment to opinions: some people would main-
tain that; I guess one of the arguments in favour of...is. The potential for face-threatening
disagreements thus appears to have a significant affect on the interaction in task 7. Us-
ing entire transcripts to reveal such features to learners can be difficult, but often short
extracts can be used to illustrate relevant features.
In all three tasks, agreement is signalled by simple utterances such as yes, ok, or
right, although that’s true also appears in all three transcripts for task 9 alone, along
with phrases such as I agree and that’s a good one. Task 7 shows speakers occasionally
choosing more complex patterns such as I don’t have a problem with that, I think we
can both go along with that, and that’s exactly what was going through my mind.
The use of fluency devices and vague language is striking in the opinion-gap tasks
(see Extract 1), but task 3 (picture story) is the only other task in which they are
notable. Silence is frequently avoided by false starts, pauses are filled with uh and you
know, and vague terms such as whatever and all of that abound. These features ap-
pear in all three opinion-gap tasks, and appear to result from the cognitive challenge
involved in formulating opinions and simultaneously finding ways to express them.
The same double focus of attention is required in task 3 (story task). This contrasts
markedly with information-gap tasks, in which speakers need only find ways to ex-
press given meanings, leading to a corresponding reduction in the use of pause fill-
ers and vague language. Again, though, trends are clearly not rules, as there are few
such examples in the transcripts for pair 2, who seem to pay more conscious atten-
tion to their choice of words (which may be connected with nationality, may reflect
personal characteristics, or may simply be a subconscious effort to speak correctly
while being recorded).
 Chapter 6.  Task structure and patterns of interaction 

Discussion and implications

Overall, a number of interactional features related to the themes of reasoning, exchang-

ing information, and making decisions can be identified by this study. These include:
– task organizers in reasoning tasks that require ordering/sorting;
– Initiation-Response-Feedback exchanges in information-gap tasks;
– many lexical chunks associated with expressing opinions found in decision-mak-
ing tasks;
– pause fillers, false starts, and vague language in decision-making tasks.
However, the study also revealed features unique to each particular task:
– A focus on past events (task 1) elicited far more grammatical variety than did a
structurally similar task with no time element (task 2).
– In ordering and sorting tasks, increased difficulty (i.e., greater uncertainty) cor-
relates with the use of I think to hedge commitment to answers (task 1).
– Task 3 (picture story) elicits many lexical patterns associated with storytelling.
– In opinion-gap tasks, hedging opinions with variations of I think is common only
when there is clear potential for disagreement (task 7).
The relationship between task design and patterns of interaction is thus more complex
than we might imagine, and even a very close look at the structural details of a par-
ticular task may often not put us in a position to make reliable predictions about the
lexical choices that task participants will make. Indeed, the data seem to confirm the
importance of considering the many different ways speakers can choose to convey
similar meanings. Furthermore, as van Lier and Matsuo (2000) note, with nonnative
speakers (NNS), the symmetry of learner interaction – turn lengths and the adoption
of leader/follower roles – will be influenced by differences in language ability. The cur-
rent data also show that unless the order of speaking turns is specified (e.g., task 4),
then even with NSs it is likely that cooperative tasks will elicit a leader/follower pat-
tern. In task 1, pairs 1 and 2 each have a leader who suggests answers, and a follower
who simply agrees or disagrees until the leader has considered all the items. Speakers’
linguistic needs may thus be quite different, even in the same task.
Findings of this study show that valuable insights can be obtained by analyzing NS
task performance, leading us to question why NSs usually appear in research data only
as task interlocutors, such that “it is rather unusual for their own performance to be
subject to scrutiny” (Foster & Tavakoli, 2009, p. 867). By the same token, the wide-
spread acceptance of Willis’s (1996, p. 88) claim that learners benefit from listening to
NS (or advanced nonnative speakers) performing tasks similar to the ones they per-
form implies that extracts of NS task performance can make effective teaching materi-
als, and that this is particularly important in EFL contexts. Indeed, data obtained from
NS task performance, as was the case in the present study, might enable L2 researchers
to obtain a deeper understanding of how task design and task selection can influence
 James Hobbs

interaction, which will enable them to make principled recommendations to materials

writers, syllabus designers, and classroom teachers for more successful TBLT imple-
mentation and practice in the classroom including the selection, design, and sequenc-
ing of tasks, and the structuring of task-based lessons. Above all, learners whose
ultimate goal is to perform tasks as a native speaker would perform them need oppor-
tunities to listen to and analyse examples of native speakers speaking freely in task-like
scenarios.

Classroom applications

The question now arises regarding how to present findings from NS recordings to
learners in ways that help them improve their own performance on tasks. In this re-
spect, it is relatively easy to isolate relevant features of NS interaction in areas where
learners have difficulty. For example, my own Japanese EFL students often neglect to
discuss progress and task procedure in English: The Japanese ikimasu, roughly trans-
lated as here I go, is common in tasks requiring distinct turn-taking (e.g., task 4), and
ordering and sorting tasks often begin with discussion of procedures in the L1 or (if the
teacher is listening) with a long silence until someone suggests an answer. Classroom
Activity 1, using extracts 2 and 3 discussed above, illustrates one possible way to ad-
dress this issue.

Classroom Activity 1
Listen to the dialogues. In each dialogue:
–  Who suggests how to start? a.  the woman b.  the man
–  How do they start? a.  They discuss each year in turn: 80, 81, 82, etc.
b. They discuss the events in the order listed on the
paper.
c.  They discuss the ones they are more certain of.
Listen again:
–  Which phrase do the speakers use to suggest how to start?
a.  Let’s... b.  Why don’t we...? c.  Shall we...?
Follow-up: Look at the transcript and find the exact phrase each speaker used.

Learners need not have performed the exact same task, as an activity like this could
benefit learners who have recently performed any task requiring discussion of task
procedure. Nor do learners always need printed handouts. In my own teaching con-
text, for instance, factors including class size, student level, and available time affect
whether I provide a handout and/or transcripts, write questions on the whiteboard, or
simply pose questions orally before playing the extract. Especially in larger classes,
 Chapter 6.  Task structure and patterns of interaction 

learners may be surprised, even shocked, to learn just how much they use the L1, and
there is much to be gained from allowing learners to compare NS recordings with re-
cordings of their own performance. The benefits can often be significant and almost
immediate, as learners reconsider the parameters of task interaction, and realize that a
simple ok or a basic follow-up question can be enough to keep the conversation flow-
ing naturally without using the L1. The following excerpt is from an opinion-exchange
task involving two learners who only a few lessons before this recording was made had
signalled each and every new question with ikimasu and had delivered most feedback
comments and follow-up questions in Japanese.
S1: Next question.
S2: OK.
S1: What are your favorite summer activities?
S2: ...First exercise.
S1: Ah ok.
S2: Second...barbeque...three.. third...eeto* ...swimming.
S1: Ohhh I like swimming too.
S2: Can/can you swim?
S1: Yes I can.
S2: Great.
S1: Thank you...next...Is travelling abroad important to you?
*Japanese pause filler
My students have benefited greatly from activities focusing on feedback (I see; Right;
That’s true, etc.), as their previous learning experiences in teacher-centred classrooms,
and their lack of opportunities to use English outside class, make this an aspect of task
performance that does not come naturally to them. As noted earlier, the nature of
feedback may vary according to the task. For learners struggling to offer much more
than ok, yes, and no as feedback in tasks requiring them to compare information, an
activity such as Classroom Activity 2 (based on part of one NS pair’s performance on
task 5) could be useful.

Classroom Activity 2
Listen to the recording, noting how speakers respond when they find similarities.
–  Which three phrases from the list below were used?
That’s true
That’s very true
That’s the same
That’s exactly the same
That’s a good one
–  If necessary, listen again and check.
–  Follow-up: Think of more ways to give feedback during this task using “That’s.....”
 James Hobbs

Even when learners are accustomed to a particular task type, activities using NS tran-
scripts can encourage them to experiment with alternative expressions. The following
extract, used with Classroom Activity 3, could be used to highlight feel/feeling as alter-
natives to think, while learners might also be able to decipher no brainer from context:
Extract 11 (pair 3, task 7)
B: number two doubling the tax on alcohol in order to discourage people from drink-
ing......well...it would seem like a no brainer. just to say yes. if you feel like it’s a simi-
lar issue.
A: is it though? I feel that...no study backs up/my feeling is that cigarettes cause more
health problems than drinking.
B: uh-hmm.....................for the person who smokes.
A: for the person who smokes but then there’s those/those costs are/are built into like
the/into the health care system. and non-smokers. that’s why I don’t feel that drink-
ing...necessarily causes the same number of health problems.

Classroom Activity 3
Listen to the extract:
–  Would the woman like to double the tax on alcohol? Why or why not?
After you listen:
–  Which word featured four times in statements of opinion?
a.  ...think... b.  ...opinion... c.  ...feel/feeling... d.  ...believe/belief...
–  Now check what the speakers said in the transcript.
–  What do you think ‘(like) a no brainer’ means?
a.  difficult b.  obvious c.  stupid d.  unusual
Follow-up:
–  Do you think the man will agree with the woman? Listen to the next part of the
discussion and find out.

Again, such an activity need not require a handout, especially if used with small groups
of motivated learners.

Limitations and suggestions for further research

Clearly, then, there are several ways in which recordings of NS task performance can
benefit research on TBLT and be used to assist teachers implementing TBLT in an EFL
setting: (a) as a way to test intuitions about the language a task will elicit, (b) as a way
to uncover other appropriate task language and identify areas of potential difficulty,
and (c) as a source of input for both meaning-focused and form-focused study within
 Chapter 6.  Task structure and patterns of interaction 

a TBLT framework. This is especially important in EFL settings, in which the class-
room is often the only place where learners can have opportunities to compare their
own performance with NS models of the kind they aspire to emulate. However, we
cannot make many broad generalizations about native speaker discourse based on the
small sample of research participants in the current study, in which factors including
age, sex, nationality, professional experience, and personality, as well as a partner’s
characteristics, may all have affected speakers’ linguistic choices in ways that were not
obvious. Moreover, we have seen that the most striking features of interaction may
often relate to factors other than the focus on reasoning, exchanging information, or
exchanging opinions. To name some:
– Is there only one possible outcome (task 4), or a range of possible outcomes (all
other tasks)?
– Is the presentation of a solution/result required (task 3)?
– Are discrete items handled in a fixed order (tasks 4, 7) or is the order negotiable
(tasks 1, 2)?
– Do speakers first have to identify the relevant information (tasks 5, 6)?
– Is there a correct solution/outcome (tasks 1, 2, 4)?
– How tight are instructions? Is there more than one way to structure the task
(tasks 3, 5, 6, 8, 9)?
Such distinctions could form the basis of further study, and there is nothing to prevent
EFL researchers and teachers with limited resources using similar research procedures
in a bid to confirm or refute some of the tentative findings presented here. However,
there is no guarantee that research would reveal patterns of interaction as clear and
predictable as those found in Hobbs (2005), and it remains possible that the interaction
elicited by question-based pair interview tasks just happens to be more predictable,
both structurally and lexically, than is the case with many other task types. Moreover,
the data presented here show that even where particular interactive moves recur, speak-
ers often use a range of phrases to perform the same function, and frequently-occurring
items are often short, fixed expressions with limited pedagogic potential. There is noth-
ing wrong with directing attention to short, simple expressions that help keep learners
in English (e.g., Let’s start; Here we go), but the main focus of instruction should be on
showing learners the range of options available and to allow them to slowly expand
their own repertoire of lexical phrases. Teachers who instruct learners to perform tasks
by reproducing the exact forms they have just listened to in an NS recording will have
completely missed the point of incorporating NS recordings into lessons.

Conclusions

TBLT researchers and language professionals seeking connections between task struc-
ture and target language can learn much from analyzing NS task performance.
 James Hobbs

Whatever aspects of interaction researchers and teachers investigate, it is likely that,

besides developing a deeper understanding of factors that influence interactive struc-
ture and lexical choices, they will also find ways to use recordings to benefit learners.
Consciousness-raising activities need not necessarily have a tight linguistic focus. If
students focus on accuracy to the detriment of fluency (i.e., speaking in complete and
accurate sentences, but with many pauses) transcripts can show that disjointed, un-
grammatical discourse is commonplace in NS conversation, especially when exchang-
ing opinions (see Extract 1). You can even challenge students to clean up a dialogue by
removing false starts and interjections, and rewriting it in complete sentences – not to
imply that whole sentences are always preferable, but to draw attention to this often-
overlooked aspect of NS discourse.
Teachers may still complain that NS recordings are too fast for learners, and that
interesting features are usually buried deep in discourse which, as a whole, is too com-
plex for learners to process. Such problems can sometimes be mitigated by pairing
older speakers with younger ones when collecting data, by recording speakers who do
not know each other well, or by recording advanced nonnative speakers. Another op-
tion is to re-record an edited version of an authentic transcript to make it more acces-
sible to learners without detracting too much from its authenticity. In these and other
ways, teachers sensitive to connections between task design and task interaction can
overcome obstacles, and can help learners by presenting examples of task interaction,
not only as models to copy, but also more importantly as opportunities to explore the
variety of communicative moves that lead to successful interaction, and the range of
lexical options available in a given situation. The three classroom activities described
above are examples, and this is exactly how writers such as Willis (1996) suggest that
recordings and transcripts be presented: after the task, in a way that allows learners to
compare their own performance with that of native speakers or more advanced learn-
ers. Teachers who can learn what to look for in transcripts of task interaction will be
surprised and encouraged by the range of possibilities that emerge for focused class-
room activities.

References

Baigent, M. (2005). Multi-word chunks in oral tasks. In C. Edwards & J. Willis (Eds.), Teachers
exploring tasks in English language teaching (pp. 157–170). Basingstoke, UK: Palgrave
MacMillan.
Becker, J. D. (1975). The phrasal lexicon. In B. L. Nash-Webber & R. Schank (Eds.), Proceedings
of the 1975 workshop on theoretical issues in natural language processing. Cambridge, MA:
Association for Computational Linguistics
Bunday, D., & Randell, N. (1996). Kick off. Tokyo: MacMillan Languagehouse.
Bygate, M. (2001). Effects of task repetition on the structure and control of oral language. In M.
Bygate, P. Skehan, & M. Swain (Eds.), Researching pedagogic tasks – Second language learn-
ing, teaching and testing (pp. 23–48). Harlow, UK: Pearson Education.
 Chapter 6.  Task structure and patterns of interaction 

Bygate, M., Skehan, P., & Swain, M. (2001). Researching pedagogic tasks – Second language learn-
ing, teaching and testing. Harlow, UK: Pearson Education.
Carter, R. (1998). Orders of reality: CANCODE, communication and culture. ELT Journal,
52(1), 43–56.
Cox, D. (2005). Can we predict language items for open tasks? In C. Edwards & J. Willis (Eds.),
Teachers exploring tasks in English language teaching (pp. 171–186). Basingstoke, UK:
Palgrave MacMillan.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Ellis, R., Basturkmen, H., & Loewen, S. (2001). Preemptive focus on form in the ESL classroom.
TESOL Quarterly 35(3), 407–32.
Foster, P., & Skehan, S. (1996). The influence of planning on performance in task-based learning.
Studies in Second Language Acquisition, 18(3), 299–324.
Foster, P., & Tavakoli, P. (2009). Native speakers and task performance: Comparing effects on
complexity, fluency, and lexical diversity. Language Learning, 59(4), 866–896.
Gilabert, R. (2007). The simultaneous manipulation of task complexity along planning time and
[± here-and-now]: Effects on L2 oral production. In M. García Mayo (Ed.), Investigating
tasks in formal language learning (pp. 44–68). Clevedon, UK: Multilingual Matters.
Hadfield, J. (1990). Intermediate communication games. Walton-on-Thames, UK: Nelson.
Hobbs, J. (2005). Interactive lexical phrases in pair interview tasks. In C. Edwards & J. Willis
(Eds.), Teachers exploring tasks in English language teaching (pp. 143–156). Basingstoke,
UK: Palgrave MacMillan.
Jones, L., & Kimbrough, V. (1987). Great ideas. Cambridge: Cambridge University Press.
Kay, S. (1997). Move up elementary resource pack. Oxford: Heinemann ELT.
Kumaravadivelu, B. (2003). Beyond methods: Macrostrategies for language teaching. New Haven,
CT: Yale University Press.
Lewis, M. (1993). The lexical approach: The state of ELT and a way forward. Hove, UK: Language
Teaching Publications.
Long, M. (1985). A role for instruction in second language acquisition: Task-based language
teaching. In K. Hylstenstam & M. Pienemann (Eds.), Modelling and assessing second lan-
guage acquisition (pp. 77–99). Clevedon, UK: Multilingual Matters.
Loschky, L., & Bley-Vroman, R. (1993). Grammar and task-based methodology. In G. Crookes &
S. Gass (Eds.),Tasks in language learning (pp. 123–163). Clevedon, UK: Multilingual Matters.
Mackey, A. (1999). Input, interaction, and second language development: An empirical study on
question formation in ESL. Studies in Second Language Acquisition, 21(4), 557–587.
Nattinger, J., & DeCarrico, J. (1992). Lexical phrases and language teaching. Oxford: Oxford Uni-
versity Press.
Nicholas, H., Lightbown, P., & Spada, N. (2001). Recasts as feedback to language learners. Lan-
guage Learning, 51(4), 719–58.
Nunan, D. (1989). Designing tasks for the communicative classroom. Cambridge: Cambridge Uni-
versity Press.
Nunan, D. (2004). Task-based language teaching. Cambridge, UK: Cambridge University Press.
Pawley, A., & Syder, F. (1983). Two puzzles for linguistic theory: Nativelike selection and native-
like fluency. In J. Richards & R. Schmidt (Eds.), Language and communication (pp. 191–226).
London: Longman.
Prabhu, N. S. (1987). Second language pedagogy. Oxford: Oxford University Press.
Samuda, V., & Bygate, M. (2008). Tasks in second language learning. Basingstoke, UK: Palgrave
MacMillan.
 James Hobbs

Skehan, P., & Foster, P. (1999). The influence of task structure and processing conditions on nar-
rative retellings. Language Learning, 49(1), 93–120.
Takahashi, S. (2005). Pragmalinguistic awareness: Is it related to motivation and proficiency?
Applied Linguistics, 26(1), 90–120.
Tavakoli, P., & Foster, P. (2008). Task design and second language performance: The effect of
narrative type on learner output. Language Learning, 58(2), 439–473.
van Lier, L., & Matsuo, N. (2000). Varieties of conversational experience: Looking for learning
opportunities. Applied Language Learning, 11(2), 265–287.
Willis, D., & Willis, J. (2007). Doing task-based teaching. Oxford: Oxford University Press.
Willis, J. (1996). A framework for task-based learning. Harlow: Longman.
 Chapter 6.  Task structure and patterns of interaction 

Appendix 1: Task 3

(Pictures from Hadfield, J. [1990])

 James Hobbs

Appendix 2: Task 6
 Chapter 6.  Task structure and patterns of interaction 

(Pictures from Kay, S. [1997])

section ii

Implementation of task-based language

teaching
 chapter 7

Patterns of corrective feedback in a task-based

adult EFL classroom setting in China

Noriko Iwashita and Huifang (Lydia) Li

The University of Queensland, Australia

This chapter reports on a study that investigated patterns of corrective

feedback observed in teacher – student and student – student interaction in
a task-based EFL class at a medium-sized university in China. Eight hours
of classroom interaction data were analyzed for various types of feedback
and uptake. Despite the large class size and the students’ unfamiliarity with a
teaching methodology that is very different from the traditional Chinese way
of learning and teaching the study found frequent interaction in the classroom
characterized by teacher feedback to the students’ non-target-like utterances and
students’ response to the feedback. These findings were interpreted in terms of
characteristics of task-based interaction observed in the study, the principles or
practices of TBLT in the context of the current study, and the factors affecting
the classroom interaction. The main implication of this study is that active
student participation was enhanced by the students’ willingness to accept new
methodologies and modes of learning that are vastly different from their past
learning experiences, from their beliefs about learning, and from the traditional
methodologies they were accustomed to.

Introduction

In response to the initiative taken by the central government of China to revise the
existing foreign language curriculum to a more communicatively oriented one, task-
based language teaching (TBLT) has been introduced in many classrooms across the
curriculum. Although teachers are aware of the need for the innovation, implementa-
tion has not always been successful. For this reason, recent English Language Teaching
research in China has mainly been concerned with the factors that hinder the imple-
mentation of this new curriculum. However, little research to date has been under-
taken in authentic settings to examine and identify the characteristics of actual
classroom interaction that may ensure the successful implementation of TBLT and
may also enhance L2 development. Within TBLT research generally, a substantial
 Noriko Iwashita and Huifang (Lydia) Li

volume of research has been devoted to corrective feedback episodes in teacher-stu-

dent and student-student interaction, and some studies have found a facilitative role of
certain types of feedback in SLA. Despite the large volume of research of this kind, to
date little work has been undertaken in Chinese classrooms where the impact of fac-
tors such as context, learner proficiency, and linguistic target on characteristics of in-
teraction are in need of consideration (e.g., Ellis & Sheen, 2006; Oliver & Mackey,
2003; Sheen, 2004). Furthermore, although many studies on corrective feedback in
classroom interaction were undertaken in TBLT-oriented classrooms, there is need for
sustained attention to the characteristics of interaction associated with task-based in-
struction. The current study sought to examine how task-based instruction may be
facilitative of learning by identifying feedback episodes in teacher – student and stu-
dent – student interaction in an EFL classroom situation at a university in China.

Communicative, task-based language teaching in China

Over the past two decades, English language education in China has witnessed waves
of top-down reforms promoting communicative, and more recently, task-based lan-
guage teaching (Adamson & Morris, 1997; Hu, 2002b; Yu, 2001). Communicative
tasks of various kinds have been presented as shaping opportunities for enhancing
classroom communication. This can be seen in several widely used textbooks, such as
Junior/Senior English for China published by the People’s Press and New College English
published in 2008 by the Press of Shanghai Foreign Language Studies (Hu, 2002b; P. Li,
2009). These textbooks include a variety of meaning-focused tasks in the format of
listening, speaking, reading, and/or writing activities (Hu, 2002b; P. Li, 2009).
Promoting communicative language teaching (CLT) through implementing TBLT
has long been discussed in the literature in second language teaching and learning
(e.g., Littlewood, 2007; Long, 1985; Nunan, 2004). The key concept of CLT, which has
been widely accepted, is that the goal of teaching is to develop learners’ competence in
real-life communication. However, interpretations of CLT have varied widely in terms
of creating optimal conditions for developing learners’ communicative competence
(Hiep, 2007). This is partly because CLT has included a broad range of methods and
curricula that “embrace both the goals and the processes of classroom learning, for
teaching practice that views competence in terms of social interaction and looks to
further language acquisition research to account for its development” (Savignon, 1991,
p. 263). TBLT is often regarded as a more thoroughly articulated, recent development
within the overall approach of CLT (Littlewood, 2007; Nunan, 2004). Communicative
tasks have been claimed to provide opportunities for learners to use the L2 to interpret,
express, and negotiate meaning, and therefore constitute a major component of teach-
ing methodology. Further, in TBLT, tasks can constitute syllabus units around which a
course may be designed (Nunan, 2004).
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

In spite of the efforts made to promote a communicatively oriented curriculum,

the prevalent teaching and learning style in China remains teacher-centred and gram-
mar-focused (Yu, 2001). This is clearly shown in the core intensive reading course
taught in most secondary schools and colleges across the country (Cortazzi & Jin,
1996a). In this course, the teacher proceeds sentence by sentence and paragraph by
paragraph in order to explain every likely grammar point or word meaning which may
arise. Teachers who are familiar with Western notions of CLT may include some com-
municative tasks which provide opportunities for negotiation of meaning, but this
usually takes a very small portion of class time and is somewhat teacher-directed or
teacher-centered (Jin & Cortazzi, 1998). For instance, two students stand in the front
of the classroom and one asks the other questions while the rest of the class listens and
is seldom engaged in spontaneous communication.
Despite the government’s strong advocacy of CLT, teachers have encountered a
number of difficulties in the successful implementation of a communicative curricu-
lum. For that reason, ELT research in China has focused primarily on investigating the
factors hindering the successful implementation of CLT and/or TBLT in the classroom
(Anderson, 1993; Hu, 2002a, 2005a; Liao, 2000; Littlewood, 2007; Yu, 2001). The fac-
tors often cited in research and commentaries are large classes, public demand for as-
sessment, the Chinese culture of teaching and learning, and teacher agency. According
to Cortazzi and Jin (2001), for instance, class size in primary and high schools in China
ranges from 20 to 80 students, with the average being 50 to 60. In a large class, class-
room management is difficult too. For example, Ng and Tang (1997) reported on
teachers’ complaints about management in a large class as follows: “We have 50 stu-
dents in a class, and if each student speaks one sentence, it will take up the whole les-
son” (p. 77). Similarly, C. Li (2003) identified a teacher’s frustration that students did
not value group work without the teacher monitoring them and considered such work
a waste of time.
The second factor is that China has a long history of public assessment, which can
be traced back to the Han Dynasty (206 BC–220 AD). In the modern society of China,
several English tests, such as the National Matriculation English Test, the College Eng-
lish Test and the Graduate School Entrance English Examination, play a key part in the
pursuit of personal well-being and future employment opportunities. According to
Cheng (2008), the high status of English tests does not support teaching but drives it.
Furthermore, Littlewood (2007) found that the public assessment system in East Asian
countries usually fails to keep pace with curricular innovation. As has been pointed
out by English teachers in China, the uniform test, which emphasizes grammar, vo-
cabulary, and reading comprehension, can be an obstacle to the implementation of
TBLT (Ng & Tang, 1997).
The third factor hindering the successful implementation of a communicatively
oriented curriculum is the Chinese culture of teaching and learning. Teaching in China
is often interpreted as a process of transmitting knowledge, information or skills from
the teacher to the student. Correspondingly, knowledge is for the student to receive
 Noriko Iwashita and Huifang (Lydia) Li

and retain. Therefore, a good teacher is expected to possess a deep knowledge of the
subject area and to be able to explain difficult points clearly to students, while a good
student is expected to listen carefully to the teacher, take notes to review later, and
memorize the knowledge transmitted by the teacher (Cortazzi & Jin, 1996b, 2001; Hu,
2002a; Rao, 1996). For instance, Hu (2002a) argues that the most serious problem
hindering the implementation of TBLT may lie in the Chinese culture of learning that
is based on deep-rooted perceptions of teaching and learning that clash with the tenets
and practices of a communicatively oriented curriculum. That is, while Chinese cul-
ture favours the epistemic mode of being mentally active in receiving knowledge, the
interactive mode of being verbally active is the crucial part of a communication-based
curriculum for English language teaching. Hu argues that, due to these fundamental
differences, any pedagogical innovation that disturbs or threatens Chinese learners’
deep-rooted belief systems about learning will not help them learn effectively.
The last factor is teacher agency. In the early stages of rebuilding English language
education at school, which began after the end of the Cultural Revolution in 1977, the
proficiency of English teachers was generally low. Due to the dire lack of English teach-
ers, Russian teachers were recruited to teach English language after receiving in-ser-
vice training that varied in length from several weeks to several months (Cowan, Light,
Mathews, & Tucker, 1979). Over the years, the government has continued to put more
effort into providing in-service training programs and revising the curriculum of
teacher training courses at universities and teacher training institutions in order to
improve teacher proficiency (Hu, 2005a). Although substantial progress has been
made, teacher qualification remains a major concern in the promotion of pedagogical
innovations such as TBLT. According to Hu (2005b), this is mainly attributed to the
pre-service teacher education curriculum. Hu points out critically that theoretical lan-
guage acquisition courses intended to improve teachers’ English proficiency take up
the largest portion of curricular time. The method employed in the majority of these
language acquisition courses is teacher-fronted and knowledge-based, which provides
few opportunities for the student teachers to develop their communicative compe-
tency. The teacher training program itself offers very few opportunities for the student
teachers to explore pedagogical and psychological issues in a domain-specific manner.
For these multiple reasons, several commentators (e.g., Xiao, 1998; Yu, 2001) have
expressed their concern that these student teachers will continue to apply in their fu-
ture teaching the method with which they themselves were trained.

Task-based interaction in the classroom and SLA

A substantial amount of research has investigated task-based interaction in both class-

room (observational) and laboratory (experimental) settings focusing on negotiation
of meaning and more recently on corrective feedback. This research agenda is largely
initiated by classroom researchers, who stress the importance of using tasks to enhance
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

communication in the foreign language curriculum (e.g., Van den Branden et al., 2009;
Pica & Doughty, 1985; Pica et al., 1989). In perhaps the first serious claim along these
lines, Long and Porter (1985) argued that task-based interaction observed in group
work enhances language opportunities and improves the quality of students’ talk.
The theoretical foundation of most studies on interaction rests in Long’s
(1981, 1983) Interaction Hypothesis. Long proposed that conversational interactions,
which occur in a variety of forms as interlocutors respond to their conversational part-
ner’s request for clarification or confirmation, promote L2 learning even though the
immediate purpose of such modifications in conversation is to make speech compre-
hensible. Based on the Interaction Hypothesis, a number of studies have examined
which tasks are more likely to generate opportunities for the negotiation of meaning.
Pica, Kanagy, and Falodun (1993) classified tasks according to types of goals, and di-
rections of communication. They claimed that closed two-way tasks provide the most
opportunities for negotiation. In closed-two way tasks, information flows in two direc-
tions as each participant is expected to share the information he/she has and there is
only one outcome to complete the task, unlike in open tasks where participants are
expected to interact with each other to obtain the necessary information or achieve the
outcome. A further finding was that one-way tasks lead to more individual input and
much less negotiation work than two-way tasks (Brown & Yule, 1983).
More recent studies on task-based interaction have focused on feedback moves
arising within the interaction, observed between a teacher or NS and students, or be-
tween students. The shift of the research focus was motivated by the increasing criti-
cisms of Krashen’s Input Hypothesis that comprehensible input alone is not sufficient
for SLA (e.g., Long, 1991; Swain, 1985). The attention to form needs to be incorporated
into meaning-focused activity which is referred as focus on form. Focus on form oc-
curs incidentally in meaning-focused interaction in form of various types of feedback
(Long, 1991). This was further articulated in Long’s updated version of the Interaction
Hyothesis. Long (1996) stressed the facilitative role of implicit negative feedback in
conversational interaction as such feedback draws learners’ attention to mismatches
between input and output.
Observational studies have examined instances of feedback (e.g., negotiation
moves, explicit corrections) given by teachers in classroom interaction and many of
the studies have also investigated learner response to the feedback. The findings of
these studies show a high frequency of recasts given by teachers (e.g., Ellis, Basturkman
& Loewen, 2001, 2002; Lyster, 1998a, 1998b; Lyster & Mori, 2006; Lyster & Ranta,
1997; Panova & Lyster, 2002; Sheen, 2006; Wei, 2002).1 However, in general, learner
responses to recasts were found to be not as frequent as their responses to other con-
versational moves, such as clarification requests.

1. Definitions and examples of different types of feedback are provided in the methodology
section.
 Noriko Iwashita and Huifang (Lydia) Li

The teaching methodology employed in classrooms where these observational

studies were conducted was communicative. A variety of communicative tasks, includ-
ing information gap, jigsaw, discussion, were used based on the assumption that task-
based conversation provides interaction. Although the recent studies above revealed
characteristics of task-based interaction and the utility of various types of feedback for
L2 development, compared with studies conducted in the 1980s and early 1990s the
recent studies, while investigating opportunities for negotiated interaction, make little
mention of tasks and task types. It appears that the recent studies have been under-
taken with the underlying assumption that communicative tasks generate opportuni-
ties for interaction in which various types of feedback are frequently provided.
Apart from the observational studies cited above, a number of factors which may
affect the occurrence of feedback during task-based interaction have also been stud-
ied: classroom vs. dyads (Oliver, 2000); EFL in Japan and immersion in Canada (Lyster
& Mori, 2006); French immersion program in Canada vs. a language school in New
Zealand (NZ) vs. a large EFL classroom in Korea (Sheen, 2004); and learner factors
such as teaching context, interactional focus (Oliver & Mackey, 2003) and linguistic
target (e.g., Lyster, 1998a, 1998b). Sheen (2004) analysed the interaction in an EFL
classroom in Korea and compared the results reported in previous studies (i.e., Loewen,
2005; Lyster & Ranta, 1997; Panova & Lyster, 2002) in four settings, that is, French im-
mersion in the earlier study, ESL in Canada, ESL in New Zealand, and EFL in Korea.
The results showed that recasts were the most frequent feedback type in all of the four
contexts, but significant variations among the four settings were observed. Interpret-
ing the results, Sheen (2004) suggested that the influence of context (e.g., meaning-
focused or form-focused, class size, course content) on corrective feedback and uptake
should be considered. Additionally, though, it should be noted that Sheen does not
mention how contextual variables in terms of use of tasks might have impacted on the
nature of the interaction.

The study

In summary, TBLT has been introduced as a way of implementing CLT principles in

many EFL classes in China, despite strong resistance from teachers. Also, because of
the difficulties teachers face in implementing a more communicatively oriented
approach, most existing research has been about identifying the issues that hinder suc-
cessful curriculum innovation rather than evaluating the newly introduced curricu-
lum. To date, little research is available documenting the classroom implementation or
effectiveness of CLT/TBLT. Within TBLT research, one approach to investigating in-
teraction is to examine various types of corrective feedback episodes in teacher-student
and student-student interaction. Despite a substantial volume of related research in the
field, few studies have been conducted in China. Nevertheless, recent research (e.g., Ellis
& Sheen, 2006; Sheen, 2004) has shown the importance of contextual variables in the
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

investigation of the nature of classroom interaction. Hence the present study was un-
dertaken to fill the gap in the research of teacher – student and student – student
interaction in a Chinese EFL TBLT class. Based on these considerations, the following
research question was addressed in this study: What are the patterns of corrective feed-
back in teacher-student and student – student interactions in a TBLT university class
setting in China?
Underpinning this research question is the fact that, despite a growing awareness of
the need to enhance learners’ communicative competence among English language
teachers in China, because of the difficulties teachers have encountered in implement-
ing CLT/TBLT, interaction between teacher and students may not be frequent, even in a
TBLT class. For that reason, the current study collected data from a TBLT classroom and
examined the instances of corrective feedback and response to the feedback in order to
establish whether and in what ways these teacher – student and student – student inter-
action patterns were consistent with the basic tenets of TBLT. As such, the study was
expected to provide empirical evidence of the challenges raised by the implementation
of TBLT in China, and provide concrete suggestions on how to meet these challenges.

Methodology

Participants

Fifty students participated in this study. They were first-year students majoring in the
English language. The participants had commenced university studies just before the
data were collected. Their age ranged from 17 to 19 years old. The ratio of females to
males was 4:1. The students had studied English for six years at middle school, where
they had received a one-hour lesson every day and a one-hour reading lesson (during
which students read aloud in order to practise pronunciation by memorizing words
and reciting a text) two or three times a week. As Hu (2003, 2005b) explains, EFL
teaching and learning in undeveloped regions2 of China is more traditional in terms of
curriculum and pedagogical practices. This was obviously the case for a majority of the
students in the current study. Because they were from remote areas of the province
where the university was located, they had benefited little from recent developments
and innovations in English language teaching led by the government’s initiative in the
middle school English curriculum. According to a substantial number of the students
in the study, English lessons in the middle school were taught in Chinese and were
mostly teacher-centred, with a heavy focus on grammar, vocabulary, and reading.

2. The developed regions are the coastal provinces and the capital cities of the inland prov-
inces. Examples of such cities are Beijing, Shanghai, and Guangzhou. The rest of the provinces
and cities are regarded as undeveloped regions. The university where the current study was
conducted is located in an undeveloped region.
 Noriko Iwashita and Huifang (Lydia) Li

The teacher of the class involved in this study (henceforth referred to as Debbie)
was born in Romania. Debbie had obtained an undergraduate degree in French and
English language in her home country and a Master’s degree in TESOL in the USA.
According to colleagues, her English proficiency was considered to be near-native. She
had been teaching for one year in the university at the time of data collection. Her
classes were chosen as a research site based on the class size (i.e., a typical Chinese
English language class of approximately 50 students) and her view of teaching and
training. Informal conversations with her revealed her belief that language learning
was for real-life communication and her strong support for a task-based approach to
teaching. This belief, Debbie said, had made her use a variety of communicative tasks
in her oral lessons that she developed from various resources, as there was no need for
her to follow a prescribed syllabus. In addition, she had no objections to having her
classes video-taped for the purposes of this study.

Instructional setting

The class observed for the study was taking an oral English course offered as part of a
four-year teacher training program in the English department of this medium-sized
university north-west of Beijing. Although the university was established to provide pre-
service teacher training for middle schools across the country, the curriculum was fo-
cused strongly on English language courses. Therefore, the program included only a few
courses that introduced basic knowledge of education, psychology, and second language
acquisition. Hu (2005c) points out that the curriculum for the teacher training program
in China has a heavy focus on language skills and knowledge. The courses in the first two
years are designed to enhance basic language skills, including intensive and extensive
reading, listening, grammar, and oral English. The subjects in the last two years of the
program focus on imparting specialized knowledge of English as a language and as a
carrier of cultural information. Topics treated in such courses can be related to lexicon,
textual stylistics, literature, and cultural knowledge about the USA and the UK.
Even with the Chinese government’s advocacy of a communication and TBLT ori-
ented curriculum, it is difficult to evaluate how closely the courses offered by the uni-
versity’s English department adhere to the basic principles of TBLT, apart from the fact
that the department uses the most current CLT-based textbook. For example, from
1999 to 2003, the department used New College English (Zhejiang University Press) in
its intensive reading course. In the textbook, the reading lessons had pre-reading tasks,
during-reading tasks, and post-reading tasks. However, as few teachers understood
the rationale for this lesson design, most reverted to the traditional method of inten-
sive reading prevalent in most classes. Although most English courses offered at the
research site followed a teacher-centered design, a practice identified by many re-
searchers (e.g., Rao, 1996; Xiao, 1998; Yu, 2001), the oral English course was a possible
exception: there was a chance that it would adopt TBLT methodology. It was usual at
the university for native or near-native speakers to teach the oral English course.
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

Furthermore, no textbook was prescribed, and the teacher was able to plan the lessons
and develop most of the class materials herself. The activities the teacher designed in
the current study included debates, role plays, and chain story-telling.

Data collection and coding

In order to investigate how the teacher and students interacted in a TBLT class, eight
hours of oral interaction lessons in total were video-taped by the second investigator
over a four-week period (i.e., two hours per week). Just before the recording, students
were informed by the teacher about the data collection. The video-taped lessons were
transcribed by two research assistants and checked again by the second investigator.
Building on empirical studies on classroom interaction (e.g., Lyster & Ranta, 1997;
Sheen, 2004), error treatment episodes were identified and coded using the coding tax-
onomy developed by Lyster and Ranta (1997). Error treatment episodes identified in
many observational studies cited earlier and in the current study incidentally occurred
during meaning-focused interaction, and are fundamentally different from error cor-
rection observed in teacher-centered classrooms. That is, quite distinct from classrooms
following a structure-based syllabus, this type of error treatment episode does not occur
unless students participate in communication activities. Although there are many other
ways to document classroom interaction, in this study we focus on corrective feedback
episodes as feedback that occurred incidentally during meaning-focused interaction to
draw learners’ attention to communicatively important forms, the basic claim being that
these episode have been found to be facilitative of L2 development (e.g., Lyster, 2004).

An episode of error treatment

An episode begins with a learner’s non-target-like utterance, which is addressed by a

type of feedback, and ends with a change in topic focus, either back to the meaning or
to another linguistic form (Ellis et al., 2001). In practice, an episode of error treatment
consisted of a non-target-like utterance by a student, the feedback from the teacher or
peer student, and a response by the student or topic continuation. Sometimes there
were several turns of teacher feedback and student response until the episode ended,
as shown in Example 1 below.
Example 1:
T: Do your parents have fun?
S: Yeah.
T: What do they do?
S: Chat.
S: Play card.
T: Play what.
S: Chest.
 Noriko Iwashita and Huifang (Lydia) Li

T: This is chest.
S: Chess.
T: Chess, chess.
In this episode, the teacher was trying to find out what the students’ parents did in
their leisure time. In response to the question, two students answered ‘chat’ and ‘play
card’. However, the pronunciation of the word in the second student’s answer was not
clear and the teacher asked for clarification. In response to the teacher’s request for
clarification, the student was finally able to say the word ‘chess’, pronouncing the word
in a way that the teacher was able to comprehend.

Definitions of feedback types

Five types of feedback can be identified in episodes of error treatment: explicit correc-
tion, teacher recast, student recast, clarification request (CR), and elicitation. Definitions
of these types are provided below and are derived from previous studies (e.g., Lyster &
Ranta, 1997; Sheen, 2004). Illustrative examples are taken from the data from the cur-
rent study.
Explicit correction refers to feedback that indicates clearly to the interacting stu-
dent which error is being corrected. It may or may not contain a correct form, as in
Example 2.
Example 2:
S: Some people say that if you eat out, it suggests that you are not full, the host is not,
hospi, er, er.
T: Hospitable, hospitality. Hospitality is the noun, hospitable is the adjective.
Recasts are reformulations of the whole or part of the learner’s erroneous utterance, but
without changing its meaning (Long & Robinson, 1998). In the present study, recasts
were distinguished according to the feedback provider. Thus, if the reformulation was
made by a peer, it was labelled as student recast, but if it was produced by the teachers,
it was labelled as teacher recast, as in Examples 3 and 4, respectively.
Example 3:
T: What?
S 1: Fruit.
→ S 2: Fruits.
T: Anything else?
Example 4:
S: My brother is forty; he’s married with two shilderen.3

3. Pronunciation of the word ‘children’ was not correct.

 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

→ T: Children
S: Children.
Clarification requests are phrases such as ‘Sorry’ and ‘I beg your pardon’ that indicate
to the student that his/her message has been misunderstood or that the utterance was
incomprehensible in some way, as in Example 5.
Example 5:
S: Is it well-known? [Pronunciation error]
→ T: er?
S: Well-known? [Pronunciation error]
Ss: Well-known.
T: Very what?
Ss: Well-known.
T: Ah, well-known, famous.
Finally, elicitation contains direct questions such as ‘Can you say that again?’, or repeat-
ing the learner’s previous utterance up to the error and waiting for the learner to supply
the correct form, as in Example 6.
Example 6:
S: yesterday we have a basketball match.
T: yesterday we
S: we had, yesterday we had a basketball match, but we couldn’t win, so it made me
sad
T: ok, good, it made me sad.

Definition of types of uptake

Uptake refers to a student’s verbal reaction to the feedback (Lyster & Ranta, 1997).
Based on the transcripts, five types of uptake were identified in the data: no chance,
ignore (no uptake), acknowledgement, successful uptake, and unsuccessful uptake. No
chance means the learner was not given a chance to show uptake, as in Example 7.
Example 7:
S: Knock the door when coming.
T: Knock at the door. Knock at the door before you come in. Something else. Think
of your daily life, what things happen, and you think they are wrong because of the
bad manner to other people.
S: Help stranger.
Ignore (no uptake) means that the learner does not show any response to the feedback,
as in Example 8.
 Noriko Iwashita and Huifang (Lydia) Li

Example 8:
S: she wears jeans.
T: a pair of jeans yah and a red...
S: she wears red.
Acknowledgement means that the learner responds with a simple ‘yes’, for instance, as
in Example 9.
Example 9:
S: because, er.
T: Because everybody else would do it.
→ S: Yeah.
Successful uptake means that the learner incorporates the linguistic information into
his/her subsequent production. Unsuccessful uptake refers to the learner’s failure to
incorporate, fully or partially, into his/her production the linguistic information that
the feedback was intended to address. It is different from acknowledge, ignore, and no
chance in that the learner makes an attempt to modify his/her initial utterance or re-
peat the teacher feedback. Successful and unsuccessful uptakes were observed in the
form of repetition or modification. Repetition is a response in which the learner repeats
the utterance in the form of a recast. Modification means that a learner modifies his/
her initial utterance in response to the teacher’s feedback. Examples of successful and
unsuccessful uptake are given below.
Example 10: Successful uptake – Repetition
S: Sometimes that our students leader will keep the will keep the, er.
T: Order.
→ S: Order.
Example 11: Successful uptake – Modification
T: Very what.
→ S: well known.
T: Ah, well known, ah, famous.
Example 12: Unsuccessful uptake – Repetition
S: Uncon, er.
→ T: Unconditionally, unconditionally.
S: Uncon, er.
Data analysis
All transcribed data were imported into the CHILDES program (MacWhinney, 1999)
and then coded according to the framework of feedback and uptake types illustrated
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

above. The frequency of occurrence of each type of feedback and uptake was calculat-
ed; further details are provided in the results section.

Results

To answer the research question about the patterns of corrective feedback in a TBLT
classroom in a Chinese educational setting, the eight hours of recorded data were tran-
scribed using standard orthography, and coded according to the feedback-and-re-
sponse framework illustrated above.

Frequency of feedback

Table 1 summarizes the frequency of feedback supplied by the teacher and the stu-
dents. The results show that more than half of the feedback consisted of recasts
(60.87%), followed by clarification requests (13.04%), and elicitations (10.63%).
These results reveal that error correction episodes occurred quite frequently in
this setting. Indeed, the results suggest that there were approximately 26 error correc-
tion episodes per hour, or one every two minutes (480 minutes divided by 207 error
correction episodes).

Frequency of feedback according to linguistic category

Table 2 summarizes the frequency of the feedback devoted given to different types of
linguistic errors. Nearly half of the teacher recasts (56 instances, 44.4%) followed an
instance of non-target-like pronunciation, followed by errors related to vocabulary
(39 instances, 31%). A similar trend was observed for student recasts. However, the
majority of clarification requests (CR) were placed in the category Other (i.e., not spe-
cific to any of the linguistic types listed here).

Table 1.  Frequency and type of feedback

Frequency % Frequency per hour

Explicit correction   16    7.73 2

Recasts 126   60.87 15.75
Student recasts   16    7.73 2
Clarification requests   27   13.04   3.38
Elicitation   22   10.63   2.75
Total 207 100.00 25.88
 Noriko Iwashita and Huifang (Lydia) Li

Table 2.  Frequency and type of feedback according to linguistic category

Syntax Morph Pron Vocab Other Total

Explicit correction 1 2 1 12 0   16
(%) 6.25 12.5 6.25 75 0 100
Teacher Recasts 5 21 56 39 5 126
(%) 4.0 16.7 44.4 31.0 4.0 100
Student recasts 0 2 7 7 0   16
(%) 0.0 12.5 43.8 43.8 0.0 100
Clarification requests (CR) 0 3 9 0 15   27
(%) 0.0 11.1 33.3 0.0 55.6 100
Elicitation 1 8 2 9 2   22
(%) 4.5 36.4 9.1 40.9 9.1 100
Total 7 36 75 67 22 207
(%) 3.4 17.4 36.2 32.4 10.6 100

Responses to feedback

Learners’ responses to the feedback are summarized in Table 3 below. Approximately

half of the instances of feedback were responded to with repetitions (62 instances,
29.95%) or modifications (77 instances, 37.2%). It should be noted that students did
not have a chance to respond to approximately 20% of the feedback, as the teacher
went on after providing the feedback.
Repetition and modification responses were further classified into two types
(successful or unsuccessful). More than 80% of the repetition responses (52 instances)
and more than 70% of the modifications were successful.

Table 3.  Frequency and type of responses to feedback

Frequency %

Repetition 62 29.95
Successful repetition 52
Unsuccessful repetition 10
Modification 77 37.20
Successful modification 55
Unsuccessful modification 22
Ignore 21 10.14
No chance 47 22.71
Total 207 100.00
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

In summary, substantial feedback instances (207 cases) were identified during the
eight hours of authentic classroom task-based interaction video-recorded for this
study. Of the 207 instances, more than half were recasts. A small amount of feedback
was supplied by students to other students. While recasts and elicitation were pro-
duced mainly in response to utterances that contained non-target-like pronunciation,
approximately half of the clarification requests were produced in response to utter-
ances which contained non-target-like features other than syntax, morphology, pro-
nunciation, and vocabulary. Students responded to many instances of teacher feedback
by repeating the feedback or modifying their initial utterance. Approximately 50% of
the responses to feedback were successful repetitions or modifications of recasts.

Discussion

The current study investigated the patterns of corrective feedback in a large, task-based
oral English class at a medium sized university in China. As reported in the previous
section, the findings of frequent teacher and student interaction characterized by a
substantial amount of recast feedback and students’ response to the feedback were
rather unexpected, due to the students’ reluctance to participate in whole class discus-
sion and also due to the large class size. Nevertheless, the findings provide further
support for the findings of previous empirical research on corrective feedback in class-
room interaction (e.g., Lyster & Ranta, 1997; Panova & Lyster, 2002; Loewen, 2004;
Sheen, 2004). In this section, these results will be discussed further by referring to
characteristics of task-based interaction, the principles or practices of TBLT in the
context of the current study, and the factors affecting the classroom interaction.

Characteristics of task-based interaction in the current study

As was reported in the results section, a substantial amount of feedback, in particular

recasts, was observed in the teacher – student and student – student task-based inter-
action. These findings were a surprise, considering the large class size and the students’
reluctance to speak, two factors which have been found to hinder the implementation
of TBLT (Hu, 2005a). In order to examine closely the characteristics of the interaction
observed in the current study, we first compared the results of the current study with
the interaction in the four studies reported in Sheen (2004). The contexts of all five
studies are very different (see Table 4). In particular, the class size in the current study
is much larger than any of the others. The L1 of the participants in the current study is
homogeneous like in the Korean research site in Sheen (2004), but while the research
site in the Korean study was a private language school where highly motivated learners
attended classes to improve their communication skills, the students in the current
study were following the course as a requirement of their English language major with-
in the teacher training program. Nevertheless, the activities that Sheen (2004)
 Noriko Iwashita and Huifang (Lydia) Li

Table 4.  Learner background in the four studies in Sheen (2004) and the current study

Length of recording hours Class size (n) age

Canada Immersion 18.3 24–30   9–10

Canada ESL 10 25 17–55
NZ ESL 12 12 18–21
Korea EFL 12 4–6 29–36
China EFL  8 50 17–19

reported the learners to be engaged in were similar, such as role plays, oral interaction
tasks, and comprehension activities.
The total number of feedback tokens given by the students in the current study is
smaller than in the two Canadian studies, but larger than in the studies conducted in
Korea and New Zealand. Because of the different total number of hours of recording,
the data reported in Sheen (2004) also calculated the occurrences of feedback given
per hour and the frequency of recast feedback (the results are summarized in Table 5).
In the EFL class in China, the third largest number of feedback token was given,
following the immersion class and EFL class in Canada. The remaining two classes
(i.e., NZ ESL and Korea) received similar amounts of feedback.
The percentage of recasts given in the Chinese EFL class is smaller than that ob-
served in the classrooms in Korea and New Zealand, but slightly larger than in the im-
mersion and ESL classrooms in Canada. In both those settings, recasts were the most
frequently given feedback. Sheen (2004) explains these results by referring to the Korean
teacher’s strong emphasis on fluency over accuracy and the advantage of recasts in main-
taining the flow of communication in the interview conducted after the data collection.
This might be the case also for the current study. In particular, we found that a large
amount of feedback (approximately 36%) was given to non-target-like pronunciation.
Recasts may be a particularly user-friendly way of providing pronunciation feedback.

Table 5.  Frequency of feedback in the current study and the four studies reported
in Sheen (2004)

Feedback Recasts

Total Per hour

Canadian Immersion 686 37.49 375 (54.7%)

Canada ESL 412 41.20 226 (54.9%)
NZ ESL 189 15.75 129 (68.3%)
Korea EFL 186 15.50 154 (82.3%)
China EFL 207 25.88 126 (56.8%)
Note. The total amount of feedback includes the instances of recasts (both teacher and students), elicitation,
clarification requests and explicit correction.
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

Table 6.  Leaner uptake in the four studies in Sheen (2004) and the current study

No

Uptake No uptake total

Immersion 376 310 686

  (%)   54.8 45.2 100
Canada ESL 192 220 412
  (%)   46.6 53.4 100
NZ ESL 152 37 189
  (%)   80.4 19.6 100
Korea EFL 153 33 186
  (%)   82.3 17.7 100
China 107 32 107
  (%)   76.98 23.02 100

As for learner uptake, the comparison with the results in Sheen (2004) is summarized
in Table 6. As in the frequency of recasts, the proportion of uptake is larger than in the
studies conducted in Canada, but smaller than in the Korean and NZ studies. Sheen
(2004) explains the higher rate of uptake found in the Korean and NZ studies com-
pared with the studies conducted in Canada by the years of formal instruction which
resulted in the likelihood of students’ attending to the teacher feedback. Also, more
form-focused recasts were provided to the students in the NZ and Korean studies. This
might be the case for the current study too, as all participants had studied English for
several years before they began their university study. This is illustrated in Example 13
below. The feedback is focused on pronunciation and the meaning of the word. In the
example, the student is not able to pronounce the word ‘unconditionally’ correctly.
Debbie, the teacher, provides the correct pronunciation with a recast, but does not give
an opportunity for the student to repeat. Instead, she explains the meaning.
Example 13
S: Uncon, er
T: Unconditionally, do you know. Unconditionally, like friends will say I give you
this, but you must give me something else. But your parents will give you this
without expecting something else in return, right?
As shown in the example below, teacher recasts were followed by unsuccessful uptake
of the elicitation feedback in the previous turn. In Example 14, Debbie provides elicita-
tion in the first turn following the student utterance in which the plural morpheme ‘s’
is missing. However, the student does not successfully modify her initial utterance, but
repeats the word without the ‘s’, to which Debbie provides recasts.
 Noriko Iwashita and Huifang (Lydia) Li

Example 14
S: yes, I have to learn many course
T: many. (Elicitation)
S: course.
T: courses.
S: courses.
S: courses, I can’t stand some boy always tell lies.
T: telling lies
S: I have a habit of drinking a lot of, a lot of
As Sheen showed in her interpretation of the recasts observed in her data, the feedback
provided in the above episodes was explicit. Recasts were often given in a single word
utterance with emphasis, which might have contributed to a high rate of uptake as
found in Sheen (2004).
In interpreting the findings in the context of task-based language teaching, it is
important to mention that Debbie’s interaction with her students during the two-hour
classes was characterized by a continuum from focus-on-forms, to focus-on-meaning,
and focus-on-form. For example, in one of the classes, where she used three tasks, she
began with a reading-aloud activity. While the students were reading the text aloud,
she corrected their pronunciation while walking around the room. This was followed
by her brief explanation of food items and questions to check the students’ compre-
hension. Before students introduced themselves, she provided 15 basic sentence pat-
terns such as ‘I’m worried about’, ‘I don’t believe in’, and ‘I’m good at’. In these phases of
the class, Debbie’s focus was heavily on forms and her approach was far from task-
based, where meaning is primarily focused on communication. However, when stu-
dents talked about their own family, Debbie’s focus was primarily on meaning, as she
asked contingent questions in order for the students to elaborate on the topic, as shown
in Example 15 below.
Example 15
S: In my father’s family tree, grandpa, grandma....
T: What’s their name?
S: My grandpa is ‘Wang Meiqiu’, my grandpa is ... Sorry I don’t know
Class: (all laugh)
T: You don’t know your grandparents’ names?
C: Yeah, yeah
T: (to the class) You also don’t know?
C: Yeah,
Although the focus of the conversation is on meaning, it is questionable whether this
kind of activity is qualified as a task, considering the definitions of task in the literature
(e.g., Long, 1985).
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

The last part of the class was group work. Students in each group performed a role
play before the whole class. The scenario was that Jessica, a 17-year-old girl, finds her-
self pregnant and is afraid to tell her parents, but Jessica’s parents learn of her preg-
nancy through her good friend. The group had to simulate a conversation between
Jessica and her parents about the pregnancy. Unlike the earlier phase of the class, the
focus was primarily on meaning, and shifts to focus-on-form whenever Debbie draws
the learners’ attention to the form through recasts, clarification requests, and elicitation.
As reported in the results section, frequent interaction was observed when stu-
dents carried out a role-play and talked about their family with Debbie. From the per-
spective of TBLT, in the first two phases of the class, where students read the text aloud
and Debbie corrected their pronunciation and provided some example sentences for
the next phase, it is debatable that such content could be considered task-based. Con-
sidering Long’s (1985) definition that a task is, more or less, the things people do in
everyday life, such as buying shoes and writing cheques, reading a text for pronuncia-
tion correction may not have a relationship to the real world. However, the reading for
comprehension represented real-life to an extent, for instance reading the newspaper
for understanding. As Littlewood (2004) suggests, it may not always be easy to iden-
tify a task with the definition of a focus on meaning. He elaborates that there is not a
dichotomy but a continuum along which students may operate, with differing focus on
forms and meaning. Furthermore, as suggested by Samuda and Bygate (2008), refer-
ring to Bygate (2000) and Shehadeh (2005), Debbie’s use of tasks in her oral English
course appears to be closer to task-supported learning and teaching than to task-based
learning and teaching. That is, tasks are not exclusively central for learning and teach-
ing, but are meaning-focused activities used for a broad range of purposes, including
providing practice, developing fluency, and raising awareness of specific language fea-
tures (Samuda & Bygate, 2008).

Factors affecting frequent classroom interaction

As discussed in the previous section, in China the factors hindering the successful
implementation of TBLT that researchers have identified (e.g., Anderson, 1993; Hu
2002a; 2005a; Liao, 2000; Littlewood, 2007; Yu, 2001), such as large class size, and the
Chinese culture of teaching and learning in which teachers are characterized as the
authority, were also observed in the classroom where the data were collected. Contrary
to our expectation, however, it appeared that the considerable amount of interaction
measured with frequency of corrective feedback and students’ response to the feed-
back was not affected by those hindering factors. While other classes at the same uni-
versity followed the traditional teacher-centred and grammar-focused method, the
frequent interaction observed in the oral English class could be attributed to Debbie’s
belief in communicative teaching, the positive relationship with her students observed
by the researchers, and the subject domain. This suggests that the students were capa-
ble of responding to alternative modes of classroom interaction.
 Noriko Iwashita and Huifang (Lydia) Li

The teacher, Debbie, expressed a strong belief that language should be learned
through communication. Her belief stemmed from her postgraduate training in the
USA, as she explained in the informal interview with the second investigator. She em-
ployed TBLT in her oral class and included a variety of tasks, such as information gap
discussions and debates on topics chosen for classroom discussion, and interaction.
Furthermore, she was open to sharing her belief with her students and encouraged the
students’ active participation in classroom interaction, as the second investigator ob-
served in Debbie’s informal interaction outside class. These factors may have contrib-
uted to the students’ participation in the classroom interaction observed in the current
study. Furthermore, the teacher’s choice of tasks was based on the relevance of the
tasks to the students’ lives, such as debating whether college students in China should
have a boy/girl friend. The teacher believed also that relevant tasks were easier for the
students to do.
Lastly, the teacher’s rapport with her students was further illustrated by her social
interaction with them outside class time, which included playing Chinese checkers
during breaks, for instance. Many students expressed very positive attitudes towards
Debbie, noting that they did not feel inhibited from making mistakes in her class. Even
if she was respected as a teacher, it appeared that they saw her as a friend. Partly, this
may have been the result of the non-threatening, face-saving and non-intrusive way in
which she corrected errors, that is, primarily through recasting.
The frequent teacher – student and student – student task-based interaction
observed in this oral English class may have been attributable also to the subject factor.
Pica (2002) found relatively few instances of modified interaction in an investigation
of the interaction patterns in several English L2 classes studying thematic units on film
and literature. As Cortazzi and Jin (1996b) commented, Chinese students, being influ-
enced by the Chinese culture of learning, tend to enjoy passive learning. This seemed
to be true of the intensive reading course and other subjects, but the students in our
study welcomed a non-Chinese near-native speaker of English to teach oral English.
This may suggest that Chinese students’ perceptions of how language should be learned
may differ depending on the focus of study. They may be able to accept a communica-
tive way of learning in the oral English course, leading to more active participation in
classroom interaction than in other types of class. However, as the present study did
not collect the students’ perceptions, what is discussed here has to remain speculation.
Nevertheless, the main implication is that these findings strongly suggest that EFL
Chinese students are capable of and willing to respond positively to an alternative
mode of classroom methodology, including a task-based approach.

Conclusion and implications of the study

The current study investigated how students and teacher interacted with one another
focusing on corrective feedback episodes in a task-based oral English class in a
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

medium-sized university in China. Despite the large class size and the students’ unfa-
miliarity with the teaching method, which is very different from the traditional Chinese
way of learning and teaching, active participation in the interaction characterized by
the students’ and frequent teacher feedback to the students’ non-target-like utterances
was observed. Furthermore, many instances of feedback were found to be recasts, to
which students responded successfully. There were also instances of recasts given by
peers.
These unexpected results were first examined in light of characteristics of the in-
teraction and then further discussed in light of the classroom teacher’s strong belief in
teaching and her rapport with the students. Furthermore, although frequent teacher
– student and student – student interaction was observed, the content of the class was
characterized as ‘task-supported’ rather than ‘task-based’, in that the communicative
tasks were introduced to promote interaction but were not treated entirely as units of
the syllabus. Nevertheless, the results provide further support for the findings of the
empirical research on corrective feedback in classroom interaction (e.g., Lyster &
Ranta, 1997; Panova & Lyster, 2002; Loewen, 2004; Sheen, 2004) and also for a number
of pedagogical implications, including classroom management and task selection.
Frequent teacher-student and student-student interaction, despite supposed dif-
ferences in learning deep-rooted in Chinese culture, may be attributed to the good
student – teacher relationship. The teacher chose activities to which students were able
to relate easily, and her rapport with the students encouraged them to participate with-
out worrying about making mistakes. It seems that the teacher’s familiarity with
communication-based methods and also her stress on establishing a good class rela-
tionship were the main driving force for the successful implementation of a communi-
cative curriculum. These aspects or skills could be taught more generally through
pre-service and in-service training.
The main implication of this study is that, even in large classes and despite these
Chinese EFL students’ unfamiliarity with a novel teaching method and their tradi-
tional views of teaching and learning, the implementation of TBLT appears to be
feasible as evidenced by the frequent and abundant teacher – student and student –
student interaction observed in this classroom. As a consequence, we believe that ap-
propriate in-service and pre-service teacher training may help develop teachers’
understanding and appreciation of the potential of TBLT and ways of applying it suc-
cessfully in similar settings in China or elsewhere. Nevertheless, the challenge remains
for instructors in teacher training to consider ways to move on from a task-supported
learning mode observed in the current study to full implementation of TBLT. In addi-
tion, we believe that establishing a good teacher – student relationship is another factor
that enhances students’ participation in the interaction, and helps to shift their beliefs
about teaching and learning from traditional, teacher-fronted methodologies to more
student-activating, task-based principles. The active student participation found in
this study was enhanced by these Chinese students’ willingness to accept methodologies
 Noriko Iwashita and Huifang (Lydia) Li

and modes of learning that are different from their past learning experience, their be-
liefs about learning, and the traditional methodologies they were accustomed to.

Limitations and suggestions for further research

There are a number of limitations to the study. First, the current study was a one-off
case study that explored teacher – student and student-student interaction in one EFL
classroom in China in terms of patterns of corrective feedback, making the findings
difficult to generalize from. It was also a context where the teacher was familiar with
TBLT methodology, even a strong supporter of it, and had a good relationship with her
students. We recognize that this might not be the case in other educational contexts.
Second, our interpretation of the findings was based on the analysis of recorded
data. We reported on the teacher’s perceptions of teaching based on one of the investi-
gator’s (Li’s) informal interviews with the teacher to provide further interpretation of
the data. Similarly, we observed that the students’ attitudes towards the teacher and
their class were positive, but we did not collect data on students’ perceptions. Survey-
ing teachers’ and students’ perceptions in addition to collecting classroom data may
provide a deeper insight into the dynamics of the classroom interactions and the fac-
tors that lead to the successful implementation of TBLT, or the lack of it.
Third, we analyzed the data following the methods employed in earlier studies
(Lyster & Ranta, 1997; Sheen, 2004) which focused exclusively on specific types of
feedback and uptake. Although the results of the analysis illustrate how the teacher
and students interacted in a large class in terms of occurrence of feedback and re-
sponse to the feedback, we acknowledge that there are other aspects of interactions
which potentially enhance L2 development as well (see Samuda & Bygate, 2008).
Lastly, although the teaching method used in the research site was task-based, a
close examination of the data showed that the methodology employed in the course
seemed more ‘task-supported’ than ‘task-based’ (Bygate, 2000; Ellis, 2000). In order to
evaluate how effectively the TBLT curriculum is being implemented in Chinese class-
rooms and the extent to which students are benefiting from the new curriculum, close
evaluation of the methodology and tasks, in addition to examining aspects of class-
room interaction, is crucial.

References

Adamson, B., & Morris, P. (1997). The English curriculum in the People’s Republic of China.
Comparative Education Review, 41(1), 3–26.
Anderson, J. (1993). Is a communicative approach practical for teaching English in China? Pros
and cons. System, 21(4), 471–480.
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

Brown, G., & Yule, G. (1983). Teaching the spoken language: An approach based on the analysis of
conversational English. Cambridge: Cambridge University Press
Bygate, M. (2000). Introduction to special issue: Tasks in language pedagogy. Language Teaching
Research, 4(3), 185–192.
Cheng, L. (2008). The key to success: English language testing in China. Language Testing, 25(1),
15–37.
Cortazzi, M., & Jin, L. (1996a). English teaching and learning in China. Language Teaching, 29,
61–80.
Cortazzi, M., & Jin, L. (1996b). Cultures of learning: Language classrooms in China. In H. Cole-
man (Ed.), Society and the language classroom (pp. 169–205). Cambridge: Cambridge Uni-
versity Press,
Cortazzi, M., & Jin, L. (2001). Large classes in China: ‘Good’ teachers and interaction. In D. A.
Watkins & J. B. Biggs (Eds.), Teaching the Chinese learner: Psychological and pedagogical
perspectives (pp. 115–134). Hong Kong and Camberwell, Australia: Comparative Education
Research Centre, Australian Council for Educational Research.
Cowan, J. R., Light, R. L., Mathews, B. E., & Tucker, G. R. (1979). English teaching in China: A
recent survey. TESOL Quarterly, 13(4), 465–482.
Ellis, R. (2000). Task-based research and pedagogy. Language Teaching Research, (special issue,
Tasks in Language Pedagogy) 4(3), 193–220.
Ellis, R., Basturkmen, H., & Loewen, S. (2001). Learner uptake in communicate ESL lessons.
Language Learning, 51(2), 281–318.
Ellis, R., Basturkmen, H., & Loewen, S. (2002). Doing focus on form. System, 30, 419–432.
Ellis, R., & Sheen, Y. (2006). Re-examining the role of recasts in second language acquisition.
Studies in Second Language Acquisition, 28(4), 575–600.
English Major Syllabus in Higher Education. (2000). Retrieved 6th January 6th, 2010 from
http://www.bfsu.edu.cn/chinese/site/gxyyzyxxw/zywj/tyyjxdg.htm.
Hiep, P. H. (2007). Communicative language teaching: Unity within diversity. ELT Journal,
67(3), 193–201.
Hu, G. (2002a). Potential cultural resistance to pedagogical imports: The case of communicative
language teaching in China. Language, Culture and Curriculum, 15(2), 93–105.
Hu, G. (2002b). Recent important developments in secondary English-language teaching in the
People’s Republic of China. Language, Culture and Curriculum, 15(1), 30–49.
Hu, G. (2003). English language teaching in China: Regional differences and contributing fac-
tors. Journal of Multilingual and Multicultural Development, 24(4), 290–318.
Hu, G. (2005a). ‘CLT is best for China’: An untenable absolutist claim. ELT Journal, 59(1), 65–68.
Hu, G. (2005b). Contextual influences on instructional practices: A Chinese case for an eco-
logical approach to ELT. TESOL Quarterly, 39(4), 635–660.
Hu, G. (2005c). English language education in China: Policies, progress, and problems. Lan-
guage Policy, 4, 5–24.
Jin, L., & Cortazzi, M. (1998). Dimensions of dialogue: Large classes in China. Interactional
Journal of Educational Research, 29, 739–761.
Li, C. (2003). A study of in-service teachers’ beliefs, difficulties and problems in current teacher
development programs. HKBU Papers in Applied Language Studies, 7, 64–85.
Li, P. (2009). The analysis of instructional design from a task-based language teaching mode.
Journal of Hubei TV University, 29(12), 142–143.
 Noriko Iwashita and Huifang (Lydia) Li

Liao, X. (2000). How communicative language teaching became acceptable in secondary schools
in China. The Internet TESL Journal, 6(10). Retrieved 30th October 2009 from http://iteslj.
org/indexPrev00.html.
Littlewood, W. T. (2004). Task-based approach: some questions and suggestions. English Lan-
guage Teaching Journal, 58(4), 319–326.
Littlewood, W. T. (2007). Communicative and task-based language teaching in East Asian class-
rooms. Language Teaching, 40, 243–249.
Loewen, S. (2005). Incidental focus on form and second language learning. Studies in Second
Language Acquisition, 27, 361–386 .
Long, M. H. (1981). Input, interaction and second-language acquisition. Annals of the New York
Academy of Sciences, 379, 259–278.
Long, M. H. (1983). Native speaker/non-native speaker conversation and the negotiation of
comprehensible input. Applied Linguistics, 4, 126–141.
Long, M. H. (1985). A role for instruction in second language acquisition: Task-based language
training. In K. Hyltenstam & M. Pienemann (Eds.), Modelling and assessing second language
acquisition (pp. 77–99). Clevedon, UK: Multilingual Matters.
Long, M. H. (1996). The role of the linguistic environment in second language acquisition. In W.
C. Ritchie (ed.), Handbook of language acquisition: Second language acquisition (vol. 2, pp.
413–468). New York, NY: Academic Press.
Long, M. H. (1991). Focus on form: A design feature in language teaching methodology. In K.
de Bot, R. Ginsberg, & C. Kramsch (Eds), Foreign language research in cross-cultural per-
spective (pp. 39–52). Amsterdam: John Benjamins.
Long, M. H., & Porter, P. A. (1985). Group work, interlanguage talk, and second Language Ac-
quisition. TESOL Quarterly, 19(2), 207 – 228.
Long, M. H., & Robinson, P. (1998). Focus on form: Theory, research, and practice. In C. Dough-
ty & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp. 15–41).
Cambridge: Cambridge University Press.
Lyster, R. (1998a). Recasts, repetition, and ambiguity in L2 classroom discourse. Studies in Sec-
ond Language Acquisition, 20(1), 51–92.
Lyster, R. (1998b). Negotiation of form, recasts, and explicit correction in relation to error types
and learner repair in immersion classrooms. Language Learning, 48(2), 183–218.
Lyster, R. (2004). Differential effects of prompts and recasts in form-focused instruction .Studies
in Second Language Acquisition, 26, 399–432 .
Lyster, R., & Mori, H. (2006). Interactional feedback and instructional counterbalance. Studies
in Second Language Acquisition, 28, 269–300.
Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake. Studies in Second Lan-
guage Acquisition, 19(1), 37–66.
MacWhinney, B. (1999). The CHILDES Project: Tools for analyzing talk (2nd ed.). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Ng, C., & Tang, E. (1997). Teachers’ needs in the process of EFL reform in China: A report from
Shanghai. Perspective, 9, 63–85.
Nunan, D. (2004). Task-based language teaching. Cambridge: Cambridge University Press.
Oliver, R. (2000). Age differences in negotiation and feedback in classroom and pair-work. Lan-
guage Learning, 50(1), 119–151.
Oliver, R., & Mackey A. (2003). Interactional context and feedback in child ESL classrooms.
Modern Language Journal, 87(4), 519–533.
 Chapter 7.  Patterns of corrective feedback in a task-based adult EFL classroom setting in China 

Panova, I., & Lyster, R. (2002). Patterns of corrective feedback and uptake in an adult ESL class-
room. TESOL Quarterly, 36(4), 573–595.
Pica, T. (2002). Subject-matter content: How does it assist the interactional and linguistic needs
of classroom language learners? Modern Language Journal, 86, 1–19.
Pica, T., & Doughty, C. (1985). Input and interaction in the communicative language classroom:
a comparison of teacher-fronted and group activities. In Gass, S. and Madden, C. (Eds.),
Input in second language acquisition (pp. 115–132). Rowley, MA: Newbury House.
Pica, T., Holliday, L., Lewis, N., & Morgenthaler, L. (1989). Comprehensible output as an out-
come of linguistic demands on the learner. Studies in Second Language Acquisition, 11(1),
63–90.
Pica, T., Kanagy R., & Falodun, J. (1993). Choosing and using communication Tasks for second
language instruction and research. In G. Crookes & S. Gass (Eds.), Tasks and language
learning: Integrating theory and practice (pp. 9–34). Clevedon, UK: Multilingual Matters.
Rao, Z. (1996). Reconciling communicative approaches to the teaching of English with tradi-
tional Chinese methods. Research in the Teaching of English, 30(4), 458–471.
Samuda, V., & Bygate, M. (2008). Tasks in second language learning. New York, NY: Palgrave
Macmillan.
Savignon, S. J. (1991). Communicative language teaching: State of the art. TESOL Quarterly,
25(2), 261–277.
Sheen, Y. (2004). Corrective feedback and learner uptake in communicative classrooms across
instructional settings. Language Teaching Research, 8(3), 263–300.
Sheen, Y. (2006). Exploring the relationship between characteristics of recasts and learner up-
take. Language Teaching Research, 10, 361–392.
Shehadeh, A. (2005). Task-based language learning and teaching: Theories and applications.
In C. Edwards & J. Willis (Eds.), Teachers exploring tasks in English language teaching
(pp. 13–30). Basingstoke, UK: Palgrave McMillan.
Swain, M. (1985). Communicative competence: some roles of comprehensible input and com-
prehensible output in its development. In S. Gass & C. Madden (Eds.), Input in Second
Language Acquisition (pp. 235–253). Rowly, MA: Newbury House.
Van den Branden, K., Bygate, M., & Norris, J. M. (Eds.) (2009). Task-Based Language Teaching:
A reader. Amsterdam: John Benjamins.
Wei, M. (2002). Recasts, noticing, and error types: Japanese learners’ perception of corrective
feedback. Acquisition of Japanese as a Second Language, 5, 24–41.
Xiao, A. (1998). An academic or a more practical approach. Forum, 36(2), 28.
Yu, L. (2001). Communicative language teaching in China: Progress and resistance. TESOL
Quarterly, 35(1), 194–198.
 chapter 8

Incidental learner-generated focus

on form in a task-based EFL classroom

Paul J. Moore
University of Wollongong, Australia

This small longitudinal study investigated oral task-based interaction in an

undergraduate EFL classroom in Japan. First, descriptive quantitative data
related to language-related episodes (LREs) and other contextual data from
four focal learners and their partners (N = 8) in two oral presentation tasks
were quantified to provide insights into the amount and effectiveness of
learner-generated focus on form in a Japanese university context. Next, a
qualitative microanalysis of one learner’s interaction with partners of similar
proficiency on two similar tasks, separated by a period of seven months, was
conducted to investigate influences of context. Much as in previous studies,
there was little focus on form in interaction and there was much variability
across dyads. Qualitative analysis revealed that the effectiveness of FOF in
interaction and performance may have been influenced by the learners’ shared
background (including L1 use), individual differences in terms of engagement
in LREs, learners’ perceptions of each other’s language proficiency, and other
interpersonally negotiated features of the interaction. The chapter concludes
with suggestions for awareness-raising activities which may improve the
effectiveness of task-based learner-generated FOF in foreign language settings.

Introduction

This chapter reports on a study into the incidence and effectiveness of learner-generat-
ed focus on form (FOF) in task-based learner-learner discourse in an English-as-a-
foreign-language (EFL) classroom in Japan. This includes a microanalysis of one
learner’s task-based interaction and performance over time and across partners, high-
lighting individual and dialogically negotiated features which may influence the ef-
fectiveness of incidental learner-generated FOF in EFL contexts.
Focus on form (FOF), initially defined as “[overtly drawing] students’ attention to
linguistic elements as they arise incidentally in lessons whose overriding focus is on
meaning” (Long, 1991, p. 46), is a major pedagogic principle in task-based language
teaching (TBLT; cf. Long 2007, p. 122). This is evidenced by its prominence in SLA
 Paul J. Moore

research from cognitive processing and sociocultural perspectives, and the several vol-
umes dedicated to its investigation and application (e.g., Doughty & Williams, 1998;
Ellis, 2001; Fotos & Nassaji, 2007; cf. also Mackey, 2007). The ability of learners to
provide each other with corrective feedback, or to collaboratively create new learning
opportunities (e.g., Donato, 1994; Storch, 2002; Swain & Lapkin, 1998) in task-based
interaction also features prominently in this research. A major unit of analysis in re-
search into task-based learner-learner discourse is the language-related episode (LRE),1
defined most succinctly by Swain & Lapkin (1998) as “any part of a dialogue where the
students talk about the language they are producing, question their language use, or
correct themselves or others” (p. 326).

Incidental learner-generated focus on form

Although it has been found that learners can provide each other with corrective feed-
back (e.g., Fujii & Mackey, 2009; Zhao & Bitchener, 2007), and that learner-learner
(L-L) dialogue can result in more uptake (immediate responses to feedback; cf. Lyster
& Ranta, 1997) than teacher-learner (T-L; Ellis et al., 2001) dialogue, only a small
number of classroom studies into FOF have investigated L-L dialogue in detail.
Seedhouse (2001) drew on conversation analysis (CA) theory and an analysis of
330 lessons from various sources to contrast interactional feedback given by teachers
to that given by other learners. CA theory (e.g., Schegloff, Jefferson & Sacks, 1977)
holds that, because of considerations of face, when conversational trouble arises, re-
pair is achieved via a series of preferential methods, starting with self-initiated self-
repair (where an individual notes an error in his/her own output and self-corrects),
followed by self-initiated other-repair (where an individual notes an error in his/her
own output and elicits correction from an interlocutor), other-initiated self-repair
(where an interlocutor highlights an error in an individual’s output and the individual
self-corrects), and finally other-initiated other-repair (where an interlocutor notes and
corrects an individual’s error; see Storch (2001) and Mennim (2005) for related analy-
ses of LRE initiation and resolution in FOF editing/revision tasks). Seedhouse found
that teachers often provide overt and direct positive evaluation when learners produce
correct responses, but avoid overt, explicit, direct negative evaluation when learners
produce incorrect responses. He argues that, while teachers are trying to create an
environment where errors are seen as a positive part of learning, avoidance of use of
the word ‘no’ sends “the interactional message that making errors is an embarrassing,
face-threatening matter” (p. 368). He cites survey research of students’ attitudes, which

1. Ellis and his colleagues (e.g., Ellis, Basturkmen & Loewen, 2001) use the term form-focused
episodes (FFEs), defining them as “very similar to ‘language related episodes’ except that they
include occasions when the teacher directed attention to form either pre-emptively or reactive-
ly” (Ellis, 2008: 831).
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

finds that learners want precisely the kind of feedback that teachers are not giving
them, and that learners do provide each other with direct negative feedback.
Williams (1999) investigated spontaneous learner-generated focus on form in
sixty-five hours of adult ESL classroom interaction data and reported the following
results:
– learners initiated LREs, but not very often;
– use of LREs increased significantly with proficiency;
– the most prevalent (by far) type of LRE was requests to the teacher, though this
decreased with increases in proficiency;
– grammar-focused tasks resulted in the most LREs and free conversation the
least;
– lexically-oriented LREs accounted for 80% of the total LREs, regardless of profi-
ciency.
Interviews with the learners revealed the perception that they were either ‘told to’ or
‘supposed to’ focus on form in one activity and fluency in another, though there was no
overt direction from the teacher to do so. Like Seedhouse, Williams argues that learn-
ers “come to this perception as a result of more subtle cues” (p. 616). She adds that
learners’ perceptions of the goal of an activity guide their focus. In conclusion, she ar-
gues that, given learners’ natural bias towards focus on vocabulary and that they “do
not spontaneously attend to formal aspects of language very frequently or consistently
across activities” (p. 619), the responsibility for encouraging focus on other aspects on
form lies with the teacher, though it is unclear whether explicit or implicit teacher in-
tervention will enable learners to focus more effectively on form.
Zhao and Bitchener (2007) compared form-focused episodes (FFEs; see note 1)
in teacher-learner (T-L) and learner-learner (L-L) interaction in unfocused informa-
tion exchange tasks, in adult migrant ESL classes in New Zealand. They found that
learners were far more likely to initiate pre-emptive FOF (where no production prob-
lem had occurred) in L-L interaction than in whole-class situations, noting that learn-
ers may be reluctant to show a lack of understanding in the latter setting. Confirming
findings from previous studies (e.g., Ellis et al. 2001), they found that recasts were the
most common form of corrective feedback used by teachers (33% of FFEs), followed
closely by explicit corrective feedback (30% of FFEs). They also found similar results
for learners, in that explicit feedback (34%) was most common, followed by recasts
(28%). In addition, for both groups the majority of FFEs were lexical, and there were
similar amounts of successful uptake. One major difference was that L-L interaction
included FFEs with no feedback (11.8%) and incorrect feedback (4.9%, with 2% in-
volving uptake). Zhao and Bitchener balanced this against the findings that there was
more uptake in L-L (72.3%) than in T-L interactions (54.4%), and that incorrect
feedback was negligible in L-L interactions, much as in earlier studies (e.g., Pica &
Doughty, 1985).
 Paul J. Moore

Research in foreign language contexts

While research in foreign language contexts aims to extend the findings of FOF re-
search in SLA, there are several distinguishing features of such research. In addition to
those outlined by Shehadeh (this volume), learners (and teachers) may be resistant to
FOF activities for social reasons (e.g., social relationships between partners; Philp,
Walter & Basturkmen, 2010), cultural reasons (e.g., face considerations; Fujii &
Mackey, 2009), or curriculum policy-related reasons (e.g., curriculum focus on ex-
plicit grammar teaching; McDonough, 2004). In addition, where learners have a shared
L1, this may play a more prominent role in their classroom interaction (e.g., Swain &
Lapkin, 2000; Moore, in press), as they may be more likely to draw on their L1 for
metalinguistic talk, for example.
In a study into learner-learner interaction in a Thai university EFL classroom,
McDonough (2004) investigated whether interaction on tasks focusing learners on real
and unreal conditional clauses improved learner production on immediate and delayed
post-tests. After separating learners into high- and low-participation groups based on
the number of negative feedback and modified output ‘episodes’, she found that learn-
ers in the high-participation group showed significant improvement on both condi-
tionals in immediate post-tests and on the delayed post-test for real conditionals, while
the low-participation learners only showed improvement on the immediate post-test
for unreal conditionals. In interviews with learners, she found that they placed little
value on feedback from other learners, and that they felt they required explicit gram-
mar instruction for upcoming examinations. Similar responses were received from
teachers, highlighting a common issue in EFL curricula in Asia: Washback from formal
written examinations can detract from a focus on language for oral communication.
Philp et al. (2010) investigated environmental and social influences on incidental
focus on form in a foreign-language French classroom in New Zealand. Seven of 31
students in an intact classroom were observed, recorded and interviewed as they per-
formed eight unfocused tasks over three weeks. Thirty-three (mostly lexical) LREs
occurred in the data, and one third of these involved the use of the learners’ L1. In
primed interviews, learners mentioned several social reasons for not focusing on form.
These included: relationships between partners (feelings of awkwardness were report-
ed between some partners); learners’ perceptions of partners’ comparative language
proficiency, or their own ability to provide corrective feedback; learners’ framing of the
task (not wanting to interrupt ‘the game’ of role plays [p. 275]); and learners’ percep-
tions of the teacher’s expectations.
Finally, Fujii and Mackey (2009) investigated interactional feedback in unfocused
tasks in a university-based EFL classroom in Japan (N = 18). In the study, feedback
followed only 6.5% of non-target-like turns, though learners did modify their output
in response to feedback, which mainly included clarification requests and confirma-
tion checks. The latter provided rich positive evidence, in that they included para-
phrases and summaries of the problematic turns. They also found that recasts provided
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

by learners were problematic in that they provide only “partially more target-like
reformulations” (p. 283). With regard to the low incidence of feedback, the authors
suggest that the learners may have avoided feedback because of their shared cultural
background, including avoidance of potential face-threatening acts. They also state
that the learners may neither have noticed errors, nor been of a level of proficiency to
provide effective feedback. Finally, the authors found several instances where LREs
involving collaborative non-target-like resolutions appeared to lead to so-called ‘mis-
learning’ on post-tests. They conclude that more research is needed into individual
differences in abilities and preferences with regard to feedback, as well as “the per-
ceived ‘authority’ of the feedback provider” (p. 293).

Sociocognitive influences on FOF

Research from a sociocultural perspective has linked the effectiveness of LREs to the
existence of mutually supportive task-focused relationships developed between learn-
ers in interaction. Such interaction may involve a variety of pedagogic moves, includ-
ing scaffolding (Wood, Bruner & Ross, 1976) where a more competent peer supports a
less competent peer, or collective scaffolding (Donato, 1994; cf. also Storch, 2002;
Swain & Lapkin, 1998) where learners provide mutual support. Forms focused on in
collective scaffolding are hypothesised to create new knowledge for all participants.
Intersubjectivity, or the ongoing dialogic negotiation of partially shared perspectives
and goals (e.g., Wells, 1998), is another important construct, linking the social and the
cognitive in task-based activity, as evidenced by learners’ engagement with each other
(e.g., through such phenomena as overlapping speech and backchanneling; cf. Brooks
& Donato, 1994; Storch, 2002) and engagement with the task. Evidence of intersubjec-
tivity (see below) has been linked with successful task completion (Brooks & Donato,
1994), as well transfer of knowledge in LREs (Storch, 2002). The negotiation of task
control (Swain & Lapkin, 1998; Storch, 2002) is also important, as dominant-passive
dyads may generate fewer opportunities for negotiation and incorporation of new
forms than dyads where both learners contribute equally to the task (Storch, 2002).
This research has found that a mixture of high levels of intersubjectivity, as well as
shared scaffolding opportunities and shared task control to be most effective in creat-
ing learning opportunities, in terms of the number, and effectiveness of LREs.
In summary, though the classroom is the main source of learning for learners in
foreign language settings, the majority of research into FOF in task-based interaction
has been carried out in second language settings (Fujii & Mackey, 2009).2 The above
research has identified challenges to the implementation of FOF in foreign language
classrooms, including influences of students’ shared cultural and linguistic back-
grounds, as well as situational and interpersonal features of the interaction on the

2. See Sato (2011: 11–13) for an interesting discussion of the use of the terms ‘foreign’ and
‘second’ language contexts in SLA research.
 Paul J. Moore

effectiveness of FOF. It has also identified links between sociocognitive aspects of in-
teraction, including intersubjectivity, task control and pedagogic roles, and the effec-
tiveness of FOF. The current study aims to add to the small, but growing number of
studies into focus on form in task-based learner-learner interaction in EFL contexts.
While this type of qualitative research into incidental focus on form is less common
than the more analytic approaches taken above, Ellis (2010) notes the “obvious merit
in more holistic, qualitative approaches that document the situated nature of [correc-
tive feedback] and the complex discoursal events where learning takes place” (p. 346).

The study

As part of a larger longitudinal study in the context of a Japanese undergraduate EFL

classroom, this study investigated focus on form generated by learners in pair work
leading up to the performance of oral presentation tasks.
The research questions for the study were:
1. To what extent do learners focus on form?
2. What links are there between language-related episodes (LREs) in interaction and
subsequent individual task performance?
3. To what extent, if any, might contextual sociocognitive features of the interaction
influence the amount and effectiveness of LREs?

Context and participants

The study was undertaken in the Faculty of Humanities in a university in Japan. The
second-year Oral Presentation class met once a week for 25 weeks over two semesters
for 1.5 hours per week. Eight students participated in the study (six male and two fe-
male; aged 19 to 33). Most learners’ English language proficiency fell within the inter-
mediate range which the class was designed for (TOEFL 450–480) with the exception
of two learners: one (Yasuko) whose proficiency was below this range and one (Mina)
whose proficiency was above.3

Data collection and analysis

Data for this study came from the interaction surrounding two oral presentation tasks,
performed seven months apart: Oral Presentation 1 (OP1) and OP3.4 OP1 was a

3. On a mock abridged version of the TOEIC test, Yasuko scored 15 and Mina scored 67 out
of a possible 100, while other participants scored between 38 and 51.
4. Another task, OP2 (held in weeks 17–18), was excluded from the study as OP1 and OP3
provided substantial comparative data and there was a need to balance data collection require-
ments against the workloads of the study’s participants.
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

ten-minute (maximum) biographical presentation. Later presentations involved more

freedom with regard to topic choice, and were also progressively longer, with the final
presentation being from 15 to 20 minutes’ duration. Topics chosen for OP1 ranged
from historical (President John F. Kennedy; Helen Keller) to more contemporary
figures (John Lydon; Mariah Carey); while OP3 involved a broad range of topics
(e.g., history of boxing; slow and fast food).
Interaction, performance and reflection data were collected for OP1 and OP3. In-
teraction data were collected in the classroom using small analogue tape recorders in
weeks 4 and 5 (OP1) and weeks 19–21 (OP3). In weeks 6 (OP1) and 22 (OP3) each
dyad practised their oral presentations in front of one or two other dyads. These pre-
sentations were audio-recorded in three adjoining classrooms. Directly after recording
their OP3 peer-practice (week 22) learners listened to their recordings to identify er-
rors in grammar, lexis and pronunciation, in an adaptation of an FOF task described
by Mennim (2003). Oral presentations were recorded on both video and audiotape
during weeks 7–8 (OP1) and week 24 (OP3). Written self-evaluations were a part of
the pedagogic approach in the course, and they were designed to gain insights into the
learners’ experience of the task-based interaction with their partners and peers, as well
as their awareness of their own language performance.
Data analysis for the larger study involved both descriptive statistical analysis and
qualitative analysis of case studies. The first stage involved categorising, coding and
descriptive quantitative analysis of the interaction data (22 transcripts in total) accord-
ing to the emergent focus of learners, classified as procedural (e.g., talk about how to
complete the task), content-creation (talk about content matter that will comprise the
oral presentation) and off-task (talk unrelated to the task at hand), paralleling de
Guerrero & Villamil’s (1994) categories of about-task, on-task and off task interaction.
Inter-rater agreement on a sample of seven transcripts was 82%. The current study
draws on data from eight of the original 12 participants, comprising four focal learners
and their interlocutors, as they had the best attendance, and therefore the most com-
plete data sets. The four focal learners (Keita [male, intermediate, 19], Mina [female,
upper intermediate, late 20s], Yasuko [female, upper-elementary/lower-intermediate,
33] and Taro [male, intermediate, 19]) were chosen for comparative analysis as they
represented the maximum range of language proficiency, age, learners from both sex-
es, as well as the maximum amount of available interaction data. Their interlocutors
(see Table 1 below) were Nao (male, intermediate, 20), Ken (male, intermediate, 23),
Daito (male, intermediate, 21) and Tomo (male, intermediate, 19).
First, descriptive quantitative analysis of the four focal learners and their inter-
locutors is presented, in terms of counts of LREs in interaction, counts of LRE forms
used in oral presentation tasks, and supporting contextual data related to share of talk-
in-interaction and use of L1. This is followed by an in-depth quantitative and qualitative
analysis of one learner’s (Keita’s) interaction and performance with two interlocutors,
Nao, in OP1, and Ken, in OP3. Keita was chosen as both of his interlocutors were of a
similar proficiency and both had known him for the same length of time prior to
 Paul J. Moore

joining the class. Qualitative data included interaction transcripts, as well as reflection
and observation data; these were analysed to investigate the influence of sociocognitive
features of the interaction on the amount and effectiveness of LREs. The analysis draws
on previous analyses of intersubjectivity, pedagogic roles and task control (e.g, Donato,
1988; Storch 2001). Evidence of intersubjectivity (or ‘mutuality’, summarised by Storch,
2001) may include repetitions (though not exclusively, cf. Extracts 1 and 2 below),
requests, collaborative completions, phatic utterances (‘um’, ‘ah’, etc.) used in backchan-
neling, acknowledgement agreement and requests, and use of the third person pro-
noun (‘we’), each of which may signal a mutual interest in the ongoing dialogue. The
following extracts provide examples of varying degrees of intersubjectivity.
Extract 1. Keita and Nao (OP1 interaction1, turns 128–139)
128 KEITA: white has ... big power
129 NAO: big power ...
130 KEITA: but black is not
131 NAO: um white=
132 KEITA: =but mm
133 NAO: hate [black people]
134 KEITA: [ah yes yes yes] ... ah Gandhi . appealed ... non violence
135 NAO: ah: OK I see
136 KEITA: (laughs)
137 NAO: ah! ... good face . good good good good man
138 KEITA: mm::
139 NAO: ah: Gandhi!
While sharing background knowledge the learners signal their intersubjectivity by us-
ing repetition (turn 129), collaborative completions (turns 131–3), latching (turns
131–2), overlapping (turns 133–4), backchanneling (turns 134–5, 138) and exclama-
tion (turns 137 and 139). Extract 1 can be contrasted with Extract 2 where both learn-
ers initially appear to make little progress in negotiating a topic and developing an
effective working relationship.
Extract 2. Daito and Mina (OP3 interaction 1, turns 12–23)
12 DAITO: nani ga ii ka na (what shall we/you do? Literally “what’s good?”)
13 MINA: nani ga ii ka (what shall we do?) (7 seconds) ((checking notes))
14 DAITO: nanka suki na mono toka nai desu ka? (mm isn’t there something
you’re interested in?)
15 MINA: iya ... nanka (ah ... mm) (27 seconds) ((reading notes))
16 DAITO: nani ga ii ka na ... nanka nai desu ka shumi (what shall we do ... do you
have any hobbies) ... ...
17 MINA: shumi (hobbies) (laughs)
18 DAITO: watashi wa kore ga suki da (I like this)
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

19 MINA: iya mo nanka mo ... ... ippai ippai tte kanji yo (no mm ... this is too much
to take (literally “I’m full”)) (laughs) ... ... ... ... ... ...
20 DAITO: yaritai mono wa nai desu ka? (Don’t you have something you’d like to
do?)
21 MINA: nanka ... nai desu ka? (mm don’t you?)
22 DAITO: boku wa betsu ni: (I don’t mind/care) ... ... ...
23 MINA: watashi mo: (me too) (laughs) mou nanka (mm) (laughs)... kangaerare-
nai tte kanji nan dakedo (I can’t really think of anything) (laughs) ... ...
Here, repetition (turns 13 and 21) is a reflection of low intersubjectivity where both
learners appear to be relying on each other to make a choice, without offering one
themselves. There is little backchanneling or latching. The use of Japanese pronouns
(both implicit and explicit) for you and I (turns 14, 16, 18, 19, 20 and 21) also reflects
the lack of a collaborative approach at this stage, as do Mina’s expression of frustration
(turn 19) and both learners’ comments that they do not mind/care if the other decides
(turns 22–3).
Task control is identified by contribution to talk-in-interaction in terms of word
counts and control over the direction of the interaction (cf. Extract 13, for example).
Pedagogic roles are identified where one or both members of a dyad provide support
in terms of the features identified by Wood, Bruner and Ross (1976) including recruit-
ing interest, simplifying the task, and marking discrepancies between actual and ideal
performance.

Results

First, descriptive quantitative findings are presented to give an overall perspective on

the occurrence and variability in LRE and other contextual data in the study. Next, the
variability will be explored in more detail through an in-depth quantitative and quali-
tative analysis of Keita’s interaction and performance with his peers in OP1 and OP3.

Descriptive quantitative data

This section addresses the first research question regarding the extent of FOF in task-
based interaction. Table 1 below provides counts of LREs in interaction for the four fo-
cal learners in the study (Keita, Mina, Yasuko & Taro) and their interlocutors (Nao, Ken,
Daito & Tomo). It also provides counts of subsequent use of forms in OP1 and OP3
presentations, and whether the produced forms were target-like or non-target-like.
Starting with basic information about the focal learners’ interlocutors (column 2),
Table 1 next provides descriptive quantitative data regarding the focal learners’ contri-
bution to pair talk over the two interactions for OP1 and three interactions in OP3
(column 3; see Table 1 for word counts). Two learners were relatively consistent with
 Paul J. Moore

Table 1.  Summary of interaction and LRE quantitative data

Focal Partner Contribution to Individual Individual/ TL forms in

learners (partner’s talk-in-interaction use of L1 total LREsa OP1 & OP3
comparative presentations/
proficiency) Total formsb

Keita OP1 Nao 1348/2488ww 14% 16/21 8/10

(similar) 54%
OP3 Ken 870/4911ww 10% 12/31 5/8
(similar) 18%
Mina OP1 Ken 1114/4315ww 18% 2/20 0/1
(lower) 26%
OP3 Daito 1946/3681ww 77% 1/11 1/1
(much lower) 53%
Yasuko OP1 Taro 514/1580ww 28% 5/11 0/0
(higher) 33%
OP3 Tomo 936/2281ww 85% 9/10 2/2
(higher) 41%
Taro OP1 Yasuko 1049/1580ww 8% 6/11 3/4
(lower) 66.5%
OP3 Nao 2487/4346ww 3% 11/24 5/5
(similar) 57%
Total number of LREs 62/139
Total number of TL LRE forms used in oral presentations 24/31
a. Individual LREs are those where the problematic form was uttered by the focal learner. b. ‘Total forms’
includes counts of both TL and IL forms used in oral presentations after being focused on in LREs.

regard to amount of talk across partners (Yasuko & Taro) while the other two varied.
Similarly, as shown in Column 4, two learners remained relatively consistent in their
L1 use across partners (Keita & Taro), while two others varied greatly (Mina & Yasuko).
Column 5 shows how many LREs were focused on the language production of the fo-
cal learners as compared to the total number of LREs (i.e., including negotiation of
problematic forms produced by the focal learners’ interlocutors). Column 6 shows
how many of these LREs resulted in a TL form being used in OP1 or OP3 presenta-
tions, compared to the total number of negotiated forms used. In Keita’s case, for ex-
ample, Column 6 shows that he produced ten forms in his OP1 performance which
were previously negotiated in LREs in interaction. Of these, eight were target-like and
the remaining two were non-target-like. Although there was variability in the occur-
rence of learner-generated LREs in interaction, this table shows that they were not
common in the study; 31/62 (50%) LREs were relevant to the focal learners’ perfor-
mance, while 24 (77.5%) of these led to TL forms being used in the oral presentations.
In summary, the effectiveness of LREs in improving task performance in this study
was progressively limited by the following:
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

– there was not a large amount of attention to form in talk-in-interaction;

– not all LREs were relevant to the focal learners’ performance (i.e., they were either
focused on the other learners’ performance or not related to the content of the
performance); and
– not all relevant LREs resulted in TL forms being used in the oral presentations.
To provide further insight into, and potential reasons for the variability in findings
above, the next section presents an in-depth case study of one learner’s (Keita’s) inter-
action with his OP1 and OP3 interlocutors.

Case study: Sociocognitive influences on FOF

This section addresses the first and second research questions in more detail, by pre-
senting an in-depth analysis of Keita’s LREs in interaction and his use of related forms
in his subsequent oral presentation. It then addresses the third research question,
which asked whether contextual sociocognitive features of the interaction (such as the
negotiation of intersubjectivity, pedagogic roles and task control) influenced the
amount and effectiveness of FOF in interaction. After providing background informa-
tion on the three learners, each case begins with a discussion and exemplification of
the dyad’s engagement in LREs. This is followed by a discussion of relevant sociocogni-
tive aspects of the interaction. Both cases are compared in terms of their effectiveness
in focusing Keita (the focal learner) on form in interaction and in improving his lan-
guage production in the ensuing oral presentation.

Background information

Keita (male, 19), Nao (male, 20) and Ken (male, 23) were all second-year students who
were familiar with each other and the teacher, having completed a first-year English
language class with the same teacher. As noted earlier, all had a similar (intermediate)
level of English language proficiency; however, Ken was exceptional in that he spoke
with high fluency, low accuracy and drew on a large academic vocabulary. Ken was
notably confident and diligent in class, often asking questions and volunteering an-
swers, interpreting teacher utterances for other students, and providing more home-
work than requested. All had presented once in English before entering the class.

Keita and Nao (OP1)

Language-related episodes (LREs) in interaction

In Keita’s OP1 interaction with Nao there was a total of 21 LREs (17 lexical, 4 gram-
matical), of which Keita initiated 16 and resolved 10. There were 16 target language
(TL) resolutions. Excluding self-initiated – self-resolved LREs, there were 14
 Paul J. Moore

opportunities for uptake (incorporation of feedback) and 12 of these resulted in up-

take. With regard to the focus of the talk in which the LREs occurred, 16 occurred
during content-related talk, one in procedural talk and four were off-task. Twelve of
the forms arising in the LREs were used in the OP1 performance. Table 2 below pro-
vides LRE data related to Keita’s OP1 performance.
This table shows the use Keita made of the forms which were the focus of his LREs.
His partner’s resolutions resulted in uptake nine out of ten times. In addition, eight of
the ten LRE forms used by Keita in OP1 were TL forms. Extract 3 provides an example
of an LRE which involved a TL resolution.

Extract 3. Keita and Nao (OP1 interaction 2, turns 37–43)

37 KEITA: she’s blind ... blind ... ... she’s blind ... ... °mimi ga kikoenai te nandarou°
(how do you say she can’t hear)
38 NAO: deaf
39 KEITA: deaf?=
40 NAO: =D-E-A-F
41 KEITA: D-F deaf
42 NAO: mm D-E-A-F
43 KEITA: blind deaf ... me ga mienai: . mimi kikoenai: . hanase[nai] (eyes can’t see:
. ears can’t hear: . can’t talk)
Although this content-related talk was initiated by Keita, it was Nao who produced the
form in the assessed presentation. Interestingly, Nao, after providing the TL resolution
to Keita, chose both an IL form “she lost sound and light,” and a TL form, in the prop-
er noun “Wright-Humason School for the Deaf ” in both the peer-practice (week 6)
and assessed presentation (week 7).
Of the 16 LREs related to Keita’s language performance, there were six LREs where
the form was not subsequently used by Keita in OP1. One (hand out) arose in a proce-
dural frame and was not required for OP1; two (introduce and seriously ill) were re-
lated to Keita’s content-creation which was ultimately performed differently in the

Table 2.  LREs related to Keita’s OP1 language performance

Type of LREa Number TL Immediate Form used

of LREs resolutions uptake in OP1

Self-self  6  5 N/Ab   4 (3TL)

Self-other  7  5 7   3 (2TL)
Other-other  3  3 2   3 (3TL)
Totals 16 13 9 10 (8TL)
a. self-self: self-initiated self-resolved; other-other: other-initiated other-resolved; other-self: other-initiated
self-resolved (cf. discussion of Seedhouse [2001] above); b. uptake is irrelevant in self-resolved LREs/self-
correction.
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

OP1 presentation by his partner, Nao; one (Braille, negotiated twice) was used unsuc-
cessfully in the peer-practice and abandoned in OP1 (see Extracts 4–5 below); and one
(cook leader) was an unrelated off-task LRE (see Extract 6 below).
Extract 4. Keita and Nao (OP1 interaction 1, turns 50–63)
50 NAO: she use [she use]
51 KEITA: [she used]
52 NAO: Braille ... tenji da yo (braille, right)
53 KEITA: and ah- ... she . she started .. ten- (Brai-) Braille
54 NAO: really?
55 KEITA: yes yes yes yes
56 NAO: ah: (laughs) ... ... ... °Braille° ... maybe
57 KEITA: ee (ah) ... ... ... ((writing)) doesn’t speak ... ((coughs)) eeto (well) ((writ-
ing)) doesn’t see ee (ah) doesn’t read ... please spell
58 NAO: Braille is B-R-A: ... I-R-E
59 KEITA: B-R-A: . [I – R]
60 NAO: I – R] ah L-E . L-E . Braille=
61 KEITA: =Braille . Braille? ... OK ... ...
In this extract, which is part of a longer negotiation regarding the meaning and form
of the term ‘Braille’, Nao explains that Keller used Braille (turns 50–2), with Keita re-
peating his utterance, though this time in the correct tense (turn 51). Turns 53–56
represent unresolved negotiation of meaning, where Keita restates Nao’s prior utter-
ance as “she started Braille”, which Nao appears to interpret as ‘invented’ (turn 54).
Keita, possibly unintentionally, reinforces this misinterpretation with his animated af-
firmative response in turn 55. In turn 56 Nao apparently rejects this proposition im-
plicitly by saying “maybe” after considering ‘Braille’ in private speech. One interesting
indication of Keita’s engagement with the topic and intersubjectivity with Nao, is that
he takes notes as he speaks, even though this is not his topic (turn 57).
During the peer-practice, Keita attempted to explain the term in detail, conclud-
ing that he was unsure how to explain it in English:
Extract 5. Keita’s performance of the term ‘Braille’
Peer practice (week 6)
KEITA: Keita: Next ah:: Sullivan began to: reading skills .. that is . Braille a:h Braille
is ... e:h Japanese is tenji: ... a:h reading skill .. reading skill lesson is .. a:h Japanese
a:h dekoboko: (sticks out) .. dekoboko: paper a:h paper a:h written on:e .. like ah: ...
like a .. ah: suuji (number) like a ...figure one .. to .. ten an:d ... jun (order) .. order!
order one to ten an:d equal (pron.: eco:l) a:h all a:ll ... order is a:h one words .. do
you understand (laughs)? wakaranai na (don’t know/understand)
Here, Keita goes to great lengths to try to explain the term in English, codeswitching
when he thinks the audience will not understand terms (‘Braille’), and when he is
 Paul J. Moore

unable to produce an English term (dekoboko). In the assessed presentation, Keita

avoided the term ‘Braille’, and any related discussion completely, later commenting
that the term was “very difficult for me.”
Interestingly, all four off-task LREs involved IL resolutions, with the majority in-
volving ‘word coinage’ (cf. Færch & Kasper, 1983), where learners created an IL term
for a Japanese word they could not translate. Extract 6 below provides an example of
such negotiation.
Extract 6. Keita and Nao (OP1 interaction 1, turns 204–207)
204 KEITA: Mr. Shimada ... top?
205 NAO: top top chouricho (head chef) ... coo- coo- cook leader
206 KEITA: cook lea- cook leader
207 NAO: yes
In addition, there was one instance where a form which was the focus of an unresolved
off-task LRE was subsequently used in OP1, suggesting that there may be some trans-
fer between off-task interaction and task-based performance.
Extract 7. Keita and Nao (OP1 interaction 1, turns 124–127)
124 KEITA: hi:s long life
125 NAO: long life
126 KEITA: is ... ah! is ... to help ... ... ... for black . man . many .. man ... ah ... black
is ... .... ah: ... ... fighted ah: ... fight white ... ah in past time
127 NAO: mm ...
In this extract, the learners are discussing Mahatma Gandhi, who was eventually re-
jected as a presentation topic, in favour of Helen Keller. In turn 126, Keita moves be-
tween two IL forms in trying to produce the past tense of the verb ‘fight’. This extract
was coded as off-task, in that, although the learners were discussing content related to
Mahatma Gandhi, it was not related to the topic chosen by the learners; however, the
IL form ‘fighted’ did appear in a different context in the learners’ performance:
Extract 8. Keita and Nao’s performance of the IL form ‘fighted’
Peer practice (week 6)
KEITA: Helen was a ... pacifist, so she fighting ... fighted for the people
Assessed presentation (week 7)
NAO: Helen loves world peace so: Helen fighted with the (pacifists).

Sociocognitive aspects of the interaction

Keita and Nao distributed their talk relatively equally in their first interaction, with Keita
contributing 53% and Nao contributing 45% of the total talk. Although Keita’s contribu-
tion remained constant in their second week of interaction (55%), Nao’s dropped to
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

34%, as Keita produced more of the content and other interlocutors interrupted the in-
teraction (i.e., the teacher and members of other dyads). The learners’ interaction over
the two weeks was marked by consistently high intersubjectivity (cf. Extracts 1 and 4, for
example). They were able to successfully negotiate task control and pedagogic roles,
with both contributing to content-construction, offering each other assistance and keep-
ing each other focused on the various aspects of the task-based activity. Interestingly, the
shift to a focus on content-creation led to an expansion of the role of L1 use in this dyad
as the learners often resorted to L1 use in lexical LREs (see Extracts 3 and 4).
When asked whether they had a “good working relationship”, the responses were
as follows:
KEITA: Maybe yes, but my partner is temperament. I confused ...
NAO: Yes, we are friendly. Keita is a hard worker.
In his interview (week 19), however, Nao expressed frustration at not being able to
meet with Keita outside class:
NAO: Keita was very busy so we couldn’t practice ... I wanted to change another
partner (laughs)
In other words, although the learners exhibited a high level of intersubjectivity during
in-class interactions, both noted that they did not feel they had enough time together.
In addition, although they both expressed that they shared the workload equally, and
worked together successfully when they met, they also noted interpersonal difficulties
which may have hindered their progress.

Keita and Ken (OP3)

Language-related episodes (LREs) in interaction

In Keita’s OP3 interaction with Ken there was a total of 31 LREs (five lexical, 25 gram-
matical and one phonological). Twenty-one of these occurred before the peer-practice,
and ten arose while the learners listened to and reflected upon their peer-practice re-
cording (FOF task, week 22). All of the former were initiated and resolved by Ken.
Fifteen of these were self-initiated – self-resolved LREs and the remaining six involved
Ken recasting Keita’s forms. Of the latter, Keita initiated three LREs and resolved four.
There were 21 TL resolutions overall; 13/21 in the interaction before the peer-practice
and 8/10 in the discussion following the peer-practice. With regard to the focus of the
talk in which the LREs occurred, 20 occurred during content-related talk,5 and 11 oc-
curred during procedural talk.

5. Ten of these occurred during the FOF task held in week 22. Given that the aim of the task
was to edit content, these are not claimed to be ‘spontaneous’ here, and they are removed from
the count in Table 3, in order to provide a more accurate comparison between OP1 and OP3
interaction.
 Paul J. Moore

Table 2.  LREs related to Keita’s OP3 language performance

Type of LRE Number TL Immediate Form used

of LREs resolutions uptake in OP1

Self-self  2 1 N/A 2 (1TL)

Self-other  1 1 0 1 (IL)
Other-other  8 5 5 4 (3TL)
Other-self  1 1 N/A 1(TL)
Totals 12 8 5 8 (5TL)

Table 2 collates the data on LREs related to Keita’s OP3 performance.

Table 2 reveals that Keita’s OP3 interaction involved fewer chances to focus on his
own forms than his OP1 interaction, and that focus was less successful in improving
his performance. His partner’s resolutions resulted in uptake five out of nine times.
Five of the eight LRE forms used by Keita in OP3 were TL forms.
Extract 9 below is an example of one such instance, where Keita is outlining the
focus of his contribution to the oral presentation.
Extract 9. Keita and Ken (OP3 interaction 2, turns 30–42)
30 KEITA: ah: today’s Japan
31 KEN: mm?
32 KEITA: hayari (popular trend)
33 KEN: today’s fashion
34 KEITA: fashion today’s fashi[on]
35 KEN: [today’s] mo- today’s mo- today’s mode on Japan
36 KEITA: research for
37 KEN: mm
38 KEITA: today’s ah today’s Japan fashion
39 KEN: [mm]
40 KEITA: [and] ah:: ...
41 KEN: ah o- I’m sorry Japanese today’s fashion
42 KEITA: Japanese today’s fashion
43 KEN: OK
In turn 32, although the general term fasshon (fashion) in Japanese is an English loan
word, Keita uses the term hayari, which carries the connotation of a current boom.
Ken (turn 33) correctly notes that fashion is also used in this way, but then attempts to
use more complex language (turn 35). Keita then uptakes this term and attempts to
create a nominal group (turn 38), which Ken attempts to correct (turn 42), using an-
other IL form. The term ‘fashion’ was used correctly by both learners several times in
OP3, though Ken used the IL terms ‘Japanese mode’ and ‘fashion mode’ (cf. turn 35).
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

Extract 10. Keita and Ken’s OP3 performance of the term ‘fashion’
Peer-practice
KEITA: Ah: today: we are ... ah: we are going to talk about ... fashionable clothes-
es ... which are: especially ah: ... especially young people ... fashion or people fash-
ion ... in modern to now ...
Assessed presentation
KEITA: Today: we are ... ah: we are going to talk about ... fashionable clothes
(pron.: kl6~δez) ... ah: which had especially used to wear: ah: young people or:
people in modern to now ...
Of the 12 LREs related to Keita’s language performance, there were four LREs where
the form was not used by Keita in OP3. As with OP1, this was mostly because they oc-
curred in non-content-related talk, which was ultimately irrelevant to the oral presen-
tation. There was also one instance where Keita’s language performance moved in the
wrong direction (from TL-IL pronunciation of the term ‘clothes’) as a result of Ken’s
corrective feedback.
Extract 11. Keita and Ken (OP3 interaction 3, turns 30–34)
30 KEITA: ... young ... young clothes (TL pron.: kl6~δz)... young ah [young gen-
eration]
31 KEN: [young cloth-es?] (IL pron.: kl6~δez) mm =
33 KEITA: = cloth-es (IL pron.: kl6~δez) ...
34 KEN: mm
As a direct result of Ken’s recast, Keita changes his pronunciation of the term ‘clothes’.
(turns 31 and 33). After this LRE, both Keita and Ken employed the mispronunciation,
which persisted, despite later correction by the teacher, until OP3 (see Extract 10
above), where he used the IL form 15 times.
The contrast between the effectiveness of Keita’s OP1 and OP3 LREs becomes
clearer if the analysis is limited to interaction before the peer-practice, as noted in
Table 3 below.
Keita’s OP1 interaction, in comparison to his OP3 interaction, had far more and
varied LREs focused on his own linguistic forms. This included far more TL resolutions,
and ultimately resulted in eight improvements to his linguistic performance, as opposed
to only one improvement (fashion) in OP3. Interestingly, the IL form which was used in
Keita’s performance was the IL pronunciation of the term clothes reported above.

Table 3.  LREs related to Keita’s performance (OP1 and OP3; prior to peer-practice)

Number of LREs TL resolutions Immediate uptake Form used in OP1

OP1 16/21 (76%) 13 (81%) 9/10 (90%) 10 (8TL)

OP3   6/21 (29%)   3 (50%)   3/6 (50%)   2 (1TL)
 Paul J. Moore

Sociocognitive aspects of the interaction

In contrast to his interaction with Nao in OP1, Keita’s contribution to talk-in-interac-

tion in OP3 was limited to a range of 17–28% over the three interactions with Ken,
while Ken’s ranged from 64–79%. In addition, in his OP1 interaction, Keita’s use of
Japanese L1 increased from 5% to 21%, whereas in his interaction with Ken it de-
creased from 21% to 7%.
As with Keita’s OP1 interactions, the learners displayed a high level of intersubjec-
tivity, with overlapping speech, laughter and exclamations showing that they were en-
gaged with each other as well as the task. There were three differences which appeared
to impact on the interaction, resulting in a relationship where Ken both controlled the
direction of the task and scaffolded Keita’s involvement. First, Keita noted that he per-
ceived Ken’s English proficiency as superior to his own; Extract 12 below shows him
praising Ken’s skills:
Extract 12. Keita and Ken (OP3 interaction 1, turns 104–5)
104 KEITA: Ken san datte reberu takain da mon (Ken, your level [of English] is
high)
105 KEN: no no no no (laughs) ... OK ...
Second, although Ken was generally careful to include Keita in discussions about the
direction of the task-based interaction, he remained in control of the task, as can be
seen in Extract 13 below:
Extract 13. Keita and Ken (OP3 interaction 2, turns 195–202)
195 KEN: °(laughs)° ... mm it depends on you ((i.e., it’s up to you))
196 KEITA: oh:: OK
197 KEN: mm ... ... ... but please choose the Western brand
198 KEITA: Western brand
199 KEN: and if you: ha- if you don’t mind .. please choose the one- one of the ..
Japanese brand ...
200 KEITA: oh:
201 KEN: and compare it
202 KEITA: OK
Ken’s comment in turn 195 that Keita can make his own decision regarding his sub-
topics, contrasts strongly with his subsequent demands (turns 197–201). Keita’s back-
channeling responses suggest that he is prepared to accept Ken’s control. It should be
noted that this extract is based on an earlier discussion of both Japanese and western
fashion, which was partly initiated by Keita.
Third, there are instances where Ken encourages Keita with comments such as
“don’t be depressed” (interaction 3, turn 219) and “never give up” (interaction 3, turn
227). Keita and Ken also both noted in their self-evaluations that they had a good
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

working relationship. Keita responded, enthusiastically, that “Ken taught in detail, he

is kind [to] me.” Interestingly, he also noted that he learned “pronunciation” from Ken.
Ken noted that Keita learned from him “that he had mistaken in grammar, so he fixed
it.” When asked if he believed Ken had learnt anything from him, Keita’s brief response
was “no ... .”
Overall, Keita’s interaction with Ken was marked by Ken’s control over the task-
based activity, combined with his attempts at scaffolding Keita’s involvement. Interest-
ingly, this scaffolding was ineffective in terms of focusing Keita on form, at least with
regard to the incidence and effectiveness of LREs.

Discussion

This study, in support of the studies reported earlier, has found that incidental learner-
generated focus on form was not common in learner-learner interaction leading up to
the performance of an oral task, that the majority of LREs were lexical, and that there
was wide variability across dyads, and across partners over time, in terms of the number
and focus of LREs they engaged in. In addition, the effectiveness of LREs progressively
diminished in the study, depending on whether there was uptake involved, whether the
feedback and/or uptake involved TL forms, whether the LRE occurred in talk which
was relevant to task performance, and whether the learner/s receiving corrective feed-
back (or collaboratively constructing new knowledge) ultimately used the forms fo-
cused on in the task performance. The qualitative analysis of the case study provided
further insights into the impact of contextual sociocognitive features of interaction
(such as the negotiation of intersubjectivity, pedagogic roles and task control [cf. Storch,
2002]), on the occurrence and effectiveness of learner-generated focus on form.
Learner perceptions appeared to play various roles in this study. Keita and Ken’s
shared belief that Ken’s English language proficiency was higher than his own may
have made Keita reluctant to provide corrective feedback to Ken (cf. Philp et al., 2010),
as well as making him more likely to accept Ken’s authority, even when this involved
‘mislearning’ of the term ‘clothes’ (cf. Fujii & Mackey, 2009). It has been noted in the
literature that such potentially negative effects of L-L interaction are rare. They must be
balanced against the more positive findings, such as Zhao & Bitchener’s (2007) finding
of more pre-emptive FOF in L-L interaction and Fujii & Mackey’s (2009) finding that
learners’ negotiation of form involved rich positive evidence in terms of reformula-
tions, both of which may indicate reduced pressure for learners in L-L as opposed to
T-L interaction (Zhao & Bitchener, 2007). Nevertheless, the issue remains that the
range of IL forms present in L-L interaction can make it unclear to learners which
forms are being focused on, and whether or not LRE resolutions involve improvement
on the original utterance (Fujii & Mackey, 2009).
McDonough’s (2004) finding that learners placed little value on feedback from
other learners, even in the face of evidence of improvement, is also of interest here.
 Paul J. Moore

While Keita, for apparent social and situational reasons, found his interaction with
Nao more problematic than that with Ken, in terms of FOF in interaction the opposite
was true. The analysis of Nao and Keita’s OP1 interaction revealed a dynamic and
somewhat effective relationship where the learners were mutually supportive (cf. Storch,
2002, for example), while Ken (in OP3) appeared to control Keita’s involvement and
focus on his own forms at the expense of Keita’s. Further research is needed to investi-
gate such mismatches between learners’ and researchers’ or teachers’ perspectives on
the effectiveness of interaction, and whether these may be better aligned. This is espe-
cially important given the contention that learners’ perception of task goals (Williams,
1999), and their perception of researchers’ or teachers’ expectations (Philp et al., 2010),
guide task-based activity.
The use of L1 was another source of variability in the study. Philp et al. (2010) note
that their learners used their L1 in negotiating one third of their LREs. While it is
logical that learners with a shared L1 background use their L1 in lexical LREs (as in
Extracts 3–6 above), it is not simply the case that engagement in LREs leads to in-
creased L1 use.
Pedagogically, the study highlights the fact that learners may need to be oriented
to, or trained in focus on form for it to be effective in influencing task performance and
learning. Kim and McDonough (2011) provide the most recent support for this strat-
egy, finding that pre-task modelling resulted in more LREs and more correct resolu-
tions for learners in a Korean EFL context. The fact that the learners were engaged in
LREs in procedural and off-task talk also shows that they were engaged in both com-
municative and pedagogic aspects of the interaction. While this is a positive finding
for classroom learning, it is an issue for task design, assessment and research, in that
learners may need more coaxing to direct their efforts toward improving their on-task
interaction, performance and learning. Researchers have investigated such issues as
‘missed opportunities’ (Storch, 2002 p. 138), where more negotiation of particular
forms may have improved performance, or dealt with issues of avoidance (e.g., Fujii &
Mackey, 2009). In Keita’s interaction with Nao, for example (cf. the discussion of Ex-
tract 3), some further discussion regarding forms related to the terms ‘deaf ’ and ‘sight’
might have improved their task performance, as such forms could be seen as essential
to their task performance.

Conclusion

While learner-generated focus on form is commonly found to have limited influence

on task performance, this study provides support for its further investigation. It has
provided evidence that learner-learner interaction can influence subsequent individu-
al language performance. Pedagogically, the study highlights the potential of attempting
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

to refine learners’ focus on form to encourage more on-task negotiation, including

forms which may be essential to task performance.
While the analysis of Keita’s interaction with different interlocutors over time pro-
vides insights into features of interaction that influence the effectiveness of FOF in
task-based interaction, the limitation of the study is that this is one example of such
influences. This study is best seen as a case from one context, providing insights into
contextual influences on focus on form, which might inspire future studies. Further
research is necessary to investigate whether the effectiveness of focus on form can be
improved by, for example, raising learners’ awareness of issues influencing the inci-
dence and effectiveness of their own focus on form, and to investigate the role of learn-
er (and teacher) perception in focus on form in other contexts.

Acknowledgements

I would like to thank the editors and series editors for their comprehensive and forma-
tive feedback on earlier drafts of this chapter. I would also like to thank Anne Burns
and Geoff Brindley for their support throughout the larger study on which this chapter
is based.

References

Brooks, F. & Donato, R. (1994). Vygotskyan approaches to understanding foreign language

learner discourse during communicative tasks. Hispania, 77, 262–74.
de Guerrero, M. & Villamil, O. (1994). Social cognitive dimensions of interaction in L2 peer
revision. The Modern Language Journal, 78, 484–496.
Donato, R. (1988). Beyond group: A psycholinguistic rationale for collective activity in second
language learning. Unpublished PhD dissertation, University of Delaware, Newark.
Donato, R. (1994). Collective scaffolding in second language learning. In J. Lantolf & G. Appel
(Eds.). Vygotskian approaches to second language research (pp. 33–56). Norwood, NJ:
Ablex.
Doughty, C. & Williams J. (Eds.) (1998). Focus on form in classroom SLA. Cambridge: Cambridge
University Press.
Ellis, R. (Ed.) (2001). Form-focused instruction and second language learning. Oxford: Blackwell.
Ellis, R. (2008).The study of second language acquisition (2nd ed.). Oxford: Oxford University
Press.
Ellis, R. (2010). Epilogue: A framework for investigative oral and written corrective feedback.
Studies in Second Language Acquisition, 32, 335–349.
Ellis, R., Basturkmen, H. & Loewen, S. (2001). Learner uptake in communicative ESL lessons,
Language Learning, 51, 281–318.
Færch, C. & Kasper, G. (Eds.) (1983). Strategies in Interlanguage Communication. London:
Longman.
 Paul J. Moore

Fotos, S. & Nassaji, H. (Eds.) (2007). Form-focused instruction and teacher education: Studies in
honour of Rod Ellis. Oxford: Oxford University Press.
Fujii, A. & Mackey, A. (2009). Interactional feedback in learner-learner interactions in a task-
based EFL classroom. International Review of Applied Linguistics, 47, 267–301.
Kim, Y. & McDonough, K. (2011). Using pretask modelling to encourage collaborative learning
opportunities. Language Teaching Research, 15, 183–199.
Loschky, L. & Bley-Vroman R. (1993). Grammar and task-based methodology. In G. Crookes &
S. Gass (Eds.), Tasks and language learning: Integrating theory and practice (pp. 123–67).
Clevedon, UK: Multilingual Matters.
Long, M. (1991). Focus on form: A design feature in language teaching methodology. In K. de
Bot, R. Ginsberg & C. Kramsch (eds.), Foreign language research in cross-cultural perspective
(pp. 39–52). Amsterdam: John Benjamins.
Long, M. (2007). Problems in SLA. New York, NY: Lawrence Erlbaum Associates.
Lyster, R. & Ranta, L. (1997). Corrective feedback and learner uptake: Negotiation of form in
communicative classrooms. Studies in Second Language Acquisition, 19, 37–66.
Mackey, A. (Ed.) (2007). Conversational interaction in second language acquisition: A collection
of empirical studies. Oxford: Oxford University Press.
McDonough, K. (2004). Learner-learner interaction during pair and small group activities in a
Thai EFL context. System, 32, 207–224.
Mennim, P. (2003). Rehearsed oral output and reactive focus on form. ELT Journal, 57, 130–138.
Mennim, P. (2005). Noticing tasks in a university EFL presentation course in Japan: Their effects
on oral output. Unpublished PhD Dissertation. University of Edinburgh, UK.
Moore, P. (in press). An emergent perspective on the use of the first language in the EFL class-
room. Modern Language Journal.
Philp, J., Walter, S. & Basturkmen, H. (2010). Peer interaction in the foreign language classroom:
What factors foster a focus on form. Language Awareness, 19, 261–279.
Pica, T. & Doughty, C. (1985). Input and interaction in the communicative language classroom:
A comparison of teacher-fronted and group activities. In S. Gass & C. Madden (Eds.), Input
in second language acquisition (pp. 115–32). Rowley, MA: Newbury House.
Sato, M. (2011). Constitution of form-orientation: Contributions of context and explicit knowl-
edge to learning from recasts. The Canadian Journal of Applied Linguistics, 14, 1–28.
Schegloff, E., Jefferson, G. &. Sacks, H. (1977). The preference for self-correction in the organi-
zation of repair in conversation. Language, 53, 361–82.
Seedhouse, P. (2001). The case of the missing “no”: The relationship between pedagogy and in-
teraction. Language Learning, 51, Supplement 1, 347–385.
Storch, N. (2001). An investigation into the nature of pair work in an ESL classroom and its effect
on grammatical development. Unpublished PhD dissertation. University of Melbourne,
Australia.
Storch, N. (2002). Patterns of interaction in ESL pair work. Language Learning, 52, 119–158.
Swain, M. (1998). Focus on form through conscious reflection. In C. Doughty & J. Williams
(eds.), Focus on Form in classroom SLA (pp. 64–81). Cambridge: Cambridge University
Press.
Swain, M. & Lapkin, S. (1998). Interaction and second language learning: two adolescent French
immersion students working together. The Modern Language Journal, 82, 320–37.
Swain, M. & Lapkin, S. (2000). Task-based second language learning: the uses of the first lan-
guage. Language Teaching Research, 4, 51–274.
 Chapter 8.  Incidental learner-generated focus on form in a task-based EFL classroom 

Wells, G. (1999). Using L1 to master L2: A response to Antón and DiCamilla”s “Socio-cognitive
functions of L1 collaborative interaction in the L2 classroom.’ The Modern Language Jour-
nal, 83, 248–254.
Williams, J., (1999). Learner-generated attention to form. Language Learning 49, 583–625.
Wood, D., Bruner, J. & Ross, G. (1976). The role of tutoring in problem-solving. Journal of Child
Psychology and Psychiatry, 17, 89–100.
Zhao, S. & Bitchener, J. (2007). Incidental focus on form in teacher–learner and learner–learner
interactions. System, 35, 431–447.
 chapter 9

Qualitative differences in novice teachers’

enactment of task-based language teaching
in Hong Kong primary classrooms

Sui Ping (Shirley) Chan

Whole Person Development Institute, Hong Kong

Diverse aspects of task-based language teaching (TBLT) and learning have

been researched for more than two decades. While there is much theoretical
discussion concerning the definition of tasks and how tasks should be designed
and implemented, there is as yet only limited understanding of how TBLT is
actually enacted in authentic classrooms. This study investigated how TBLT was
enacted in primary ESL classrooms in Hong Kong, focusing on the way teachers
manage the linguistic, cognitive, and interactional demands of tasks. Adopting a
multiple-case study approach, the data set included a total of 20 lessons taught by
four teachers on the same topic, individual lesson plans and teaching materials,
as well as interviews with these teachers. Tasks completed by the students were
also collected and analysed. Findings of the study showed that teachers differed
in enacting TBLT in their classrooms along six dimensions: (1) strategic use
of visual support to manage task demands; (2) contextualizing input to make
connections between old and new knowledge; (3) simultaneous attention to task
demands for progression in complexity; (4) provision of scaffolding through task
sequencing and adjustment of task variables; (5) creating conditions for noticing
form and salient features; and (6) creating conditions for restructuring to occur.
These findings imply that what is most important in shaping learning in the
TBLT classroom is not the task per se, but rather the interweaving of pedagogic
strategies at various levels of complexity as teachers respond to students’ needs
in the immediacy of the classroom environment.

Introduction and background

Task-based language teaching (TBLT) has generated worldwide interest over the past
25 years. While there is much discussion concerning the definition of tasks and how
tasks should be designed and implemented in the classroom, there is as yet limited un-
derstanding of how TBLT is actually enacted in the classroom (Samuda, 2007). Studies
 Sui Ping (Shirley) Chan

of task-based language instruction and second language acquisition suggest that it is

important for teachers to be aware of and able to manage different types of task demands
to facilitate language learning (e.g., Robinson, 2001; Skehan, 1998, 2007; Samuda, 2001,
2007). This study investigated how TBLT was enacted in primary ESL classrooms in
Hong Kong. It attempted to determine whether there were qualitative differences be-
tween four primary ESL teachers when they introduced TBLT in their classes, and in
what ways and to what extent they afforded different learning experiences.
Language education in Hong Kong has undergone significant changes since the
introduction of the Target-Oriented Curriculum (TOC)1 in 1995. To sustain and fur-
ther refine the curriculum changes as stipulated in the new curriculum, a series of of-
ficial documents and guidelines has been produced. The notion of task-based language
teaching has become a subject of contemporary interest in Hong Kong because it was
officially recommended in TOC. Teachers are expected to design teaching and learn-
ing activities that “help learners to achieve communicative competence, supported by
the development of linguistic competence and the mastery of skills and language de-
velopment strategies” (Candlin, 2001, p. 233). While publishers have produced “task-
based” textbooks for teachers to use, teachers in the study reported that they still have
to adapt the textbook materials to suit the needs of their students.
Unlike the more traditional approaches to language teaching, which adhere to a
structural view of language with explicit focus on teaching grammar, the task-based
approach to language teaching and learning in Hong Kong emphasizes the communi-
cative nature of language, with a focus not primarily on grammatical and structural
features of the language but on the functional and communicative aspects of the target
language (Richards & Rodgers, 2001). In implementing the new English Language
curriculum framework, it is recommended that a task-based approach be used be-
cause “Tasks are purposeful and contextualized activities in which learners draw to-
gether a range of elements in their framework of knowledge and skills to fulfill the task
set” (CDC2, 2004, p. 117).
The Curriculum Guide further suggests that to achieve communicative compe-
tence, more interactive and communicative use of language through social interaction
should be stressed. The intended curriculum adopts an interactional view of language
learning that emphasizes the interactional nature of learning between the learner and a
knowledgeable interlocutor. However, when it comes to pedagogical decisions regard-
ing the actual implementation of the task-based approach to language teaching and

1. Target-oriented Curriculum (TOC) was a modified version of Targets and Target-related

Assessment Curriculum (TTRA) which was modeled on the new National Curriculum in Brit-
ain introduced during the period 1989 to 1993. TOC was officially introduced in Primary 1
classes in 1995.
2. CDC is the abbreviation of ‘Curriculum Development Council” which is a free-standing
advisory body appointed by the Chief Executive of the Hong Kong Special Administrative Re-
gion to give advice to the Government on matters relating to curriculum development for the
local school system.
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

learning, there are numerous possibilities that the intended curriculum leaves open to
teachers (Candlin, 2001, p. 241). Hence, the aim of this study is to investigate the qual-
itative differences in teachers’ enactment of TBLT in Hong Kong primary schools.

Task-based language teaching in Hong Kong

TBLT attracted attention in Hong Kong when the Target-oriented Curriculum (TOC)
was introduced in 1995. The TOC, a large-scale, territory-wide curriculum renewal for
all school subjects, encouraged teachers not to follow a rigid pre-determined language
syllabus in planning their teaching (CDC, 1999). Over the past decade, though the
language curriculum has been subject to revision, TBLT remains the recommended
teaching and learning approach in the latest English Language Curriculum for both
the primary and secondary levels (CDC, 1999; 2004; 2007). Consequently, teachers are
expected to choose or design tasks that best suit their learners’ needs and abilities. In
other words, the exercise of designing, selecting, sequencing, and grading tasks, which
had traditionally been considered the province of syllabus designers, now falls within
the purview of the classroom language teacher. However, in actual implementation,
Carless (2004) has found that the teachers in his study were practising only a weak
form of TBLT with a focus on gaining control over individual skills (i.e., pronuncia-
tion, grammar, vocabulary) before applying them into communicative tasks.
Numerous studies (Carless, 2003, 2004; Morris, Adamson, Au, K. Chan, W. Chan,
Ko, Lai, Lo, Morris, Ng, Wong, & Wong, 1996; Zhang, 2005), together with our own
experience of related in-service courses, suggest that many teachers in Hong Kong find
the concept of TBLT difficult to grasp. Even though some claim to be implementing
TBLT in their classrooms, they still find it difficult to conceptualize what TBLT stands
for. The need for further empirical studies of task design and implementation is, there-
fore, central to the current TBLT research agenda (Samuda, 2007; Van den Branden,
2007). Designing a task-based curriculum involves making decisions about task selec-
tion, grading, and the choice of specific methodological procedures for implementing
each task. This requires teachers to build up competence, develop task-based work-
plans, and implement them (Samuda, 2007).

Task complexity/difficulty in task design and task implementation

Task design and task implementation require teachers to possess a sound understand-
ing of the demands of tasks and the ability to relate these to students’ skills and needs.
The challenge of understanding the cognitive, communicative, and linguistic complex-
ities of tasks, which are often implicit rather than explicit, may render it difficult
for teachers to make informed choices. Besides addressing the issue of linguistic
 Sui Ping (Shirley) Chan

complexity, which usually receives the greatest attention in the learning context, the
cognitive demands together with the performance conditions of the setting in which
the task is undertaken represent vital dimensions that contribute to task complexity,
and ultimately to language learning via tasks. Recent studies on task complexity indi-
cate that a number of different factors impact task complexity (Ellis, 2003; Nunan,
2004; Robinson, 2001, 2007; Skehan, 2007). However, while the literature enumerates
multiple task factors affecting performance, little empirical evidence exists on how
these may relate to the teaching of young learners. In Hong Kong, the teaching of young
learners is particularly important because English is taught as early as the age of 3.
In this study, the notion of task refers to pedagogic tasks that teachers use in the
classroom. These will progressively approximate to target tasks which are real-world
related (Long & Crookes, 1993). In order to help students complete these tasks, the
teacher needs to assist students in developing specific skills through their participation
in certain preparatory activities. In this study these activities will be referred to as
pedagogic tasks. The choice of pedagogic tasks and the sequence in which they are to
be performed will have significant consequences for the role of the learner in assimi-
lating the language encountered in the classroom (Robinson, 2001).
Previous studies on task complexity identify three main dimensions of complexi-
ty: code complexity, cognitive complexity, and communicative stress (Candlin, 1987;
Ellis, 2003; Skehan, 1998). However, the main question that arises here is: what types
of demands will be imposed on learners in the process of learning, knowing that task
demands affect learners’ conscious and unconscious decisions in channelling their at-
tentional resources to achieve a particular language learning goal? I will discuss three
main types of demands embedded in the task, namely linguistic demands, cognitive
demands, and interactional demands. For convenience, I will analyse these separately,
but it does not mean that they stand alone. Rather, they are inter-related and should be
dealt with as an integrated whole.
In the field of teaching young learners, whether a linguistic code is difficult for the
learners or not is largely determined by the way the code is presented to them (White,
1998; Robinson, 2001). The nature of input provided and the outcomes required are
also important elements contributing to linguistic demands.
The cognitive demands on learners are determined by the familiarity of the topic,
discourse genre, and task type. When learners are familiar with the topic, task type, or
genre, they can release attentional resources to focus on language. In terms of cogni-
tive processing of the target language input, learners need to make sense of the infor-
mation to be processed. This process involves the employment of a set of skills ranging
from simple to complex (Ellis, 2003). Thus, it is more manageable to begin with simple
cognitive processes such as classifying, grouping, or ordering concrete objects than
reasoning or opinion giving, which require working with concepts. Studies have shown
that different types of information pose different levels of challenge to the learners
(Brown, Anderson, Shillcock, & Yule, 1984; Skehan, 1998; Robinson, 2001). The struc-
ture, clarity, and sufficiency of information given are also crucial in determining the
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

cognitive processing load which contribute to the cognitive demands on learners

(Ellis, 2003; Robinson, 2001; Skehan, 1998).
When performing tasks, learners will engage in interaction with the target audi-
ence or readers. Hence, the interactional relationships, interactional requirements,
interactional goals, and outcome options will contribute to task demands (Ellis, 2003;
Pica, Kanagy, & Falodun, 1993). It should be noted the three types of demand are not
independent of one and another. By adjusting one type of demand, the teacher can
increase or reduce the demand in another type. The following section will illustrate
how to manage demands: linguistic demands, cognitive demands, and interactional
demands.

Managing demands

To manage the demands of a task in actual implementation, we need to consider the

pedagogical decisions that teachers make to facilitate language learning. First, we can
examine what teachers can do to make language accessible to learners. For instance,
the provision of visual support can free up the attention that is needed by students to
retain the information in memory. Sequencing linguistic complexity by adjusting the
task variables in the input data or adjusting the task outcome is another form of sup-
port. The modality of the task, that is whether a task is spoken or written, is another
form of management of linguistic demands The teacher has a range of options to make
the language more or less accessible to the learners. Such conscious pedagogical deci-
sions can increase or reduce the linguistic demands on the learners.
The management of cognitive demands requires conscious pedagogical decisions
by the teacher to establish topic familiarity and create conditions for noticing and re-
structuring to occur. Topic familiarity can be established by activating background
knowledge, providing foregrounding, and building in progression in cognitive com-
plexity from simple to complex, from concrete to abstract, and so on. In addition, the
cognitive processing load can be made manageable if new knowledge is presented
from concrete to abstract or from familiar to unfamiliar, by relating it to students’
knowledge of the world or prior experience of the information (or discourse) type. The
use of strategies which engage learners in cognitive processes such as guessing from
context, studying the word parts, making predictions, and forming associations will
help bring about noticing (Schmidt, 1990). Restructuring is another important process
which enables the learner to produce progressively more complex language. For re-
structuring to occur, conditions should be created to enable learners to notice the gap
in their existing resources by being exposed to and engaging in the use of the target
language in different contexts (Ellis, 1997, 2002; Lee & Van Patten, 1995; McLaughlin,
1990; Richards, 2002; Robinson, 2001; Samuda, 2001; Schmidt, 1990; Skehan, 1998).
The teacher should create conditions to make learners more ambitious in what they try
to say and do with the target language.
 Sui Ping (Shirley) Chan

The management of interaction demand refers to the conscious pedagogical deci-

sions made by the teacher to maximise the opportunities for pupils to take up an active
role in classroom exchanges. This process involves the use of strategies to create the
conditions for moving away from teacher-led to pupil-led communication and from
producing restricted to open responses. These strategies include the modification of the
interactional features to render the interactional demand more or less demanding. For
instance, the teacher could begin with one-way communication and then scaffold the
interaction to move towards two-way or multiple channels/ways of communication.
As mentioned earlier, it is important to remember that while these three strands
are discussed separately, they should be handled in an integrated way in actual class-
room situations.

Research question

The literature reviewed above on types of task demands sets directions for teachers to
consider when making decisions on task design and implementation. Based on the
purpose of the study and these considerations, the present investigation sought to an-
swer the following research question:
How do novice teachers manage the task demands in their enactment of task-
based language teaching in Hong Kong primary classrooms?
The study sought to answer this question with reference to the way teachers vary
in their task design and implementation by way of addressing the requirements of task
demands in their teaching practice.

Method and data collection

An initial data collection framework was derived largely from Skehan (1998), but with
modifications and extensions drawn from Ellis (2003) and other work in the literature
(Candlin, 1987; Nunan, 1989; Willis, 1996). The preliminary analysis of the data gath-
ered through the initial instrument was then conducted and the results used to further
develop the framework in order to make it more powerful and capable of greater and
more in-depth analysis. The analytical framework for task demands is presented in
Table 1.
This analytical framework on task demand was adopted in this study to analyse
how these demands are actually managed by teachers at pre-, while, and post- task
stages. As the three types of demands are intertwined in contributing to the overall
task complexity, it is of equal importance to examine how the teachers in this study
(or teachers in general) organize the pedagogical activities and tasks in the way that
serve as scaffolding in task implementation.
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

Table 1.  Task demand and its components

Task demand Task components

Linguistic demands Nature of input

–  With or without visual support
–  Context dependency
–  Familiarity of information
–  Frequency of occurrence and recycling
Nature of expected outcome
–  Medium
–  Scope
–  Complexity
Cognitive demands Cognitive familiarity
–  familiarity of topic, discourse genre and task type
Cognitive processing
–  cognitive processes
–  single or dual demand
Information type and information structure
–  amount of information
–  type
–  structure
–  clarity
Interactional demands Communicative stress
–  Interactional relationship- one-way or two-way
–  Interactional requirement- required vs. optional
Task characteristics
–  Goal orientation- convergent vs. divergent
–  Outcome options- closed vs. open

Teacher participants

The four teachers in this study, Kathy, Linda, Clare, and Maria (pseudonyms), were
participants in an in-service teacher development course offered by the Hong Kong
Institute of Education. The programme was a four-month, full-time, credit-bearing,
in-service course for teachers of English in primary schools. It aims to keep teachers
abreast of the latest developments in language teaching in general and to invite them
to apply what they have learnt in the classroom with a view to enhancing their teaching
effectiveness. The four teachers hold qualified teacher status and are teaching English
in primary schools. Clare and Maria each had thirty years experience at the time of
data collection; Kathy had twenty years experience while Linda had five. All could,
therefore, be classified as experienced teachers for the purposes of this study. Their
 Sui Ping (Shirley) Chan

academic qualifications were not considered as a criterion for selection since there ap-
pears to be no direct relationship between teaching effectiveness and academic quali-
fications (Tsui, 2003).
However, it is worth mentioning that Kathy was undertaking studies for a part-
time degree in English Language Teaching at the time of data collection. Clare held a
Bachelor’s degree in Special Education. Maria was a degree holder in Education. Linda
held a teacher’s certificate. It seems that the other three teachers were academically
more qualified than Linda, as she had not started her degree study at the time of data
collection. Linda, however, graduated from the four-month in-service training courses
with better academic results than Maria. In terms of experience in adopting a task-
based approach to teaching, the four teachers stated in the interviews that they had
never or seldom tried out TBLT in their classroom although the TOC had been intro-
duced since 1995. They all considered themselves very textbook-bound with a main
focus on teaching grammar to their students.
The four teachers chose the same topic “Weather” when conducting their school-
based investigation projects on TBLT with their pupils (aged between7–9) at lower
primary levels. The schools from which the teachers came were relatively homoge-
neous in terms of the family and academic backgrounds of the students, in that all
schools were located in public housing estates. Since the students mainly came from
low socio-economic family backgrounds, the parents had limited knowledge of Eng-
lish, and generally limited support was provided at home.

Tasks

The following is a summary of the tasks that the teachers designed and implemented
as part of the programme requirements in trying out TBLT in their school-based in-
vestigation projects. Details of the overall task design of the four cases are presented in
Appendix A.
Kathy designed a unit on ‘What’s the weather like?’, the final task of which re-
quired children to work in pairs to produce a small booklet on the ‘Four seasons’ in the
form of simple rhymes. They were required to write about the change of weather con-
ditions, the clothes they wore during each of the four seasons, and to draw pictures to
illustrate their rhymes.
Linda designed a unit on ‘Weather and activities they do’, and the final task re-
quired children to work individually to make a booklet on ‘My favourite activities in
different weather(s)’. The topic Clare chose was ‘Seasons and clothes’, and she designed
a task of writing a new ending for a story by suggesting ways of keeping warm. The fi-
nal task was to write about clothes that the children wear in winter and summer. Maria
chose the topic ‘Weather and Seasons’. The final task required students to write a four-
line poem about their own ‘favourite’ season similar to the one in their course book.
All four teachers allocated five lessons of 35–40 minutes per lesson to the unit, and
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

they designed a series of pedagogical activities which required the children to integrate
their knowledge of weather and aspects related to the different weather conditions.

Data collection

Data were collected from several sources: (1) lesson observations of classroom teach-
ing with lessons recorded; (2) semi-structured interviews with the teachers; and (3)
documentations such as lesson plans, teaching materials, and students’ work. A total of
20 lessons, 5 for each teacher, were observed and video-taped. Each teacher taught
about their chosen topic (the weather) of which they designed the tasks and activities
for the study of this topic. The analysis of the enactment of TBLT by these four teachers
went through an iterative process in which an initial framework (presented earlier)
was applied to the data but was revised a number of times in response to new dimen-
sions and features that emerged from the data and re-applied to the data.

Data analysis

To examine how learning is organized in the classroom in an attempt to answer the

research question that guided this study (see above), it is necessary to examine the pat-
tern of verbal interactions: who spoke about what, in response to what, and with what
effect in the implementation of task-based teaching in the language classroom. Hence,
classroom observations with field notes and classroom discourse data were video-re-
corded and transcribed. Secondly, semi-structured interviews with teachers were held
before and after the unit was taught, and they also involved stimulated recall. Each
interview was based on a common interview schedule which was drawn up based on a
review of the literature on TBLT (Nunan, 1989, 2004; Samuda, 2001), and on personal
notes developed from prior experience in supervising in-service teachers during their
practicum (see Appendix B for the interview questions). Stimulated recall involves
asking teachers to comment on what was happening at the time that the teaching and
learning took place by looking at their lesson plans, transcriptions of the lessons, and
students’ work.

Findings

Findings of the study suggest that there seemed to be six dimensions on which the
teachers differed in enacting TBLT in their classrooms: (1) strategic use of visual sup-
port to manage task demands; (2) contextualizing input to make connections between
old and new knowledge; (3) simultaneous attention to task demands for progression in
complexity; (4) provision of scaffolding through task sequencing and adjustment of
task variables; (5) creating conditions for noticing form and salient features; and (6)
 Sui Ping (Shirley) Chan

creating conditions for restructuring to occur. These will be illustrated and discussed
separately below.

Strategic use of visual support to manage task demands

All four teachers used a lot of visual aids, including real objects, pictures, and word
cards. However, the ways in which they made use of visual aids differed in a number of
ways. First, Kathy and Clare were better able to provide visual support to reduce the
linguistic demands made on the students so that they could focus on specific aspects
of language use, such as semantics and pronunciation. For example, Kathy made use of
the strategy of building semantic links to facilitate noticing. Thus, she developed stu-
dents’ vocabulary knowledge of clothes by showing the clothes that a cut-out boy and
girl wore in a particular season. By presenting a set of summer clothes at the same
time, the teacher built a lexical network to form a semantic cluster or lexical set: for
example, sunglasses, T-shirt, and a pair of shorts as typical summer garments; while
gloves, coat, and trousers were typical winter wear. Building semantic networks in a
meaningful context may facilitate the intake of new vocabulary, which is essential to
the completion of the shared writing task: “Composing a rhyme to share the clothes
one wears in different seasons and weather.”
By contrast, when Linda guided students to see the relationship between days of
the week, weather, and activities through the use of a weather chart and stories, the
relationship was arbitrary. It was not easy for the researcher to see the criteria for se-
lecting an activity. For instance, one story showed what Bobby, one of the students,
liked doing in different sorts of weather on each day of the week. However, the ‘story’
simply provided a list of activities that Bobby did each day. For instance, in the story,
the weather on Tuesday was cold and cloudy, and Bobby liked drawing pictures. How-
ever, the weather of Wednesday was also cool and cloudy, but Bobby liked playing
basketball. While both days shared much the same weather, it was not clear why the
activity ‘drawing pictures’ was preferred on a cold and cloudy Tuesday while the activ-
ity ‘playing basketball’ was chosen on a cool and cloudy Wednesday. Further confusion
arose as another ball game, ‘playing football’, was chosen on a warm and windy Thurs-
day. The story failed to provide a solution stage showing the relationship between
weather and activities. Hence, it was probably unfamiliar to students and they would
find it difficult to understand or make sensible predictions of the relationship between
the activity and the weather of the day.
Second, visual aids were used not only to reduce but also to increase the linguistic
demands on pupils. For example, while Maria and Linda introduced vocabulary items
with pictorial representations to aid comprehension, Kathy and Clare further elicited
vocabulary items which are associated with the pictorial representations in the form of
a semantic network. For example, a picture of the item ‘snowing’ was presented as a
stimulus for eliciting the word ‘winter’ (season) or the word ‘cold’ (weather condition)
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

in Kathy’s case. Clare used pictures which contained activities and food items that
students had learnt before and required them to transfer such knowledge to another
context of giving advice to keep warm in cold weather, using sentences such as ‘Have a
hot bath’. By contrast, Maria used real clothes to elicit students’ knowledge of weather
by deduction. Linda mainly used word cards to label the picture of a particular weath-
er condition or activity. No association with another concept was required.

Contextualizing input

The selection of discourse genre and topic is equally important to the cognitive de-
mands made on the learners. Although the four teachers emphasized in their inter-
views the importance of presenting target language in context and used stories as input
materials, it seems that Kathy and Clare were much better able to do this than Linda
and Maria. The story used by Kathy contained many features of a good story. Kathy
based her story on ‘Cinderella’ which was familiar to the pupils, and it contained the
usual elements of a story or fairy tale (i.e., orientation, problem, resolution, and a hap-
py ending). The story was about a stepmother who asked Mary to complete impossible
tasks such as getting plants or fruits that could not be found in a particular season.
By getting pupils to think about what could be found in a particular season and
what could not be found, Kathy focused their attention on the features which are
distinctive of that season. For example, she asked the students to use their knowledge
of the relationship between weather and season learnt in the lesson, and to relate it to
their existing knowledge of plants and fruits. Kathy ‘problematised’ the situation by
asking the students what is not available in a certain season. In the story ‘Four broth-
ers’, Mary, the story character, faced the problem of being asked by the stepmother to
get some flowers in winter and then was rescued by the four brothers who changed
the weather from winter to spring. Students were able to relate to the unique features
of winter and could discard irrelevant or impossible elements, such as getting flowers
in the cold season. Moreover, she engaged her pupils in co-constructing the story
with her and their peers by drawing on their linguistic resources to make meaning.
For example, Excerpt (1) shows how Kathy guided the pupils to learn the newly ac-
quired linguistic knowledge through interpretation and making inferences at a dis-
course level.
By contrast, the stories in Maria’s and Linda’s classes were by no means authentic.
For example, the ‘story’ that was presented to students in Linda’s lesson was what the
boy did when the weather was hot or cold. This lesson, then, lacked a problem which
could show a sequence of unexpected and disruptive events leading to a crisis point, as
well as the character’s reactions, feelings, and opinions about these events. Although
the story seemed to include the stage of providing solutions, it was not at all stimulat-
ing. Because of the contrived contextualisation of the linguistic items, the task ended
up becoming a structural drill.
 Sui Ping (Shirley) Chan

Excerpt (1) Problematising the situation in the story

Situation Teacher-Student interaction

1 *T: Because outside it is ...
LLL: Cold and snowy.
T: Can Mary get some flowers?
LLL: NO! (Students replied loudly)
:
:
T: What will happen when Brother Spring holds the magic stick?
What will happen? Winter? Or autumn? What will happen? ...
You don’t know?
LLL: No
T: It’s spring. It is spring. It’s warm and rainy.
LLL: /warm and rainy.
T: Can Mary get some flowers now [T points at pictures of
flowers]?
LLL: Yes!
2 T: Mary gets some flowers now. OK. Let’s see. [T turns the page
and points at the words] Stepmother says, “No. Unless you get
me some ...”
LLL: Apples!
T: Can Mary get some apples in the Winter?
LLL: No.
T: Then, what would Mary do? ... Ask for help. Who would help her?
L: Four Brothers.
T: Four Brothers. Very good... [T turns the page] Mary went back
to the Four Brothers. Now, Brother Winter says, it’s easy.
Brother Summer, you hold the magic stick. What will happen?
It is winter? Or it is spring? Or it is ... it is ...
LLL: summer
T: [T turns the page] summer now. That’s good. It is summer now.
The sun is very hot. [T points at pictures in the book] And all the
trees ... Look at the apple trees. See? Many many apples. Can
Mary get the apples now?
LLL: YES!
T: Mary can get the apples. So ... [T points at words and reads with
Ls] It’s summer. It’s hot and sunny. Mary picks some apples and
goes home.
*(see Appendix C for transcription conventions)
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

Simultaneous attention to task demands and progression in complexity

To allow progression in complexity, teachers need to attend to different types of task

demands simultaneously. They have to be skilful in varying the demands, not in a lin-
ear manner but by interweaving them with a gradual spiral trend of going from less to
more demanding.
Kathy and Clare started with pedagogic tasks that were mainly teacher-led with
convergent goals and closed responses such as picture descriptions. These activities
were followed by storytelling activities which invited more open responses with con-
vergent goals. This created opportunities for negotiation of meaning between the
teacher and students, because the rich storyline elicited imagination and the creative
use of language to express personal thoughts in the search for a consensus. They then
moved back to using pseudo-communicative activities like board games, competi-
tions, and question and answer as a form of scaffolding to support further student-led
activities.
In addition, Kathy provided progression in the use of different types of tasks,
namely static, dynamic, and abstract. She started with static tasks with information to
be processed remaining the same, such as matching game. Then she proceeded to use
a dynamic task, storytelling, which contained changing events and activities. Through
storytelling, learners were required to apply their new knowledge to make restructur-
ing possible. After this, she used a static task, a board game, to consolidate the forms
used to describe the weather. A slightly more challenging task, the information-gap
activity, was used to prepare students for the abstract task of choosing clothes for dif-
ferent seasons. This is considered as an abstract task as students needed to use the in-
formation to express their opinions. In other words, pseudo-communicative activities
were used as a means not an end to the development of students’ communicative com-
petence. Kathy and Clare were able to focus on both accuracy and fluency during these
stages. They provided scaffolding not just by going from the less to the more demand-
ing activities but also by moving back and forth between closed (or manipulative) re-
sponses and open (or spontaneous) responses, with a general progression from the
former to the latter.
By contrast, Maria and Linda focused largely on accuracy. There was no further
progression from pseudo-communicative to a semi-controlled or free activity. In other
words, pseudo-communicative activities were used as an end in themselves. For ex-
ample, in Linda’s case, after students had matched pictures of different weather patterns
with the correct words and stated the words correctly in Lesson 1, another matching
activity for group work in the following lesson was designed to push the students to
engage in a dialogue. They were given a set of cards and asked to find the pairs, such as
a picture of ‘the sun’ being matched with the word card ‘sunny’. The other group mem-
bers were required to ask ‘What’s the weather like?. The one who correctly matched the
most pairs was the winner. Although there was an attempt to move students’ language
production from single word responses to sentence level through matching activities,
 Sui Ping (Shirley) Chan

the requirement to produce sentences was confined to contrived contexts. Moreover,

the responses were selected from the word cards provided. There was little evidence of
conditions being created in which students had to construct personal responses which
could be described as pushed output, except for the final individual writing task.

Provision of scaffolding through task sequencing and adjustment of task variables

Scaffolding involves making pedagogical decisions on how students can be supported

in such a way that learning is not simply made easier but also becomes more challeng-
ing. Kathy and Clare sequenced the tasks to scaffold pupils’ learning by alternating
tasks of different levels of demand strategically. For instance, Kathy and Clare started
with tasks that were cognitively, linguistically, and interactionally less demanding and
then moved to more demanding ones with adjustments in task variables to keep the
demand on other aspects low. When the cognitive demand was increased, Kathy less-
ened the linguistic and interactional demands by accepting single word responses
through teacher-led elicitation to make the task easier. For example, in Excerpt 2 stu-
dents were first asked to identify the clothes that the cut-out boy and a girl put on and
then deduce the season the boy or girl was in.
Excerpt 2

T: OK. [T puts up another transparency and points at clothes while asking ques-
tions] Let me see another pictures. See whether you can tell me ... To see
whether you can tell me the seasons, or the weather. OK. You see, this boy is
wearing a pair of sunglasses. And he puts on T-shirt with short sleeves. And
he puts on a pair of ...?
LLL: /Shorts.
T: Shorts. And he put on ... shoes. And the girl is wearing a pair of sunglasses too.
And she is wearing a dress with no sleeves. Tell me, which season is this?
L1: Summer.
T: Summer. Very good. How does they, sorry, how do they feel?
L2: Hot.
T: They feel ...
LLL: Hot.

Similarly, Clare reduced the linguistic demands by asking for responses at word level
when she engaged students in cognitively demanding tasks. For example, to scaffold
the progression in linguistic complexity, Clare structured the activities from simple to
difficult. For instance, in Lesson 1, students were asked to name the clothes at word
level. Then they guessed the clothes the boy would wear in response to the change in
the weather from hot to cold using the pattern ‘Put on (a scarf)’. The medium was
verbal but supported by the use of pictorial representations in the story. Then, a quiz
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

was designed to push students to write down the clothes that the boy put on in accor-
dance with the sequence in the story. The sequence moved from putting on trousers to
a coat, then socks and boots, and finally mittens followed by a scarf, when the boy in
the story kept telling his mother that he was cold. Although the outcome was closed,
this required students to produce a written statement that showed the logical link be-
tween vocabulary such as socks and boots. This was linguistically more demanding
because the students were asked to link up those items logically than simply recalling
the clothes that the story character wore. Kathy and Clare constantly adjusted three
kinds of demands as a means of scaffolding in task sequencing.
Furthermore, to scaffold progression in terms of cognitive complexity, Kathy first
asked students to recognize types of weather or clothes, which basically required recall-
ing and then associating them with a particular season. When the pupils’ knowledge of
weather and clothes was well established, she challenged them cognitively to deduce
the season from its associated aspects such as weather conditions and clothes, which
involved a higher cognitive process of applying knowledge. Clare allowed for progres-
sion in cognitive demand by starting with the familiar and straightforward relation-
ship between clothes, weather, and seasons and then proceeded to get students to think
about circumstances where this relationship does not apply. For example, she asked
the students to identify workers who do not follow the established norm, such as put-
ting on mittens in winter. Both Kathy and Clare scaffolded the transition by engaging
students in progressively more demanding cognitive processes as a form of working
towards the performance of the final task. These kinds of scaffolding were less evident
in Maria’s and Linda’s cases.

Creating conditions for noticing form and salient features

All four teachers tried to bring about the noticing of form in salient features of the
target input. Kathy, Clare, and Linda employed the strategy of making a contrast when
presenting new concepts. For instance, Kathy contrasted hot weather with cold weath-
er by showing a boy with thick clothes in cold weather and a boy with thin clothes in
hot weather. Similarly, Clare contrasted hot and cold weather by showing the thick
clothes that a story character, Bobby, put on and the summer clothes that the students
were putting on. The contrast between the thick clothes and the thin summer clothes
helped to bring about the noticing of the distinctive features of hot and cold weather.
Linda employed the same strategy by telling the following ‘story’ to present the
four adjectives, namely ‘hot’, ‘cold’, ‘cool’, and ‘warm’.
Today is Thursday. He is very hot. Then he walks into the room and switches on
the fan. He feels cool now. On the next day, the weather has changed. He becomes
a snowman. So, he walks into the room and switches on the heater. He feels warm.
He is warm now.
 Sui Ping (Shirley) Chan

She experienced some difficulty in making an explicit contrast between the adjectives
‘warm’ and ‘cool’ which were used to describe sensations while ‘cold’ and ‘hot’ were
used to describe weather conditions. When presenting ‘hot’ and ‘cold’, Linda focused
on the weather conditions. However, when presenting ‘warm’ and ‘cool’, she used the
pictures which included two contexts, weather conditions and sensations mediated by
a heater and a fan respectively. Students got confused because they were not sure if the
teacher wanted them to talk about the weather or sensations. In other words, the vari-
ations that the teacher tried to exploit were not related to the same dimension: one had
to do with physical sensations mediated by weather phenomena and the other was
mediated by home appliances. The use of the two different contexts from which the
meanings of the words were derived in fact confused the students instead of focusing
their attention on the salient features. The variations that Linda exploited seemed to be
unprincipled and consequently confused students.

Creating conditions for restructuring to occur

Kathy and Clare engaged students in making deductions, problematising the situa-
tions, and using task repetition which can facilitate noticing and restructuring. To fur-
ther challenge their students, Kathy and Clare problematised the situations by making
students respond to the changing conditions. For instance, Clare ‘problematised’ the
situation by showing hot weather while a boy had thick clothes on. This created the
need to talk about taking off clothes. Then, she varied the weather from hot to cold,
creating the need to put on the clothes again. To further develop the concept of keep-
ing warm in cold weather, Clare asked students to give other suggestions related to
food and activities. Although some of the examples were not always appropriate such
as ‘Have a packet of fries’, this provided the opportunities for students to restructure
their knowledge of food and activities and relate them to the concept of keeping warm
using the imperatives such as ‘Have (a bowl of congee)’. Table 2 below summarizes the
situations created by teacher Clare to help learners notice the salient features in the
target input.

Table 2.  Problematising the situation

Situation Weather Clothes/food/activities Solution (Use of imperatives)

1 Hot Winter clothes: a coat, boots, trousers, Take off the clothes
mittens, a scarf
2 Cold Winter clothes: a coat, boots, trousers, Put on the clothes
mittens, a scarf
3 Cold Food: a chicken leg; a bowl of noodles Have ...; Go to bed
Activity: a hot bath; a bowl of congee
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

Students’ work

I also analysed the students’ output. Their writing showed that they were able to use
imperatives creatively to give suggestions, albeit with some spelling mistakes. For in-
stance, Kathy invited the pupils to solve a problem in the story about the stepmother.
She problematised the situation by asking students what plants were not found in a
particular season. In order to solve the problem, the four brothers in the story had to
change the season in order to get the plant or fruit demanded by the stepmother. In
other words, the teacher restructured students’ knowledge of weather and seasons by
getting them to reflect on their relationship with plants and fruits.
Kathy and Clare also recycled target structures and vocabulary in a new context
and introduced them incrementally so that students could focus their attentional re-
sources on the new element. For instance, Kathy recycled the phrasal verbs ‘put on’ and
‘take off ’ in different contexts, moving from a teacher-led whole class activity about
what clothes the boy should take off in very hot weather to a student-led information
gap activity about the clothes that the boy should put on, using ‘Put on...’ with free
choice of vocabulary for clothes. The use of imperatives was then repeated in different
contexts. This required students to engage in syntactic processing to consider what
items to replace when the contexts changed. The use of similar description tasks en-
abled students to focus their attention on the language when the task type was familiar
(Skehan, 1998). This served as a form of cognitive scaffolding.
Clare also recycled the language for giving advice and expressing preferences in a
new context. For instance, to help students get familiar with the language for giving
advice, she first asked students to interpret the weather, explain the problem of inade-
quate clothes, and identify clothes that can keep us warm when telling a story ‘Brrr!’.
Once students got used to the use of the imperative ‘put on’, she presented a second
task which required students to suggest other ways of keeping warm. This is a form of
task repetition of giving advice in a new context. The requirement of giving other sug-
gestions compelled students to use other imperative forms such as ‘Have a...’ and relate
other familiar domains such as food and activities. Hence, the vocabulary items for
food and activities were recycled and introduced incrementally for a different com-
municative function.
To introduce the expression of preference, Clare also adopted a similar strategy of
task repetition in a new context. She divided the task of expressing preference into
smaller and manageable sub-tasks for students to complete. For instance, she con-
ducted an activity ‘Colour and label’ which required students to identify their favourite
clothes. Then, she asked them to write down three names of clothes that they liked.
This involved selection. After that, the two activities were integrated in an individual
writing task ‘My favourite clothes’. The task was made cognitively familiar to the stu-
dents because they had been exposed to expressing preferences in different contexts.
The previous sub-tasks, namely colouring and labelling and selecting three favourite
clothes items, prepared students cognitively for expressing preferences. The discourse
 Sui Ping (Shirley) Chan

of expressing preference was incrementally introduced and recycled in a new context.

There was a high degree of cognitive familiarity which provided cognitive scaffolding
to reduce cognitive processing load and help channel the attentional resources to the
target form. Task repetition in a new context is a powerful way of drawing students’
attention to the new elements in a familiar yet different context.
By contrast, Maria and Linda largely repeated tasks which required students to
revise existing knowledge rather than teaching new knowledge, and they tended to ask
their students to work mostly in pseudo-communicative contexts that might have pre-
vented them from experimenting with new forms creatively. The findings show the
importance of re-interpreting old knowledge in new contexts in second language
learning. Further analysis of students’ work shows that student outcomes tended to be
limited in scope with closed responses at a single word/phrase level.

Discussion

Interweaving focus on form and on meaning

To allow progression in complexity, teachers need to attend to different types of task

demands simultaneously. They have to be skilful in varying the demands not in a linear
manner but by interweaving them in a general spiral trend of going from less to more
demanding. The findings of this study suggest that providing opportunities for students
to engage in fluency tasks in which they are required to stretch their language compe-
tence to construct free responses is important. It helps students to restructure their prior
knowledge of words or concepts and the relationship between them (Robinson, 2001).
The findings indicate the importance of helping students to move along the con-
tinuum of focusing on form to focusing on meaning in communicative contexts by
starting with controlled and semi-controlled activities and then extending the context
of language use to facilitate the production of free or open responses, as evident in
Kathy’s and Clare’s cases. It also shows the importance of interweaving focus on form
and focus on meaning at various stages of a task so that students are equipped with the
linguistic resources to complete a task.

Scaffolding by interweaving task demands and task sequencing

Scaffolding involves making pedagogical decisions on how students can be supported

in such a way that learning is not simply made easier but is also more challenging and
hopefully more effective. The findings show that a progressive increase in task de-
mands requires teachers to constantly adjust the task variables when sequencing their
activities to ensure that students’ attention is drawn to the necessary knowledge or skill
required for the final task completion. Progression in complexity should be attended
to but complexity is determined by all three dimensions of task demand, namely lin-
guistic demands, cognitive demands, and interactional demands, and these interact.
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

So an increase in complexity in one dimension may require an adjustment to the com-

plexity in the other dimensions.
Teachers need to have a framework for understanding the demands of the tasks
that they give to students, to be aware of the relationship between the three types of
demands, and to be sensitive to pupils’ needs so that they can adjust the demands ac-
cordingly. They need to make pedagogical judgments moment by moment and appre-
ciate the dynamics of task demand. Kathy and Clare achieved this goal by starting with
tasks that were cognitively, linguistically, and interactionally less demanding, and then
they moved to more demanding ones with adjustments in task variables to keep the
demands on other aspects low. For instance, when the cognitive demand was increased,
Kathy kept the linguistic and interactional demands by accepting single word respons-
es through teacher-led elicitation to make the task easier.
Similarly, Clare reduced the linguistic demands by asking for responses at word
level when she engaged students in cognitively demanding tasks such as reconstructing
the sequence of clothes. They constantly adjusted three kinds of demands as a means of
scaffolding in task sequencing. While research on TBLT has emphasised the impor-
tance of sequencing tasks according to complexity, the picture in real-life situations is
quite complex. It is the interweaving of tasks and activities of different demands and
the teachers’ awareness of the simultaneous demands made on students that matters.

Creating conditions for noticing form and salient features

The findings of this study suggest that in order for learners to focus their attention on,
that is, to ‘notice’ the object of learning, it is necessary to ensure the other aspects of
the object are held constant. For instance, to present the relationship between clothes
and seasons, Kathy varied the clothes by first drawing students’ attention to the sum-
mer clothes that the two figure cut-outs put on. She helped students notice what was
common in the clothing and accessories of the two figure cut-outs. Then, she changed
the clothes and accessories from those worn in the summer to winter. The change of
clothing made students aware of the season, which was initially kept invariant.
In contrast, the variations that Linda exploited seemed to be unprincipled and
consequently confused students. Students’ attention was therefore not focused on the
critical aspects of the language and rendered learning ineffective. Hence, teachers
should be aware of the strategies for ‘noticing’ so that they would be able to make in-
formed decisions about their pedagogical practices.

Creating conditions for restructuring to occur

The literature on SLA suggests that for restructuring to occur, we need to create the
proper conditions for it. These conditions include the need for producing output and
provision of practice or task repetition. Language learning is not simply a cumulative
process of adding new knowledge to what learners already know. Swain (1985, 1995,
 Sui Ping (Shirley) Chan

2000) argues that while input may have the function of triggering a reorganization of
existing structures, it is the process of producing something by experimenting with
their existing language resources that causes learners to notice the gap between cur-
rent knowledge and required knowledge.
In support of these arguments, the findings of this study also suggest that restruc-
turing is likely to occur when students are engaged in making deductions and
problematising situations. Kathy and Clare were able to challenge their students by
problematising the situations. For instance, Clare changed the weather to hot when the
boy put on thick clothes so that pupils had to give advice on taking off the thick clothes
and putting on summer clothes. Similarly, Kathy problematised the situation by asking
students what plants were not available in a particular season when telling the story. This
helped students to restructure their knowledge by getting them to solve the problems.

Conclusions and implications

There has been wide discussion of what a task is and what it is not in the TBLT literature.
The discussion contributes to the clarification of the features of ‘a task’ and the identifi-
cation of the value of pedagogical activities. However, in the enactment of TBLT in the
classroom, the clustering of pedagogical activities and tasks, and pedagogical consider-
ations of task demand and task sequencing, are crucial. This study shows the impor-
tance of developing teachers’ awareness of the intricate relationships between task
demands and task variables and their ability to interweave task features of different lev-
els of demand while constantly adjusting task variables as a mean of scaffolding. It also
shows that scaffolding is an important concept in task-based pedagogy. It indicates that
strategic advance planning through careful design and sequencing of activities is vital.
The findings also imply that task sequencing, as a form of scaffolding, is not linear
but cyclical and interwoven. The progression of complexity relies on the teacher’s pro-
fessional judgment as to how to manage the three dimensions of task demand
(linguistic, cognitive, and interactional) at different stages of learning through con-
stant adjustment of the task variables to make the task more or less demanding. Very
often, the moment-to-moment adjustments and the teachers’ judgments are crucial as
the learners’ orientation to tasks is unpredictable.
Local studies and my personal interaction with in-service teachers seem to sug-
gest that teachers find the concept of TBLT difficult to grasp even after taking an in-
service training course which prepared them to incorporate tasks in their educational
practice. There are several ways in which the case studies in this study may be of rele-
vance to teacher education. The critical differences between teachers identified in the
study illustrate the complexities underlying the enactment of TBLT. Teacher educators
can make use of the detailed description of the four cases in this study to present con-
crete examples of practice to help teachers understand the complexities involved in
TBLT. Self-reflection and critical peer evaluation help teachers articulate their thoughts
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

and justify their actions. As Tsui (2003) suggests, case studies will help teachers raise
their awareness of their own actions, and help make their tacit knowledge explicit.
The framework for the analysis of TBLT that emerged from this study can be used
to help teachers to investigate their own task-based teaching in a holistic manner.
Alternatively, teachers can focus on any one of the six areas which are found to be
critical in the enactment of TBLT: (a) strategic use of visual support to manage task
demands; (2) contextualizing input to make connections between old and new knowl-
edge; (3) simultaneous attention to task demands for progression in complexity;
(4) provision of scaffolding through task sequencing and adjustment of task variables;
(5) creating conditions in noticing form and salient features; and (6) creating condi-
tions for restructuring to occur. As noted in the study, although these features are
prominent in Kathy’s and Clare’s teaching, the teachers seemed not to be aware of this
and doubted their own effectiveness in enacting TBLT at both the design and imple-
mentation stages. Hence, teacher educators can make use of this six-dimensional
framework to help teachers to make their tacit knowledge of these features explicit.
The study also implies the need of drawing upon different perspectives or theories
to understand the complexities underlying the enactment of TBLT. Such complexities
cannot be simply understood or interpreted by drawing on a single theory. As a teach-
er educator and researcher, I had to engage in a rigorous and critical process of review-
ing and reflecting on how the teachers in my study translated their tacit knowledge
into classroom practices. For instance, when visual support was found common in the
four cases, two teachers seemed to have used the support more effectively than the
other two. This led me to investigate what contributed to the differences. The psycho-
linguistic view of language learning helped me understand the importance of creating
conditions for noticing to occur. However, the understanding of how ‘noticing’ hap-
pens and how teachers can structure learning to bring it about seems not to have been
adequately addressed in the TBLT literature.
This further drove me to draw on another theory which adopts a very different
perspective on the nature of learning from an information processing approach to
learning, to explain how noticing could be brought about. The theory of variation
helped me explain how the provision of visual support varied qualitatively in the four
cases. Hence, the study implies that teachers need to be critical of their beliefs in lan-
guage and language learning and be aware of their limitations. They need to be willing
to seek other theories to address issues that could not be explained by their existing
beliefs or understandings. In other words, teacher educators need to help teachers to
develop a critical stance on their own beliefs in language and language learning and
challenge teachers to reshape them.
Apart from helping teachers critically reflect on their own beliefs, this study also
shows the importance of developing teachers’ awareness of the intricate relationships
between task demands and task variables, and their ability of interweaving tasks of
different levels of demand and constantly adjusting task variables as a mean of scaf-
folding. Teacher educators can make use of the analytical framework of this study as a
 Sui Ping (Shirley) Chan

planning and evaluative tool to facilitate teachers to design, implement and evaluate
their own enactment of TBLT. As Kumaravadivelu (2003, 2005) states, we are moving
from methods-based pedagogies to post-method pedagogies to help teachers develop
their own theories of practice, awakened to the complexity of teacher beliefs, multi-
plicity of learner identities, as well as vitality of macrostructures- social, cultural, po-
litical, and historical that shape and reshape our pedagogy.

References

Brown, G., Anderson, A., Shillcock, R., & Yule, G. (1984). Teaching talk: Strategies for production
and assessment. Cambridge: Cambridge University Press.
Candlin, C.N. (1987). Towards task-based learning. In C. Candlin & D.F. Murphy (Eds.), Lan-
guage learning tasks (pp. 5–22). Englewood Cliffs, NJ: Prentice Hall International.
Candlin, C.N. (2001). Afterword: taking the curriculum to task. In C. Candlin & D.H. Murphy
(Eds.), Researching pedagogic tasks: Second language learning, teaching and testing
(pp. 229–243). Essex, UK: Pearson Education.
Carless, D. (2003). Factors in the implementation of task-based teaching in primary schools.
System, 3, 485–500.
Carless, D. (2004). Issues in teachers’ reinterpretation of a task-based innovation in primary
schools. TESOL Quarterly, 38(4), 639–662.
Curriculum Development Council. (1999). Syllabuses for secondary schools: English Language
Education (Secondary 1–5). Hong Kong, SAR: Printing Department.
Curriculum Development Council. (2004). English Language Curriculum Guide (Primary 1–6).
Hong Kong, SAR: Government Logistics Department.
Curriculum Development Council. (2007). English Language: Curriculum and Assessment Guide
(Secondary 4–6). Hong Kong, SAR: Government Logistics Department.
Ellis, R. (1997). Second language acquisition. Oxford: Oxford University Press.
Ellis, R. (2002). Does form-focused instruction affect the acquisition of implicit knowledge? A
review of the research. Studies in Second Language Acquisition, 24(2), 223–236.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Kumaravadivelu, B. (2003). Macrostrategies for language teaching. New Haven, CT: Yale Univer-
sity Press.
Kumaravadivelu, B. (2005). Learner perception of learning tasks. Paper presented at the Interna-
tional conference on task-based language teaching, Leuven, Belgium. 23 September –
25 September, 2005.
Lee, J.F., & Van Patten, B. (1995). Making communicative language teaching happen. New York,
NY: McGraw-Hill.
Long, M. H., & Crookes, G. (1993). Units of analysis in syllabus design: The case of task. In G.
Crookes & S. M. Gass (Eds.), Tasks in pedagogical context (pp. 9–54). Clevedon, UK: Multi-
lingual Matters.
McLaughlin, B. (1990). Restructuring. Applied Linguistics 11(2), 113–128.
Morris, P., Adamson, B., Au, M.L, Chan, K.K., Chan, W.Y., Ko, P.Y., Lai, W., Lo, M.L., Moriris, E.,
Ng, F.P., Ng, Y.Y., Wong, W.M., & Wong, P.H. (1996). Target Oriented Curriculum Evalua-
tion Project: Interim Report. Hong Kong: INSTEP, The University of Hong Kong.
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

Nunan, D. (1989). Designing tasks for the communicative classroom. Cambridge: Cambridge Uni-
versity Press.
Nunan, D. (2004). Task-based language teaching. Cambridge: Cambridge University Press.
Pica, T., Kanagy, R., & Falodun, J. (1993). Choosing and using communication tasks for second
language instruction. In G. Crookes & S. Gass (Eds.), Tasks and language learning: Integrat-
ing theory and practice (pp. 9–54). Clevedon, UK: Multilingual Matters Ltd.
Richards, J. C. (2002). 30 years of TEFL/TESL: A personal reflection. RELC Journal 33(2),
1–35.
Richards, J., & Rodgers, T. (2001). Approaches and methods in language teaching. Cambridge:
Cambridge University Press.
Robinson, P. (2001). Task complexity, cognitive resources and syllabus design. In P. Robinson
(Ed.), Cognition and second language instruction (pp. 287–318). Cambridge: Cambridge
University Press.
Robinson, P. (2007). Rethinking-for-speaking and L2 task demands: The Cognition hypotheses,
task classification, and sequencing. Point-counterpoint plenary presentation at the second
International Conference on Task-based Language Teaching, Honolulu, Hawaii, 20 Sep-
tember – 22 September 2007.
Samuda, V. (2001). Focus on form through collaborative dialogue: exploring task effects. In M.
Bygate, P. Skehan, & M. Swain (Eds.), Researching pedagogic tasks: Second language learning,
teaching and testing (pp. 110–14). Essex, UK: Pearson Education.
Samuda, V. (2007). Tasks, design, and the architecture of pedagogic spaces. Plenary presentation
at the second International Conference on Task-based Language Teaching, Honolulu,
Hawaii, 20 September – 22 September 2007.
Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics,
11,129–158
Skehan, P. (1998). A cognitive approach to language learning. Oxford: Oxford University Press.
Skehan, P. (2007). Tradeoff and cognition: Two hypotheses regarding attention during task-based
performance. Point-counterpoint plenary presentation at the second International Conference
on Task-based Language Teaching, Honolulu, Hawaii, 20 September – 22 September 2007.
Swain, M. (1985). Communicative competence: some roles of comprehensible input and com-
prehensible output in its development. In S. Gass & C. Madden (Eds.), Input in second lan-
guage acquisition (pp. 235–253). Rowley, MA: Newbury House.
Swain, M. (1995). Three functions of output in second language learning. In G. Cook & B. Seidl-
hofer (Eds.), Principles and practice in Applied Linguistics (pp. 125–144). Oxford: Oxford
University Press.
Swain, M. (2000). The output hypothesis and beyond: Mediating acquisition through collabora-
tive dialogue. In J.P. Lantolf (Ed.), Sociocultural theory and second language learning
(pp. 97–114). Oxford: Oxford University Press.
Tsui, A.B.M. (2003). Understanding expertise in teaching. Cambridge: Cambridge University
Press.
Van den Branden, K. (2007). Task-based language education: From theory to practice... and back
again. Plenary presentation at the second International Conference on Task-based Lan-
guage Teaching, Honolulu, Hawaii, 20 September – 22 September 2007.
White, J. (1998). Getting the learners’ attention: A typographical input. In C. Doughty & J. Wil-
liams (Eds.), Focus on form in classroom: Second language acquisition (pp. 85–113).
Cambridge: Cambridge University Press.
Willis, J. (1996). A framework for task-based learning. London, UK: Longman.
 Sui Ping (Shirley) Chan

Zhang, Y.F. (2005). Implementing task-based approach in primary school ELT in Mainland China.
Paper presented at the International Conference on Task-based Language Teaching, Leuven,
Belgium, 23 September – 25 September 2005.

Appendix A

Table 1.  An overview of Katherine’s five lessons

Task type Description

Picture description Revision of the previous knowledge (Knowledge of weather, season and
clothes) and presentation of new adjectives for describing weather (e.g.
sunny, rainy, foggy, cloudy and snowy)
Story telling Integrating seasons, weather and clothes and identify unique features of
different seasons such as things we can get in a particular season
Reading aloud Reading aloud the story, new vocabulary items or structures related to
weather description, clothes or seasons
Matching Matching picture cards & word cards of season, weather, clothes
Board game Asking and providing information about the weather using ‘What’s the
weather like?’ and ‘It’s (cold)..’
Opinion giving Giving advice on what to wear in response to changing weather
conditions
Information gap Providing information about the clothes that the boy or the girl in the
picture puts on
Shared writing task Composing a rhyme to share the clothes one wears in different seasons
and weather

Table 2.  An overview of Linda’s five lessons

Task type Description

Singing Singing a song ‘Days of the week’ to revise the previous knowledge of
days of the week
Picture description Revision of the previous knowledge of weather and presentation of new
vocabulary for activities; describing the weather or activities on a par-
ticular day of the week
Reading aloud Reading aloud the stories, new vocabulary items or structures related to
weather, days of the week and activities
Weather chart Using symbols to record the weather conditions of a week
Matching Matching picture cards and word cards about weather
Story telling Guessing the activities that the character likes doing in different weather
conditions
Individual writing Write four sentences about the activities that one likes doing in different
weather conditions
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

Table 3.  An overview of Clare’s five lessons

Task type Description

Picture description Revision of previous knowledge (Knowledge of clothes, food,

activities, seasons and occupations) and presentation of new
vocabulary for summer and winter clothes (e.g. coat, jacket,
shorts)
Reading aloud Reading aloud the story, new vocabulary or structures related to
clothes and occupations
Storytelling Integrating weather and clothes by showing how the teacher gave
advice to Willy, the story character, to keep warm using ‘Put on (a
scarf).’
Giving advice Giving advice on keeping warm using the imperatives ‘Have a (hot
bath).’ by asking students to rewrite a story ending
Completing a story quiz Recalling the sequence of clothes that the boy in the story put on.
Question and answer Identifying the word or picture cards of clothes such as ‘a jumper’;
deducing the season from clothes; integrating seasons, weather
and clothes to describe one’s preference for what clothes to wear
Classifying Classifying clothes into summer and winter clothes by putting the
cut-outs of clothes into the appropriate box
Matching Putting the cut-outs of the clothes on the story character’s body
parts, such as a scarf on the neck
Bingo Matching the picture cards for clothes at hand with the ones that
the teacher wanted
Individual writing Describing the clothes that one usually puts on in summer or
winter
Completing a Providing information about the different clothes worn by
substitution table different workers in summer and winter
Information gap Identifying differences in the clothes different workers wear in
summer and winter by exchanging information orally
 Sui Ping (Shirley) Chan

Table 4.  An overview of Maria’s five lessons

Task type Description

Story telling
Picture description
Reading aloud
Matching
Completing weather chart
Singing
Information gap
(Pseudo-communicative)
Poem reading and writing
Telling the activities the five green monsters in the story played
Introducing different sorts of weather such as ‘rainy’, ‘cloudy’,
‘windy’ and the sensations using the pattern ‘I can feel (the
wind).’; deducing season from clothes or weather
Reading aloud the picture book, new vocabulary or structures
related to weather description and seasons.
Matching pictures and sentence strips to describe the weather
in the picture using the patterns ‘What’s the weather today?’
‘Today is (windy).’
Recording the weather for a week on the weather chart
Asking and providing information about the weather using
‘What’s the weather like today?’ and ‘Today is (windy).’
Asking and providing information about the weather and
sensations
Integrating weather, seasons and feelings of a season; Describ-
ing the weather in a season

Appendix B

Semi-structured Teacher Interview

  1. What did you intend to teach this lesson?
  2. How did you organize your lesson to achieve your goal?
  3. How did you make use of your pupils’ personal knowledge (both world knowledge
and linguistic knowledge) to learn the new language?
  4. How can you ensure your lesson is well structured and logical? What consider-
ations have you made when sequencing the teaching events?
  5. What did you think the pedagogical value of each activity has in the lesson? Why
did you plan your lesson in this way?
  6. Why do you use the following activities? What are the benefits for you and the
students by using them? (List the activities)
  7. Did you encounter any difficulties in planning and teaching this lesson? What are
they? Why?
  8.  What is unexpected to you after teaching the lesson?
  9.  To what extent do you think you could achieve the goal of this lesson?
10.  Is there a part that you think you want to re-teach the next lesson? Why?
11. Based on today’s lesson, do you think you need to further revise your teaching
plan for the next lesson? If yes, what needs to be revised, adapted or changed to-
tally? Why?
 Chapter 9.  Task-based language teaching in Hong Kong primary classrooms 

Appendix C

Transcription conventions
T: Teacher
LL: Group of learners choral
LLL: Whole class choral
L1, L2 etc = identified learner
[In italics] = commentary
... = pause
/ = overlapping speech
CAPITALISATION = emphasis
C-P-I-T-A-L-I-S-A-T-I-O-N = word spelt letter by letter
chapter 10

Implementing computer-assisted task-based

language teaching in the Korean secondary
EFL context

Moonyoung Park
Iowa State University

Task-based language teaching (TBLT) is a comprehensive approach to

language education that emphasizes tasks in each stage of the program’s design
and implementation (Long & Norris, 2000). Although TBLT has attracted
considerable attention since the 1980s, little research has been conducted on its
actual implementation in secondary EFL contexts. This study sought, firstly, to
illustrate how computer-assisted TBLT lessons can be designed and implemented
in a conventional English classroom in a Korean school setting, and, secondly,
to investigate students’ L2 development in writing as well as their perceptions
of TBLT. Thirty Grade 7 students at a Korean middle school participated in
the study. An online needs analysis survey of the students and the teachers was
first conducted to reveal their needs and perceptions about the general English
curriculum. Based on the triangulated analysis of the participants’, societal,
and institutional needs, a series of computer-assisted TBLT lesson plans for
two instructional units was designed. An experimental group (N = 30) was
taught with the TBLT lesson plans, while a control group (N = 31) was taught
in a conventional teacher-centered and forms-focused approach. For each unit,
two task-based writing tests (pre/post-test) and a conventional unit test on
grammar and reading comprehension were administered. A paired sample t-test
of the two groups revealed that the mean scores of the experimental group were
significantly higher than those of the control group. The experimental group
also exceeded the control group in the conventional unit tests. The findings
lend support to the interpretation that the task-based approach can be effective
in improving communicative competence while not hindering form-focused
L2 learning. Students and teachers both found the TBLT lessons are effective
and motivating. This study has implications for administrators, curriculum
designers, materials writers, and, in particular, EFL teachers, and can assist in the
creation of innovative and experiential learning environments to complement
the existing English curriculum in EFL contexts.
 Moonyoung Park

Introduction

Task-based language teaching (TBLT) has become one of the most commonly dis-
cussed language teaching approaches in the field of instructed second language acqui-
sition since the 1980s, and it has recently led to diverse impacts on English education
policy, curriculum design, materials development, and classroom teaching. Many re-
searchers, syllabus designers, and educators have paid substantial attention to TBLT,
and thus numerous studies have been conducted proposing a variety of instructional
ideas for using tasks (e.g., Lee, 2000; Willis & Willis, 2007).
Recently, Norris (2009) defined TBLT as an approach to second or foreign lan-
guage education that integrates theoretical and empirical foundations for good peda-
gogy with a focus on tangible learning outcomes in the form of “tasks” – that is, what
learners are able to do with the language. While researchers differ in regard to what
characterizes a task and how it is applied in the classroom, all seem to agree that learn-
ers can best acquire the target language by engaging in activities that they will likely
encounter in real-world communicative contexts. Somewhat parallel to the introduc-
tion of TBLT in language education have been technological advancements, including
the increasing use of computers and the Internet, for language learning and teaching.
Several prominent voices have highlighted the potential of computer-assisted lan-
guage learning (CALL) based on SLA principles and associated instructional prac-
tices, such as TBLT, to make language teaching more effective (Chapelle, 1998, 2001;
Ortega, 2009).
Recognizing the importance of communicative-oriented teaching generally speak-
ing, Korea’s National Curriculum emphasizes communicative competence and fluency.
More specifically, the Ministry of Education (MOE) has encouraged Korean EFL
teachers to develop students’ communicative competence by incorporating both CALL
(Ministry of Education and Human Resources, 2002) and student-oriented task-based
instruction (Ministry of Education and Human Resources, 2007). However, English
classes in public education settings in Korea do not yet meet these national curricula
expectations for several reasons, including challenges to the feasibility of TBLT in
Asian contexts (McDonough & Chaikitmongkol, 2007) and poor incorporation of stu-
dents’ and teachers’ needs in the English curriculum, lesson plans, and teaching mate-
rials (Yoon, 2004).
Nevertheless, the implementation of computer-assisted task-based language teach-
ing (CATBLT) may be able to foster a synergistic effect amongst EFL learners, tasks,
and technology to create authentic, innovative, and immersive language learning envi-
ronments that complement existing English classes in Asian EFL contexts, including
the Korean EFL context (Thomas & Reinders, 2010). Still, the feasibility of TBLT needs
to be examined in Asian contexts in which different cultural and educational back-
grounds exist (McDonough & Chaikitmongkol, 2007). In addition, little empirical
TBLT research based on needs analysis that considers students, teachers, and school
administrators has been conducted in secondary schooling settings in Asia.
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

The current study began with and is based on a needs analysis for a Korean middle
school EFL curriculum (reported in detail in Park, 2010). The findings of that needs
analysis suggested that both students and teachers recognized the importance of pre-
paring for high-stakes exams as well as acquiring communication skills; in other
words, they strongly expressed the necessity of formal L2 accuracy and communica-
tive fluency at the same time. To achieve the goals of both fluency and accuracy, this
study attempted to employ TBLT as an overarching pedagogical framework, with the
inclusion of technology-mediated instruction as a way of providing authentic target
language and enhancing students’ motivation and affect. Prior to detailing the peda-
gogical methods and study findings, I briefly situate the status quo for implementing
and research TBLT in the Korean EFL context.

Task-based language teaching in a Korean EFL context

As discussed by Norris (2009) and Van den Branden, Bygate, and Norris (2009), TBLT,
as developed since the early 1980s, has attempted to harness the benefits of a focus on
meaning via adoption of an analytic syllabus, while simultaneously, through use of
focus on form, to deal with its shortcomings. These shortcomings include, potentially,
a slow rate of development and incompleteness where grammatical accuracy is con-
cerned. Today, TBLT is being promoted in many countries as a potentially powerful
language teaching approach (e.g., Van den Branden, 2006), and the scope of TBLT has
become broader as well, especially in the design, implementation, and evaluation of
task-based language teaching courses and programs. For over two decades, then, re-
searchers and language educators have discussed the task-based approach, which
stems from communicative language teaching (CLT) (e.g., Bygate, Skehan, & Swain,
2001; Ellis, 2003; Long, 1985; Prabhu, 1987; Skehan, 1996). Summarizing some of the
major principles discussed in relation to TBLT, Nunan (1991, p. 279) proposed five
features:
1. An emphasis on learning to communicate through interaction in the target
language
2. The introduction of authentic texts into the learning situation
3. The provision of opportunities for learners to focus, not only on language, but also
on the learning process itself
4. An enhancement of the learners’ own personal experiences as important contrib-
uting elements to classroom learning
5. An attempt to link classroom language learning with language activation outside
the classroom
However, according to some previous research and commentary (e.g., Anderson, 1993;
Li, 1998), such pedagogical features have been considered difficult to put into practice
in Asian EFL contexts due to several contextual limitations. These include big class
sizes, heavy focus on preparation for college entrance exams, rigid national curricula,
 Moonyoung Park

traditional teaching methods, lack of student motivation, and lack of teacher compe-
tence. The implementation of TBLT in Korean EFL settings has faced these difficulties
as well, but only a few studies have sought to examine the possibilities for TBLT-­related
pedagogic approaches in Korean EFL settings.
In the first decade of the 2000s, publications on TBLT in Korea primarily ques-
tioned the definition and value of the task-based approach, with writing and speaking
as the two most frequently explored areas (Hahn, 2008; Kim, 2009; Ko, 2004; Ko, 2008;
J.-W. Lee & K.-J. Lee, 2005; M.-J. Lee, 2006; S.-M. Lee, 2005; Shin, 2001). While quite
a few researchers have taken interest in the feasibility of TBLT in the Korean EFL con-
text, it is only recently that the research focus seems to have shifted from simply intro-
ducing and examining the possibility of task-based approaches to exploring issues of
actual implementation in Korean school settings. However, the majority of studies on
the actual practice of task-based approaches have been conducted at the college level,
where course syllabi and teaching materials tend to be more flexible than in secondary
school settings. There is thus a need for more research based on needs of younger EFL
learners and focusing on implementation and evaluation of TBLT at both the middle
and high school levels, in order to know how best to implement task-based teaching
and learning at secondary school in Korea.

Computer-assisted language learning in a Korean EFL context

The advancement of computer technology makes CALL a viable alternative to tradi-

tional teaching methods; some hope that it will revolutionize the way students learn a
second or foreign language. Such hopes are based on the assumption that CALL can
expedite language learning and become an important instructional medium (Singhal,
1998). Since the advent of CALL in Korea, a handful of researchers have investigated
how to enhance Korean students’ English ability using computers and the Internet.
Studies have found that computers and the Internet help Korean EFL learners develop
their speaking ability (Shin, 2001), listening comprehension (Park, 2001), writing abil-
ity (Shin & Kwon, 1999), and reading comprehension (Yoon & Lee, 2001). The results
of these studies also suggest that English classes that integrate computers and the In-
ternet become more learner-centered. When they have access to innovative learning
realia, learners seem to devote more time and effort to authentic reading, writing, lis-
tening, and speaking activities.
A glance at the literature on CALL indicates that the evidence supports the effec-
tiveness of using computers and the Internet for enriching language learning in Korean
contexts. However, as pointed out as early as Garrett (1991), the use of computers and
technology does not constitute a method or approach. Rather, it is a medium in which
a variety of approaches, methods, and pedagogical philosophies can be implemented.
It is no wonder that this new medium of language teaching and learning constitutes a
challenge for EFL teachers and learners. This study emphasizes how CALL-integrated
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

English lessons can be developed within a task-based framework and implemented in

otherwise highly structured secondary school settings.

The study

The aims of the current study were to illustrate how computer-assisted TBLT lessons
can be designed and implemented in a conventional English classroom in an EFL sec-
ondary school setting and to investigate the linguistic development in students’ writ-
ing as well as their perceptions of TBLT. Based on the results of the needs analysis,
CATBLT lessons were developed, and then they were embedded and delivered within
the regular English course at a Korean school. Units 1 and 2 were developed and im-
plemented for 30 first-year Korean middle school students (roughly seventh-graders
in the North American system). Two task-based writing assessments were also con-
ducted to measure and compare the learning outcomes of the experimental group
(CATBLT class) and the control group (conventional class).
This study focused on the following research questions (RQs):
RQ 1: How do the learners perform on the task-based writing assessment following
lessons delivered via two different teaching approaches? Is there any meaning-
ful difference?
RQ 2: How much do the learners from the two different teaching approaches achieve
in the traditional unit-review test developed by the school teachers and fo-
cused on vocabulary, grammar, and reading comprehension? Are there any
meaningful differences?
RQ 3: What are student and teacher opinions of the CATBLT lessons?

Method

Participants

The research site was a private boys’ middle school in Korea. Out of the 10 first year
(7th grade) English classes, two intact classes were chosen for the study. The student
participants were male, Korean, middle school students (N = 61) from 13 to 14 years
old who were enrolled in required English courses at the school. All of them had four
years of previous English instruction in the form of required EFL classes in elementary
school. As the school follows a level-differentiated English curriculum, students were
assigned to classes based on the results of an English placement test administered by the
school, which focuses on reading comprehension and grammar. The two selected class-
es were for most advanced-level students with similar class mean values in the English
placement test (control group, M = 94.67; experimental group, M = 94.29). Each class
 Moonyoung Park

was assigned to one of two conditions; one served as a control group (non-CATBLT
group, N = 31) and the other, an experimental group (CATBLT group, N = 30).
The teacher participant who taught both groups was a 37 year old Korean teacher
of English. He had eight years of teaching experience (three years at a private language
institute; five years at the middle school). He taught five advanced-level, first year Eng-
lish classes including the control and experimental groups, and he volunteered to ex-
periment with TBLT for this study.

Materials
Lesson plans. The same school-mandated English textbook was used in both groups;
however, as a part of the experimental treatment, task-based lesson plans for two units
of the textbook were developed by the author and then implemented by the volunteer
teacher in the experimental group. For each unit, three task-based lesson plans and
one task-based writing assessment were implemented over four class periods (three
hours of total instructional hours with one hour of performance assessment). Task-
based lessons were designed to elicit communicative interaction between the students,
so that students were able to accomplish authentic and real-life tasks using productive
skills of speaking and writing. Target tasks, including writing a self-introduction mes-
sage on line, introducing an e-pal to class, and giving directions to e-pals on transpor-
tation, were used to operationalize the lessons in the two units based on the needs
analysis (Park, 2010), in which both students and teachers identified the strong need
for conversation skills.
The design of the task-based lessons involved the following steps. In the pre-task
phase, students were provided (and practiced) target task models that featured key
linguistic items. In the during-task phase, students undertook the task itself, usually in
a small group. Next, students planned how to report their findings or achievements
back to the class. Finally, in the post-task phase, students made reports in a different
group setting. The teacher occasionally reminded the students of the target structures
and grammar at the end of the task sequence.
Based on the textbook topic and target forms, the example lesson plan in Figure 1
is designed to provide learners with an opportunity to understand how to introduce an
e-pal through a pre-task, find an e-pal online in during-task, and introduce her or him
to the class through the post-task. To better serve students’ needs as well as the na-
tional English curriculum guidelines for the use of computers in English learning,
computer-assisted task-based language teaching lesson plans were developed that op-
timized use of personal computers and online sources. Learners in the experimental
group took their English lessons in the school computer lab, and explored actual E-pal
websites and searched for their E-pals by themselves, then uploaded their self-intro-
duction essay on the website. Students also used Internet search engines to find au-
thentic information to complete their tasks, and they used word processing software
and presentation software (e.g., PowerPoint) to type their essays and prepare for group
presentations.
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Lesson #1: (in the computer lab)

Target Task: Introduce an E-pal to class
Target Forms: I like… / He likes…/ My name is…/ Her name is…/ She is from…

– Warm-up activity:
- Show an authentic third-person introduction video from a website and ask the students to guess the
relation of the interviewees in the video (http://www.real-english.com/reo/6/unit6.html).[Worksheet #3]
(5min)

– Pre-Task:
- Have the class watch the teacher’s introduction of his E-pals focusing on the forms in third-person
introduction. (He is…, she likes…, she is from…)[worksheet #4] (Explain that the students will read three
self-introduction messages from an E-pal website. Have the class do a brief analysis of the forms and
structures used in the self-introductions, and ask them to brainstorm what makes a good E-pal self-
introduction, and what doesn’t. (5 min)
- Ask them to choose one different E-pal from the textbook and introduce the person to their partner in
English after completing Step 2 in Worksheet #4. Remind the students how to do a third-person
introduction. [Pair work; the messages are in the textbook, pp. 16–17] (7min)

– During-Task:
- Introduce an E-pal project to the class, encouraging them to find their favorite E-pal for the next e-mail
correspondence. Have them connect to the E-pal website, and introduce the features of the website. (3min)
- Ask them to search for their E-pals by selecting age, gender, nationality [http://penpalsnow.com], and
complete Step 1 in Worksheet #5. (17min)

– Post-Task:
- Have the class move into 8 groups of 4 students each, and ask them to introduce their E-pal to their group
in English. Complete Step 2 in Worksheet #5. (8min)

– Homework:
Answer questions on p. 15 in workbook for reading comprehension.

Figure 1.  A Sample Task-Based Lesson Plan in Lesson 1

For the control group, the volunteer teacher was asked to document his lesson content
and sequences, but teach with the same textbook and target forms as in the other, ex-
perimental English class. Overall, his lessons in the control group were considerably
teacher-centered and grammar- and vocabulary-focused (see Figure 2).

Lesson #3: (in the classroom)

Target Forms: third person singular, subjective, possessive, and objective pronoun

– Warm-up Activity:
- Vocabulary quiz (10 min)

– Main Activity:
- The teacher helped students translate the content in the “Enjoy Reading”section into Korean by
explaining new vocabulary, sentence structure, and grammar. (20 min)

– Review:
- The teacher summarized the target grammatical rules with example sentences and reminded students of
the correct use of them. (10 min)

Figure 2.  A Sample Lesson Sequence in the Control Group

 Moonyoung Park

Learners in the control group were presented target grammar, sentence structure,
and vocabulary in the reading section first by the teacher, and then led to practice
them by translating the reading texts and solving questions in the textbook. There
were few chances for students to interact in pairs or groups. Instead, they engaged in
individual activities to solve grammar or reading comprehension exercise drills in the
textbook.
Pre-test, post-test. To compare the two groups and to measure change resulting
from experimental treatments (i.e., the task-based approach), a pretest-posttest design
was used. Following guidelines for designing performance assessment (Norris, Brown,
Hudson, &Yoshioka, 1998), two sets of task-based writing tests were developed – one
set for Unit 1 and one set for Unit 2 – to measure participants’ performance in ap-
proximations of real-life, authentic writing tasks.
The performance assessment was designed to meet three requirements: (a) exam-
inees must perform some sort of task, (b) the tasks must be as authentic as possible,
and (c) the performances are scored by qualified raters (see Shohamy, 1995; Wiggins,
1989.) The theme, target structures, and grammar of each unit from the English text-
book were also considered in the design of the tests. A sample pre- and post-test for
unit 1 is provided in Figure 3.
The pre- and post-test writing data were scored by two trained raters, who scored
performances using an analytic rubric which was developed based on Norris et al.’s
(1998) research on performance assessment. The criteria consisted of four subcatego-
ries: content, language, and organization (20 points each), and task completion
(40 points), with 100 points as the highest score. As the given test focused on achieving
a specific goal, task completion comprised 40% of the score. Pearson’s correlation coef-
ficient was computed to assess the interrater reliability between the two raters. The
results indicated a strong correlation between the two raters: (r = 0.95, n = 244,
p = 0.000). To check the reliability of the pre- and post-tests at separating students into
distinct performance ability levels, they were first administered to an additional class
(N = 28) at the same advanced level. For this additional reliability test, Cronbach’s al-
pha values for the two sets of pre- and post-tests were 0.82 and 0.97, deemed sufficient
for the current study.
After the implementation of CATBLT, a retrospective written survey was conduct-
ed in order to see students’ and the teacher’s reaction toward this English curriculum
with the embedded CATBLT. The survey focused on the effectiveness of CATBLT in
English language skills, preference of participation style in CATBLT, preferred lan-
guage learning activities and materials, and an overall evaluation. I also wanted to in-
vestigate the feasibility of CATBLT in the Korean EFL secondary school context. For
the quantitative data, descriptive statistics revealed students’ evaluative feedback on
the implementation of CATBLT. A five-point Likert-scale was used in the closed-re-
sponse questions (1: Strongly disagree, 2: Disagree, 3: Neither agree nor disagree, 4:
Agree, and 5: Strongly Agree; or 1: Not at all likely, 2: Somewhat unlikely, 3: Neither
unlikely or likely, 4: Somewhat likely, and 5: Very likely). The qualitative data from the
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Pre-Test

Task 1. Writing an e-mail to American Host Family

- Imagine you just received an e-mail from John, a member of your American host family who lives in the
U.S. army base in our hometown.

Dear (your name),

Hi! I hope you still remember our dinner gathering this Sunday. There is one more seat available, so you
can bring your best friend Minsu. Please let us know his age, hometown, hobby, favorite food, and food he
doesn’t like for better preparation. Hope to hear from you soon!
Sincerely,
John

- After reading the e-mail, you sent a text message to Minsu asking his age, hometown, hobby, favorite
food, and food he doesn’t like. You just received a text message from Minsu with his personal information.

Name Age Hometown Hobby Favorite food Food he doesn’t like

Minsu 12 Bokhyun-dong Playing computer games Pizza Fried chicken

- Now it is time for you to send an e-mail to Mr. John Smith introducing Minsu with his personal
information. Please complete the e-mail in the blank below in 20 minutes.

Post-Test

Task 1. Writing an e-mail to your E-pal

- Imagine you just received an e-mail from your E-pal, Mike, from Canada.

Dear (your name),

Hello! Thank you for your kind e-mail! I am also happy to be your E-pal. Oh, you said your favorite person
is your brother, right? Please let us know more about your brother. Hope to hear from you soon!
Sincerely,
Mike

- After reading the e-mail, you interviewed your brother and got his personal information.

Name Age Appearance Favorite Sport Favorite Food Food he doesn’t like
Minho 15 Tall and thin Swimming Kimchi Pizza

- Now it is time for you to send an e-mail to Mike introducing your brother, Minho, with his personal
information. Please complete the e-mail in the blank below in 20 minutes.

Figure 3.  Sample Pre- & Post-Test in Unit 1

students were translated, categorized, and tabulated. In terms of the teacher’s data, as
there was only one teacher participant who actually implemented CATBLT, individual
quotes were simply used to explain his perceptions.

Procedures

On the first day of the English lesson, which occurred during spring semester in 2010,
participants of both control group and experimental group were told that they were
taking part in a study on the implementation of a task-based teaching approach in
general English curriculum settings and given information describing the study. They
 Moonyoung Park

were then given up to 20 minutes to complete the pre-test. From the second to fourth
class hours, participants in the experimental group were taught using the task-based
materials. When units 1 and 2 ended (one unit requires five class meetings), both
groups were given 20 minutes to complete the post-test after each unit.
All participants also completed a unit-review test at the end of each unit, which
consisted of 20 multiple-choice questions focusing on vocabulary, reading compre-
hension, and grammar for a given unit.

Data analysis

All test data were analyzed for descriptive statistics to identify whether there were pre-
post differences for either group, and whether the groups differed from each other. To
investigate whether any observed differences were statistically significant,, an analysis
of variance with repeated measures was also conducted. The between-subject factors
were the control group (traditional approach) and the experimental group (task-based
approach), and the within-subject factors were test scores on the pre-test and the post-
test for both Unit 1 and Unit 2. An alpha decision level of p < .05 was set for all
inferential tests. To examine participants’ reactions to CATBLT, an open-ended online
survey was administered to the experimental group participants.

Results and discussion

RQ 1. How do the learners perform on the task-based writing assessment following les-
sons delivered via two different teaching approaches? Is there any meaningful
difference?
Pre- and post-tests were used to identify any meaningful difference of participants’
task-based writing performances. The pre- and post-tests were delivered before and
after each unit. The data regarding possible effects of CATBLT were analyzed using
descriptive statistics and repeated measures analysis of variance (ANOVA). Descrip-
tive statistics of the pre- and post-tests from unit 1 indicate that the TBLT group
(M = 82.17, SD = 14.12) and the control group (M = 82.42, SD = 12.31) had similar
mean values in the pre-test (see Table 1), which indicates both groups earned mean
scores as high as 82 points out of 100 points in the test. On the post-test, the TBLT
group (M = 90.67, SD = 14.12) performed better than the control group (M = 86.77, SD
= 8.06). Although the TBLT group performed slightly better than the control group,
the overall scores from pre- to post-test in both groups did not differ very much.
Table 2 shows ANOVA results for unit 1 pre- and post-tests. There was a main ef-
fect for time (pre- and post-test), which means participants experienced a significant
increase in the task-based writing assessment scores regardless of their group. And
there was a non-significant result for Time × Group (p = .09), which indicates that the
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Table 1.  Descriptive Statistics for Pre- & Post-test Writing Assessments

Unit Test Group N M SD

Unit 1 Pre-test TBLT 30 82.17 14.12

Control 31 82.42 12.31
Post-test TBLT 30 90.67 11.20
Control 31 86.77   8.06
Unit 2 Pre-test TBLT 30 66.17 21.92
Control 31 67.90 14.71
Post-test TBLT 30 83.50 13.01
Control 31 69.52 17.86

Table 2.  Analysis of Variance with Repeated Measures for Unit 1 Tests

Source SS df MS F Sig

Between Subjects
Group    100.99  1   100.99    .45 .51
Error 13269.50 59   224.91
Within Subjects
Time (pre/post)   1259.67  1 1259.67 28.34 .00*
Time × Group    130.98  1   130.98   2.95 .09
Error   2622.30 59    44.45
* p < .05

two groups’ test performances did not differ, statically speaking. There are several pos-
sible reasons for the similar test performance of the two groups. First, the task itself
seemed to be quite easy for both groups, exhibiting a ceiling effect in overall scores.
Second, the CATBLT treatment during unit 1 was only three hours of total instruc-
tional hours, which may have been too short a period of time to impact task perfor-
mance. Nevertheless, the task-based group did perform on average somewhat better
than the comparison group on the post-test.
According to the ANOVA results for unit 2 tests (see Table 3), the significance
value of Time × Group interaction was p = .002; therefore, the null hypothesis – that
CATBLT made no difference to learning gains in task-based writing assessment –
could be rejected, and the alternative hypothesis – that significant improvement did
take place and that group membership made a difference – could be accepted. That is,
the CATBLT group improved substantially (a 17 point improvement) and outper-
formed the traditional group (with a 14 point difference on the post-test), both to
statistically significant degrees. The experimental group students’ improvement in the
task-based writing assessment after three lessons of CATBLT treatment was encouraging
 Moonyoung Park

Table 3.  Analysis of Variance with Repeated Measures for Unit 2 Tests

Source SS df MS F Sig

Between Subjects
Intercept 628271.29  1 628271.29 1533.63 .00*
Group    1143.42  1    1143.42     2.79 .10
Error   24170.11 59     409.66
Within Subjects
Time (pre/post)    2736.33  1    2736.33    15.05 .00*
Time × Group    1883.87  1    1883.87    10.36 .002*
Error   10728.01 59     181.83
* p < .05

for stakeholders at the middle school. Participants received TBLT instruction through
unit 2. Thus, the significant improvement for the CATBLT group on the post-test after
unit 2 can most likely be attributed to the additional TBLT instruction and practice of
various communicative tasks.
RQ 2. How much do the learners from the two different teaching approaches achieve in
the traditional unit-review test developed by the school teachers and focused on
vocabulary, grammar, and reading comprehension? Are there any meaningful
differences?
After covering each unit, all first year students, regardless of their level, were required
to take a school unit-review test, which focused heavily on vocabulary, grammar, and
reading comprehension related to the unit. To investigate possible differences in per-
formance between the TBLT group and the control group on the two unit-review tests,
descriptive statistics (see Table 4) and independent t-tests (see Table 5) were used.
When comparing the two groups’ scores on the two unit-review tests out of
100 points, both groups performed well overall; however, the TBLT group performed
much better in Unit 2 based on the mean value.

Table 4.  Group Statistics in the Unit-Review Tests

Unit Group N M SD

Unit 1 TBLT 30 95.33   5.40

Control 31 93.23   5.85
Unit 2 TBLT 30 87.50 10.15
Control 31 77.90 11.60
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Table 5.  Independent Samples Test

Levene’s Test for t-test for Equality

Equality of Variances of Means

F Sig. t df Sig Mean

(2-tailed) Difference

Unit 1 Equal variances assumed 1.97 .17 1.46 59 .15 2.11

Equal variances not assumed 1.46 58.87 .15 2.11
Unit 2 Equal variances assumed   .46 .50 3.43 59 .001 9.61
Equal variances not assumed 3.44 58.42 .001 9.61

The t-test finding shows that there was no statistically significant difference between
the groups for unit 1, but there was a clear statistically significant difference in favor of
the TBLT group for unit 2. This result may support the positive correlation between
CATBLT lessons and form-focused test performance; moreover, it may be a promising
sign in that the task-based approach does not necessarily sacrifice students’ language
accuracy while leading to apparent benefits for learning to do communicative tasks.
RQ 3. What are student and teacher opinions of the CATBLT lessons?
After 10 class hours’ of implementation of CATBLT, a retrospective written survey was
conducted in order to obtain the students’ and the volunteer teacher’s reactions to the
use of CATBLT. The purpose of inquiring into the participants’ reactions was to better
understand the feasibility of innovations like CATBLT in the Korean EFL secondary
school context.

Students’ reactions

The students were asked to provide written feedback of their opinions of the CATBLT
lessons via an online survey. Thirty students were asked to rate their preferences for
various aspects of CATBLT instruction on a scale from 1 to 5. Tables 6, 7, and 8 show
survey results for participants’ (a) perceptions of the effectiveness of CATBLT for de-
veloping English language skills (Table 6), (b) preference of participation style (Table 7),
and (c) opinions about activities and materials (Table 8). Their written feedback on an
open-ended survey item was categorized into advantages, difficulties and concerns,
and suggestions, and is summarized, with the frequency of responses in each category,
in Table 9.
Table 6 shows the results from the questions asking students to rate their opinion
on the effectiveness of CATBLT for English language skills on a scale of 1 (“not effec-
tive at all”) to 5 (“most effective”). It is worth noting that the ratings are overall pretty
high on average. Students indicated CATBLT is most effective in improving learning
 Moonyoung Park

Table 6.  Students’ Views on the Effectiveness of CATBLT in English Language Skills

Item N M SD Rank

Practical Communication 30 4.33   .48 1

Writing 30 3.93   .79 2
Listening 30 3.87   .97 3
Reading 30 3.87 1.01 4
Speaking 30 3.70   .99 5

Table 7.  Students’ Preference of Participation Style in CATBLT

Item N M SD Rank

Teacher-centered 30 4.27 .53 1

Individual Work 30 3.90 .76 2
Pair Work 30 3.67 .76 3
Group Work 30 3.50 .97 4

Table 8.  Students’ Preferred Language Learning Activities and Materials

Item N M SD Rank

Computer (Internet) 30 4.37 .67 1

Pre/Post Test 30 4.30 .65 2
Worksheet 30 4.17 .71 3
Presentation 30 3.90 .76 4
Homework 30 3.83 .59 5
Word Processing 30 3.80 .66 6
Textbook 30 3.60 .81 7

and practicing communication skills (M = 4.33, SD = 0.48). Writing skills were ranked
second highest, followed by listening and reading skills with the same mean value
(M = 3.87). As for speaking skills, students viewed CATBLT as relatively less beneficial
to speaking skills than other skills despite the fact that they marked practical commu-
nication skills as benefiting the most from CATBLT instruction. This is probably be-
cause the target tasks focused on writing rather than speaking.
There was no difference in the students’ most preferred participation style in the
CATBLT evaluation, nor in the needs analysis survey administered four months prior
(See Table 7). Their most preferred classroom participation style was teacher-centered
(M = 4.27, SD = 0.53), followed by individual work (M = 3.90, SD = 0.76), pair work
(M = 3.67, SD = 0.76), and group work (M = 3.50, SD = 0.97). It is possible that a more
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Table 9.  Students’ Overall Comments on CATBLT

Summary of students’ comments Frequency

Advantages
Students could actually correspond with E-pals in English. 20
Students could introduce themselves in English. 14
Students were highly motivated and interested in the use of computer for language 12
learning and practice.
As tasks were very authentic, students became confident in real-life communica- 11
tion in English.
Difficulties/Concerns
Implicit grammar learning made students nervous about midterm and final exams.  8
Students were not sure how well or poorly they performed in each task.  5
Students were confused about task instructions.  4
Students were distracted due to unnecessary web surfing.  3
Suggestions
It may be more effective if native English teacher can teach in task-based way.  4
Teacher’s feedback on students’ writing may be helpful to them.  3

gradual transition from communicative activities toward performing target tasks, in

conjunction with careful teacher’s guidance, may help students adapt their preference
for participation styles and feel more confident and comfortable in TBLT.
Students were also asked to rate their preferred language learning activities and
materials used in the CATBLT classes on a scale of 1 (“not preferred at all”) to 5 (“most
preferred”) (Table 8). Students acknowledged the use of computers (and the Internet)
as the most preferred language learning tool (M = 4.37, SD = 0.67). It was surprising
that students ranked the pre- and post-task-based writing assessment as the second
most preferred learning tool. This finding indicates the potential value of task-based
language assessment as a learning and motivational mechanism in future implementa-
tion of CATBLT. The textbook was identified as the least preferred tool (M = 3.60, SD
= 0.81). These results can serve as a reminder to English teachers to be mindful in their
use of the textbook and to consider learners’ interests and needs.
Students provided numerous positive comments on CATBLT (Table 9). The ques-
tionnaire contained an open-ended question asking for students’ overall opinions of
CATBLT after they participated in the series of CATBLT classes. The most prominent
advantage of CATBLT was providing students with communicative tasks that are not
only motivating and reflecting their needs but also closely related to their curriculum
and textbook. The implementation of the task-based approach incorporated with com-
puter technology appeared to be a good match for language learning and teaching.
Students’ main concerns were that implicit learning of English grammar and key struc-
tures might negatively affect their scores on high-stakes tests and that there was little
 Moonyoung Park

individual feedback from the teacher, which is easy to fix in a fuller version of TBLT.
Students also pointed out that task instructions should be clearer. Another concern
was that inappropriate use of computers and the Internet in English class could dis-
tract them. As for suggestions, a few students mentioned that task-based approaches
might work better if a native English teacher implemented the class. They also ex-
pected individual feedback on their writing tasks from the teacher.

Teacher’s reactions

After implementing CATBLT over 10 class hours, the teacher participant was also
asked to complete the same survey. His overall impression of the advantages of CAT-
BLT was that students appeared to be highly motivated by the real-life tasks. Introduc-
ing authentic pictures and video clips to students enabled them to experience more
real-life tasks and gain more confidence in task performance. For the teacher, the most
serious limitation was the lack of individual feedback he could give to students, espe-
cially during small group activities. He also pointed out the need for preparatory ses-
sions for both English teachers and students to better prepare for the implementation
of TBLT. Like the students, the teacher worried that the computer and the Internet
could distract students with unnecessary web surfing or computer games. In terms of
pair or group activities, as students were not familiar with small group activities, the
atmosphere in the interaction became somewhat slack according to the teacher. Lastly,
he also expressed a concern about the potentially heavy work load in CATBLT lesson
development.

Conclusion

This study investigated the effectiveness of CATBLT in the secondary general English
curriculum in Korean EFL contexts. The findings indicate that CATBLT may be an
effective pedagogical approach and tool for helping young EFL students under a gen-
eral English curriculum. Students’ opinions of CATBLT were mainly positive, although
they did mention a few concerns. Nevertheless, student participants noted more ad-
vantages of CATBLT than problems. The teacher also expressed positive views, noting
that the task-based approach has a lot of potential. This study was part of a larger effort
to build program innovation and evaluation into the development and delivery of an
effective and needs-based English language program. As part of a language program
development and evaluation and sequence, this CATBLT implementation should not
just be considered theory testing or a demonstration of program effectiveness; its re-
sults should be used to improve language programs as they are put into practice.
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

The current study has limitations that should be acknowledged. Because this study
focused on a specific instructional context (i.e., private middle school students in Ko-
rea) the findings are not generalizable to other contexts or L2 populations. Further, the
small number of participants and short experimental period are a major limitation.
Studies should continue to expand the empirical basis of CATBLT by including more
diverse EFL school contexts and learners. Finally, in order to examine the feasibility of
CATBLT in the Korean EFL context, broadly speaking, more longitudinal studies need
to be conducted throughout a semester or academic year so that accumulated findings
can contribute to building a valuable basis for implementing CATBLT at national cur-
riculum levels in the Korean educational system.

References

Anderson, J. (1993). Is a communicative approach practical teaching English in China? Pros and
cons. System, 21, 471–480.
Bygate, M., Skehan, P., & Swain, M. (Eds.). (2001). Researching pedagogic tasks: Second language
learning, teaching, and testing. London: Longman.
Chapelle, C. (1998). Multimedia CALL: Lessons to be learned from research on instructed LA.
Language Learning & Technology, 2(1), 22–34. Retrieved March 25, 2010, from http://llt.
msu.edu/vol2num1/article1/index.html.
Chapelle, C. (2001). Computer applications in second language acquisition. Cambridge:
Cambridge University Press.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Garrett, N. (1991). Technology in the service of language learning: Trends and issues. Modern
Language Journal, 75(1), 74–101.
Hahn, J.-W. (2008). Challenges of task-based language teaching: With reference to Korean EFL
teachers’ beliefs. English Language & Literature Teaching, 14(3), 49–68.
Kim, Y.-J. (2009). The role of task complexity and pair grouping in the occurrence of learning op-
portunities and L2 development. Unpublished Doctoral dissertation, Northern Arizona
University. Retrieved June 15, 2010, from Dissertations & Theses: Full Text. (Publication
No. AAT 3370628).
Ko, K.-S. (2004). Task-based language teaching applications for network-based language teach-
ing. Modern English Education, 5(2), 39–53.
Ko, M.-H. (2008). Korean college students’ and a teacher-participant’s reaction to TBLT. English
Teaching, 63(3), 25–44.
Lee, J. F. (2000). Tasks and communicating in language classrooms. Boston, MA: McGraw-Hill.
Lee, J.-W., & Lee, K.-J. (2005). A model of the learning materials for middle school multi-pur-
pose English classes through TBL framework. English Language & Literature Teaching,
11(4), 335–363.
Lee, M.-J. (2006). The effect of L1 on task-based foreign language learning of Korean learners.
Foreign Languages Education, 13(2), 257–282.
Lee, S.-M. (2005). The pros and cons of task-based instruction in elementary English class. Eng-
lish Teaching, 60(2), 185–205.
 Moonyoung Park

Li, D. (1998). “It is always more difficult than you plan and imagine”: Teachers’ perceived diffi-
culties in introducing the communicative approach in South Korea. TESOL Quarterly, 32,
677–703.
Long, M. H. (1985). A role for instruction in second language acquisition: Task-based language
teaching. In K. Hyltenstam & M. Pienemann (Eds.), Modeling and assessing second language
acquisition (pp. 77–99). Clevedon, UK: Multilingual Matters.
Long, M. H., & Norris, J. M. (2000). Task-based teaching and assessment. In M. Byram (Ed.),
Routledge encyclopedia of language teaching and learning (pp. 597–603). London: Routledge.
McDonough, K., & Chaikitmongkol, W. (2007). Teachers’ and learners’ reactions to a task-based
EFL course in Thailand. TESOL Quarterly, 41(1), 107–132.
Ministry of Education and Human Resources. (2002). The 7th national curriculum for English
education [online]. Retrieved May, 2010 from http://www.moe.go.kr.
Ministry of Education and Human Resources. (2007). The revised 7th national curriculum for
English education [online]. Retrieved May, 2010 from http://www.moe.go.kr.
Norris, J. M. (2009). Task-based teaching and testing. In M. Long & C. Doughty (Eds.), Hand-
book of language teaching (pp. 578–594). Cambridge: Blackwell.
Norris, J. M., Brown, J. D., Hudson, T., & Yoshioka, J. (1998). Designing second language perfor-
mance assessments. Honolulu, HI: University of Hawai‘i Press.
Nunan, D. (1991). Communicative tasks and the language curriculum. TESOL Quarterly, 25,
279–295.
Ortega, L. (2009, September). Tasks and technology in language learning: Elective affinities and
(dis)encounters. Plenary delivered at the 3rd International Task-Based Language Teaching
Conference, Lancaster, UK.
Park, G. P. (2001). The effects of linguistic ability, background knowledge, and question types on
EFL listening comprehension ability. English Teaching, 56(2), 245–263.
Park, M. (2010). A needs analysis for a Korean middle school EFL general education curricu-
lum. In J. Davis & G. Lamb (Eds.), Proceedings of the 14th College-Wide Conference for
Graduate Students in Languages, Linguistics, and Literature, (pp. 12–24). Honolulu, HI:
University of Hawaii, Mānoa.
Prabhu, N. S. (1987). Second language pedagogy. Oxford: Oxford University Press.
Shin, H. & Kwon, C. (1999). Computer mediated instruction in English education: Practice and
implications. Journal of the Applied Linguistic Association of Korea, 15(2), 201–218.
Shin, J. (2001). Adaptation of TBL approach for English class in primary schools. Primary Eng-
lish Education, 7(2), 164–185.
Shohamy, E. (1995). Performance assessment in language testing. Annual Review of Applied Lin-
guistics, 15, 188–211.
Singhal, M. (1998). The Internet and foreign language education: Making the most of the infor-
mation superhighway. Language Education, 26(2), 125–39.
Skehan, P. (1996). A framework for the implementation of task-based instruction. Applied Lin-
guistics, 17, 38–62.
Thomas, M., & Reinders, H. (2010). (Eds.). Task-based language learning and teaching with tech-
nology. London: Continuum.
Van den Branden, K., Bygate, M., & Norris, J. M. (Eds.). (2009). Task-based language teaching: A
reader. Amsterdam: John Benjamins.
Wiggins, G. (1989). A true test: Toward more authentic and equitable assessment. Phi Delta
Kappan, 70, 703–713.
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Willis, D., & Willis, J. (2007). Doing task-based teaching. Oxford: Oxford University Press.
Yoon, K. (2004). CLT theories and practices in EFL curricula: A case study of Korea. Asian EFL
Journal. 6(3), Article 1. Retrieved March 25, 2010, from http://www.asian-efl-journal.com/
september_04_yke.php.
Yoon, M. & Lee, C. (2001). A study of using CALL software for the improvement of university
students reading speed in English. Multimedia-Assisted Language Learning, 4(1), 86–121.
 Moonyoung Park

Appendix A

Task-Based writing assessment

Task 1–1. Writing an E-mail to American Host Family
Task Guidelines for Students

Your writing will be assessed on the following:

Content The content needs to be relevant and sufficient.

Relevant means the content is meaningful to the task topic, and
Sufficient means that there is enough content (i.e. not too little and not too
much).
Language – You need to make use of a range of grammatical and sentence structures
accurately.
–  You need to use a variety of vocabulary and expressions accurately.
–  Your punctuation will be assessed.
–  Your spelling needs to be accurate.
Organization Content/Ideas should be presented logically and grouped together or
separated in meaningful ways.
Task Completion You need to follow the task requirements and complete the goal. For
example, a task requires including two possible options. Therefore read and
follow directions carefully.
Leave enough time to proofread your writing.
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Task 1–1. Writing an E-mail to American Host Family (Answer Sheet)

 Moonyoung Park

Task 1–2. Writing an E-mail to E-Pal

Task Guidelines for Students

Your writing will be assessed on the following:

Content The content needs to be relevant and sufficient.

Relevant means the content is meaningful to the task topic, and
Sufficient means that there is enough content (i.e. not too little and not too
much).
Language – You need to make use of a range of grammatical and sentence structures
accurately.
–  You need to use a variety of vocabulary and expressions accurately.
–  Your punctuation will be assessed.
–  Your spelling needs to be accurate.
Organization Content/Ideas should be presented logically and grouped together or
separated in meaningful ways.
Task Completion You need to follow the task requirements and complete the goal. For
example, a task requires including two possible options. Therefore read and
follow directions carefully.
Leave enough time to proofread your writing.
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Task 1–2. Writing an E-mail to E-Pal (Answer Sheet)

 Moonyoung Park

Task 2–1. Giving Transportation Information

Task Guidelines for Students
 Chapter 10.  Task-based language teaching in the Korean secondary EFL context 

Task 2–2. Giving Transportation Information

Task Guidelines for Students
 Moonyoung Park

Task-Based Writing Assessment Rating Criteria

Points 20 15 10 5

Content Meaningful Either content to Somewhat Inadequate

content to the the topic, or ideas inadequate content to the
topic, enough to the target task content to the topic, irrelevant
content, relevance is presented in a topic, less relevant ideas to the
of ideas to the less relevant way ideas to the target target task
target task task
20 15 10 5
Language Essentially error Mechanical and Repeated Mechanical &
free. Evidence of usage errors that weaknesses in usage errors so
superior control do not interfere mechanics and severe that
of grammar, with meaning usage. Pattern of writer’s ideas are
spelling, punctua- flaws hidden
tion and target
structures
20 15 10 5
Organization Content and ideas Content and ideas Either content or Neither content
presented presented in a less ideas was not nor ideas was
logically and logically and presented presented
well-organized grouped in less appropriately and appropriately
comprehensible was not grouped and was not
ways in comprehensible grouped together
ways in meaningful
ways
40 30 20 10
Task Superior Completion of the Partial comple- Minimal
Completion completion of the task, missing one tion of the task in completion of
task in an or two elements a less effective the task, missing
effective way of the target task way, missing most elements of
including all several elements the target task
elements of the of the target task
target task
 chapter 11

Task-based language teaching through

film-oriented activities in a teacher
education program in Venezuela

Carmen Teresa Chacón

The goal of this study is to explore ways of enhancing EFL students’ oral skills
using task-based language teaching (TBLT) through film-oriented activities in a
teacher education program in Venezuela. Students’ diaries, recordings, and focus
group interviews were used to gather data from 50 third-year students enrolled
in the program. The study was carried out over a ten-week period and consisted
of three stages: pre-task (planning), task (delivering an oral report in class), and
post-task (producing film-based activities). The study found that implementing
TBLT through cooperative learning projects using films was successful and
beneficial for L2 learning in multiple ways, including improvements in the
students’ fluency and intelligibility in L2, their listening comprehension, and
their vocabulary building skills. The project was also a useful source that
exposed students to different accents of the English language and fostered
collaboration among students, which ensured more authentic and purposeful
communication. In line with the recent EFL teaching curriculum reforms in
Venezuela, which recommend the adoption of TBLT for teacher education
programs in the country, the main implication of the findings of the study is that
through TBLT, these would-be teachers were able to both develop their English
competence as learners, and at the same time gain firsthand knowledge about
the TBLT methodology and practice in order to implement it successfully in
their future teaching situations.

Introduction and context of the study

In Venezuela, English is taught as a Foreign Language (EFL), and it is a required aca-

demic course to earn a secondary school diploma. During secondary education, stu-
dents attend English classes three times a week. Classes last 80 minutes per session and
class sizes are usually between 38–40 students. The new national curriculum for sec-
ondary schools was launched in 2007 by the Ministry of Popular Power for Education.
Since then it has changed from an objective-oriented and teacher-centered syllabus to
 Carmen Teresa Chacón

a task or project-based syllabus. Syllabus negotiation and student-centeredness are

major changes in the recent curriculum. For foreign language learning, the mandated
goal as stated by the Ministry of Popular Power for Education (2007) is “to use oral and
written language as a means for communication with the rest of the world and as a
means of accessing scientific and humanistic knowledge” (p. 15).
In line with this reform, curriculum planning for teaching foreign languages –
English, in our case – demands a shift towards alternative ways to teach foreign lan-
guages involving students in social interaction and integrating into the curricula the
competences on learning how to know, how to do, how to be, and how to live and co-
exist together (Ministry of Popular Power for Education, 2007). To comply with these
demands, teacher education programs have been required to focus on developing the
prospective teachers’ competence to speak, read, write, and understand English, and
this shift has motivated the introduction of a task-based language teaching (TBLT)
approach into these programs. It was hoped that through TBLT, these would-be teach-
ers would (a) develop their English competence, and (b) become familiar with this
new approach in order to implement it in their future teaching situations.
The course Experiences for Developing Oral Expression in English II is mandatory
for third-year EFL students enrolled in the teacher education program where the pres-
ent study took place. The main goal of the course was to improve students’ intelligibil-
ity and fluency in English. In order to achieve this goal, TBLT was utilized to achieve
the following two specific objectives: (1) to engage prospective EFL teachers in social
interaction through tasks created using films as real-life activities; and (2) to help them
improve their fluency and intelligibility in spoken English. It is important to point out
that although fluency and intelligibility were the main goals of the course, tasks were
constructed to integrate the four skills of listening, speaking, reading, and writing in
the L2 in order to achieve these outcomes.

TBLT in EFL settings and authentic, purposeful communication

Over the past decade of work in language education, there has been a shift from tradi-
tional methods to a post-methods era. As Richards (2002) explains, the post- methods
era changed the focus from a search for the best method to the exploration of pro-
cesses involved in teaching and learning based on reflective teaching and action re-
search. Thus, among the common recommendations for current views of language
education, teaching and learning should rely on approaches that are student-centered,
offer a mix of bottom-up and top-down processes, integrate technology, and support
students’ autonomy. Under this paradigm, TBLT has been of great interest for second
language acquisition (SLA) researchers interested in instruction as well as for language
teachers interested in educational innovation (see, e.g., Brooks & Donato, 1994;
 Chapter 11.  Task-based language teaching through film-oriented activities 

Coughlan & Duff, 1994; Edwards & Willis, 2005; Ellis, 2009; Nunan, 1995; Shehadeh,
2005; Van den Branden, 2006; Van den Branden, Bygate, & Norris, 2009).
For instance, TBLT research suggests that dialogic interaction in collaborative
tasks may go a long way towards facilitating language learning (Brooks & Donato,
1994; Coughlan & Duff, 1994; Coulson, 2005; Pinter, 2005). Poupore (2005), for ex-
ample, found that on collaborative problem-solving tasks, learners have “more free-
dom to control the task and to control the language that they want to use” (p. 252).
Poupore argues that collaborative problem-solving tasks give learners “more opportu-
nities to experiment with their language and to naturally discuss and negotiate ele-
ments related to task content, procedures, and personal experiences” (p. 253).
In English-as-a-foreign-language contexts in particular (as was the case with the
current study), where English is usually taught in formal classroom settings without
real advantages of classroom-external input or sufficient communication opportuni-
ties, learners need to be exposed to different types of authentic input and interaction
in the classroom in order to learn the L2 more effectively (Ellis, 1997). Ellis argues
that this can be accomplished through the use of authentic tasks. That is, different
types of input and interaction which include group work and pair work, and in which
both form and meaning are emphasized, can foster collaboration and facilitate lan-
guage learning through the use of authentic communicative tasks in the classroom.
Based on this approach, tasks are understood to be “the kinds of activities in which
people engage in order to attain some non-linguistic objective, and which involve or
necessitate the functional use of language” (Van den Branden, 2009, p. 269). Van den
Branden’s definition also implies that authentic task outcomes cannot be entirely pre-
dicted or controlled as in the traditional teacher-fronted classroom because such
tasks resemble real-life situations, and these are dynamic, not static, and largely
unpredictable.
In light of these considerations, the study reported below sought to create collab-
orative and authentic tasks based on these principles of TBLT in an EFL teacher educa-
tion program setting in Venezuela. The researcher (also the teacher of the course) used
films as a source to set up narrative and description tasks designed to expose students
to the English-speaking culture, and to provide them with the input necessary to im-
prove their fluency in the L2. The researcher sought to engage these students in an
extensive listening activity for communicative tasks, so that they use both bottom-up
and top-down processes to decode the messages and “make use of the information
provided in the spoken text, not as an end itself, but as a resource to use in order to
achieve a communicative task” (Morley, 1991, p. 84). Besides being an authentic source
to encourage learners to train their ear for native speakers’ input and “English-like
‘blurrings’ into their speech” (Kenworthy, 1987, p. 79), films serve as a tool for expos-
ing students to different accents and making them aware of the subtleties of language
use and cultural diversity, while at the same time manipulating the required linguistic
items like features of connected speech, linkage, and structures.
 Carmen Teresa Chacón

Purpose and rationale of the study

The main purpose of this study was to explore ways of enhancing a group of EFL stu-
dent-teachers’ oral fluency and prosodic features of language, focusing on discourse
(or connected speech) and suprasegmental aspects as important features for improv-
ing fluency, using film-oriented activities. The rationale for using task-based language
teaching with films as input is based on three reasons. First, in our context, students
rarely interact with native speakers and, therefore, they need input from authentic
sources such as films to train their ears for suprasegmentals. Second, task-based lan-
guage teaching emphasizes fluency and communication. Third, in our context, ELT is
more focused on grammar (form) rather than meaning (communication), so students
are more likely to achieve accuracy than oral fluency (Chacón, 2005). Thus, task-based
teaching through film-oriented activities was chosen to meet these objectives: (1) to
make students aware of fluency-related features in native speakers’ speech; (2) to en-
gage students in social interaction in order to perform communicative tasks in English;
(3) to develop social skills to work in cooperative teams; and (4) to encourage the use
of strategies for facilitating listening comprehension and oral production. These objec-
tives were motivated by a number of considerations.
First, in his review of research on task-based learning and instruction, Ellis (2009)
notes that most studies on task-based performance are based on quantitative research
conducted in laboratories or test situations (see also Van den Branden, 2006 for a
similar observation). In response to this concern, the present study sought to create
film-oriented activities to be carried out in part as an-out-of class assignment, giving
learners extended time for advance planning and preparation before they delivered an
oral report about their film in class. With this goal in mind, the study used a qualitative
research methodology aimed at examining TBLT for oral production in a non-con-
trolled, less threatening atmosphere where students could proceed at their own pace
with no time constraints.
Second, TBLT principles were investigated here to see whether and to what degree
they foster students’ interaction in the L2 in this EFL setting. Meaning-oriented tasks
can draw on films as materials to address the need to provide learners with the neces-
sary exposure to the L2. This is quite important in the EFL case of Venezuela, where
English is usually taught following a traditional, linear curriculum in formal class-
rooms, where there is hardly any time or opportunity for interaction or negotiation of
meaning, and where teachers face numerous constraints such as lack of materials and
school facilities, tight schedules, and large classes (Chacón, 2005).
Third, films were popular among these students. I therefore drew on film content to
create a project-based task (PBT) that enabled students to show their understanding of
the film (Appendix A). Through paraphrasing and reformulating the input received,
students completed two-way open-ended tasks based on narratives and descriptions.
Students in cooperative groups of four were free to choose the film they wanted to
watch, adapt the tasks to their own interests and needs, and develop other film-activities
 Chapter 11.  Task-based language teaching through film-oriented activities 

aimed at improving their English oral fluency and prosodic features. As an outcome,
students delivered an in-class oral report of their own version of the film watched.
Fourth, research shows that sometimes the suprasegmental aspects of language
play a more important role than segmental ones in enhancing intelligibility (Field,
2005). In particular, Field argues that prosodic features contribute to intelligibility
while “the occasional insertion of nonstandard phoneme should not grossly disrupt
communication” (p. 402). Although intelligibility is considered a key aspect of pro-
nunciation, it is a ‘slippery’ notion because intelligibility is often judged using the na-
tive speaker as the point of reference (Field, 2005; Morley, 1991). In the present study,
Kenworthy’s (1987) definition of intelligibility as “being understood by a listener at a
given time in a given situation” (p. 13) has been adopted because for nonnative speak-
ers the goal is not necessarily acquiring a native-like pronunciation.
Finally, prosodic features of language in authentic speech like the ones used in
films are fundamental to foster comprehensibility. Therefore, English learners need
training in prosody and instruction in listening for linkage, intonation, and word stress
so that they notice these phonological features and raise their consciousness about
them (Kenworthy, 1987; Morley, 1991).

The study

Working within a qualitative research paradigm, this study was framed using an action
research methodology (Carr & Kemmis, 1986) following the cyclical process of plan-
ning, acting, observing, and reflecting throughout the implementation of the TBLT
approach in this teacher education program setting in Venezuela.

Participants

Fifty third-year EFL prospective teachers, 32 females and 18 males, participated in the
study. All participants were native speakers of Spanish who ranged in age from 21 to
26 and who were enrolled in the course Experiences for Developing Oral Expression in
English II, in a teacher education program in Venezuela. Class met twice a week for
90 minutes each session. The majority of the participants had no experience travelling
to or living in an English-speaking country. Their English proficiency level was consid-
ered high intermediate by the program’s diagnostic test administered at the beginning
of the year. A student at this level is able to read, write, understand, and speak in
English intelligibly but without complete control of structure and pronunciation.

Research questions

The following questions guided this study:

 Carmen Teresa Chacón

– How do film-oriented tasks foster language learning?

– What abilities can students improve through film-oriented tasks?
– What are the students’ perceptions about cooperative learning?

The project-based task

At the beginning of the course, the students brainstormed several projects (e.g., role-
plays, watching films, storytelling, drama, using music) they could do to improve their
fluency and intelligibility. Films were chosen as meaningful and relevant to the par-
ticipants’ interest. They agreed to watch a film of their preference as an out-of-class
activity, get together as groups for planning and preparation, and then deliver an oral
report in class. The oral report included open-ended tasks. The rationale for using this
type of task was based on the fact that students’ responses were not fixed or limited; on
the contrary, open-ended tasks allowed spontaneous unstructured responses about the
film watched (see Appendix B for examples). In addition to the tasks I designed, each
team was in charge of designing two tasks for their classmates to accomplish during
the oral report delivery (see Appendix C for examples). These tasks-within-tasks were
achieved in class as a way to involve the whole group in the tasks set out for each film.
Following Kenworthy (1987), the open tasks included both analytical listening
and self-critical listening. For instance, in these tasks, the students transcribed film
segments of their choice and predicted features of connected speech (e.g., intonation,
linking, assimilation); then, they compared their predictions with the actual native
speakers’ talk (see appendix D). According to Kenworthy (1987), “[i]n doing analytical
listening activities learners discover what native speakers of English do; in self-critical
listening, the learners check their own performance against a native-speaker model
and their own pronunciation goals and priorities” (p. 78). For both analytical and self-
critical listening activities, students received in-class instruction about the features of
connected speech like intonation patterns, linking, assimilation, and stress.
Students also performed problem-solving tasks to decide what other activities they
could create based on the film, besides narration, description of main characters, and
sequencing the film events. During the pre-task activities, the instructor monitored the
students’ interaction and helped them clarify difficulties and questions about the proj-
ect. Directions and procedures were explained in class so that the teams understood
what was expected of them. The project was conducted during a ten-week period and
consisted of three stages: Pre-task (planning), Task (delivering an oral report in class),
and Post-task (producing film-based activities), following Willis’s (1996) framework.

Pre-task: Planning

The planning phase prepared learners to carry out the PBT. The follow-up tasks took
about six weeks to complete. During the pre-task phase, most of the negotiation and
interaction was carried out in class. Learners were encouraged to select films of
 Chapter 11.  Task-based language teaching through film-oriented activities 

personal interest from a list provided by the instructor containing issues of social rel-
evance. Right after that, they received explicit training about listening strategies they
could use to aid listening comprehension. However, they were also encouraged to use
any other learning strategies they found appropriate to them.
A worksheet (see Appendix A) with the criteria for the PBT was distributed and
discussed in class to make sure that students understood each stage. The teams allo-
cated mini-tasks among members, planned the activities together, and carried out
most of them as at-home activities. In-class time was devoted to teacher and peer input
and assistance with the project.

Task: Oral report

Once students watched the film and summarized the events, they were instructed to
rehearse before delivering the oral report in class. Teams drew on the pre-task activi-
ties to move on to the next learning goals of the PBT. The allocated time for the oral
report, including the showing of film excerpts and peer-activities, was 25 minutes per
team, with each team member speaking around five minutes. The oral presentation
included the following elements.
A narrative task. This task required learners to watch the whole film and then sequence
the events chronologically in order to organize them in a summary of the storyline.
The fact that they had to recall and retell the development, climax, and resolution of
the story made them work together to figure out meaning from context as well as use
paraphrasing to put the storyline together in a summary. A description of main char-
acters and their roles was also part of the oral report.
Language focus. Teams were required to select two specific film segments of one to
two-minutes in length and burn them on a CD for peers to listen to and work with.
Then, they were asked to create film-oriented activities for listening comprehension,
vocabulary building, and practice of features of connected speech taken from the film
characters’ speech – linking, assimilation, elision, reduced forms, and intonation. Most
importantly, all activities were related to comprehension, vocabulary, and features of
connected speech which were analyzed and compared as to how these features were
articulated by the students in Spanish and in English.

Post-task: Review

Teams also created post-viewing, film-oriented tasks and used them with their peers
right after they delivered their oral report. According to Skehan’s (2007) recommenda-
tion for the post-task stage, teachers should plan for activities that allow students to
practice new language and build upon it to internalize it and consolidate the new
structures. In the post-task stage, the participants put into practice the new language
 Carmen Teresa Chacón

knowledge and integrated it into their schema. Despite the focus on phonological fea-
tures, the students worked on accuracy as well, as they had to carry out more form-
focused tasks, and on aspects of language complexity too, as they had to elaborate
more to produce useful language structures.
In completing the above tasks, students used films as the primary source of input
to manipulate language, make sense of unfamiliar vocabulary and idiomatic expres-
sions, and understand the different registers used in the film. Watching the film was an
authentic task that allowed students to find new ways of expressing what was happen-
ing in the film. The tasks got students involved in problem solving and collaborative
activities in preparation for the oral presentation. Team members had to improvise,
paraphrase, rehearse, and reorganize language before delivering their oral report. In
addition, writing, rewriting, and editing were skills they needed in order to write a
coherent summary and create film-based activities for their peers.

Assessment and evaluation

Throughout the PBT, learners were engaged in self-assessment and peer-assessment so

that they assessed their own and each other’ strengths and weaknesses as they ad-
vanced in the completion of tasks. Students were also individually assessed by the in-
structor based on their performance during the oral presentation. Rubrics were used
for individual and peer assessment. The criteria evaluated for individual assessment
were the use of pauses, hesitations, repetitions, reformulations, presence of linking,
blending, reduced forms, and intonation, while peer assessment criteria were based
on peers’ oral performance involving intelligibility during the oral presentations
(see Appendix E).

Data collection

Data were collected from three main sources: focus group interviews, diary entries
with students’ reflections, and audio-recordings of the students’ presentations in the
manner illustrated below.
Focus group interviews. Focus group interviews were conducted to gather data about
the strategies used to complete the task. Teams were invited to meet with the teacher
to talk about their project. The following questions were asked to each focus group of
twelve learners (three teams per session): What was it like for you to participate in this
project? What strategies did you use to understand the film? Focus-group interviews
were recorded and later transcribed for analysis purposes. Pseudonyms are used
throughout to protect students’ identity and confidentiality.
Audio-recordings. I audio-recorded the students’ oral performances during each in-
class report, took notes, and collected all materials used – worksheets, film segments,
 Chapter 11.  Task-based language teaching through film-oriented activities 

and overheads. Twelve recordings (one per group) as well as students’ documents
(worksheets, PW presentations, CDs) were the products I collected from each team’s
performance to analyze aspects related to fluency and the use of suprasegmentals in
connected speech.
Language learning diaries. Diaries are a commonly used tool for gathering data on
students’ experiences, insights, and perceptions in acquiring an L2 (Bailey, 1990). Stu-
dents in this study were prompted to write diary entries every week to keep track of the
learning language strategies they used while doing the PBT. They were guided to re-
flect upon their individual and group experience of watching the film and the steps
they followed to understand what they saw and heard in the film.

Procedure of analysis

All data were analyzed thoroughly, through examination and coding for any emergent
themes and recurrent patterns. Through open coding the collected data were broken
into units of analysis, which were words and phrases. Then, the data were reassembled
through axial coding in order to find meaningful relationships between the codes from
open coding. The extracted categories went through conceptual selective analysis for
the selected categories. As a result of the codification process a total of 100 units of
analysis were identified, and the outcomes of the analysis are further described below.

Findings

Data from the focus group interviews and students’ diaries were analyzed using the
software ATLAS-ti version 6.0 (Muhr, 2006). The categories and subcategories shown
in Table 1 emerged from the ATLAS-ti analysis of the diaries and interviews. Docu-
ments (groups’ worksheets, CDs, and overheads) were also analyzed to compare them
with what students’ reported both in the diaries and in the interviews. Table 1 below
shows the frequency results, in terms of the number of times each category and sub-
category was mentioned.
Table 1 reveals that the category language competence including listening, speak-
ing, and the four skills has a subtotal frequency of 47 out of the 100 units of analysis
while the category strategies development for learning and teaching, that is, cognition,
autonomy, and the use of audiovisual aids, has a subtotal frequency of 34. The least
frequency mentions appeared for the category cooperative learning with 19 out of 100.
Within each subcategory, listening was the ability with most frequent mentions, reach-
ing a frequency of 16, whereas the use of audiovisual aids has a frequency of 15, and
group work has 14 mentions within cooperative learning. Of all, the most frequent
mentions were for listening (16) and the least frequent mentions was for principles of
cooperation (5).
 Carmen Teresa Chacón

Table 1.  Distribution of frequencies by categories and subcategories

Categories Subcategories frequency

Strategies Cognition    7
Autonomy   12
Audiovisual aids   15
Subtotal 1   34
Audio-Oral    8
Competence Four skills   11
Speaking   12
Listening   16
Subtotal 2   47
Cooperative Learning Principles of Coop    5
Group Work   14
Subtotal 3   19
Total 100

In the following section, I explain in more detail the areas of improvement as self-re-
ported by the group of prospective teachers who participated in this study.

Areas of reported teaching and learning improvement

Figure 1 shows the number of times participants reported improvement in some as-
pects of their own skills, which included not only language abilities but also teaching
skills.
The numbers in the figure represent the frequency in each case. As Figure 1 sug-
gests, listening and speaking were the language skills that had the highest frequency of
prospective teachers reporting improvement. Autonomy, group work, and use of au-
diovisual aids were also reported quite frequently as areas of improvement.
In the next section, I provide indicative testimonies gathered during the focus
group interviews and expressed in the diaries. These testimonies particularly referred
to improvement in language abilities (listening, speaking, vocabulary building), strat-
egy development for learning and teaching (autonomy, use of audiovisual aids, cogni-
tion strategies), and student perceptions towards the use of cooperative learning as an
approach for achieving the PBT.

Language abilities

In regard to language abilities, listening and speaking, most participants commented

positively on their progress in those areas. One of the students said: “I improved all
 Chapter 11.  Task-based language teaching through film-oriented activities 

Speaking 12

Principles of coop 5

Listening 16

Group work 14
Subcategories

Categories
Strategy
Four skills 11 Competence
Cooperative
learning
Cognition 7

Autonomy 12

Audiovisual aids 15

Audio-oral 8

0 5 10 15 20
Frequency

Figure 1.  Frequency of mention of different areas of improvement

skills but particularly my listening and speaking because when we analyzed each seg-
ment, we had to focus our attention to decode and interpret the message for the dis-
cussion.” Another stated: “I trained my ear to listen and my speaking improved with
the help of my peers.” Students also reported gains in practicing features of connected
speech by analyzing film dialogues. One student wrote in her diary: “I trained my ear
to recognize different accents and understand English at natural speed. Now, I can
understand my peers and teacher better.” Another one, Francys, wrote: “I think I im-
proved my fluency, I can communicate more easily. Also I understand faster when I
listen to people speaking English in films, TV, news, etc.” Similarly, Alexandra, wrote:
“I learned to identify and recognize falling and rising intonation and some reduced
forms, and linking.” Some diary entries also showed students’ progress on the four
skills. For instance, José wrote: “I improved all skills because I had to read and write
while doing research about the film and listening to the film and watching it in prepa-
ration for the presentation.”
Vocabulary building and pragmatics were also mentioned as gains. Learning slang
and idioms was, according to the participants’ testimonies, of particular interest to
them. It is worth noting that language in textbooks does not generally expose students
 Carmen Teresa Chacón

to ‘authentic’ speech, while films do, for they are intended for native speakers. Dia-
logue in films is much richer and more varied in conversational exchanges. Thus, most
of the students reported that they learned vocabulary they had never found in their
textbooks. In a focus interview, Ana, for example, stated: “I learned a lot of new words
that were not part of the colloquial English I use.” Furthermore, learning new vocabu-
lary from spoken input accompanied by contextual clues seemed to motivate students
to learn vocabulary and incorporate it into their schema. Film segments can be re-
wound to allow for the repeated retrieval of words and phrases which learners can
figure out or guess using contextual clues or look up in dictionaries while they watch
the film. In fact, most students reported that they benefited from looking up words and
searching for meaning and pronunciation in online dictionaries to understand film
messages and unfamiliar or new words. During a focus-group interview one student
said: “We worked together to write the script of the discussion and to design the ac-
tivities for the class. Also we used a lot of internet and dictionaries to look for new
words, expressions, and IPA [International Phonetic Alphabet] transcriptions.”

Strategy development for learning and teaching

The tasks demanded extensive listening preparation outside the classroom. In com-
pleting these tasks, teams developed their own learning and teaching strategies as they
solved problems and interacted to comprehend the film and design the activities for
the class. Based on the group work, it can be said that positive interdependence was
built while they carried out the tasks, in particular the oral report delivery. In the dia-
ries and during the focus interviews, the participants commented on the strategies
they used to aid comprehension and task achievement. The following section explains
the strategies the students said they used and which were reported as gains towards the
achievement of the PBT.
Incorporating Information and Communication Technology (ICT). Most of the students
said that they searched the Web to find information about the film they were working
on. In an effort to comply with the tasks, the students developed their own strategies
and managed to find information using search engines and websites such as http://
www.imdb.com, http://movies.yahoo.com, http://www.blockbuster.com, http://www.
youtube.com, http://www.filaffinity.com, www.movies.nytimes.com, http://www.cap-
tionswap.com . They reported reading film reviews on the internet and searching for
clips, trailers, and scene scripts. A team member explained: “We watched the film over
and over to select the segments we would use. After selecting the scenes, we down-
loaded the scripts and watched it again.” Another one stated: “The first thing we did
was looking the film on the internet, and then watched it individually. A few days later
we met to talk about the film, plot and characters and agreed on the scenes for class.
We met again to practice our performance.”
 Chapter 11.  Task-based language teaching through film-oriented activities 

Using captions. The use of captions in English differed among students. Captions re-
fer to on-screen English text designed for hearing impaired viewers that can be acti-
vated to facilitate students’ comprehension of what is being said. Subtitles are direct
Spanish translations of film dialogues. Some students reported that reading captions
helped them understand while others refused to read captions and preferred to con-
centrate their attention on listening and predicting, taking notes and inferring based
on the nonlinguistic clues provided by the context. One student said: “We saw the
film with no subtitles and then we put on the captions to compare if our predictions
and comprehension were right. We realized that what we missed were expressions
totally new to us.”
Using L1. During the focus groups interviews, some students mentioned that they used
L1 to read film subtitles and to negotiate meaning particularly during pre-task plan-
ning. For some, it was important to read subtitles because as one team member said,
“we consider that is very important not to lose the essence of the plot. We first saw it
with subtitles in Spanish and then in English with no subtitles.” Other teams, however,
preferred not to use L1 subtitles: “we covered the subtitles in Spanish to listen to
English and identify vocabulary”; “we watched the film at least two times, listened to
the dialogues without subtitles; we transcribed by our own hand the scenes selected, to
show to our classmates”; “we saw the film without the subtitles to practice our listen-
ing. We listened four times to a specific scene and wrote the script. We tried to write
everything we heard and check up unknown words in the dictionary.”
Pausing and rewinding the film. Students also reported that watching the film at home
allowed them autonomy to work at their own pace and make the necessary adjust-
ments to meet their particular needs. Those for whom it was hard to understand ex-
pressed that pausing and rewinding the film were helpful strategies that enabled them
to check comprehension, predict and paraphrase. For instance, Rosa commented in
her diary: “When I listened to the scenes the first time I didn’t understand, but I re-
peated and repeated the scene and I could understand the dialogue between the actors.
I think this strategy was very good and effective.”

Student perceptions towards the project-based task

In their diaries, the students expressed their views about the cooperative PBT. Figure 1
indicates positive attitudes towards cooperative learning and group work. As a result
of peer-to-peer assistance and mutual support to fulfill task demands, students re-
ported improvement in their oral production in their diaries and focus interviews. In
reflecting upon the gains, during the focus interviews, one student explained: “we had
to work so hard on the film project, but we enjoyed spending time together to do that.”
Other students expressed similar views:
 Carmen Teresa Chacón

– Cooperative learning is an interesting experience. It allows us to learn from each

other to see our language’s strengths and weaknesses through our partners’ eyes
(Isabel).
– I think this is a very effective method because it creates bonds among team mem-
bers. Personally, it helped me because I worked with students that had a higher
level of comprehension in English than me, and because of that, I felt obligated to
work harder so my team could have good grade (Alex).
– I believe this kind of activity is very interactive because it keeps students’ attention
on the activity itself (Virginia).
– In doing the project, we worked together to write the script of the discussion, and
to design the activities for the class (Maria).
– It’s motivating because when someone selects a film it means he likes the film, and
that means he will keep himself focused on the task and share this with his team
members (Alejandro).

Peer assessment

Another way to report improvement was through peer assessment (see Appendix E).
Figure 2 shows assessments of oral presentation skills based on items evaluated during
peer performance using a scale from 1 “Needs Improvement” to 3 “Very good”. The
majority chose “Very Good” for items 1 through 4 meaning clear, intelligible, and well-
organized speech. Figure 2 shows that the highest Mean (2.62) was awarded for the
oral delivery of the PBT which suggests that students rated their peers’ speech as clear
and intelligible.

2
Mean

2,44 2,62
2,42 2,42
1

0
Gave an interesting Presented a Presented info. Spoke clearly,
introduction clear summary in an acceptable intelligibly,
manner confidently

Figure 2.  Peer Assessment of participant oral improvement

 Chapter 11.  Task-based language teaching through film-oriented activities 

Discussion

Based on self-reported data and self-assessment, results of the study suggest that the
cooperative task-based project employed in this study did foster student collaboration
and interaction. First, using films in this cooperative project as a basis for open-ended
tasks served to expose students to authentic language and extensive listening activity,
and it supported their autonomy and creativity. Collaboration gave students control
over the task and encouraged them to use their own strategies to cope with and over-
come language difficulties. In these two-way communicative tasks, performance was
not controlled, that is, the tasks were not created to elicit specific responses; on the
contrary, the students set up their own goals for the PBT, developed strategies to un-
derstand what they saw and heard, and to understand one another.
As reported in the findings, the participants had plenty of opportunities to ex-
change and negotiate meaning through decision-making and problem-solving activi-
ties. Through TBLT, they experimented with the language and used their own stories
and assumptions about the task and context. They re-structured the language to un-
derstand, narrate, describe, and deliver an oral report and to create new film-oriented
activities, all of which contributed substantially to enhancing the students’ fluency and
intelligibility in the L2.
Additionally, the participants reported that working on this cooperative TBLT
project was an informative and enriching experience in multiple ways. According to
their self-assessment, participants expressed that they improved their fluency and in-
telligibility in English, which were the main goals of the study. They reported making
progress in listening comprehension; they also said they learned new vocabulary and
developed awareness of different accents and connected speech features in native-
speakers’ discourse. Further, designing task-based activities for peers helped the par-
ticipants experience TBLT both as learners and as future teachers. They planned and
decided which tasks to use with their classmates.

Implications of the study

EFL in Venezuela, as in several other EFL settings, is taught in formal classroom situ-
ations. In these settings, learners do not have sufficient opportunities to interact in
English in the classroom in part due to large class sizes. In addition, teachers rely heav-
ily on explicit grammar teaching and dedicate little time to authentic oral skills, in-
cluding speaking and listening activities. It was necessary, therefore, to explore other
ways to bring elements of authentic communication into the classroom. Using this
film-oriented activity based on the principles of TBLT was one way of providing learn-
ers with such communication opportunities. As was shown here, utilizing TBLT
through film-oriented activities was reported to improve learners’ speaking fluency
and intelligibility in the L2, and it exposed them to a series of extensive listening
 Carmen Teresa Chacón

activities. On the other hand, the study also indicates the participants’ positive atti-
tudes towards the whole project. For instance, almost all the participants commented
on the importance and usefulness of working in pairs or groups to achieve communi-
cative ends through tasks. Further, they commented that getting together to create
their own film-oriented tasks was an enriching pedagogical experience on how to use
TBLT in the future.
In line with the recent FL teaching curriculum reforms in Venezuela introduced
earlier, which recommend the adoption of TBLT for teacher education programs in the
country, the main implication of the findings of this study is that through TBLT, these
would-be teachers are able to both develop their English competence as learners, and
at the same time gain firsthand knowledge about the TBLT methodology and practice
in order to utilize it (hopefully successfully) in their future teaching situations.

Constraints of the study and conclusion

The main constraint of the study related to building teams. This was a real challenge
because these students were not accustomed to cooperative learning. To combat this, I
created opportunities for in-class teambuilding and observed and supported team
members along the process, taking into account each member’s opinion and reaching
an agreement to complete the learning outcomes. In class, I devoted a good amount of
time to checking the progress of the projects and provided teams with the necessary
guidance to complete the tasks successfully.
Another constraint was that a few students did not like the group formation pro-
cedure followed. At first, some students resisted being assigned to specific teams
because they were used to selecting the group in which they wanted to work, and
therefore, they complained. It took patience and careful planning to figure out how to
bring cooperative learning principles into play, especially during the first weeks when
team members needed to build mutual support and social skills. Although it took
some students some time to get used to working in teams, they soon learned how to
solve problems and negotiate among each other so that every team member’s contribu-
tion was incorporated into the final report. After the first two weeks, I re-formed some
groups at some students’ requests. Regrouping was necessary because a few members
of some groups were frequently absent and were thus unable to contribute actively or
fully to the project-based tasks.
These constraints should be borne in mind when designing and implementing
cooperative project-based tasks in the future in other EFL settings similar to the one
investigated in this study. Despite these and any other challenges that the investigator
faced, like class size and unfamiliarity with TBLT (see introduction), the main finding
of this study was that implementing TBLT through cooperative learning projects using
films was reported by the participants as a beneficial experience for L2 learning in
multiple ways. According to their self-assessment perceptions, through TBLT they
 Chapter 11.  Task-based language teaching through film-oriented activities 

seemed to improve fluency and intelligibility in the L2, as well as listening comprehen-
sion and vocabulary building skills. These positive findings indicate at a minimum the
potential contribution of a task-based approach to language learning and to teacher
education in this, and other, EFL settings.

References

Bailey, K. (1990). The use of diary studies in teacher education programs. In J. C. Richards &
D. Nunan (Eds.), Second language teacher education (pp. 215–225). New York, NY: Cambridge
University Press.
Boyle, D. (Director) (2006). Slumdog Millionaire [Motion Picture] UK.
Brooks, F.B., & Donato, R. (1994). Vygostkian approaches to understanding foreign language
learner discourse during communicative tasks. Hispania, 77(2), 262–274.
Chacón, C. (2005). Teachers’ perceived efficacy among English as a foreign language teachers in
middle schools in Venezuela. Teaching and Teacher Education, 21, 257–272.
Carr, W., & Kemmis, S. (1986). Becoming critical: Knowing through action research. Geelong,
Australia: Deakin University Press.
Coughlan, P., & Duff, P. A. (1994). Same task, different activities: Analysis of SLA task from an
activity theory perspective. In J. P. Lantolf & G. Appel (Eds.), Vygostkian approaches to sec-
ond language research (pp. 173–193). Norwood, NJ: Ablex.
Coulson, D. (2005). Collaborative tasks for cross-cultural communication. In C. Edwards & J.
Willis (Eds.), Teachers exploring tasks in English language teaching (pp. 127–138). New York,
NY: Palgrave McMillan.
Demme, J. (Director) (1993). Philadelphia [Motion Picture]. United States: TriStar Pictures
Edwards, C., & Willis, J. (2005). (Eds.). Teachers exploring tasks in English language teaching.
Hampshire, UK: Palgrave Macmillan.
Ellis, R. (1997). SLA research and language teaching. New York, NY: Oxford University Press.
Ellis, R. (2009). The differential effects of three types of tasks planning in the fluency, complex-
ity and accuracy in L2 oral production. Applied Linguistics, 30(4), 474–509.
Field, J. (2005). Intelligibility and the listener: The role of lexical stress. TESOL Quarterly, 39(3),
399–423.
Haggins, P. (Director) (2004). Crash. United States: Lions Gate
Kenworthy, J. (1987). Teaching English pronunciation. New York, NY: Longman.
Ministerio del Poder Popular para la Educación. Subsistema Educación Secundaria Bolivariana.
Currículo y Orientaciones Metodológicas. [Ministry of Popular Power for Education.
Bolivarian High School Subsystem. Curricula and Methodological Orientations].
Morley, J. (1991). The pronunciation component in teaching English to speakers of other lan-
guages. TESOL Quarterly, 25(3), 481–520.
Muhr, T. (2006). Atlas-ti version 6.0. Berlin: Scientific Software Development
Nunan, D. (1995). Language teaching methodology. A textbook for teachers. New York, NY:
Phoenix ELT.
Pinter, A. (2005). Non-proficient children using interactive tasks in pairs. Paper presented at the
1st TBLT Conference, Leuven, Belgium.
Polanski, R. (Director) (2002). The Pianist [motion picture] United States: Universal Studios.
 Carmen Teresa Chacón

Poupore, G. (2005). Quality interaction and types of negotiation in problem-solving and Jigsaw
tasks. In C. Edwards & J. Willis (Eds.), Teachers exploring tasks in English language teaching
(pp. 242–255). Hampshire, UK: Palgrave McMillan.
Richards, J.C. (2002). 30 years of TEFL/TESL: A personal reflection. RELC Journal, 33(2),
1–36.
Shehadeh, A. (2005). Task-based language learning and teaching: Theories and applications. In
C. Edwards & J. Willis (Eds.), Teachers exploring tasks in English language teaching
(pp. 13–30). Hampshire, UK: Palgrave McMillan.
Smith, J. N. (Director) (1995). Dangerous Minds. [Motion Picture]. United States: Hollywood
Pictures.
Skehan, P. (2007). Tradeoff and Cognition: Two hypotheses regarding attention during task-
based performance. Paper presented at the 2nd International Conference on TBLT, Hono-
lulu, Hawaii.
Van den Branden, K. (Ed.). (2006). Task-based language education: From theory to practice.
Cambridge: Cambridge University Press.
Van den Branden, K. (2009). Mediating between predetermined order and chaos: The role of the
teacher in task-based language education. International Journal of Applied Linguistics, 19(3),
264–285.
Van den Branden, K., Bygate, M., & Norris, J. (Eds.). (2009). Task-based language teaching: A
reader. Amsterdam: John Benjamins.
Weir, P. (Director) (1989). Dead Poets Society [Motion Picture]. United States: Touchtone
pictures.
Willis, J. (1996). A framework for task-based learning. Harlow, UK: Longman Addison- Wesley.
Zwick, E. (Director) (2006). Blood Diamond [Motion Picture]. United States: Warner Bros.

Appendix A

Guidelines for the Project-Based Task: Your Favorite Film

Directions: In group, select a film of your preference from the list given. Arrange the
time and place to watch it together. Do the following tasks:
Task 1
a. Based on the title, make some predictions about the film.
b. Brainstorm possible vocabulary you expect to find in the film.
c. Watch the film and take notes about the main ideas or theme of the film.
d. Identify the main characters and describe their roles.
e. Summarize the plot – main events, climax, and resolution of the plot.
Task 2
a. Transcribe two segments (scenes) of your preference – one minute to two-minute
in length.
 Chapter 11.  Task-based language teaching through film-oriented activities 

b. Identify features of connected speech – linking, assimilation, word stress, etc. in

each segment to present in class.
c. Create two film-oriented activities to use with your peers.

Appendix B

Samples of open-ended tasks

Task 1. Summarizing the plot of the film
This narrative corresponds to the team that presented about the film “Blood Dia-
mond”.
The plot of BLOOD DIAMOND is developed in SIERRA LEONA (AFRICA). This
movie is based on the struggle that Solomon Vandy suffered to join up his family that
was brutally torn apart from him. Solomon was kidnapped and forced to work in a
mine field where he discovered a big pink diamond. Solomon hid the diamond and in
that moment a series of events started to develop. As the events went on, new charac-
ters appeared to make things harder for him. One of them is Danny Archer a south-
African smuggler soldier who heard about the pink diamond while he was in prison.
For Archer, the diamond was his ticket to leave Africa. Another character, a journalist
called Maddy Bowen played by the beautiful Jennifer Connelly was a journalist who’s
trying to write the big story. So as the movie developed, she made a deal with Danny
Archer. Maddy was willing to help Archer if he gave her the names of the diamond
merchants (big bosses) to write her story. Thus, they needed one another to achieve
their goals. So, these characters started a dangerous passage across the territory con-
trolled by the rebellious troops. Solomon’s main goal was to reunite with his family but
this would not be easy. The rest of his family was in the second largest refugee camp in
Africa, and his son had been recruited by the rebels while Solomon was gone. That’s
when Danny Archer and the pink diamond became his last hope to reunite his family.
Task 2. Expressing opinions and feelings about the films
The following sample is a comment on the film “Crash.”
It is a controversial movie because there are many scenes of discrimination and
racism. People are connected by the actions they do. We have to learn that we cannot
offend a person because of their color, their religion or their social conditions. This
movie teaches people to see the world in a different way. For example, in Venezuela
and specifically in Táchira State, there are citizens from several countries, and if we
discriminate those people we are not human beings because they are like us, and they
have feelings, and as human beings we are not different from them. The only difference
we can see is the nationality, the language they use and some costumes.
The following is an individual reflection about the film “Dead Poets’ Society.” In
this case, the team handed in individual reflections about the film.
 Carmen Teresa Chacón
 Chapter 11.  Task-based language teaching through film-oriented activities 

Appendix C

Samples of film-oriented tasks each team designed for classmates

Task 1: You will be shown a segment of the film “The Pianist” three times. Fill in the
blanks with the words or phrases you hear. The first time, listen; the second time fill in,
and the third time check your responses.
I hereby order that all Jews in the Warsaw district will wear visible emblems when
out of doors. This decree will come into force on the   1st of December 1939 and ap-
plies to all Jews over twelve years of age. The emblem will be worn on the right sleeve
and will represent a blue star of David on a white background. The background must
be sufficiently large for the star to measure eight centimeters from point to  point.
The width of the arms of the star must be one centimeter. Jews who don’t respect this
decree will  be severely punished.
Government of Warsaw District, Dr. Fisher
Task 2. Filling in the blanks with the missing words.
You will be shown a segment of the film “The Pianist.” Watch the scene and mark the
intonation using arrows
Kitty: Open this door at once or we’ll call the police.
Are you from the flat in there? You’re not registered
Szpilman: It belongs to a friend of mine. I came to visit but I must have just missed
him
 Carmen Teresa Chacón

Kitty: Have you got your identity card?

Let me see your identity card. I want to see your identity card
He’s a Jew! He’s a Jew! Stop the Jew!
Don’t let him out!
Task 3 Part A: Identifying reductions
Watch the following scene from “Dangerous minds” and fill in the blanks with the re-
ductions you hear.
Task 3 Part B: Identifying linking
Watch the scene again and listen for linking. Mark it using a ç .
Callie: I realized. . . Thisç is my last chance andç I decided. . .
We decided that we’re not gonna just letç you leave like that.
Raul: Yeah, we realized like the poem said: “You can’t giveç in, you can’t go
gentle, you’ve got to rageç against the dyingç of the light.”
Student A: Yeah, you gotta go for yours. You know that, right?
Louanne: Waitç a minute, waitç a minute. No, no, wait, wait. I’m not givingç in.
Thisç is my choice. I have no reason “to rageç against the dyingç of the
light.”
Callie: You’re not the one who’s raging. We’re the ones who are raging. See, cause
we seeç you as beingç our light.
Louanne: What?
Student B: You’re our Tambourine Man.
Louanne: I´m your drug dealer?
Student C: You’re our teacher. You got what we need. It’s the same thing.
Raul: Começ on, Miss “J” all the poems you taughtç us say you can’t giveç in,
you can’t giveç up. We ain’t giving youç up.
Student C: No way. Now, listen baby we’re gonna haveç to tie you down to the chair
and getç you 'cause we know we wantç you to stay.
Task 4: Creating questions
Based on the summary of the movie “The Pianist,” discuss in your group the answer to
the following questions:
1. What do you think about discrimination?
2. What do you think about Wladyslaw’s courage?
Task 5: Predicting events
Based on the title of the film, students asked questions for predictions before present-
ing each oral report.
For the film “Crash” questions were:
 Chapter 11.  Task-based language teaching through film-oriented activities 

What does Crash mean for you? What do you expect from this film?
For the film Philadelphia, the questions were:
What do you know about AIDS? Who has watched the film?
Post viewing activities
The following two samples were used as Post viewing activities. Team 1 created a post-
er for the film Slumdog Millionaire while team 4 created a newspaper about Dead
Poets’ Society to share with peers.

Another example of a Post viewing activity was the game, who wants to be a million-
aire? created by the team 1 to play it with classmates.
 Carmen Teresa Chacón

Who is the main character of the movie? Young Jamal worked as a tourist guide at…?

A: Jamal B: Salim A: Himalaya B: Mumbai

C: Danny D: Jashan C: Taj Mahal D: Jaipur

Where was the movie filmed? How many brothers and sisters does Jamal have?

A: Madurai B: New Delhi A: One brother B: One brother and one sister

C: Madras D: Mombay C: Two Brothers D: Five Brothers

On an American one hondred dollar bill there Who is the director of the movie ?
is a portrait of which American Statesman?

A: George Washington B: Benjamin Franklin A: Steven Spielberg B: David Lean

C: Abraham Lincoln D: Franklin Roosevelt C: Danny Boyle D: Martin Scorsese

 Chapter 11.  Task-based language teaching through film-oriented activities 

Appendix D

Transcribing movie segments to identify features of connected speech

The following sample transcript from the film Philadelphia shows intonation,
linking, and word stress.
 Carmen Teresa Chacón

Appendix E

Peer Evaluation
Directions: For each of the following, select from 1 to 3 to evaluate your peer’s perfor-
mance during the oral presentation.

Very Good Satisfactory Needs

Improvement
3 2 1
  1.  Gave an interesting introduction
  2.  Presented a clear summary of film events
  3.  Presented information in an acceptable order
  4.  Spoke clearly, intelligibly and confidently
  5.  Maintained eye contact
  6.  Presentation was interesting
  7.  Film activities were well-planned and engaging
  8.  Film segments were thought-provoking
  9.  Used visual/audio aids well
10. Offered a concluding summary

Comment:
 chapter 12

Task-based language teacher education

in an undergraduate program in Japan

Daniel O. Jackson
University of Hawai‘i at Mānoa, USA

Task-based language teacher education can assist participants in adopting (a)

roles and responsibilities based on their needs, (b) collaborative approaches
to professional development, and (c) communicative language teaching
practices. The classroom study presented in this chapter examines aspects of
novice teacher cognition among 15 participants in a one-semester, task-based,
undergraduate seminar on language teaching methods. A mixed-methods
investigation included retrospective comments, classroom discourse, and survey
results as data on the effectiveness of the task-based teacher education approach.
Findings of the study indicated that seminar participants both gained and shared
knowledge related to teaching practice through classroom tasks. In addition,
participants and their peers rated various recommended instructional practices
from the task-based language teaching (TBLT) literature highly. The chapter
discusses the potential of task-based teacher training to support curricular
innovation initiatives like TBLT designed to enhance language education in the
Japanese EFL setting.

Introduction

This chapter explores task-based undergraduate L2 teacher education in the context of

an intensive English program in Japan. In order to justify the application of task-based
instruction to this particular context, it begins with a review of the impact of policy
shifts by the Japanese government on public school educators. To demonstrate the
implementation of task-based instruction in teacher preparation courses, a seminar on
TESOL methods, designed with Japanese schoolteachers’ needs in mind, will then be
described. Findings from a mixed-methods study conducted in the seminar will be
discussed in terms of specific knowledge and attitudes displayed by participants. Based
on these findings, it is suggested that task-based teacher training may prepare the next
generation of schoolteachers to face the challenge of reforming English language edu-
cation in Japan.
 Daniel O. Jackson

Policy background

In order to address the need for global communication skills, the Japanese Ministry of
Education, Culture, Sports, Science and Technology (MEXT) announced a strategic
plan in 2002 aimed at improving English abilities among Japanese citizens (MEXT,
2002). One initiative within the plan was the establishment of Super English Language
High Schools (SELHi) intended to aid in the promotion of exemplary models of
English education (MEXT, 2003). A recent comparison study showed that SELHi stu-
dents outperformed students in regular Japanese high schools on measures of their
confidence in using English as well as their listening, reading, and writing proficiency
(Yoshida, 2008). Task-based language teaching (TBLT) is among the curricular inno-
vations being implemented at SELHi schools, such as Kawagoe High School in Mie
Prefecture, where tasks provide a way for learners to use the English they are studying
(Matsuzawa, 2004). However, the policy’s emphasis on enhancing communicative
English abilities poses a number of unique challenges for classroom teachers in the
Japan schooling context.
Scholarly views vary regarding the impact of the subsequent Action Plan to Culti-
vate Japanese with English Abilities (MEXT, 2003) on teachers. Butler and Iino (2005)
discussed the plan’s seven target domains of improvement, which include English
classes, teachers’ abilities, motivation for learning English, evaluation, elementary
school English, Japanese language abilities and practical research. They argued that the
plan reflects conflicting ideologies, raising questions about whether its aims are to
promote English versus multiculturalism, language education versus international un-
derstanding, and egalitarian versus individualized forms of education. At the same
time, they commented favorably on the greater teacher autonomy afforded by the plan
and anticipated teachers’ active participation in shaping language policy as a result of
its introduction.
Aspinall (2006) focused on how the innovations proposed by the ministry may be
thwarted at the local level, by preoccupation with university entrance exams (which
traditionally do not test communicative skills), educational values that seek to avoid
placing students who struggle with language learning at further disadvantage, and
day-to-day issues such as maintaining an acceptable level of noise in the classroom so
as not to disturb neighboring classes. He suggested that small cultures (Holliday, 1999)
of stakeholders at the local level may be needed to support the changes initiated by
new policies. Finally, Gottlieb (2008) located the plan in the wider context of the de-
velopment of language planning and policy in Japan. She cited positive responses by
students, parents, and teachers to various measures implemented in connection with
the action plan, viewed the initial results from the SELHi project as encouraging, and
discussed perhaps the most controversial aspect of the plan: English as a compulsory
subject in elementary school. The broader implications of this move remain to be seen,
yet research has shown that practicing elementary school teachers in South Korea,
Japan, and Taiwan struggle with defining the goals of communicative language teaching
 Chapter 12.  Task-based language teacher education in Japan 

in local contexts, matching foreign language activities to students’ developmental lev-

els, and situating foreign language learning and teaching within their specific contexts
(Butler, 2005).
This brief overview suggests that the ministry’s action plan may have a better
chance of succeeding if teachers entering the profession are willing to take on new roles
and responsibilities, develop a collaborative culture within their schools valuing the new
initiatives, and consider ways of implementing communicative teaching practices with
younger learners. While it would no doubt help, it may be unrealistic to expect novice
teachers to undergo training at the master’s or doctoral level, since certification ob-
tained through a bachelor’s degree program is the minimum requirement for teachers
in Japan1. The remainder of this chapter therefore focuses on undergraduate language
teacher training in Japan with a view toward illustrating how TBLT can be leveraged to
address the primary challenges facing new teachers under the Action Plan to Cultivate
Japanese with English Abilities. As Van den Branden (2006) observed in his discussion
of task-based teacher training in the context of the large-scale implementation of TBLT
in Belgium, “various partners will only be prepared to support...educational innova-
tion if each of them can find their own ‘truth’ and their own ‘worth’, in the innovative
concepts” (p. 248).

Rationale

The rationale for the classroom project described here is that task-based instruction
offers opportunities to develop language skills in concert with the strategic abilities
Japanese teachers need to attend to responsibilities, cultures, and practices, which,
crucially, involve classroom communication in the target language (Butler, 2004).
TBLT provides for, on the one hand, language development, and on the other, content
mastery and professional skill development. Also relevant are the growing number of
reports on putting tasks into practice in various educational settings in Japan
(e.g., Benevides & Valvona, 2009; Konoeda & Watanabe, 2008; Willis & Willis, 2007),
and the underlying holistic nature of both the case made for task-based learning
(Samuda & Bygate, 2008) as well as the educational philosophy behind Japanese school
teaching (Shimahara, 2002), which indicate that TBLT is perhaps increasingly and
aptly finding a home in Japan (see also Izumi, 2003).

The integrated English seminar: Approaches to TESOL

In this section, the TESOL seminar that formed the context for this study will be intro-
duced. In the first and second years of a four-year degree program, the English

1. However, a survey conducted in 1997 indicated that 61 percent of prefectural boards

of education were interested in increasing the number of teachers with master’s degrees
(Shimahara, 2002).
 Daniel O. Jackson

department at the university where this course was offered provides elective courses
that develop language skills in its Integrated English Program. These seminars typi-
cally follow a model of content-based instruction; however, syllabus design is negoti-
ated between instructors and program coordinators, such that a task-based framework
could be superimposed. Due to space limitations, the focus here will be on a sequence
of teaching tasks that students performed during weeks nine though twelve, although
there were other elements that lead to a characterization of the seminar as task-based.
The steps taken to develop this four-week phase of the seminar course can be de-
scribed in terms of six elements comprising task-based education presented in Norris
(2009), which include: needs analysis, task selection and sequencing, materials devel-
opment, teaching, assessment, and evaluation.
First, a triangulated needs analysis was conducted. To begin, a purpose for using
tasks was established through the identification of culturally relevant target tasks in
the literature on Japanese education (Butler & Iino, 2005; Sato & Asanuma, 2000;
Shimahara, 1998), prior to the start of the semester. Task implementation was then
planned in consultation with program coordinators who lent their insights on the use
of tasks and technology in seminar courses, among other areas. Subsequently, to vali-
date the foregoing activities and appraise students’ ability to successfully perform the
tasks, information on seminar participants’ language learning and teaching experi-
ences was gathered through a survey administered at the beginning of the course.
Second, task selection was primarily driven by the previously described literature
search but was, importantly, also constrained by the 15-week duration of the seminar,
the students’ workload, and the survey results, which indicated that few participants
had teaching experience. Planning a lesson, conducting a teaching demonstration, ob-
serving the lessons, and carrying out a debriefing session were thus found to be feasible
and appropriate target tasks for these developing teachers. Because it may be unlikely
for certain natural sequences of real-world tasks to adhere to a progression from sim-
ple to complex, in terms of cognitive criteria such as those found in Skehan (1996) or
Robinson (2001), the teaching tasks described here were sequenced according to com-
binability (Candlin, 1987), or their usual order of occurrence during in-service teach-
er development.
Third, materials were developed to make use of available classroom resources, in-
cluding computers mounted on each desk. Permission had been sought from students
in the previous year’s TESOL seminar to use video recordings of their teaching dem-
onstrations as models. Two of these videos, along with a collection of handouts related
to the target tasks were uploaded to an online course management system, which also
enabled (a) drop boxes, so that students could submit lesson plans to the instructor for
feedback, and (b) uploading and password-protection of the class members’ video-
taped teaching demonstrations, which they viewed privately to perform their observa-
tions. Students were also given assistance during class in creating their own teaching
materials.
 Chapter 12.  Task-based language teacher education in Japan 

Fourth, teaching involved actively planning and/or seeking out teachable mo-
ments. The teacher (i.e., the current author) modeled the language of the tasks though
the aforementioned videos and live, in-class demonstrations, but in addition offered
recasts and explanations of target language forms which learners were acquiring, for-
matively assessed task progress at each stage in the sequence, and guided students as
they managed various resources to complete the tasks.
Fifth, task-based performance assessment (e.g., Brindley, 1994) was conducted.
Twenty ‘can-do’ statements formed the basis of the scoring criteria for the planning,
teaching, observation, and debriefing tasks. These statements were presented to stu-
dents at the beginning of, and referred to intermittently throughout, the four-week
period so as to guide performance. The rubric was then later used by the teacher to
give summative feedback to all students upon completion of the target tasks.
Lastly, though recommended as an ongoing part of program development, evalu-
ation will not be addressed here except to mention that course evaluations in the previ-
ous year had indicated high levels of student satisfaction with a similar approach.
In sum, the seminar course combined a focus on content knowledge with a proce-
dural focus on teaching practice in weeks nine through twelve. During this time, class-
room tasks highlighted the real-world demands of teaching in the Japanese context via
attention to specific job characteristics, including collaboration in planning, deliver-
ing, observing, and reflecting on lessons. By doing so, these four tasks offered a space
in which participants could enhance their understanding of the course content by at-
tempting to apply that knowledge, but also one in which they could generate new
knowledge.

The present study

Though second language teacher education research has shown that pre-service ESL
teachers’ emerging beliefs and practices may be based primarily on their formal lan-
guage learning experiences (Johnson, 1994), few studies at the undergraduate level
have been published to date. The relative scarcity of research on undergraduate lan-
guage teacher training is perhaps surprising in light of Barduhn and Johnson’s (2009)
observation that, in general, a bachelor’s degree is sufficient to procure employment as
a language teacher in many settings. However, some examples of research with novice
teachers in EFL settings do exist. In one summer school project carried out in Japan,
an experiential, team-teaching model was used to enhance pre-service, university-
trained English teachers’ responsibility, empathy, self-esteem, communication skills,
teacher development, and English knowledge (Tanaka & Fukada, 2004). Experiential
models, including TBLT, may thus provide a valuable transition in offering future
teachers opportunities to participate in new roles, collaborate with peers, and acquire
skills for managing communicative classrooms (see also Chacón, this volume).
 Daniel O. Jackson

The study reported here was based on classroom research conducted in the semi-
nar course described above. The project was undertaken with several purposes in
mind, which included (a) experimenting with task-based instruction as a method of
language teacher training (Van den Branden, 2006), (b) exploring the kind of practical
knowledge acquired when tasks address not only language learning needs but societal
needs as well (Norris, 2009), and (c) investigating future teachers’ attitudes toward a
situated task-based approach, in which contextual factors interact with task-based
principles (Carless, 2007). Based on these considerations, the following research ques-
tions were adopted to gain insight into the broader construct of teacher cognition
(Borg, 2006) in this context:
1. In terms of practical knowledge, what outcomes do participants say they gained
through each of the teaching tasks?
2. What do classroom discussions suggest about the nature of practical knowledge
gained by participants?
3. How do participants’ attitudes toward task-based language teaching compare with
those of peers not enrolled in the TESOL seminar?

Methods

Participants

The participants in this study were second-year English majors at a private university
based in Tokyo. There were nine females and six males in the group, whose mean age
was 21. They had studied foreign languages for an average of 11 years, and all but one
of the 15 students had studied abroad, under various circumstances including home
stays and living abroad with family. Often, but not always, this time abroad was in
countries where the dominant language is English. Thus, their general proficiency in
English and their communicative ability could be characterized as advanced. In regard
to language teaching experience, one had some experience working at a cram school
and 5 others reported tutoring either friends or relatives in Japanese or English. At the
beginning of the semester, 7 participants indicated that they were interested in teach-
ing as a profession and this number increased to 12 at the end of the term. Addition-
ally, 15 second-year English majors not enrolled in the TESOL seminar also partici-
pated as a comparison group, as described later.

Procedures

One week of instructional time was devoted to each of the four previously mentioned
target tasks. Their implementation entailed various pedagogic tasks and student roles.
First, to design a lesson plan, students decided the focus and content for an interactive
 Chapter 12.  Task-based language teacher education in Japan 

lesson, downloaded a lesson plan template, completed the template in small groups,
and uploaded their finished plans to the course website for review. Second, to conduct
teaching demonstrations, students developed, rehearsed, performed, and, as called for,
participated in demonstration lessons based on their plans. Third, peer observations
incorporated tasks completed both in and outside of class. In class, students read and
discussed a bilingual (English and Japanese) account of peer observation as a method
of language teacher development (Croker, 2007), agreed on observation goals in small
groups, and exchanged information necessary to view password-protected online vid-
eos of their demo lessons. Outside class, they observed these lessons via the Internet
and took notes, based on the observation focus their group had chosen during class.
Fourth, for the debriefing task, students rejoined their groups in the next week’s class,
discussed their observations, exchanged views and opinions, and summarized these
discussions in brief oral reports to the entire class.
The classroom study reported here followed an adapted version of the conver-
gence model of triangulation design discussed in Cresswell and Plano Clark (2007).
Data collection involved mixed sources, including written retrospective comments,
classroom discourse, and a questionnaire. These are described in detail below.
Written retrospective comments. In order to investigate participants’ practical
knowledge, or the knowledge teachers generate from reflecting on experience
(see, e.g., Meijer, Verloop, & Beijaard, 1999), retrospective comments on each of the
tasks (i.e., lesson planning, teaching demonstration, observation, and debriefing) were
collected immediately upon their completion. A form was used to elicit comments
about each task, asking “what outcomes do you feel you have achieved so far?” Re-
sponses were written in English, the main language of communication between the
students and teacher. In total, 53 responses were gathered over a period of four weeks.
To assist further analysis, the teacher-researcher identified common themes found in
the data to derive a set of key categories, which were then used in coding. Comments all
related to one of five areas: classroom technique, learning from others, implementing
plans, language skills, and learner contributions. Each of these categories was then de-
fined and a coding scheme was developed. The instructor and a trained rater then cod-
ed the responses independently, leading to an inter-rater reliability of 91% simple agree-
ment. Discrepancies were discussed and the instructor made final coding decisions.
Classroom discourse. The class members, who were split into three groups, recorded
their discussions while debriefing at the end of the four-week task sequence. Three dig-
ital recorders were used to collect the oral data, which was subsequently transcribed
using CLAN software (MacWhinney, 2000). Each of the three discussions lasted ap-
proximately ten minutes, leading to a total of 30 minutes of data. For the purpose of this
study, which is primarily concerned with participants’ thoughts, knowledge, and beliefs,
the frequency of words and phrases related to each group’s observation focus was ana-
lyzed using the FREQ command. To identify segments of discourse representative of
participants’ knowledge, the KWAL (keyword and line) command was also employed.
Searches were conducted based on each group’s chosen observation focus (e.g., gestures);
 Daniel O. Jackson

however, transcripts were reviewed in their totality to identify exchanges related to the
co-construction of practical knowledge and interconnections between topics.
Tasks in language teaching questionnaire. Multiple steps were taken to develop the
Tasks in English Language Teaching questionnaire (TELT-Q) used in this study (items
are displayed in the results section). First, a review of the literature on task-based in-
struction was conducted. Based on the literature, a list of statements characterizing
TBLT was drawn up and then translated into Japanese by a native speaker who held a
master’s degree in language education. Further piloting and analyses followed proce-
dures discussed in Brown (2001). English majors from two classes – the TESOL semi-
nar and a comparison group of 15 students from a writing course at the same institution
– responded to the survey. Respondents were asked to imagine designing a course for
EFL learners in Japan and to indicate opinions on 17 items regarding course planning.
To establish informed consent, participants signed an agreement attached to the
questionnaire explaining the purpose of the study and guaranteeing that personal in-
formation would be withheld. The same assurance was given in class with regard to the
use of the written evaluations and oral data described above. Participants were encour-
aged to express any concerns they had about the study either directly to the researcher,
or if they preferred, to the English department.

Results

What outcomes do participants say they gained through each of the teaching tasks?

To answer the first research question, definitions and examples for each category of
comment elicited during the 4-week period of instruction, encompassing the lesson
planning, teaching demonstration, observation, and debriefing tasks, are displayed in
Table 1. The examples were taken from the data set and are intended to be representa-
tive of other comments made by the participants across the four tasks, although all
comments differed in detail. The task which students had worked on prior to each re-
flection is shown in parentheses. There was a tendency to reflect on outcomes using
verbs such as notice, share, get, learn, achieve, and understand, suggesting that partici-
pants themselves were active in constructing practical knowledge about teaching and
valued the experiences the tasks offered.
The number of comments in each category and the way these were patterned
across the four tasks varied as well. In Table 2, more detail is provided about the nature
of perceived outcomes under different task conditions. As might be expected from this
novice group, classroom techniques were a primary concern (34% of all comments).
The nature of the tasks, which required students to collaborate to design and imple-
ment a lesson, may have led to the large proportion of comments on learning from
others. The emphasis on lesson plan writing, teaching, and reflecting as a holistic pro-
cess, rather than discrete stages, may have encouraged the focus on implementing
 Chapter 12.  Task-based language teacher education in Japan 

Table 1.  Representative retrospective comments in response to the question:

“What outcomes do you feel you have achieved so far?”

Category Definition Example*

Classroom technique Knowledge related to techniques, I have noticed that being cheerful
practices, or activities used by is one of a very important
teachers in a classroom setting elements of teacher quality...I
have learned a variety of gestures
such as pointing, nodding, and
clapping. (debriefing)
Learning from others Knowledge related to processes and [I’ve gained] other point of views
products of discussing teaching and that I’ve never had before
learning with others because by talking with a lot of
people about same topic and
share some ideas, we get new
informations (debriefing)
Implementing plans Knowledge related to making, I’ve learned the importance of
adapting, or using lesson plans preparations in advance and at
the same time being flexible
(observation)
Language skills Knowledge related to the availability I think I’ve achieved some
of one’s language skills for use, speaking skills or communication
especially in a classroom setting skills (observation)
Learner contributions Knowledge related to the learner’s proper level of a lesson to proper
contribution or needs, whether age – we need to understand the
affective, linguistic, or otherwise student level and provide for the
class (teaching demonstration)
*Tasks after which these comments were written are given in parentheses.

plans across three of the four tasks (planning, demonstration, and observation, in or-
der of occurrence).
It is also interesting to note that although the debriefing task elicited the greatest
number of reflective comments (28%), not all comment categories were represented
here. Thus, considering the importance of reflection in teacher development more
generally, this finding supports the practice of inviting reflection during various stages
in a sequence of teaching tasks, not only upon concluding a lesson. Also, important
opportunities for restructuring practical knowledge in specific areas might be lost if
reflection is directed solely at what has happened during the lesson stage, since, as
shown, comments collected after the teaching demonstration amounted to only 25%
of the entire data set and did not include more than a single comment for three of the
five areas of practical knowledge investigated.
The sparse number of comments on learner contributions might reflect the fact
that the demonstrations were conducted with peers sometimes taking the role of
 Daniel O. Jackson

Table 2.  Frequency of comments elicited after each task

Task

Comment Category Debriefing Observation Teaching Lesson Total

Demonstration Planning

Classroom technique 4 5 4 5 18 (34%)

Learning from others 9 4 0 0 13 (25%)
Implementing plans 0 2 7 4 13 (25%)
Language skills 2 3 1 0   6 (11%)
Learner contributions 0 0 1 2   3 (5%)
Total 15 (28%) 14 (26%) 13 (25%) 11 (21%) 53 (100%)

students (but other times simply functioning as an audience for the demonstration).
While it was not feasible to arrange for demonstrations to be conducted at a primary
or secondary school in this case, on-site teaching practice is common at the later stag-
es of teacher training at the undergraduate level in Japan (see Yonesaka, 1999, for a
helpful overview).

What do classroom discussions suggest about the nature

of practical knowledge gains?

The next research question addressed the potential for classroom conversation to lead
to practical knowledge gains. Student talk during only the debriefing task is analyzed
below in order to provide a window on participants’ ongoing learning, since this task
came later in the sequence and was preceded by an observation task designed to focus
groups on features of the teaching demonstrations they believed to be important. In
addition, as noted above, it elicited the most written comments. Each of the three
groups selected an observation focus: gestures, check questions, and directions to stu-
dents (later, introductions) prior to watching the teaching demonstration videos for
homework, then regrouped in the next class meeting to talk about what they had seen.
In order to confirm that these topics were actually discussed during the approximately
10-minute debriefing task, the coded data were analyzed using CLAN. Explicit refer-
ence to the topic was made in the gestures (30 instances), check questions (10 instanc-
es), and directions (five instances) groups, although the latter group shifted their focus
to introductions, referring to this eight times.
Several excerpts from the conversations demonstrate that practical knowledge is
often not confined to individuals, but instead, shared across a group in ways that allow
developing teachers to elaborate, synthesize, or critique particular aspects of teaching
practice.
Excerpt 1 illustrates how students extended and then restricted their notion of the
role of gestures in classroom communication. S1 clearly believes that a variety of
 Chapter 12.  Task-based language teacher education in Japan 

paralinguistic elements in teacher talk serve to direct attention, maintain alertness,

and facilitate comprehension on the part of students. As seen in S2’s comment, there is
no single most important gesture; every gesture can be important. Finally, there is also
mention of the naturalness of gestures and their embodiment, which seems to suggest
that to take advantage of gesture as a tool for student-teacher interaction requires flu-
ency in using gestures and a willingness to use one’s hands to function communica-
tively. In excerpt 1, the contributions of S1 and S2 led to much greater elaboration than
was typically provided in written comments.
(1) S1: I think clapping, well the gesture of clapping is an important, because first
of all the students eh the students can get their attention and well they’ll
wake up if they get bored or something and also as everybody said the
pointing gesture is very important as well, because they’ll understand
better.
S1: what’s your important (.) huh what’s your favorite?
S2: I think every gesture’s important.
S2: I don’t know there’s no particular one.
S3: mmm.
S2: it depends on what comes naturally.
S1: on the hands.
Check questions, defined in class as questions used by a teacher to ensure that instruc-
tions are understood, were the focal point of the next group’s debriefing. This excerpt
comes from the middle of the conversation, at a point where the group is attempting to
summarize their discussion.
(2) S4: well you have to think about making gestures and using whiteboards and
and try not to make the students get bored and you need to make the bal-
ance between of t-t-t [teacher talking time] and s-t-t [student talking
time] so there’s many tasks to for teachers.
S5: yeah.
S4: but that’s the ideal thing for teacher (.) for good teachers they can a make
all those stuff bal all balance.
S6: okay.
S4: and check questions have to be have to be included in those parts in those
tasks in order to be good teachers.
S5: yeah, that’s pretty much all.
Here we see how participants were able to link the previous group’s concern for em-
ploying gestures with the main topic of their discussion, check questions (along with
other topics such as maintaining student interest, using classroom resources, and bal-
ancing classroom talk). S4’s initial comment shows that she considers these to be tasks
in their own right. Importantly, this comment combines elements of teaching practice
that were explicitly part of the performance criteria given to the class (i.e., strategic
 Daniel O. Jackson

language use and balanced student-teacher talk) with other goals valued by the semi-
nar participants (i.e., the use of gestures, classroom resources, and techniques for pre-
venting boredom). Task-based instruction combined with assessment criteria which
reference a minimum of the essential characteristics of performance may encourage
this kind of flexibility and participation more than tests aimed at discrete knowledge.
The group also affirms the interpretation of teaching practice as multi-faceted, offered
by S4, who continues on to note the importance of balancing this range of factors.
Finally, the last group, which initially chose directions to students as their focus but
subsequently revised this to introductions, debated over the role of the teacher. In the
demonstration video discussed here, the teacher employed a warm-up in which stu-
dents took turns writing English words on the board using each letter of the alphabet in
sequence. They had three minutes to do so. The teacher sat down and assumed the role
of timekeeper, monitoring the students’ production. After three minutes, the teacher
stopped the activity, which culminated in a list of words ending in X-ray. Then, after
confirming with the students that all words had been spelled correctly, the teacher in-
troduced a second activity by asking students to make sentences using the words. Thus,
while this might be considered an example of focus on forms, rather than focus on
meaning, or focus on form (Long & Robinson, 1998), the learners also produced much
more language than the teacher did. In the following excerpt, S8 was the teacher.
(3) S7: are wa chotto dame da to omotta. are wa ichiban dame da to omotta. [that
was a little bad, I thought. That was the worst, I thought]
S8: why?
S7: uh, because you didn’t teach anything the students.
S9: the students were told to do something and you were just sitting.
S10: yeah yeah yeah yeah yeah I remember now.
S8: I read with the students.
S9: there was no t-t-t.
S11: no t-t-t.
S9: student talking time too much.
S10: s-t-t-t-m [laughter]!
S9: demo saa kou sensei wa isyoukenmei oshiete seito ga kotaenaino to sensei
ga tekitou ni oshiete seito ga sugoi gambaru teiuno doushitara ii ka to
omou [but, well, which do you think is better: to have a teacher work
hard and students not respond much, or to have the teacher teach with-
out much effort and expect the students to try hard?]
S8: seito ni gambattemoraitai to omoimasu [I think I want to have the stu-
dents try their best].
Apart from S7 and S9, whose remarks single out the teacher, these undergraduates
generally spoke about performances in objective terms. Nevertheless, the teacher here
actively defends the lesson discussed. Instead of accepting either of the options S9
presents in the penultimate line of this excerpt, S8 appeals to the notion of students
 Chapter 12.  Task-based language teacher education in Japan 

trying their best. Doing so offers a cultural perspective on student-teacher roles that is
also in alignment with contemporary language teaching practice, in which students
are expected to produce language to convey personally relevant meanings (even if the
demonstration lesson only approximates this). However, the specific nature of the
teacher’s role remains unresolved; classroom management under the education minis-
try’s new initiatives is likely to require similar reflection.

How do seminar participants’ attitudes toward task-based language

teaching compare with those of their peers?

The 17-item TELT questionnaire described earlier was administered after all tasks
were completed to evaluate the attitudes of students enrolled in the TESOL seminar
(trainees) alongside those of peers who were not enrolled, but who were also second-
year English majors at the university (non-trainees). Responses were based on a four-
point Likert-type scale (0 = “strongly disagree” and 3 = “strongly agree”). Internal
consistency of responses was assessed using Cronbach’s alpha, indicating that the
instrument met standard criteria for construct measurement (α = .85). Multiple inde-
pendent t­-tests, with an adjusted alpha of .003 revealed no statistically significant dif-
ferences in the means of the two groups. However, as shown in Table 3, there were
several items on which the two groups diverged in terms of mean scores.
The largest difference in means concerned whether it is necessary to teach linguistic
items prior to a task. Non-trainees were much more likely to agree with this item. Sec-
ondly, more non-trainees believed that learning activities should focus on target lan-
guage use related to classroom learning. In contrast to this, the third largest mean
difference dealt with whether English should be taught for specific purposes outside of
the classroom. Here, the opposite trend was observed; trainees were more likely to
agree with specific purpose English courses. These findings suggest that the trainees
may have felt that focus on form can occur at any stage during a lesson, and that they
also felt more strongly than their peers that language education should correspond to
real-world purposes. Although these notions are fundamental to TBLT, the trainees
agreed less than non-trainees with an item regarding the effectiveness of TBLT for
learning (Item 1). However, average agreement was still high for this item (in fact, all
agreed it was effective, although more in the non-trainee group strongly agreed), and
any difference may reflect the view that the effectiveness of language teaching ap-
proaches cannot be determined a priori. High levels of agreement on the part of the
trainees were found for several other core principles in TBLT, including sequencing
tasks according to a gradual progression (Item 8) and negotiating meaning to make
input comprehensible (Item 10).
Several open-ended comments on the questionnaire indicated that students in the
seminar had already thought through the particulars of task-based teaching in the
Japanese context. Responding in Japanese to a prompt that asked for impressions or
 Daniel O. Jackson

Table 3.  Means comparisons for trainee and non-trainee attitudes toward TBLT

Trainees Non-trainees
(n = 15) (n = 15)

No. Item M SD M SD Diff.*

11 Before a task takes place, it is necessary to teach idioms 1.93 .70 2.60 .51 –0.66
or fixed phrases necessary for communication.
 9 Classroom activities should consist of language-learn- 2.13 .52 2.46 .64 –0.33
ing activities that are important to individuals as
students.
 4 English courses should be designed with specific 2.13 .83 1.87 .83   0.26
purposes in mind. Specific purposes can mean
Business English or English for Study Abroad.
 1 I think that task-based education is an effective 2.13 .35 2.40 .51 –0.26
learning method.
 7 The most important goal of English education is the 2.33 .72 2.60 .51 –0.26
teaching of effective ways of communicating in
English.
16 Teachers should determine content based on their 2.20 .56 2.47 .52 –0.26
understanding of students’ existing knowledge.
 6 Many different actual English expressions often contain 2.13 .64 1.93 .59   0.20
similar language use. For example, to make a dinner
reservation and to fill out a form both require stating
the date.
 8 The content of learning and activities should gradually 2.47 .64 2.67 .49 –0.20
move from easy to difficult in a step-by-step progres-
sion.
12 Learning should integrate various language items. For 1.66 .62 1.80 .68 –0.13
example, learners’ attention should be drawn to phrasal
verbs and compound nouns.
17 Classes should be divided by ability level and content 2.28 .61 2.40 .63 –0.11
should be prepared in accordance with student ability.
 2 Course content should be decided based on student 2.33 .82 2.27 .46   0.07
needs (the content they wish to study). For example,
content can be determined though surveys or
interviews to understand student needs.
 5 English courses should enable learners to use English 2.33 .72 2.40 .63 –0.07
freely in daily activities. For example, activities like
making a restaurant or hotel reservation, or filling out
a form.
13 Learning should focus on grammar and sentence 1.33 .74 1.20 .68 –0.07
patterns.
 Chapter 12.  Task-based language teacher education in Japan 

 3 To understand student needs, there are many other 2.13 .64 2.10 .80 0.03
methods apart from the surveys and interviews in item
2 above.
10 During English communication between students, it is 2.40 .63 2.40 .51 0.00
necessary to confirm whether there is mutual under-
standing. For example, through the use of expressions
such as, ‘What does that word mean?’ etc.
15 Learning should include classroom interaction through 2.33 .49 2.33 .49 0.00
group and pair work.
14 If there are errors in students’ English usage, a teacher 1.80 .68 1.80 .94 0.00
should point them out even during conversation.
*Items are ranked here according to mean differences between the two groups

opinions of TBLT, one emphasized the importance of the social context in task-based
teaching, commenting that Tokyo differs from rural parts of the country in terms of
opportunities for speaking English. A second mentioned balancing skill development
and the importance of raising awareness of gaps in individual students’ abilities for
furthering language study. Another student replied that educators must come to grips
with communicative teaching approaches, voicing strong support, as well, for the
adoption of English classes in elementary schools.

Summary and conclusions

In this final section, the main findings from the study are summarized in terms of the
three research questions, and some preliminary insights on pre-service teacher cogni-
tion in this context are offered. With regard to research question one, trainees, who
were asked to report practical knowledge outcomes in the context of a task-based ap-
proach to language teacher education, mainly commented on (a) classroom tech-
niques, practices, or activities, followed by (b) processes and products of learning from
others, and (c) making, adapting, or using plans. Concerning question two, analyses of
trainees’ classroom discussions suggested that the debriefing task enabled participants
to elaborate, synthesize, and critique practical knowledge and ideas that arose in con-
nection with the teaching tasks, in particular, the lesson observations. Question three
differed from the first two research questions in that it explored attitudes toward TBLT
(see also Chacón, this volume). While means comparisons between the trainee group
and a non-trainee peer group showed no significant mean differences, questionnaire
results revealed that trainees differed from the non-trainees in ways consonant with
practices advocated within TBLT. In addition, both groups surveyed rated various rec-
ommendations from the TBLT literature highly (see Table 3).
 Daniel O. Jackson

By using a mixed-methods design, this project has shown the possibility that task-
based undergraduate second language teacher education may be construed as an
important stage of professional development during which individuals have the oppor-
tunity to rethink their experiences and realign themselves with classroom practices in-
formed by contextual factors, their own schooling, and professionally-oriented course-
work, components which together form a model of teacher cognition (Borg, 2006).
Although this project cannot address all of the components in Borg’s model, it does offer
some insight into two important aspects of it: practical knowledge and attitudes.
Written comments indicated that the categories of practical knowledge that train-
ees were concerned with shifted across the tasks used in this study. In fact, no task
contained all five of the categories, which means that in order to obtain a fuller picture
of teacher cognition, it may help to elicit reflections at various stages of teaching prac-
tice. In terms of future studies, it might be useful to research whether a well-planned
task sequence can enhance the quality or quantity of written reflections. In contrast,
the debriefing task discourse showed that practical knowledge among teacher trainees
was manifested in ways unobservable in written comments. Therefore, comments
alone may obscure the richness of teacher cognition at the early stages. Studies should
thus acknowledge the role of tasks as well as the complementarity of data sources such
as individual reflections and group discussions when examining teacher cognition.
The survey data gathered for this study also permit a balanced conclusion in that
results generally did not show large differences between trainees and non-trainees in
regard to attitudes toward TBLT. Finding no significant differences in the mean scores
of trainees and non-trainees should not be interpreted as a shortcoming of task-based
teacher training. Rather, this result suggests that even those undergraduate students
who do not work through teaching tasks, but merely read a definition and examples of
task-based teaching, may favor many of its core principles (perhaps reflecting an over-
all shift in attitudes of emerging generations of future teachers). Additionally, there
was a trend toward seminar students agreeing to a greater extent with the flexibility
offered by TBLT in terms of the timing of form-focused instruction and the emphasis
that task-based approaches place on needs analysis. In their open-ended comments on
TBLT, these future stakeholders in English language education in Japan posed impor-
tant questions of nationwide implementation, added value, and teaching communica-
tion to younger learners.
Mertens (2005) points out characteristics of transformative mixed-methods re-
search, noting that it gives primacy to value-based and action-oriented dimensions
and aims to promote change (p. 297). Consistent with the values and actions inherent
in the Japanese Ministry of Education, Culture, Sports, Science and Technology’s plans
for revitalizing English education described at the beginning of this chapter, task-
based second language teacher education offers opportunities for novice teachers to
explore roles and responsibilities, develop a collaborative culture around language
teaching, and plan and implement communicative teaching practices designed for
school-based learners (see also Chacón, this volume for a similar argument on EFL
 Chapter 12.  Task-based language teacher education in Japan 

teacher education programs in Venezuela). However, relying on economic, social, or

political shifts will hardly guarantee such training opportunities. Teacher educators in
Japan and elsewhere must pay closer attention to formal training in language teaching
at the undergraduate level. Since influences on teacher cognition are clearly operating
before an individual’s official entry into the profession, the undergraduate years are a
good starting point for improving second language education.

Acknowledgements

Funding to present a version of this paper at the 3rd Biennial International Conference
on Task-Based Language Teaching at Lancaster University, UK (2009), was awarded
through the Ruth Crymes Scholarship Fund, Department of Second Language Studies,
University of Hawai‘i at Mānoa. I am indebted to Gregory Strong, Joseph Dias, James
Sick, Junko Hondo, Manami Suzuki, Misa Nagatsu, and John Norris for their support
at various stages of this project. Naturally, the views expressed here and any remaining
errors are my sole responsibility.

References

Aspinall, R. W. (2006). Using the paradigm of ‘small cultures’ to explain policy failure in the case
of foreign language education in Japan. Japan Forum, 18, 255–274.
Barduhn, S., & Johnson, J. (2009). Certification and professional qualifications. In A. Burns & J.
C. Richards (Eds.), The Cambridge guide to second language teacher education (pp. 59–65).
Cambridge: Cambridge University Press.
Benevides, M., & Valvona, C. (2009). Special issue: Task-based language teaching in Japan. The
Language Teacher, 33(3), 1.
Borg, S. (2006). Teacher cognition and language education: Research and practice. London:
Continuum.
Brindley, G. (1994). Task-centered language assessment in language learning: The promise and
the challenge. In N. Bird, P. Falvey, A. Tsui, D. Allison, & A. McNeill (Eds.), Language and
learning: Papers presented at the Annual International Language in Education Conference
(pp. 73–94). Hong Kong: Hong Kong Education Department.
Brown, J. D. (2001). Using surveys in language programs. Cambridge: Cambridge University
Press.
Butler, Y. G. (2004). What level of English proficiency do elementary school teachers need to
attain to teach EFL? Case studies from Korea, Taiwan, and Japan. TESOL Quarterly, 38,
245–278.
Butler, Y. G. (2005). Comparative perspectives toward communicative activities among elemen-
tary school teachers in South Korean, Japan and Taiwan. Language Teaching Research, 9,
423–446.
Butler, Y. G., & Iino, M. (2005). Current Japanese reforms in English language education: The
2003 “action plan”. Language Policy, 4, 25–45.
 Daniel O. Jackson

Candlin, C. N. (1987). Towards task-based language learning. In C. N. Candlin & D. F. Murphey

(Eds.), Language learning tasks (pp. 5–22). London: Prentice Hall.
Carless, D. (2007). The suitability of task-based approaches for secondary schools: Perspectives
from Hong Kong. System, 35, 595–608.
Cresswell, J. W., & Plano Clark, V. L. (2007). Designing and conducting mixed methods research.
Thousand Oaks, CA: Sage.
Croker, R. (2007). Teacher-directed reciprocal peer observations: “Looking with” not “looking
at”. Learning Learning, 14, 31–39.
Gottlieb, N. (2008). Japan: Language planning and policy in transition. In R. B. Kaplan & R. B.
Baldauf (Eds.), Language planning and policy in Asia, Vol. 1: Japan, Nepal, Taiwan and
Chinese Characters (pp. 102–169). Bristol, UK: Multilingual Matters.
Holliday, A. (1999). Small cultures. Applied Linguistics, 20, 237–264.
Izumi, S. (2003). EFL education in Japan from the perspective of second language acquisition
research. Sophia Linguistica, 50, 3–14.
Johnson, K. E. (1994). The emerging beliefs and instructional practices of preservice English as
a second language teachers. Teaching and Teacher Education, 10, 439–452.
Konoeda, K., & Watanabe, Y. (2008). Task-based critical pedagogy in Japanese EFL classrooms:
Rationale, principles, examples. In M. Mantero, P. Chamness Miller, & J. L. Watzke (Eds.),
Readings In Language Studies, Volume 1: Language Across Disciplinary Boundaries (pp. 45–
72). St. Louis, MO: International Society For Language Studies.
Long, M. H., & Robinson, P. (1998). Focus on form: Theory, research, and practice. In C. Dough-
ty & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp. 15–41).
Cambridge: Cambridge University Press.
MacWhinney, B. (2000). The CHILDES project: Tools for analyzing talk (3d ed.). Mahwah, NJ:
Lawrence Erlbaum.
Matsuzawa, M. (2004, March 30). SELHI in action/School focuses on task-based approach.
Daily Yomiuri, pp. 1–3.
Meijer, P. C., Verloop, N., & Beijaard, D. (1999). Exploring language teachers’ practical knowl-
edge about teaching reading comprehension. Teaching and Teacher Education, 15, 59–84.
Mertens, D. M. (2005). Research and evaluation in education and psychology: Integrating diver-
sity with quantitative, qualitative, and mixed methods (2nd ed.). Thousand Oaks, CA: Sage.
Ministry of Education, Culture, Sports, Science, and Technology (2002). Developing a strategic
plan to cultivate “Japanese with English abilities”. Retrieved November 30, 2009, from
http://www.mext.go.jp/english/news/2002/07/020901.htm.
Ministry of Education, Culture, Sports, Science, and Technology (2003). Regarding the estab-
lishment of an action plan to cultivate “Japanese with English abilities”. Retrieved Novem-
ber 30, 2009, from http://www.mext.go.jp/english/topics/03072801.htm.
Norris, J. (2009). Task-based teaching and testing. In M. H. Long & C. J. Doughty (Eds.), The
Handbook of language teaching (pp. 578–594). Malden, MA: Blackwell.
Robinson, P. (2001). Task complexity, task difficulty, and task production: Exploring interac-
tions in a componential framework. Applied Linguistics, 22, 27–57.
Samuda, V., & Bygate, M. (2008). Tasks in second language learning. Hampshire, UK: Palgrave
Macmillan.
Sato, M., & Asanuma, S. (2000). Japan. In P. Morris & J. Williamson (Eds.), Teacher education in
the Asia-Pacific region: A comparative study (pp. 107–131). New York, NY: Falmer.
Shimahara, N. K. (1998). The Japanese model of professional development: Teaching as craft.
Teaching and Teacher Education, 14, 451–462.
 Chapter 12.  Task-based language teacher education in Japan 

Shimahara, N. K. (2002). Teaching in Japan: A cultural perspective. New York, NY: Routledge.
Skehan, P. (1996). A framework for the implementation of task-based instruction. Applied Lin-
guistics, 17, 38–62.
Tanaka, H., & Fukada, Y. (2004). Application of team-teaching techniques to teacher training
courses. Paper presented at the The Fifth Pan-Asian Conference on Language Teaching at
FEELTA, Vladivostok.
Van den Branden, K. (2006). Training teachers: Task-based as well? In K. Van den Branden
(Ed.), Task-based language education: From theory to practice (pp. 217–248). Cambridge:
Cambridge University Press.
Willis, D., & Willis, J. (2007). Doing task-based teaching. Oxford: Oxford University Press.
Yonesaka, S. (1999). The pre-service training of Japanese teachers of English. The Language
Teacher, 23(11), 9–15.
Yoshida, K. (2008). The effects of teaching practice on the development of students’ English
proficiency. In T. Ogura, H. Kobayashi, S. Inagaki, M. Hirakawa, S. Arita, & Y. Terao (Eds.),
Studies in language sciences 7 (pp. 23–37). Tokyo: Kurosio.
 chapter 13

Incorporating a formative assessment

cycle into task-based language teaching
in a university setting in Japan

Christopher Weaver
Toyo University, Japan

This chapter explains how a formative assessment cycle can inform the
development, implementation, and evaluation of pedagogical tasks in a task-
based syllabus by reporting the results of a study that examined 46 Japanese
university business students’ competence to deliver a PowerPoint presentation
in English. A Many-Facet Rasch analysis of student ratings of the PowerPoint
presentations combined with a discourse analysis of the presentations revealed
a gap between the students’ English presentation and their PowerPoint design
skills. A gap also existed between the students’ descriptive and explanatory skills.
A formative assessment cycle can thus identify specific competencies that need
further development to ensure a smooth transition from an EFL classroom to a
real-world use of English in a business situation. The study presents a successful
example of how a formative assessment cycle can provide the teacher and
the students in an EFL TBLT program setting with informative feedback that
identifies students’ level of competence to perform a given task, the difficulty
of that task, and ways in which students can develop towards target task
performances.

Introduction

Although assessment is an integral component of task-based language teaching

(TBLT), teachers face a myriad of issues when deciding how to best design, imple-
ment, and evaluate assessment practices within a TBLT curriculum. At its core, task-
based language assessment (TBLA) involves evaluating the degree to which language
learners can use their L2 to accomplish given tasks. A well-designed and implement-
ed assessment can also provide teachers and language learners with a detailed ac-
count of task performance that can inform future task-based instruction and L2 de-
velopment. The purpose of this chapter is to examine how assessment can be
 Christopher Weaver

conceptualized as a cyclical framework aiming to maximize the synergy between task

performance and TBLT.
Assessment within a TBLT curriculum can fulfil two broad purposes: summa-
tive assessment or formative assessment (Norris, 2009). Summative assessment pri-
marily involves the use of tasks to make high-stakes decisions concerning L2 learn-
ers’ level of competence to perform specific tasks or task-types. An example of
summative assessment is the Occupational Foreign Language Test which assesses L2
listening and speaking competence as it relates to the tourism and hospitality indus-
try (Brown, 1993). Although typically occurring at the end of a TBLT curriculum,
summative assessments can be an important means of evaluating program design
and implementation.
Formative assessment, on the other hand, involves the use of tasks to provide feed-
back to language learners and teachers throughout a TBLT curriculum. The analysis of
task performances in formative assessments can help teachers identify specific gaps in
learner competences as well as track learner L2 development over time. A formative
assessment cycle thus helps teachers to be more responsive to learners’ internal sylla-
bus, which is exposed through the performance of tasks (Long & Norris, 2000).

A formative assessment cycle in TBLT

Incorporating a formative assessment cycle into a TBLT curriculum involves a number

of steps outlined in Figure 1. This particular formative assessment cycle draws upon
the evidence-centred design framework (Mislevy, Steinberg, & Almond, 2002) in ad-
dition to other evidence-based approaches to TBLT assessment (e.g., Long & Norris,
2000; Norris, 2009; Norris, Brown, Hudson, & Yoshioka, 1998), as well as a series of
empirical studies that investigated the design process of high-stakes university en-
trance examinations in Japan (Weaver & Sato, 2008).

Task selection Feedback

Task definition Assessment

implementation

Assessment criteria Task performance

Figure 1.  Framework for a formative assessment cycle in TBLT

 Chapter 13.  Task-based language teaching in a university setting in Japan 

The first step in the formative assessment cycle involves task selection. Typically in
formative assessments this selection involves pedagogical tasks designed to help learn-
ers develop the competences needed to perform target tasks or task-types identified by
a needs analysis as being important to the learners (Long, 2005).
The second step entails defining the task to be used in the formative assessment.
The definition of tasks, which has taken on a life of its own in the TBLT literature, can
be quite overwhelming for some teachers and sometimes hinder the adaptation of
TBLT in the language classrooms. From a formative assessment perspective, task defi-
nition should involve identifying five task features (Ellis, 2003): (a) the goal or the
general purpose of the task, (b) the verbal or non-verbal input the task supplies to
language learners, (c) the conditions under which information in the task is presented
(ideally these conditions are representative of target tasks (e.g., Norris et al., 1998)),
(d) the methodological procedures for performing the task, and (e) the task’s predicted
outcomes such as the linguistic and cognitive processes required by task performance
and the product coming out of it.
The third step involves establishing the assessment criteria. When defining the
assessment criteria, teachers need to keep focused on what they would like to learn
from the learners’ performance of the task (Mislevy et al., 2002). Assessment criteria
can involve the processes undertaken while performing a task and/or features of the
final product after completing the task. When evaluating task processes, teachers
should evaluate L2 competences required of target task performances (for example, a
sociolinguistic measure that examined the use of address terms in a German speaking
test (Norris, 2001)). From a task product perspective, the assessment criteria may sim-
ply be defined as the successful completion of a given task, for example locating a
journal article about a particular topic in a university library (Robinson & Ross, 1996).
Regardless of the choices teachers make, language learners should have access to the
assessment criteria so that they will have a clear idea of how they will be evaluated.
This access can also deepen learners’ understanding of the task because the assessment
criteria highlight the essential qualities of a successful task performance (e.g., Byrnes,
2002). Moreover, teachers should take advantage of a wide range of data collection
techniques such as discourse analysis methods, raters utilizing scoring rubrics
(e.g., Bonk & Ockey, 2003), and/or self-assessments (e.g., Matsuno, 2009) to provide
learners with the most informative feedback possible.
The fourth step is the language learners’ performance of the task. The double-
ended arrow located between task description and task performance in Figure 1 cap-
tures the interactional nature existing between these two steps. In some cases, task
performances reveal a re-interpretation of the task (e.g., Loschky & Bley-Vroman,
1993) or learner discourse that is interwoven in the task performance, but is not di-
rectly related to the task (e.g., Thornborrow, 2003). In order to capture this level of
detail, video and audio recordings of oral task performances are essential for
creating a refined account of learner L2 output. Tasks designed for other modes of
 Christopher Weaver

communication such as writing will require different kinds of documentation that can
serve as a useful resource in post-task feedback sessions with students.
The fifth step is the application of the assessment criteria. Similar to task perfor-
mances, the double-ended arrow signifies the interactional nature existing between
raters, the assessment criteria, and the task description (Bachman, 2002). In addition,
there is always the possibility that scores assigned to learners may reflect not only their
task performance, but also rater bias and/or limitations of the assessment criteria
(Skehan, 2001). Teachers thus should utilize a variety of measures and data collection
techniques that can be triangulated to provide a valid and reliable measure of learners’
task performances.
The sixth step of the assessment cycle entails providing feedback to learners. Ulti-
mately, the aim of formative assessment is to provide feedback from a variety of pos-
sible perspectives, in order to inform improvements in teaching and learning, with
learners obviously constituting the primary recipients of assessment information.
Although teachers may be tempted to provide learners with comprehensive feedback
on their task performance, at least in formative assessments used at the lesson or unit
level, they should keep their feedback focused to help language learners not only
identify current gaps in their competence, and track their L2 development over time
(Norris, 2009).
After providing learners with focused feedback, teachers return to the first step of
the formative assessment cycle to select the most effective pedagogical task to address
current gaps in L2 competence and advance learners towards target task performanc-
es. The formative assessment cycle thus helps teachers and language learners maximize
opportunities for L2 development in TBLT by strengthening the interdependence that
exists between task selection, task and assessment criteria definitions, task perfor-
mance, application of assessment criteria, and reflections on task performances.

An empirical investigation of a formative assessment cycle

in a Japanese context

To illustrate how these steps work together in a formative assessment cycle in practice,
this chapter reports the results of an empirical study that investigated the competence
of 46 Japanese university business students to deliver a PowerPoint presentation in
English aiming to persuade their peers to invest in a particular company listed on the
New York Stock Exchange. The study was guided by three research questions:
Research Question 1: To what extent does students’ use of assessment criteria to evalu-
ate their peers’ PowerPoint presentations vary?
Research Question 2: To what extent do students vary according to their level of com-
petence to deliver a PowerPoint presentation in English?
 Chapter 13.  Task-based language teaching in a university setting in Japan 

Research Question 3: To what extent do the different task requirements for each sec-
tion of the PowerPoint presentation vary in difficulty? Moreover, how difficult is it to
convince other business students to invest in a company?
Although the answers to these research questions primarily address the implementa-
tion of the assessment criteria and the evaluations of the students’ task performances,
they also provide teachers with the quantitative and qualitative information necessary
to identify which types of competences students need to develop further in order to
progress towards target task performances. Moreover, these research questions create
an opportunity for teachers to reflect upon the effectiveness of the formative assess-
ment cycle in a TBLT program.

Method

Participants

The participants in this study were 46 Japanese first year Business Administration stu-
dents attending a prestigious private university located on the outskirts of Tokyo, Ja-
pan. The 18 females and the 28 male students, who ranged from the ages of 19 to 21,
were taking a year-long business presentation course organized around a series of ped-
agogical tasks designed to develop students’ oral English presentation and PowerPoint
design skills. The tasks in the curriculum are sequenced according to task difficultly,
which is defined by a number of factors such as the purpose of the presentation
(e.g., sales pitch), the type of speech required to achieve that purpose (e.g., persuasive
speech), the content knowledge required for the presentation, the length of the presen-
tation, and the number of slides required in the PowerPoint.

The task

The students were required to give a PowerPoint presentation in English about a pub-
licly traded company listed on the New York Stock Exchange that they thought would
be a good investment (see Appendix A for the task instructions). The presentation had
three sections. Each section had specific task requirements defining the type of infor-
mation that the students needed to convey in their oral presentation and on their
PowerPoint slides.
The first section of the presentation required students to describe the company
they chose and explain the reason for their choice. The second section required stu-
dents to describe how the stock price of the company had changed over a four-week
period (from December 2006 to January 2007) and provide a possible reason for the
variability in the stock price. The third section required students to describe three ar-
eas of future growth for the company and explain why they thought the company
 Christopher Weaver

would make a good investment opportunity. Although the task description and assess-
ment criteria defined the parameters of the task, the students had complete control
over the content of their PowerPoint presentations. It was up to each student to deter-
mine the most effective way to construct a PowerPoint presentation that would
convince their peers to invest in their company. The delivery of the PowerPoint presen-
tations ranged from three to five minutes.
This particular task was designed and incorporated into a year-long business pre-
sentation course for a number of reasons. Business students need analytical skills to
research and evaluate which companies would be a good investment of their time
when they look for a job after graduation. Moreover, Japanese business students in-
creasingly need practical English communicative competence and polished Power-
Point presentation skills to be successful in Japan’s more competitive job market. The
use of tasks such as this one, however, creates a number of challenges in the Japanese
educational context. The highly competitive university entrance examinations influ-
ence the manner in which English as a foreign language is taught and learnt in Japanese
junior and senior high schools (e.g., Guest, 2000; Watanabe, 1996). The use of the
yakadoku method, which relies heavily on translation, creates the expectation that the
teacher is the primary source of knowledge within the classroom and students should
be the consumers of this knowledge (Gorsuch, 1998). These roles and related respon-
sibilities, however, have begun to shift recently with the introduction of communica-
tive language teaching in senior high school English classrooms in Japan (Nishino,
2008; Sakui, 2004, 2007). Some post-secondary institutions in Japan have also begun
redefining EFL education from a requirement for graduation to an opportunity where
students develop the necessary competences needed to perform real-world tasks.

Materials

The students evaluated their peers’ PowerPoint presentations using assessment criteria
written in Japanese (see Appendix B for an English translation). The assessment crite-
ria were the end product of a previous pedagogical task in which the students watched
a video of five PowerPoint presentations delivered in the previous academic year. In
small groups, the students ranked the five PowerPoint presentations from the most to
the least convincing. Then they identified what aspects of the students’ task perfor-
mances led to a convincing or non-convincing presentation. The student groups re-
ported back to the entire class, leading to a detailed list of factors thought to influence
the students’ task performances. These factors were then organized into two broad
categories: speaking skills and PowerPoint design skills. Finally, the teacher and the
students selected two speaking skills and two PowerPoint design skills for each section
of the presentation. The speaking skills were selected on the basis of whether or not
they emphasized the interactive nature of a PowerPoint presentation where the goal of
the presenter is to have the audience clearly understand the company being profiled
and why this particular company is a good investment opportunity. The PowerPoint
 Chapter 13.  Task-based language teaching in a university setting in Japan 

design skills were selected to highlight the importance of constructing PowerPoint

slides that presented a clean and focused message. In addition, students were asked to
provide a holistic evaluation of the entire presentation from the perspective of two
speaking skills and two PowerPoint design skills. Using a 4-point Likert scale, the stu-
dents completed the evaluations by indicating the extent to which they felt the pre-
senter and the PowerPoint slides met the task requirements defined in the assessment
criteria. The points of the scale were labelled in Japanese “1. Strongly Disagree,”
“2. Disagree,” “3. Agree,” and “4. Strongly Agree”. Finally, students had to decide
whether or not they would invest in a company based upon their peer’s presentation.
This “call to action” was evaluated using a 2-point dichotomous scale labelled in Japa-
nese “1. No, this presentation has not convinced me to invest in this company.” and
“2. Yes, this presentation has convinced me to invest in this company.”

Procedure

One of the goals of the year-long business presentation course was to involve students
in the assessment process. The rationale for this goal was to help students develop an
ownership of the TBLT-based course and help them gain a deeper understanding of
the tasks as well as the competences and the level of performance required to success-
fully complete the tasks. At the beginning of the course, the teacher explained how the
formative assessment cycle (Figure 1) would guide the creation and the delivery of the
six PowerPoint presentations outlined in the course description. The task of presenting
a company listed on the stock exchange was the second of three presentations that the
students delivered in the first semester of the business presentation course.
Following the formative assessment cycle, the students first watched a video of a
PowerPoint presentation from the previous academic year as an introduction to the
task. This introduction led to a collaborative process where the teacher and the stu-
dents defined what the task entails. As previously discussed students then watched five
videos of past PowerPoint presentations and collaboratively constructed the assess-
ment criteria with the teacher. The students then evaluated another five videos of Pow-
erPoint presentations with the assessment criteria that they constructed. This step
helped the students become comfortable using the criteria in a real time assessment
situation as well as allowed for the opportunity for criteria refinement if needed. After-
wards the students were then randomly selected to deliver their PowerPoint presenta-
tion, which were video and audio recorded. Halfway through the presentations, the
students watched another video of three PowerPoint presentations. The second set of
videos acted as a norming session to check if the students had changed the manner in
which they were using the assessment criteria (e.g., Myford & Wolfe, 2003; Wilson &
Case, 2000).
The students’ ratings of the PowerPoint presentations were then analyzed using
Rasch measurement theory. With the results of this analysis, the teacher first provided
students with an overall account of their task performances compared to the speaking
 Christopher Weaver

skills and the PowerPoint design skills featured in the assessment criteria. Afterwards,
the teacher conducted one-on-one feedback sessions with students to review the rat-
ings that each student received for his/her speaking skills and PowerPoint design skills.
The next step involved students transcribing their PowerPoint presentations. The tran-
scriptions were then used in teacher-class and teacher-student feedback sessions as a
qualitative means to identify specific linguistic and/or behavioural competences that
influenced students’ task performances. From start to finish, the progression through
the entire formative assessment cycle took five 90-minute class periods.

Analysis

The students’ evaluations of their peers’ presentations were analyzed with a Many-Facet
analysis. A Many-Facet analysis involves an extended version of the Rasch measurement
model (Rasch 1960/1980) to analyze rating data that has multiple facets of interest. In
the case of this study, the four facets of interest were: the students as raters, the students
as presenters, the difficulty of the task requirements for each section of the presentation,
and a holistic “call to action” evaluation of the presentation. A Many-Facet Rasch analy-
sis determines the probabilities of ordered-category ratings for these different facets and
calibrates them on a shared frame of reference known as the logit scale (Linacre, 1989).
As a result, it is possible to make direct comparisons between the different facets.
The students used a version of conversation analysis (Tsui, 1994) when they tran-
scribed their task performances (see Appendix C for the conversation analysis conven-
tions). Unlike the transcriptions presented in this chapter, which provide a very
detailed account of task performance, the students focused upon the level of fluency in
which they delivered their presentation. This feature of speech was identified in the
previous formative assessment cycle as being a factor that mediated students’ effective
delivery of their PowerPoint presentations. Of special interest was when and where
students paused when they spoke. Looking for the occurrences of micro-pauses and
prolongation of sounds, students examined whether pauses in their speech occurred at
major syntactic boundaries, referred to as “juncture pauses”, or within syntactic units,
which are called “hesitation pauses” (Lounsbury, 1954). The aim of this analysis was
thus to reveal how features of fluency can mediate students’ ability to deliver a con-
vincing argument of why their company is a good investment opportunity.

Results

Figure 2 shows the results of the Many-Facet Rasch analysis of the students’ evalua-
tions of their peers’ PowerPoint presentations. Before reporting the results, a few re-
marks explaining how to interpret Figure 2 are in order. The first column in Figure 2 is
the logit scale, which acts as the common frame of reference when interpreting each
 Chapter 13.  Task-based language teaching in a university setting in Japan 

facet and the relationships between the different facets. For this study, the logit scale
ranges from –2 to 4 logits.
The second column is “the students as raters” facet. The students are ordered ac-
cording to the severity with which they used the assessment criteria to evaluate their
peers’ presentations. The one student located above zero logits is more severe in their
evaluations compared to the students located below zero logits. The gender of the rat-
ers is indicated with the number one for females and the number two for males.
The level of 46 business students’ competence to deliver a PowerPoint presentation
in English is shown in the third column. The students are identified by their company’s
ticker symbol and a number indicating their gender (i.e., 1 is female and 2 is male). For
example, the student located at the bottom of the third column has the code MMM2,
which means that this male student gave a presentation about the 3M Company being
a good investment opportunity. Students located above zero logits were rated as being
more competent to deliver a presentation in English whereas students located below
zero logits were evaluated as being less competent.
The fourth column lists the difficulty of the different task requirements for each
section of the presentation. The different task requirements are first indicated with a
number specifying the section of the presentation (i.e., 1 is the first slide of the

Logit Severe raters More competent students More difficult task requirements Invest
2

PG2 2SE price explanation

1 KDE2
TOT2 HSC convincing argument
SLW2 1SE company explanation
CAJ1 FDX2 SKM2 ZNH2 3SE strengths explanation HPI needed information
AFL1 DCM1 DCM1 HMC1
DCM1 DCM2 MTU2 2SD price description
DCM1 KNM2 MSFT2 MTU1 SNP2 UBS2 HSD information description Yes
1 CHU2 KNM1 KNM2 MDCO1SNE2 VIP2
0 2 2 2 2 2 SNE1 SNE1 1PP company picture 3SD strengths description HPE powerpoint enhancement
IKR1
1 2 2 2 MITSY2 MO2 NWA2 1SD company description NO
2 2 GU2 2PP stock price 3PB brief points
1 2 2 GFI2 NYT1 RLH2 XOM1 3PS company strengths
1 2
2 XOM1
SNE1 SNE2
–1 1 1 1 2 2 DCM2 HIT1
1 1 1 2 DIS2 ENT2
2 2PC price change
2 2 2
1 MMM2
1 1 2 2
2
1 2 1PN company name
–2 2
1 1
2

–3

–4

Logit Lenient raters Less competent students Less difficult task requirements Invest

Figure 2.  Many-Facet Rasch analysis of the student PowerPoint presentations

 Christopher Weaver

presentation, 2 is the second, and 3 is the third) or the letter “H” for the holistic evalu-
ations of the entire presentation. The task requirements are then identified with either
an “S” indicating a speaking skill or a “P” indicating a PowerPoint design skill. Follow-
ing that is a short two-word description of the task requirement. Task requirements
that have a score exceeding zero logits are more difficult than task requirements with a
score below zero, which are less difficult to perform.
The final column on the right side of Figure 2 is the holistic “call to action” evalu-
ation of the presentations defined in terms of whether or not students would invest in
the company based upon their peer’s presentation. The decision to invest in a company
(i.e., “Yes”) is located at 0.29 logits. Students would thus probably invest in the compa-
nies listed in column 3 above this logit score. On the other hand, students would prob-
ably not invest in companies that had a logit score lower than –0.29 logits (i.e., “No”).

The application of the assessment criteria

The severity with which the students used the assessment criteria to evaluate their
peers’ PowerPoint presentations varied considerably from .07 logits (i.e., the most se-
vere rater) to –3.41 logits (i.e., the most lenient rater). The students as raters tend to
cluster into three groups along the continuum of rater severity (see the second column
in Figure 2). The first cluster is a group of 17 students who were slightly lenient in their
ratings of their peers (i.e., they have a logit score ranging from 0 to –.62 logits). The
next cluster of 14 students is more lenient with their ratings (i.e., their logit scores
range from –.96 to –1.7 logits). The last cluster of 12 students is the most lenient in
their use of the assessment criteria (i.e., their logit scores range from –1.93 to –2.25
logits). A fixed chi-square value of 2348.6 with 45 degrees of freedom indicates a .001
probability, after allowing for measurement error, that the students could be consid-
ered equally severe. The separation index for rater severity is 7.44, which suggests that
amongst the 46 students there are ten statistically distinct levels of rater severity
(Fischer, 1992). This clear indication that the raters differ significantly in their severity
provides a strong rationale for utilizing a Many-Facet Rasch analysis because this sta-
tistical approach takes into account differences between raters in determining students’
level of competence to deliver a PowerPoint presentation in English.

The range of student competence to deliver a powerpoint presentation

According to Figure 2, student competence to deliver a PowerPoint presentation in

English ranged from –1.5 logits (i.e., the least competent student) to 1.3 logits (i.e., the
most competent student). Remembering that the score of zero logits differentiates the
less competent from the more competent students, 18 students were located below
zero logits while the remaining 28 students were located above this point. The students
tend to cluster in a couple of positions along the continuum of competence. The first
cluster is located around –1 logits (i.e., a group of seven less competent students). The
 Chapter 13.  Task-based language teaching in a university setting in Japan 

next cluster is just above zero logits (i.e., a group of 12 slightly competent students) and
the last cluster is located at .55 logits (i.e., a group of eight more competent students).
A fixed chi-square value of 1693.6 with 45 degrees of freedom indicates a .001 proba-
bility, after allowing for measurement error, that the students have an equal level of
competence to deliver the PowerPoint presentation in English. The separation index
for student competence is 6.11, which suggests that amongst the 46 students there are
eight statistically distinct levels of student competence.

The difficulty level of the task requirements

Table 1 shows the level of difficulty students had performing the different task require-
ments of the PowerPoint presentation. Once again, task requirements that have a score
exceeding zero logits are more difficult than task requirements with a score below zero.
A fixed chi-square value of 2543 with 15 degrees of freedom indicates a .001 probabil-
ity, after allowing for measurement error, that the different task requirements have an
equal level of difficulty. The separation index for the task requirements is 13.69, which
suggests that there are eighteen statistically distinct levels of difficulty amongst the
16 task requirements.
The different task requirements tend to cluster together in four groupings along
the continuum of difficulty (see the fourth column in Figure 2). The first cluster is lo-
cated around .7 logits (i.e., a group of four quite difficult task requirements composed
of three speaking skills and one PowerPoint design skill). The next cluster is at .33
logits (i.e., a group of two difficult speaking skills). The third cluster is located at zero

Table 1.  Level of difficulty of the task requirements

Less Difficult More Difficult

Task Requirement Logit Mean Sq Task Requirement Logit Mean Sq

1PN company name –1.91 1.12 1PP company picture   .03 1.14
2PC price change –1.19 1.09 HSD information   .30   .73
description
3PS company strengths   –.56 1.05 2SD price description   .33   .99
2PP stock price   –.39 1.07 3SE strengths explanation   .62   .93
3PB brief points   –.38 1.07 HPI needed information   .64   .63
1SD company description   –.27 1.04 1SE company explanation   .70 1.02
3SD strengths description   –.05   .80 HSC convincing   .86 .7
argument
HPE PowerPoint   –.03   .89 2SE price explanation 1.29   .95
enhancement
 Christopher Weaver

logits (i.e., a group of three task requirements composed of one speaking skill and two
PowerPoint skills) and the last cluster is located at –.39 logits (i.e., a group of four less
difficult task requirements composed of one speaking skill and three PowerPoint de-
sign skills). There are also a number of gaps between the clusters indicating a marked
difference in the level of difficulty amongst the different task requirements. In other
words, the PowerPoint design skill of making the first slide displaying the company’s
name (–1.91 logits) is significantly easier than making the second slide displaying the
change in the company’s stock price (–1.19 logits).
In terms of the specific task requirements for the three sections of the presenta-
tion, the less difficult task requirements were largely composed of the five PowerPoint
design skills. The two less difficult speaking skills involved describing the company
(1SD) and three of its strengths (3SD). The more difficult task requirements were the
three speaking skills requiring students to explain the underlying rationale for choos-
ing the company (1SE), the reason behind the company’s stock price changes (2SE),
and the three areas of future growth that make the company a good investment (3SE).
Choosing a picture or an image to represent the company’s business (1PP) was the
only difficult PowerPoint design skill.
Examining the holistic evaluations of the students’ presentations, the easiest task
requirement (–.03 logits) was the PowerPoint design skill of making three slides that
visually supported the overall message of the presentation (HPE). Making a Power-
Point presentation that provided enough information about the company (HPI) was
more difficult (.64 logits). The speaking skill of delivering a presentation with enough
descriptive information about the company (HSD) was moderately difficult (.30 log-
its). The most difficult overall speaking skill (.86 logits) was to provide a convincing
argument about why to invest in the company (HSC).
A Many-Facet Rasch analysis also produces fit statistics that indicate the extent to
which the 16 task requirements work together to define a meaningful variable. Infit
mean-square values greater than one indicate more variation in the ratings for a task
requirement than what the Rasch model expects. The high mean-square values for
task requirements 1PN and 1PP thus suggest that high ratings for these requirements
do not seem to correspond to high ratings for the other task requirements; similarly,
lower ratings do not seem to correspond to low rating for the task requirements. This
finding suggests that there is potential variability amongst students’ task performances
and/or how students applied these rating criteria. The smaller mean-square values of
HPE, HPI, HSD, and HSC also suggest that these holistic evaluations are not function-
ing independently from the other task requirements. As a result, a case could be made
for dropping the holistic evaluations when revising the assessment criteria for future
use because they do not seem to make an independent contribution to the measure-
ment of students’ competence to deliver a PowerPoint in English.
 Chapter 13.  Task-based language teaching in a university setting in Japan 

The holistic “call to action” evaluation

The difficulty that students experienced presenting a convincing argument to invest in

a company listed on the New York Stock Exchange is also reflected in the “call to ac-
tion” facet. Figure 2 shows that the positive response of “yes” to the presented invest-
ment opportunities is located at 0.29 logits. In other words, students were probably
willing to invest in only 21 out of the 46 companies presented.

A more refined account of students’ task performances

Thus far, the teacher can provide the students with an overall account of their compe-
tence to deliver a PowerPoint presentation in relation to the different task require-
ments and the holistic “call to action” evaluation. From a formative assessment per-
spective, the teacher can determine that the speaking skills pose a greater challenge for
the students than PowerPoint design skills. More specifically, the students need help
with their explanatory speaking skills compared to their descriptive skills. Students, on
the other hand, can compare their overall level of competence to their peers and more
importantly they can determine whether or not they successfully convinced their peers
to invest in their company. This level of feedback, however, does not provide the teach-
er or the students with a detailed account of what students need to do to progress
towards more target task performances. Student FDX2, for example, received a com-
petence score of .61 logits for his PowerPoint presentation about the Fed Ex Corpora-
tion being a good investment opportunity. Although his logit score alone has no peda-
gogical meaning, it can be quite informative when mapped against the expected and
observed ratings that he received from his peers (see Figure 3).
The straight line of numbers in Figure 3 represents the Student FDX2’s level of
competence (.61 logits). In other words, the numbers are the ratings that the Rasch

–5 –4 –3 –2 –1 0 1 2 3 4 Task requirements
.4. 1PN company name
(1) 3 1PP company picture
.4. 1SD company description
.1. 2 1SE company explanation
3 .4. 2PP stock price
.4. 2PC price change
3 .4. 2SD price description
2 (4) 2SE price explanation
.4. 3PS company strengths
.3. 4 3PB brief points
.4. 3SD strenghts description
3 .4. 3SE strenghts explanation
.3. HPI needed information
3 .4. HPE powerpoint enhancement
3 .4. HSD information description
3 .4. HSC convincing argument
–5 –4 –3 –2 –1 0 1 2 3 4

Figure 3.  Expected and observed ratings of Student FDX2’s PowerPoint presentation
 Christopher Weaver

measurement model expects Student FDX2 to receive from his peers based upon his
level of competence. For example, the Rasch model expects that this student should
receive the rating of 4 for the PowerPoint design skill of clearly displaying the company
name on the first slide of the PowerPoint presentation (i.e., 1PN). The dots on either
side of the four (i.e., .4.) indicate that four was the observed rating Student FDX2 re-
ceived from his peers. The next number in the line indicates that the Rasch model
expects a rating of three for the PowerPoint design skill of selecting an image represen-
tative of the company and its business (i.e., 1PP). The student, however, received a
rating of one. The brackets on either side of this rating indicate that this rating is high-
ly unexpected for a student who has a level of competence of .61 logits.
Continuing down the line of numbers, there are two more occasions (i.e., 1SE and
3PB) where Student FDX2 received a rating lower than expected; five instances (i.e.,
1SD, 2PP, 3PS, 3SD, and HPI) where he received the rating expected from the Rasch
model; six instances (i.e., 2PP, 2SD, 3SE, HPE, HSD, and HSE) where he received a rat-
ing exceeding the expectations of the Rasch model, and one instance (2SE) where he
received an unexpectedly high rating for his explanation of the change in the company’s
stock price. In sum, Student FDX2 received a mixture of positive and negative evalua-
tions on his PowerPoint design skills and generally positive evaluations on his descrip-
tive and explanatory skills. These evaluations were also accompanied by higher than
expected holistic evaluations of his PowerPoint presentation. The difference between
the expected and the actual ratings of Student FDX2 task performance is also reflected
in a mean-square infit statistic of 1.27, which indicates that the two unexpected ratings
for the task requirements 1PP and 2SE do not seem to fit with the other ratings of this
student’s task performance. Yet, what still remains unknown is what was lacking in
Student FDX2’s PowerPoint design skill in the first slide of his presentation (1PP) and
how he was able to provide the unexpected explanation for the variability of the FedEx’s
stock price (2SE). Incorporating a qualitative aspect into the formative assessment cycle
can thus provide another source of feedback for the teacher and the students.

A discourse analysis of student FDX2’s task performance

Student FDX2’s delivery of his PowerPoint presentation is a combination of scripted

and non-scripted speech. The task description stated that students would not able to
read off a piece of paper when they delivered their presentations. Students often cir-
cumvented this task requirement by reading off their PowerPoint slides. This is illus-
trated by the excerpt below.
Excerpt 1 Student FDX2’s presentation of the first PowerPoint slide
1. Well I’d like to introduce Fed Ex- ((presenter turns away from the audience and
2. reads off the PowerPoint slide)) Fed Ex Corporation. This company provides a
3. portfolio of transportation and eco e commerce and business service through
4. companies.
 Chapter 13.  Task-based language teaching in a university setting in Japan 

Fed EX Corporation
• FedEx provides a portfolio of
transportation, ecommerce
and business services through
companies.

Figure 4.  Student FDX2’s first PowerPoint slide

Stock price
$96
$94
$92
$90
$88
Stock price
$86
$84
$82
$80
$78
07–Dec 14–Dec 21–Dec 28–Dec 04–Jan

Figure 5.  Student FDX2’s second PowerPoint slide

In Excerpt 1, Student FDX2 turns away from his audience in line 1 after a false start
with the company’s name and continues to read off the PowerPoint slide in lines 2 to 4
with relative ease, though he has a little trouble with the word “e-commerce” in line 3.
His fairly fluent scripted presentation, however, lacks an explanation of why he se-
lected the Fed Ex Corporation as a company worthy of investment. Moreover, his first
PowerPoint slide (see Figure 4) lacks any visual representation of the Fed Ex Corpora-
tion and its business. These two omissions led to lower than expected ratings of Stu-
dent FDX2’s performance in the first section of the presentation (i.e., a rating of one for
1PP and 1SE).
The second PowerPoint slide (see Figure 5) provides less of a written script for
Student FDX2. As a result, his speech rate varies with noticeably lengthened enuncia-
tion of “next” and “and” in line 5 of Excerpt 2. Changes in fluency, however, can in-
crease the audience’s level of comprehension of the presentation. For example in line 6,
Student FDX2’s micro-pause followed by a slow delivery of the current stock price of
84 dollars helps to make this information more salient. As a result, his peers give him
the highest rating of four for his description of the company’s stock price.
Excerpt 2 Student FDX2’s presentation of the second PowerPoint slide
  5. And n:: next the stock stock price is going down and down. And:: it it used to be 120
  6. dollar. But it is going down to (.) <84 dollar>. I don’t know why. December has a
 Christopher Weaver

  7. Christmas ((presenter turns away from the PowerPoint slide and looks at the
  8. audience)) Christmas:::: ya Christmas ((presenter looks back to PowerPoint slide))
  9. and every people (.1) send a present but I don’t know why the price going down.
10. But (.2) I think the (.) it’s bottom pr price. So the stock price ((presenter turns away
11. from the PowerPoint slide and looks at the audience)) will be going up.
Student FDX2’s orientation of his body during this section of the presentation has a
similar effect of drawing the audience’s attention to specific pieces of information. In
lines 7, he turns away from the PowerPoint slide and looks at the audience repeating
that Christmas is the reason why he thought the price of the Fed Ex Corporation would
increase. Student FDX2 turns to the audience once again in line 9 to assure his audi-
ence that the price of the company’s stock will increase in the future. This direct en-
gagement with the audience contributed to the unexpectedly high rating (i.e., a rating
of four for 2SE) for his explanation of the variability in the company’s stock price.
The third slide of the presentation (see Figure 6) once again provides Student
FDX2 with a comprehensive script of his presentation. The amount of text on his third
PowerPoint slide, however, resulted in a lower than expected rating from his peers
(i.e., a rating of three for 3PB). In line 12 of Excerpt 3, he turns his body back towards
the PowerPoint slide and states that there are three reasons to invest in the Fed Ex
Corporation. The unscripted nature of this introduction is apparent with the false start
of “you” in line 12 and the lengthened enunciation of “this” in line 13.
Excerpt 3 Student FDX2’s presentation of the third PowerPoint slide
12. And ((presenter looks back to PowerPoint slide)) three reason why y- you should
13. buy this:: stock. The first is the stock price is cheaper than other months. And (.)
14. second the stock price is bottom (.) so maybe the price going up. And third (.) come
15. Easter soon people use portfolio transportation. ((presenter turns away from the
16. PowerPoint slide and looks at the audience)) (.3) Finished thank you.
From lines 13 to 15, Student FDX2 reads off his PowerPoint slide in rapid succession
explaining why the Fed Ex Corporation is a good investment opportunity. Once again,
he uses micro-pauses to aid audience comprehension. This time he pauses slightly
between the three different reasons for investment. Once finished, Student FDX2 turns

• The stock price is cheaper than

other months.
• The stock price is bottom. So
maybe the price will going up.
• Come easter soon, people use
portfolio transportation.

Figure 6.  Student FDX2’s third PowerPoint slide

 Chapter 13.  Task-based language teaching in a university setting in Japan 

back towards the audience in line 15, waits three-tenths of a second, and then con-
cludes his presentation with “finished thank you.” His delivery of the third section of
the PowerPoint presentation received the expected rating of four for his description
of the company’s strengths (3SD) and a higher than expected rating for his explanation
of why to invest in the company (i.e., a rating of four for 3SE).
Overall, his peers gave Student FDX2 a higher than expected rating of four for his
PowerPoint design skills (HPE), his explanatory skills (HSE), and his ability to deliver
a convincing argument (HSC). However, the audience identified that the student
needed to improve upon his skills of creating a PowerPoint presentation that featured
the necessary information needed to invest in the Fed Ex Corporation (i.e., a rating of
three for HPI). The audience also responded positively to his “call to action” to invest
in the Fed Ex Corporation.

Discussion

The results of this empirical study may prompt some teachers interested in TBLT to
question the need and/or the feasibility of incorporating a similar formative assess-
ment cycle into their own classrooms. The discussion of the results will thus focus
upon how the quantitative and qualitative data arising from this study was incorpo-
rated in the teacher-student feedback sessions, the pedagogical choices made in the
business presentation course, and the ongoing refinements taken to improve the ef-
fectiveness of the formative assessment cycle.
The first research question focused upon the variability of students’ use of the
assessment criteria to evaluate their peers’ PowerPoint presentations. The second col-
umn of Figure 2 shows that rater severity ranged from –.07 logits to 3.41 logits mean-
ing that students’ use of the assessment criteria composed ten statistically distinct
levels of rater severity. This large discrepancy between the student raters raises a couple
of pedagogical issues. The first is the possibility that the differences in rater severity
could reflect real differences in how the audience evaluated the PowerPoint presenta-
tions, which in turn could lead to a larger variety of recommendations of how the
presenter could develop towards a target task performance. Weaver (2011), for exam-
ple, found that Japanese university students who were rated by their peers as being the
most competent in a job interview conducted in English were also the most severe rat-
ers of their peers. Follow-up interviews with these students revealed that their severe
ratings reflected an evaluation of their peer’s level of grammatical competence, a crite-
rion that was not even included in the task’s assessment criteria. Teachers can attempt
to reduce the range of rater severity with training sessions aiming to standardize the
use of the assessment criteria; however, within a classroom setting that may be imprac-
tical and consume valuable classroom time (Luoma, 2004). In the context of this study,
a Many-Facet Rasch analysis is used to help the teachers take account of these
 Christopher Weaver

differences and control their influence on the evaluation of task performances (Linacre
& Wright, 2002).
The use of peer evaluations, however, does have a number of potential pedagogical
benefits. Learners may deepen their understanding of the task requirements as they
use the assessment criteria and observe their peers comply with these requirements to
various degrees of success. Peer assessment also has the additional benefit in that stu-
dents are engaged and attentive when other students are delivering their PowerPoint
presentations. Moreover, the responsibility of evaluating one’s peers helps strengthen
students’ sense of ownership over the formative assessment process and the TBLT pro-
gram as a whole.
As learners progress towards target task performances, there is also an increasing
need for the inclusion of real-world assessment criteria along with the involvement of
raters from the target context in order to define what competent communication en-
tails in real-world contexts (Grove & Brown, 2001; Jacoby & McNamara, 1999). For
example, Student FDX2 received higher than expected ratings for his explanation of
the Fed Ex Corporation’s variable stock price. Yet, the discourse analysis of his presen-
tation revealed in Excerpt 2 line 9 that he did not know the reason behind the com-
pany’s declining stock price during the Christmas season. One would expect that raters
from the business world would evaluate this students’ explanation differently from
students who also had a difficult time explaining the reason underlying the variability
in their own companies’ stock prices. Achieving reliable formative assessment feed-
back is thus intertwined with issues of validity and the need to explicitly define the
characteristics of a target task performance.
The second research question of this study focused upon determining the stu-
dents’ level of competence to deliver a convincing PowerPoint presentation. The
formative assessment cycle in this chapter draws upon a variety of quantitative and
qualitative techniques to define and refine what student competence entails. The
graphical output from the Many-Facet Rasch analysis (Figure 2) inevitably causes a
certain degree of uncertainty amongst teachers and students the first time they try to
interpret the results of the analysis. Teachers and students are more likely expecting a
number that represents a certain level of competency. By consulting the first and sec-
ond columns in Figure 2, it is possible for students to determine their level of compe-
tence measured with the logit scale. Yet similar to other number-based representations
of competence, a logit score has little meaning without context. Figure 2 provides
teachers and students with a number of different contexts to define task competence.
In the beginning of the first teacher-class feedback session, the teacher asked stu-
dents to identify themselves in the second column of Figure 2. Once students circled
their company’s ticker symbol, they can quickly see how they compare to their peers.
The teacher, however, quickly redirected the students’ focus to the fourth column of
Figure 2 with the message that competence cannot be defined solely by comparing
oneself to others in the class. Equally important is students’ level of competence to suc-
cessfully complete the task at hand. In the context of this study, 21 out of 46 students
 Chapter 13.  Task-based language teaching in a university setting in Japan 

were rated by their peers as being able to deliver a convincing PowerPoint presentation
in English. At this point in the teacher-class feedback session, the teacher directed the
students’ attention to the third column in Figure 2. In this column, the teacher and the
students identified the relative difficulty of the different task requirements. The 46
business students could see that their PowerPoint design skills were significantly stron-
ger than their English speaking skills. The take-away is that when students have time
to prepare a written text, generally they are able to effectively summarize the necessary
information. In terms of speaking skills, students experienced a greater level of diffi-
culty when they needed to explain something about their respective companies com-
pared to when they had to describe something.
Being able to provide more convincing explanations, however, may entail a deeper
level of content knowledge in addition to the linguistic competence required to effec-
tively deliver this information. Student FDX2, for example, was at a loss why the Fed Ex
Corporation stock price would decline during the busy Christmas holiday season when
people could be expected to massively use courier services to deliver gifts. What this
student may have failed to realize is that the Fed Ex Corporation faces stiff competition
from the U.S. Postal Service (2009, December 14). Providing an explanation of a com-
pany’s stock price (2SE) thus not only requires students to draw upon specific language
skills such as hedges, but also their general knowledge of the world and more specifi-
cally their understanding of the economic conditions that influences stock prices.
It must be remember, however, that Figure 2 provides an overview of student com-
petency relative to the different task requirements. An effective formative assessment
cycle must also take into account the possibility of individual differences. In the case of
Student FDX2, the feedback that he received in his teacher-student feedback session
was in many ways different from the feedback given to the entire class. Figure 2 and
Table 1 show that the explanatory speaking skill requirements (i.e., 1SE, 2SE, and 3SE)
pose a significant challenge for the students as a group. Because Student FDX2’s overall
level of competence (.61 logits) is lower that the level of difficulty of task requirements
1SE (.7 logits), 2SE (1.29 logits), and 3SE (.62 logits), Figure 2 and Table 1 might lead
the teacher to provide erroneous feedback to this student. Figure 3, however, shows
that Student FDX2 received higher than expected ratings for his ability to explain a
possible reason for changes in the company’s stock price (2SE) and why the three fu-
ture growth areas make the company a good investment opportunity (3SE). Student
FDX2, however, needs to focus on providing a better explanation at the beginning of
his PowerPoint presentation of why he selected FedEx (1SE). The combination of the
teacher-class and teacher-student feedback sessions can thus help teachers take into
account important individual differences when they select the next pedagogical task to
help students advance towards target task performances.
The third research question of this study aimed to determine the relative difficulty
of the different task requirements and the difficulty of convincing one’s peers to invest
in a company listed on the New York Stock Exchange. These points of interest can help
teachers gain a deeper understanding of the selected task and how the task and/or the
 Christopher Weaver

task procedures might be improved in the future. For example, when the PowerPoint
presentation about companies listed on the New York Stock Exchange was used again
in the next academic year, the four holistic evaluations (i.e., HPI, HPE, HSD, and HSC)
were replaced with a set of assessment criteria that evaluated students’ level of engage-
ment with the audience during the three sections of the PowerPoint presentation. An-
other adjustment made was that students were asked to focus on one type of industry
(e.g., consumer goods) for the entire year-long business presentation course so that
they could develop a more comprehensive level of content knowledge about that par-
ticular industry, which in turn helped them when they needed to provide explanations
about factors that could potentially influence the performance of the companies they
selected.
The graphical output of a Many-Facet Rasch analysis, however, can take teachers
only so far. In the first teacher-student feedback session with Student FDX2, the teach-
er told him that, “Your explanation of the FedEx’s stock price was really good, but you
need to explain why you choose this company in the first section of your PowerPoint
presentation.” This type of feedback gives this student a direction in which to focus his
efforts. Yet, Figure 3 provides no information about what made his explanation con-
cerning the stock price (2SE) effective compared to his explanation of his motivation
for choosing FedEx (1SE). A Many-Facet Rasch analysis thus naturally leads to a fo-
cused qualitative inquiry (Smith 2000). In the case of this particular formative assess-
ment cycle, a discourse analysis using conversation analysis techniques was used to
define specific behaviours of interest. Although discourse-analytic methods have been
discounted as a practical means of conducting planned formative assessments
(e.g., Ellis, 2003), a guided discourse analysis can be effectively integrated into a forma-
tive assessment cycle when students undertake the transcribing duties and examine
specific features in their transcriptions through teacher-class and teacher-student in-
teractions. The benefits of doing so are numerous. A guided discourse analysis of task
performances can help teachers and students clarify the connection between the rat-
ings students received from their peers and the students’ actual task performance.
Teachers can also use the transcriptions to monitor the extent to which students
are incorporating lessons learnt from previous task performances. For example, dur-
ing the discourse analysis-based feedback sessions, the students examined the number
and the location of micro-pauses in their first and second PowerPoint presentations.
The aim was to draw students’ attention to how pauses within syntactic units can
sometimes create the impression of a lack of fluency, but can also increase the saliency
of important information in a presentation such as when Student FDX2 described the
drop in FedEx’s stock price in line 6 of Excerpt 2. By repeating this type of analysis
throughout the year-long business presentation course, the students not only produced
a very refined account of how their level of fluency developed over time, but they also
gained a deeper understanding of how hesitation pauses can be effectively used to
achieve a specific communicative purpose. Finally, a discourse analysis can help teach-
ers identify areas of competence that were not featured on the task assessment criteria,
 Chapter 13.  Task-based language teaching in a university setting in Japan 

but need to be incorporated in future formative assessment cycles (e.g., Leedham,

2005). For example, Student FDX2’s PowerPoint presentation, much like the majority
of his peers, featured large sections of highly scripted speech where he read directly off
his PowerPoint slide. Although the fluency of his delivery increased during these times,
his connection with the audience suffered because he had his back turned to them. As
a result, the task assessment criteria for the next PowerPoint presentation included
items that evaluated of the speakers’ location relative to the audience.

Conclusion

A formative assessment cycle can help teachers establish a framework for systemati-
cally implementing TBLT in their classrooms. As the teacher and students progress
through the different steps of the formative assessment cycle, there are numerous op-
portunities to gain a deeper understanding of the componential features of the task at
hand and the requirements that need to be satisfied in order to successfully perform
the task. In this particular study, a Many-Facet Rasch analysis identified that students’
explanatory skills was one area of competence requiring further development. The
follow-up guided discourse analysis also revealed how students could use hesitation
pauses to increase the saliency of important information in their presentations and the
need in future formative assessment cycles to evaluate the interaction between highly
scripted speech and the presenter’s level of engagement with the audience.
The combination of a Many-Facet Rasch analysis and a discourse analysis used in
this study, however, may not meet the specific needs of other educational contexts.
Teachers can draw upon a number of other quantitative-qualitative combinations to
evaluate their students’ task performances from a variety of informative perspectives.
The whole idea is to set-up a dynamic and responsive feedback mechanism that can
provide students with personalized opportunities for learning in a meaningful context,
where gaps in L2 competence arise from the performance of a task according to learn-
ers’ internal syllabus rather than originating from an external syllabus. In sum, a
formative assessment cycle is an iterative process in which L2 use and informative
feedback can help transform hindsight into foresight, which in turn advances students
closer to target task performances that are meaningful to them.

References

Bachman, L. (2002). Some reflection on task-based language performance assessment. Lan-

guage Testing, 19(4), 453–476.
Bonk, W., & Ockey, G. (2003). A many-facet Rasch analysis of the second language group oral
discussion. Language Testing, 20(1), 89–110.
 Christopher Weaver

Brown, A. (1993). The role of test-taker feedback in the test development process: Test-takers’
reactions to a tape-mediated test of proficiency in spoken Japanese. Language Testing,
10(3), 277–301.
Byrnes, H. (2002). The role of task and task-based assessment in a content-oriented collegiate
foreign language curriculum. Language Testing, 19(4), 419–437.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Fischer, W. J. (1992). Reliability statistics. Rasch Measurement Transactions, 6(3), 238.
Gorsuch, G. (1998). Yakudoku EFL instruction in two Japanese high school classrooms. JALT
Journal, 20(1), 6–32.
Grove, E., & Brown, A. (2001). Tasks and criteria in a test of oral communication skills for first-
year health science students: Where from? Melbourne Papers in Language Testing, 10(1),
37–47.
Guest, M. (2000). “But I have to teach grammar!”: An analysis of the role “grammar” plays in
Japanese university English entrance examinations. The Language Teacher, 20(11), 23–29.
Jacoby, S., & McNamara. (1999). Locating competence. English for Specific Purposes, 18(3),
213–241.
Leedham, M. (2005). Exam-orientated tasks: Transcripts, turn-taking, and back-channeling. In
C. Edwards & J. Willis (Eds.), Teachers exploring tasks in English language teaching
(pp. 93–102). Basingstok, UK: Palgrave Macmillan.
Linacre, J. (1989). Many-facet Rasch measurement. Chicago, IL: MESA Press.
Linacre, J., & Wright, B. (2002). Construction of measures from many-Facet data. Journal of Ap-
plied Measurement, 3(4), 486–512.
Long, M. (Ed.). (2005). Second language needs analysis. Cambridge: Cambridge University
Press.
Long, M., & Norris, J. (2000). Task-based teaching and assessment. In M. Byram (Ed.), Encyclo-
pedia of language teaching (pp. 597–603). London: Routledge.
Loschky, L., & Bley-Vroman, R. (1993). Grammar and task-based methodology. In G. Crookes
& S. Gass (Eds.), Tasks and language learning: Integrating theory and practice (pp. 123–167).
Clevedon: Multilingual Matters.
Lounsbury, F. (1954). Transitional probability, linguistic structure and systems of habit-family
hierarchies. In C. Osgood & T. Sebock (Eds.), Pyscholinguistics: A survey of theory and re-
search problems (pp. 92–101). Bloomington, IN: Indiana University Press.
Luoma, S. (2004). Assessing speaking. Cambridge: Cambridge University Press.
Matsuno, S. (2009). Self-, peer-, and teacher-assessments in Japanese university EFL writing
classrooms. Language Testing, 26(1), 75–100.
Mislevy, R., Steinberg, L., & Almond, R. (2002). Design and analysis in task-based language as-
sessment. Language Testing, 19(4), 477–496.
Myford, C., & Wolfe, E. (2003). Detecting and measuring rater effects using many-facet Rasch
measurement: Part 1. Journal of Applied Measurement, 4(4), 386–422.
Nishino, T. (2008). Japanese secondary school teachers’ beliefs and practices regarding commu-
nicative language teaching: An explanatory study. JALT Journal, 30(1), 27–50.
Norris, J. (2001). Use of address terms on the German speaking test. In K. Rose & G. Kasper
(Eds.), Pragmatics in language teaching (pp. 248–282). Cambridge: Cambridge University
Press.
Norris, J. (2009). Task-based teaching and testing. In M. Long & C. Doughty (Eds.), The hand-
book of language teaching (pp. 578–594). Cambridge: Wiley-Blackwell.
 Chapter 13.  Task-based language teaching in a university setting in Japan 

Norris, J., Brown, J., Hudson, T., & Yoshioka, J. (1998). Designing second language performance
assessments. Honolulu, HI: Second Language Teaching & Curriculum Center, University of
Hawai’i at Manoa.
Rasch, G. (1960/1980). Probabilistic models for some intelligence and attainment tests (Copenha-
gen, Danish Institute for Educational Research, expanded edition (1980) with foreword and
afterword by B. Wright ed.). Chicago, IL: The University of Chicago Press.
Robinson, P., & Ross, S. (1996). The development of task-based assessment in English for aca-
demic purposes programs. Applied Linguistics, 17(4), 455–476.
Sakui, K. (2004). Wearing two pairs of shoes: Language teaching in Japan. ELT Journal, 58(2),
155–163.
Sakui, K. (2007). Classroom management in Japanese EFL classrooms. JALT Journal, 29(1),
41–58.
Skehan, P. (2001). Tasks and language performance assessment. In M. Bygate, P. Skehan & M.
Swain (Eds.), Researching pedagogical tasks: Second language learning, teaching and testing
(pp. 167–185). Harlow, UK: Longman.
Smith, E., Jr. (2000). Metric development and score reporting in Rasch measurement. Journal of
Applied Measurement, 1(3), 303–326.
Thornborrow, J. (2003). The organization of primary school children’s on-task and off-task talk
in a small group setting. Research on language and social interaction, 36(1), 7–32.
Tsui, A. (1994). English Conversation. Oxford: Oxford University Press.
United States Postal Service. (2009, December 14). Postal service ready for busiest mailing day.
Retrieved. from http://www.usps.com/communications/newsroom/2009/pr09_110.htm.
Watanabe, Y. (1996). Does grammar translation come from the entrance exams? Preliminary
findings from classroom-based research. Language Testing, 13(3), 318–333.
Weaver, C. (2011). Interpretations and Implementations of a task assessment criteria. Manuscript
submitted for publication.
Weaver, C., & Sato, Y. (2008). Tracking and targeting: Investigating item performance on the
English section of a university entrance examination over a four year period. JALT Journal,
30(1), 105–128.
Wilson, M., & Case, H. (2000). An examination of variation in rater severity over time: A study
in rater drift. In J. M. Wilson & G. Engelhard (Ed.), Objective measurement: Theory into
practice (Vol. 5, pp. 114–118). Stamford, CT: Ablex.
 Christopher Weaver

Appendix A

You are a junior analyst at a brokerage firm. You need to deliver a three-slide Power-
Point presentation about a company listed on the New York Stock Exchange that you
think is a good investment opportunity. Your PowerPoint presentation should include
the following information:

Slide PowerPoint Presentation

One Name of the company Describe which company you chose and
Picture representing the company’s the type of business they do
business Explain why you chose this company
Two Current stock price of the company Describe the current stock price
Stock price from December 2006 to Explain a possible reason for changes in
January 2007 the company’s stock price
Three Three future growth areas for the Describe three future growth areas for the
company company
Explain why these future growth areas
make the company a good investment
opportunity

Your peers will evaluate your presentation using the Assessment Criteria below. You
will not be able to read off a piece of paper while you are delivering your PowerPoint
presentation.
 Chapter 13.  Task-based language teaching in a university setting in Japan 

Appendix B

Assessment criteria

Slide 1
PowerPoint Design Skills It was easy to see the name of the company on the PowerPoint
slide.
It was easy to understand what kind of business the company is
involved in from the picture on the PowerPoint slide.
Speaking Skills I could easily understand the presenter’s description about what
type of business this company does.
I could easily understand the presenter’s explanation why he or she
chose this company.
Slide 2
PowerPoint Design Skills I could easily see the price of the stock on the PowerPoint slide.
I could easily see how the price of the stock changed on the graph
on the PowerPoint slide.
Speaking Skills I could easily understand the presenter’s description about how the
price of the stock changed.
I could easily understand the presenter’s explanation why the price
of the stock changed.
Slide 3
PowerPoint Design Skills I could easily see the three areas for future growth on the Power-
Point slide.
There was not too much text used to describe the future growth
areas on the PowerPoint side.
Speaking Skills I could easily understand the presenter’s description of three areas
for future growth.
I could easily understand the presenter’s explanation of why these
future growth areas make this company a good investment.
The PowerPoint Presentation on the Whole
PowerPoint Design Skills The PowerPoint presentation enhanced the presenter’s oral presen-
tation. The presenter’s PowerPoint gave me enough information so
that I could decide whether or not to invest in this company.
Speaking Skills The presenter’s description of the company gave me enough infor-
mation to decide whether or to invest in this company. The pre-
senter’s explanation of why I should invest in this company was a
very convincing argument.
Call to Action
– Based upon this presentation, I would invest in this company.
 Christopher Weaver

Appendix C

Transcription conventions
length of silence
(.) micro-pause
- sudden cut-off of the current sound
:: prolongation of the previous sound. The number of colons indicate the length of
prolongation.
? rising intonation
. falling intonation
> < increase in tempo
chapter 14

Language teachers’ perceptions of a task-based

learning programme in a French University

Julie McAllister, Marie-Françoise Narcy-Combes

and Rebecca Starkey-Perret
University of Nantes, France

This study is based on a large-scale research project concerning a task-

based blended language learning programme for first-year Business English
undergraduate students in a university in Northwest France. The programme is
entering its fourth year (it was extended to third-year students in January 2011).
The project is funded by the Région des Pays de la Loire and involves a team of
ten researchers.
The team have found encouraging results concerning learner involvement
and satisfaction. However, as teachers play a key role in the successful
implementation of any innovative programme, this study aims to explore the
teachers’ attitudes to their changing and increasingly complex roles within that
environment. It examines teachers’ self-perceptions and attitudes to learning and
teaching, after two years of implementing the programme, involving as it did a
change from face-to-face teacher-centred approaches to computer-mediated and
task-based teaching.
Qualitative data derived from interviews with 14 teachers involved in
the task-based blended learning programme are presented and discussed.
Results indicate that most of the teachers accept and are adapting to their new,
multifaceted role. Group dynamics and teamwork have contributed to this,
apparently playing a key role in developing a common culture in terms of their
approach to language teaching. However, some teachers have reservations,
particularly about the increased workload associated with the provision of more
personalised support for students, and to some extent about the shift away
from a transmission-based approach to teaching, which is implied in a TBLT
programme. Institutional and cultural factors are also highlighted as a major
constraint.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Theoretical background and context of the study

Introduction

The implementation of a Task-Based Learning and Teaching (TBLT) programme at

the University of Nantes (France) was triggered by the need to solve a number of prob-
lems inherited from an obsolete and inadequate system. The system needed reorganis-
ing to ensure greater coherence and cohesion within the Business English syllabus of
the Faculty of Foreign Languages and Cultures, better adaptation of the syllabus to the
real world, greater efficiency of learning of both content and language, and a reduction
in the drop-out rate. The innovation process began in 2007, with the mobilisation of a
team of fourteen (including two technicians) to develop a new learning programme.
Apart from one individual, the teaching staff involved in the team were not conversant
with English Teaching methodology, second language acquisition (SLA) theories, or
TBLT. However they were aware of the problems with the existing programme, and
complained about having to cope with large groups of students, and about the ineffi-
ciency of teaching in such conditions.
After a year of regular voluntary meetings and training sessions, the team came up
with an innovative task-based programme coupled with a technological dimension (as
a means of dealing with large numbers of students). The team started designing tasks
in 2008 and gained approval from the university governing body for implementation
in 2009 with the financial support of the regional government authority (Région des
Pays de la Loire).
Accountability to the regional government authority required a complete evalua-
tion of the implementation of the system, including the impact on students’ language
learning and adoption of the system (Narcy-Combes & McAllister, 2011; Buck &
McAllister, 2011), as well as teachers’ involvement, the subject of the present study. The
evaluation process was begun in 2009. The purpose of the present study, which is part
of the wider evaluation, is to examine the teachers’ views of the TBLT programme after
two years of experience with it. Specifically, it aims to investigate teachers’ cognitions
relating to the basic principles underpinning the programme, to understand their be-
liefs about the nature of their pedagogic roles and practices, and to examine their views
about the advantages and limitations of the programme. One reason for this focus is
that teachers’ engagement with the programme and with its development would likely
be affected by their orientation to the principles that it was based on. It was also thought
important to understand how they felt about the programme, whether they were new
to it or early adopters, so as to make decisions about future developments. In this in-
stance, the researchers’ role did not merely include evaluating and giving recommen-
dations on what to do for the programme implementation, as this could have been
resented by the teachers. Two of the researchers were also involved in the programme
as teachers, partly in order to enhance the credibility of the project in the eyes of the
teachers, and partly to increase the validity of the study itself.
 Chapter 14.  Task-based learning programme in a French University 

This chapter explores the teachers’ views through an analysis of a corpus of inter-
views with fourteen teachers involved in the programme in 2011. However, to better
interpret the study, we first describe the institutional context and how the innovation
came to be introduced. We then describe the innovation, in terms of the rationale, the
programme design, and the resulting redefinition of the role of the teachers.

Managing resistance to change in the French context

Implementing such a programme in a French university proves a challenging task in

itself. Due to its dual educational approach where an élite system exists together with
an “egalitarian” one, French universities stand apart from the rest of the world. The
brightest students do not go to university, but to highly selective Grandes Ecoles,
which attract the major share of private and public funding. In contrast, French uni-
versities do not select their students on entry. Any student with a baccalauréat1 has
the right to take up a state university place, and more and more students are obtaining
the baccalauréat due to government policy. The fees are also quite low by interna-
tional standards. The result is that France’s public universities are overcrowded and
under-funded. Only one in seven students will survive to complete the full course of
studies from year one to year four: the drop-out rate is appalling. Universities are also
still very conservative and do not accept innovation easily, though this is slowly
changing. As Joffe (2000, p. 4) puts it, “traditional universities are filled with tradi-
tional instructors” who are often not open to novel ideas and programmes. She adds
that the novelty of a new teaching method can be “overwhelming and frightening” for
some teachers (Joffe, 2000, p. 13). In this case, the task-based and technology dimen-
sions of the programme represented a potentially threatening new way of doing things
for the teachers.
Despite the fact that the proposed learning environment was grounded in innova-
tive educational approaches, supported by theory and research findings, and had been
carefully constructed, its implementation faced unforeseen difficulties, thus highlight-
ing the fact that what might be widely accepted in one context might not be easily
transferable to another. When presented for the first time to the university’s governing
body the project originally triggered hostile reactions and fierce resistance. It was nec-
essary at that stage to set out to understand what underlying phenomena were in-
volved, and we thought that the literature on work psychology and organisational
studies could help to shed light on the situation.
For example, work psychology suggests that resistance to change is a self-protec-
tion mechanism that can take the form of avoidance strategies, rebellion, opposition,
and rejection, whether that change comes from the students themselves, the teachers,

1. The baccalauréat is an academic qualification which French students take at the end of
secondary education.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

or the institution (Goguelin, 1989). As such, it should not be interpreted as a nega-

tive phenomenon, but seen as an integral part of the innovation process, which itself
is destructive by nature. In this sense, it should be considered as a parameter to be
questioned, analysed, and integrated into the project itself (Bareil, 2004). The pro-
posed innovative course was a paradigmatic revolution from instruction to autono-
my, from transmission to construction of knowledge, and from teacher-centred to
learner-centred learning and individualisation (Soubrié, 2008). The teachers in-
volved were likely to have to leave their comfort zone and face uncertainty, triggering
anxiety as their interpersonal and intrapersonal balance was disrupted. And indeed
they expressed fears of losing face due to technophobia and of losing their peda-
gogical freedom. They also thought that the students might refuse to do the tasks,
that their relations with the students might be altered for the worse, and that the in-
novation would mean an overall increase of their workload. Even so, they could im-
mediately see the benefits of dealing with smaller groups of students and shared the
widespread concerns about the drop-out rate. This provided one cornerstone for the
innovation.
In the French context in general resistance should be expected, whether active or
passive, individual or collective, conscious or unconscious. For Michel Crozier
(1963), who adapted Max Weber’s model to the French context, France has a bureau-
cratic type of organisation which implies a high degree of control, rigid and complex
regulations, lengthy and slow procedures, and a disregard for individual needs, which
does not facilitate the implementation of new projects. Change will be accepted only
if individuals can see how they can benefit from it, and if they can be certain that they
will be allowed to go back to the old ways if desired (Bareil, 2004; Rogers, 2003;
Fullan, 2007). The researchers had to acknowledge these facts, listen to the fears ex-
pressed, and resist the temptation to minimise them. Then action had to be taken.
Seminars were organised every month to explain what was new, reassure, and also
listen to objections and offer the participants an opportunity to take part in the proj-
ect, which implied flexibility and pragmatism. Training sessions were provided on
request to reduce anxiety due to technophobia. All along, the watchword has been
“communicate”.
Common sense holds that a minority cannot influence an overwhelming majority
of people. Social psychology, however, shows that common sense can be overridden if
the minority group has a strong conviction of the validity of their views, and takes a
cogent and well-founded stance, and this may lead to deep and durable change (Anzieu
& Martin, 2000). This approach implies theoretical and technical rigor, but also time
– to communicate with and inform participants and decision-makers – and the prag-
matism needed to avoid conflict. The whole process also requires patience. This was
the other main cornerstone that enabled the innovation to be introduced. Having
sketched out the broader context for the innovation, we now turn to the specifics of the
programme itself.
 Chapter 14.  Task-based learning programme in a French University 

Rationale for the TBLT programme

The TBLT programme, in which this research was conducted, involves more than 800
students enrolled in their first year of a three-year undergraduate degree course in
Languages and International Trade at Nantes University. Students can study up to
three languages, including English, which is compulsory, and Spanish, German, Italian,
Portuguese, Russian, Arabic, Chinese, or Japanese. They also study business-related
subjects including law, economics, and management. Despite the programme attract-
ing increasing student numbers in the first year (up by 29% between 2008 and 2010),
it also experiences a 40% + drop-out rate at the end of the first year, a trend which
begins at the end of the first semester.
This state of affairs has raised questions for the team of teachers as to why students
drop out and what can be done to reduce the problem. The size of language classes,
which can vary from 45 to 60 students per class, is a contributing factor to student
failure, because personalised follow-up proves difficult in such conditions. In addition,
placement tests show that students’ language levels are heterogeneous, with only 25%
attaining the required B2 level on entering university (see Table 1).
As a response to these problems, a task-based approach was developed in con-
junction with a distance learning facility provided through a learning platform
(Moodle), and with tutorials, all with a view to coping with large groups of often de-
motivated students, many of whom choose this course as a last resort. The task-based
approach would enable us to provide students with real-life tasks, directly linked to the
students’ career prospects, potentially triggering their motivation and involvement,
and hopefully fostering a degree of learner autonomy. In addition the programme de-
sign would enable the face-to-face tutorials to be organised in smaller groups, favour-
ing more peer and student-teacher interactions, which had the potential to promote
more effective second language (L2) acquisition processes.
The programme was initially designed for first-year students studying English
within a programme of Languages and International Trade. It has since been extended
to third-year students. For institutional and practical reasons, the blended element of
the programme is only implemented during the second semester, although the TBLT
programme as a whole is used throughout both first and second semesters.

Table 1.  1st Year Student Proficiency Levels at the Beginning of Semester 2, 2010

CEFR Language Level Number of students (n = 285) Percentage

A1    9   3%
A2   68 24%
B1 135 47.5%
B2 (target level)   55 19%
C1/C2   18   6.5%
Note: CEFR is the Common European Framework of Reference.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Programme design

Following on from the broader rationale outlined above, a number of guiding princi-
ples were adopted as a working basis for the design of the programme, drawing on a
broad range of literature (Bygate, Skehan, & Swain, 2001; Ellis, 2003a; Willis & Willis,
2007; Willis, 2009). This led towards a holistic approach to language learning, a learner-
centred programme, and seeking a balance between focus on meaning and focus on
form. There are many definitions of what constitutes a task, but the team chose to work
with Van den Branden’s (2006) definition: “A task is an activity in which a person en-
gages in order to attain an objective, and which necessitates the use of language” (p. 4).
An underlying assumption is that the development of any programme is crucially
influenced by teachers’ beliefs about L2 acquisition, following Chapelle (2003), who ar-
gued that the implementation of a learning environment depends on the adoption of a
congruent language acquisition theory. In our case, the development of the programme
was grounded in a socio-constructivist and cognitivist approach to language learning,
combining social tasks and individual tasks for metareflection and practice (Narcy-
Combes, J.-P., 2012; Bertin, Gravé, & Narcy-Combes, 2010). Given the nature of the TBLT
project, it follows that there is special interest in exploring the beliefs about L2 acquisition
and about the teaching practices of teachers who have been involved in the programme,
and to relate the findings to the assumptions that underpinned the innovation.
According to socio-constructivist and cognitivist approaches to language learning
and evaluation (Bruner, 2000; Robinson, 2001), learning is not transmitted from
teacher to learner, but rather constructed by the learner through social interaction and
individual involvement and reflection (Kintsch, 2009; Lantolf, 2000; Little, 2007). Ac-
cordingly, the programme proposed social tasks which were complex holistic tasks
leading learners to interact and collaborate, without direct interference from the
teacher (see below for an example). In addition, self-training, in the form of repetition
and practice, plays a key role in ensuring long term acquisition (Bygate, 2009). Hence,
focused tasks were proposed to learners for that purpose.
A TBLT programme requires a certain degree of autonomy in order for a learner
to function. However, the idea of autonomy, in the sense defined by Holec (1981, p. 3),
as “the ability to take charge of one’s own learning”, represents a challenge to the cul-
tural and educational tradition in France, where learners are accustomed to a more
passive role in the traditional classroom setting and are often reluctant to learn in a
wholly self-directed manner, while teachers themselves are not always ready to grant
learners the autonomy they would require to perform the tasks (for a full up-to-date
discussion of differences in the field across cultures see Ehrenberg, 2010). Thus, the
question of learner autonomy represents a significant challenge in the French context,
and the issue could not be ignored.
The programme combines two hours of face-to-face teaching with 2–3 hours of
autonomous group work and asynchronous computer-mediated communication
(CMC). The latter are self-paced, so the actual duration depends on the rhythm of the
 Chapter 14.  Task-based learning programme in a French University 

students. The face-to-face teaching is broken down into two distinct classes: students
meet with the teacher for one hour as a whole class of 45 students and one hour in
tutorials of fifteen students. The plenary session is used for feedback, focus on form,
and vocabulary or skills-based focus, while the tutorials are used for task performance
(for types of tasks, see below) and to encourage meta-reflection based on personalised
feedback and support from teachers (in accordance with Swain’s (2000) work on the
role of “noticing” and “meta-reflection” in second language acquisition).
The CMC part of the course is based on Moodle, an open source virtual learning
platform, and comprises three distinct spaces:
– a space containing lesson materials, that is, documents with detailed instructions
relating to the preparatory work to be carried out and task objectives;
– a self-training centre which contains interactive grammar, vocabulary, and listen-
ing exercises to help students work on those morphosyntactic, phonetic, and lexi-
cal areas they have most difficulty with as highlighted by the teacher’s feedback;
– a space reserved for the teacher’s specific group where students can upload their
written work for correction and where the teacher provides feedback; it includes a
class forum to promote the development of a community of learners.
Students work on eight tasks during a twelve-week semester. The tasks are designed to
be student-centred and involve collaborative learning and problem solving. They are
in the form of real-life business scenarios (Roots-Buck, 2005) with an overriding focus
on meaning and authenticity, however a beneficial focus on form has also been inte-
grated. The topics of the tasks relate to the syllabus of the course which focuses on
business and economic issues (for example, creating a company, outsourcing, and
counterfeiting). The fact that the tasks are relevant to students’ interests is a key factor
in fostering motivation. Dörnyei (2005) describes task motivation as a complex pro-
cessing system comprising three interrelated mechanisms: firstly, the engagement of
learners in task execution; secondly, the continuous appraisal process of task perfor-
mance; and finally action control (pp. 81–82). He particularly emphasises the role of
focused engagement in an activity which can lead to improved performance on a task.
This can be exemplified as follows for an oral presentation. The students are given a
brief about a specific theme, here “outsourcing”.

TASK A: OUTSOURCING
Oral Presentation
Brief:
InterState is currently considering outsourcing all or part of its call-centre in order
to reduce its operating costs.
Several groups of managers have been asked to research different host countries in
order to host a pilot project.
The management teams will be meeting later on in the day to present their recom-
mendations.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Students take an active role in gathering and evaluating information, discussing ideas,
and reaching an agreed outcome. They work in groups of three to four students, which
they organise themselves outside official class time. Thus, the learners are empowered
to take responsibility for their own learning experience to help them progress towards
greater autonomy. Preparatory tasks include building vocabulary relevant to the issue,
gathering data in the form of a webquest, listening and reading for information, and
reactivating useful forms (e.g., here they might need to make predictions about the
future by referring to situations that may possibly happen). The purpose of prepara-
tory tasks is to make sure that they have access to the input necessary to deal with the
subject. Typical outcomes include a jointly prepared oral presentation of the students’
work, role-plays involving negotiation of meaning, or a collaborative written product
in the form of a business report. The tasks have no expected “correct” answer and so
are open to the learner’s interpretation.
Individual feedback is provided by the teacher to cater for individual differences.
Learners can then log on to the self-training centre for individual training. Several
authors such as Dörnyei (2005), Ellis (2004), Robinson (2002), and Skehan (1991)
have highlighted the importance of individual differences (IDs) among language learn-
ers in terms of their success in mastering a second language (L2). Dörnyei (2005, p. 7)
mentions individual factors such as personality, ability/aptitude, motivation, learning
styles, and language learning strategies as being key predictors of L2 learning success.
The tasks described above were written collaboratively by all the teachers initially
involved in the programme working in tandem, with meetings of the whole team once
a month to discuss, review, and where necessary modify each task. The other compo-
nents of the programme – remedial exercises, links to external sites, and a teacher’s
resource bank – were also created collaboratively. In addition, the team met regularly
for ICT training, specifically for the use of the Virtual Learning Environment. This
kind of teamwork is uncommon in a French university context and constitutes a di-
mension of change for the teachers involved in the programme.

Redefinition of the teacher’s role

The TBLT approach outlined above, along with the transition from a traditional face-
to-face learning environment to a blended one, imply a fundamental change in the
teacher’s role in the classroom and in teacher-student relationships. As noted above,
traditionally, in the French context, foreign language classes can often be very teacher-
directed and a transmission approach dominates. Samuda (2009, p. 380) reminds us
that in this type of context the teachers operate as “provider[s] of bite-sized input,
supplier[s] of feedback and engineer[s] of controlled progressions of classroom activ-
ity”, leading them to adopt roles such as “controller”, “organiser”, and “assessor”. In the
sociocultural and task-based perspective, learners become more empowered by taking
greater responsibility for their own language learning, and the teacher’s main role is,
 Chapter 14.  Task-based learning programme in a French University 

according to Van den Branden (2009), “to act as a true interactional partner, negotiat-
ing meaning and content with the students, eliciting and encouraging their output,
focusing on form when appropriate and offering them a rich, relevant and communi-
cative input” (p. 401) (a principle that would of course be endorsed by some general
educationists within the French context). In this sense, the teacher complements and
supports the interactions and form-meaning relationships stimulated by the task in
order to favour L2 development (Samuda, 2009, p. 398). This leads the teacher to adopt
a very different set of roles including those of “monitor”, “language adviser”, and “chair-
person” according to the different stages of the task cycle (Willis, 2009).
In addition, Hampel (2006, p. 112) emphasises the “democratic and learner-cen-
tred features” inherent in many online environments which further support this shift
away from hierarchical and transmission learning approaches. She contends that “the
role of the tutor is less that of a traditional instructor than that of a facilitator, support-
ing student learning” and interactions (p. 113). In the same vein, Lamy and Goodfellow
(1999) stress equality in the classroom whereby the instructor has “to relate to learners
on a personal basis and more often as a colleague than an authority” (p. 461). They
claim that some situations will require instructors to take greater control of the class-
room, particularly when encouraging meta-reflection skills among students, but in-
structors should also know when to withdraw support from learners as they become
more autonomous.
Understanding the redefined roles of the teacher in this type of programme ex-
plains our interest in exploring teachers’ beliefs about the nature of their pedagogic
roles and practices, and how their beliefs relate to the design features and principles of
the programme. Teachers are key figures in the success of the whole project. Whatever
the learning project, the beliefs and representations of the primary agents in its imple-
mentation will be key. Strikingly, the part teachers play in a task-based learning envi-
ronment has tended to be overlooked until recently (Samuda, 2009). Van den Branden
(2009, p. 404) stresses the link between teachers’ perceptions and actions in the class-
room and how they in turn influence and are influenced by students’ perceptions and
actions. This means that, when implementing a new learning system, it is particularly
important to evaluate the impact of the programme on teachers as well as learners
(Narcy-Combes M.-F., 2008).
Two years of implementation of the programme provided a valuable opportunity
to investigate teachers’ thinking to shed light on the innovation, on the demands that
TBLT can place on teachers, and on the extent to which their thinking reflected the
impact of both the support processes and the hands-on experience of the programme.
There had been seminars, interactions during meetings, teamwork on task designing,
training to use ICT, and tutorials with smaller numbers of students. Would a common
culture emerge from it? Would teachers’ views converge with each other, with the as-
sumptions underpinning the new programme, and indeed with its key design features?
These questions motivate the study reported in the next section.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Empirical study

Aims of the study

The main aims of this study were to investigate the beliefs of teachers on a TBLT pro-
gramme about the nature of language learning, and about their pedagogic roles and
practices, in order to relate them to the theoretical assumptions of the design team,
and to the design features of the programme, so as to gain insights into factors affect-
ing its appropriation and development.
The study focused on the following research questions:
Following two years’ experience of working on the new TBLT programme, what
are teachers’ beliefs about (RQ1) second language acquisition and (RQ2) their
own roles and practices as language teachers

The underlying assumptions motivating this study are that:

(1) Representations of L2 acquisition that do not correspond to the current theo-
ries could potentially hinder the successful implementation of the TBLT pro-
gramme.
(2) Resistance from teachers to the adoption of new roles and teaching practices
could have an impact on their appropriation and usage of the TBLT pro-
gramme, which could in turn negatively impact students’ acceptance and
adoption of the programme.
The findings will be discussed in relation to the assumptions and principles informing
the design of the programme.

Methods

Participants

Fourteen teachers agreed to contribute to this study, including eleven teachers already
involved in the TBLT programme, a further two who were about to start teaching in
the programme at the time of the interviews, and one who only taught in the tradi-
tional face-to-face first semester programme. The study provides a representative and
balanced view of teachers’ attitudes to the programme as only one teacher involved in
the TBLT programme refused to be interviewed.
The group of fourteen teachers includes twelve women and two men of whom ten
are French natives and four are native English speakers. The majority are experienced
teachers who have worked between six and twenty or more years in higher education,
while four are relatively new teachers with less than three to four years’ experience. As
previously mentioned, eleven have taught in the TBLT programme since it was first
 Chapter 14.  Task-based learning programme in a French University 

implemented in 2009 and so had two years of experience with it at the time of the
study. This includes six teachers who were involved in the initial development of the
programme in 2008, and four who were specifically recruited for the programme. In
terms of their professional status at the university, two are professors, six are senior
lecturers, three are qualified teachers, and three are doctoral students on temporary
contracts.

Interviews

The study (data collection, data analysis, and interpretation) was carried out by three
researchers, the authors of this paper, who will be referred to as “the researchers”. They
all taught in the programme at some point. One instigated the programme and was
involved in its development and another was specifically responsible for carrying out
its evaluation. The experience and views of the sample group of teachers were explored
via in-depth, semi-structured interviews. This method allows participants to guide the
interview, so that key themes raised by the participant are discussed, and not just those
planned in advance by the researchers (Blanchet & Gotman, 2007).
Fourteen interviews were conducted lasting approximately 30 minutes each and
generating a total of 7 hours and 38 minutes of recorded material. All interviews took
place in French using the same interviewer to ensure consistency. The recordings were
transcribed, and hesitations and speech meanders were retained so that the final tran-
scription reflected accurately the teachers’ verbalisations. For the purpose of this chap-
ter, teachers are referred to as T1, T2 etc. in order to protect their identity, and their
comments have been translated from French to English.
The overall corpus count was 73,792 words, and the data analysis was conducted
using two different methods. Firstly, the three researchers undertook a content analy-
sis, following Bardin (2007). Secondly, to ensure objectivity and methodological trian-
gulation, a statistical textual analysis was conducted with specific software (Alceste)2.
The first step involved a quantitative thematic analysis (Bardin, 2007, p. 130) by
manually examining the fourteen individual interview transcripts for salient and recur-
ring themes. Each of the three researchers conducted this analysis independently, then
compared their results, and reached agreement concerning the themes to retain. In-
stances of the identified themes were counted and frequencies ranked across all inter-
views. From this step, the researchers identified the common themes which emerged
across the individual accounts and grouped them under 14 broad categories. Thus, the
categories and subthemes were not predetermined by the researchers, but derived in-
ductively from the analysis of the teachers’ discourse. For the purpose of this study, we

2. Alceste is a textual data analytical software tool incorporating sophisticated statistical pro-
cessing developed by France’s Centre National de la Recherche Scientifique (C.N.R.S.) with the
support of the Agence Nationale pour la Valorisation de la Recherche (ANVAR).
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

have retained 5 of the 14 themes which specifically relate to teachers’ views about L2
acquisition and the teachers’ pedagogic roles and practices (i.e., those relating to the
research questions). The themes were assigned code numbers and the researchers then
went through each transcript assigning the codes to the relevant sections when any of
the themes were mentioned. The next step involved cross-checking results and reaching
consensus. Bardin (2007) refers to the data segments resulting from this analysis as
«context units» (CUs) (p. 138) which can range in length from a phrase to a paragraph.
In parallel, the researchers used the Alceste software to analyse the corpus in order
to cross-check the themes obtained manually. Alceste is based on three approaches:
lexicometrics, content analysis, and data analysis. The analysis works by applying a set
of statistical clustering techniques to the text in order to obtain a ranking of its seman-
tic units, which is derived from the recurrent distribution of words in the text. The
software first identifies and counts the different words and forms of language in the
corpus. Then, it automatically identifies and segments the text into sequences of sen-
tences termed «elementary context units». Finally, cluster analysis is used in order to
structure the corpus after its segmentation and to eliminate non-relevant data. Cluster
analysis uses the presence/absence of words in order to construct the clusters. Alceste
carries out successive splits of the text in order to find the strongest vocabulary links
(co-occurrences) and then to extract categories of representative terms. As a result,
each cluster, which can be considered as a semantic context, is characterised by its
specific and related vocabulary. The final output is a hierarchical classification of the
clusters with calculations of the coefficient of association between a word and its clus-
ter (using Chi² statistic to measure the strength of the links). Alceste confirmed the
themes identified by the manual content analysis. However, although Alceste can be
used for content analysis, it does not provide such a fine-tuned analysis; for example,
it is unable to distinguish between positive and negative statements linked to a given
theme. For that reason, in the present study, the researchers chose to refer to the data
output obtained manually for content analysis while Alceste was used solely for verifi-
cation of the themes.

Results

Main themes reported by teachers

Out of the 14 main themes identified by the manual content analysis of the interviews
and subsequently confirmed by Alceste, we have retained five relevant to the research
questions addressed in this study: (1) views on L2 acquisition, (2) teachers’ roles,
(3) group work, (4) investment (of time and effort, etc.), and (5) autonomy. In what
follows, we present the results of the manual content analysis relating to the two re-
search questions (RQ). The first theme relates to RQ1 and deals with teachers’ beliefs
about L2 acquisition. This theme comprises six subthemes focusing on the SLA theories
 Chapter 14.  Task-based learning programme in a French University 

which teachers claimed to either endorse or oppose. Then in relation to RQ2, we ex-
plore the teachers’ beliefs about their pedagogical roles in the programme followed by
their perceptions of the practical implications of the programme which include group
work, student and teacher investment, and dealing with student autonomy.

Beliefs about second language acquisition

We will first examine the teachers’ own perceptions about L2 acquisition. This is im-
portant in helping us understand to what extent their beliefs are convergent with the
SLA-based assumptions underpinning the programme and to what extent they are
convergent with each other. Although it may be difficult to say that all the teachers
agree on the interpretation of the concepts they mention, we assume that an agreement
on the principles underpinning the programme would tend to show that a common
culture might have emerged, whereas predominant diverging views might endanger
the teachers’ long term commitment. Table 2 shows the subthemes relating to SLA
theories documented in the teachers’ interviews, together with the number of teachers
who endorsed the theory or expressed views contrary to the theory. The results are
presented in descending order of frequency in terms of the number of teachers who
mention a given subtheme. Overall, the most frequently mentioned subthemes relat-
ing to SLA theories, which were highlighted by more than half of the 14 interviewees,
were the themes of interactions/co-construction (11 teachers), meta-reflection (8), and
practice and repetition (7). We will now examine these and the other subthemes in
more detail.
Interactions/co-construction. One of the key underlying assumptions of this programme
is that learners play an active role in constructing their linguistic knowledge through
social interactions with their peers and teachers (Lantolf, 2000). Table 2 shows that the
majority of teachers (11) in this study endorsed this view. This subtheme comprises the
most comments (27 CUs) and was particularly important for five teachers (T2, T3, T4,
T5, and T12), whose comments represent 70% of the total number of CUs. Teachers
highlighted the importance of student participation, interactions, dialogic processes
(such as scaffolding), and proactiveness when performing a task as helping to promote
L2 acquisition.
Illustration of teachers’ views relating to interactions/co-construction
1. “It’s about learning from others and helping each other...they will learn if they are
active, not passive.” (T12)
2. “The learner becomes a social actor within his group. It’s about the co-construc-
tion of a project, co-construction of performance. Uh it’s all the instances of scaf-
folding, negotiation, clarification, and repetition, uh all that. That’s the strong
point for me.” (T2)
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Meta-reflection. A further assumption underlying this programme is the emphasis on

the role that “noticing” and meta-reflection play in L2 acquisition (Schmidt, 1995;
Swain, 2000; Van Patten, 1994). More than half the teachers (8) highlighted the impor-
tance of providing feedback to students on their task performance to focus their atten-
tion on problem areas in order to encourage students to reflect critically on the
language being studied, as well as on their own learning experience. Meta-reflection is
the second most important subtheme in terms of the number of teachers endorsing it
and the number of CUs. This was particularly important for one teacher (T6) whose
comments account for almost a third of the total CUs in this subtheme. The remaining
mentions are spread evenly across teachers. However, some teachers may interpret this
concept differently depending on other views they hold with respect to language teach-
ing and learning, for example T6, T7, and T10 also support transmission approaches,
so their assumptions about instruction and acquisition may be very different from the
other teachers. One teacher’s (T11) views are not in line with the assumptions of the
programme, as she questions the role of correction in L2 acquisition.
Illustration of teachers’ views relating to meta-reflection
3. “I find that when we highlight a problem... and tell the student that it has ap-
peared three times in his work and that the same problem keeps occurring de-
spite us having worked on it. uh, then ask why does that keep occurring, so they
reflect on it.” (T9)
4. “So I try, but it doesn’t always work (laughter), to make them aware of what is caus-
ing them a problem and often it has something to do with grammar. So that they are
capable of having a kind of critical reflection on what they are doing uh.” (T14)
Practice and repetition. Swain (2000) argues that learner output is key to L2 acquisi-
tion. In line with this view, the new programme assumes that producing the target
language, in the sense of using the language, enhances fluency (Ellis, 2003b) and rep-
etition of tasks, as highlighted by Bygate (2009), leads to improved performance. Seven
teachers endorsed this view mentioning the importance of regular practice, particu-
larly oral production, for favouring language acquisition. They highlighted the benefits
of students performing tasks which encourage them to participate more actively in
class and speak spontaneously. Four teachers emphasised the role of repetition and
how it helps students not only to become more confident at speaking in class, but also
get better at the task. However, one teacher (T6) in this group interpreted this concept
in a different way from the other teachers. The comments of this teacher, who supports
transmission approaches, focused on the use of repetitive drill exercises in class in a
behaviourist way rather than in a way that might fit the underpinning theory of the
TBLT programme.
Illustration of teachers’ views relating to practice and repetition
5. “They have more opportunities to practise, more opportunities to produce.” (T1)
 Chapter 14.  Task-based learning programme in a French University 

6. “It’s by doing that they will learn. And given that they are doing a lot there, well,
inevitably, I think it makes them more active. They have to produce a lot of Eng-
lish.” (T14)
Transmission of knowledge. An underlying assumption of the programme is that learn-
ers are not simply receiving information given by the teacher, but are actively engaged
in building their linguistic knowledge (Kintsch, 2009). A transmission model of edu-
cation is therefore relegated to a non-central role within the TBLT programme.
Ten teachers mentioned this subtheme, although there is a notable divergence in
their beliefs, with four teachers explicitly stating that they were against transmission
approaches (T1, T2, T4, T8) and four who were clearly in favour (T6, T7, T10, T11)
(see Table 2). Particularly favourable towards the use of transmission modes was T6,
whose comments account for almost half of the CUs (6 out of 13), and T11. On the
other hand, the other two teachers exhibited a more balanced view in that they equal-
ly favoured other SLA theories underpinning the programme. A further two teachers
had a more ambivalent view as they favoured transmission approaches under certain
circumstances (for example, grammar instruction), but in other respects were gener-
ally against this approach.
Thus, these results suggest that two teachers may have given more importance to
a transmission approach than assumed within the design of the programme, while
four others are less categorical. This shows that for some teachers there is no clear-cut
positioning, which is indeed compatible with the views of many regarding TBLT
(e.g. Willis & Willis, 2007). However it is also of course possible that there are incon-
sistencies in how people think and speak about a teaching approach, and it is also pos-
sible that views on one aspect of language teaching and learning could influence other
aspects of their thinking. Thus, those who were well-disposed towards the use of trans-
mission approaches may also have interpreted key terms such as practice, repetition,
and meta-reflection in different ways from the other participants.
Illustration of teachers’ views relating to transmission of knowledge
7. “I don’t believe in transmission. Because as long as we are transmitting, we are in
fact in an extreme narcissistic position and everything we transmit is about our-
selves and uh there is no exchange and no-one benefits from that.” (T4)
8. “I like to share ideas, communicate them...I inspire rather than teach. I like the fact
of giving ideas to someone, inspiring someone.”(T6)
The role of input. Input, in terms of the language content a learner is exposed to, is one
part of the processing model for L2 acquisition that forms the underlying assumption
of the programme (Ellis, 2003b). The developers of the programme took the view that
input should be contextualised. Four teachers recognised that exposure to meaningful
and authentic input is essential to L2 acquisition and they actively encouraged stu-
dents to get exposure outside the classroom. Only one teacher (T6) supported the ex-
plicit instruction of grammar in a non-contextualised way by focusing the learners’
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

attention on forms and structures coupled with practice. This strategy is not entirely in
contradiction with the theoretical positioning of the programme as it was used within
a broader task-based approach in which grammatical features were systematically in-
troduced within clear familiar contexts.
Illustration of teachers’ views relating to the role of input
  9. “I try to encourage the students to work outside the university context and to take
advantage of all opportunities available to them, such as watching a film or series
in English with or without subtitles. Also to read as much as possible in English
and to meet with English students at the university.” (T1)
10. “I explain to students that they have to surround themselves with the language and
that way they’ll discover gaps in their knowledge and progress.” (T9)
Individual differences. In line with authors such as Dörnyei (2005), Ellis (2004),
Robinson (2002), and Skehan (1991), the programme assumes that individual differ-
ences between learners are key to determining L2 learning success and have to be
taken into account by teachers in the classroom. In this study 4 teachers recognised the
need to cater for individual differences, pointing specifically to differences in learning
abilities and styles. One teacher highlighted the difficulty of dealing with heteroge-
neous language levels in the classroom when working on tasks and wondered if she
should continue at a certain level and risk leaving some students behind.
Illustration of teachers’ views relating to individual differences
11. “I find that we have to arbitrate between the very, uh, very diverse levels [of stu-
dents].” (T14)
12. “..it depends on a lot of parameters. I think it depends on their individual profile.
Uh particularly in that no student learns the same way.” (T1)

Table 2.  SLA Theories Reported by the 14 Teachers

SLA theory subthemes Number of Number of Number of CUs

teachers endorsing teachers with in subtheme
key principles reservations about
key principles

1.  Interactions/co-construction 11 27
2.  Meta-reflection  8 1 22
3.  Practice and repetition  7 1 20
4.  Transmission of knowledge  6 6 20
5.  The role of input  4 1  8
6.  Individual differences  4  4
 Chapter 14.  Task-based learning programme in a French University 

Table 3.  Teachers’ Stance on SLA principles underpinning the programme

Number of teachers Number of teachers Number of teachers

expressing positive views expressing negative views expressing mixed views

7 2 4

Overview by teacher. Having seen to what extent teachers’ beliefs are convergent with
the SLA-based assumptions underpinning the programme, we will now explore to
what extent their beliefs are convergent with each other. Table 3 sums up the position-
ing of 13 teachers with respect to the SLA theories underpinning the programme. It
shows that the views of 7 teachers are uniformly in line with the theoretical assump-
tions of the programme, whereas the views of two teachers (T6 and T11) oppose these
assumptions as they strongly favour transmission accounts of SLA. A further 4 teach-
ers expressed mixed views. One teacher (T13), who was a new contractor teacher at
the time of the study and who has since left the programme, did not express her views
on L2 acquisition. Closer examination of the beliefs held by the 7 teachers who support
the theoretical position of the programme revealed that they all share the idea that
social interaction and co-construction are key to L2 acquisition. Similarly, the 4 teach-
ers who expressed mixed views all share a cognitive perspective on L2 acquisition and
agreed on the importance of the role of meta-reflection in L2 acquisition.

Teachers’ beliefs about their pedagogical roles

We will now examine the teachers’ beliefs about their roles to understand if they are in
line with the pedagogic requirements of the TBLT programme. An underlying as-
sumption of the programme is the shift from a purely transmission role, where the
teacher directs and controls language learning, to multiple roles such as monitor, ad-
viser, tutor, and facilitator, where the teacher supports interactions and student learn-
ing (Hampel, 2006; Lamy & Goodfellow, 1999; Willis, 2009). Specifically, Willis (2009)
identifies three roles for the teacher depending on the stage of the task cycle, “monitor”
at the task stage, “language adviser” at the planning stage, and finally “chairperson” at
the report stage. Willis and Willis (2007) suggest a broader range of six roles, but the
account that follows limits itself to Willis’ approach (originally published in 1996).
Table 4 identifies 8 roles endorsed by the teachers in this study and presents the
results in descending order of frequency in terms of the number of teachers who en-
dorse a given role. The majority of teachers appear in several categories with some
teachers explicitly recognising the multifaceted nature of their role. We will first exam-
ine those roles identified by the teachers – six in number – which are in line with
Willis’ original account. These were endorsed by 12 teachers. Then we will review a
further role which emerged from the data but does not have a counterpart in the
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

classification defined by Willis (2009). This is a “social role” which focuses on the
teachers’ belief that they have a social responsibility to help prepare students for life
and for their future career. Finally, we will review the “transmitter” role, a potentially
controversial aspect within a TBLT programme.
The task stage – teacher as facilitator and guide. According to Willis (2009, p. 227) at
this stage students work on the task in small groups while the teacher monitors from a
distance observing and encouraging students, but without intervening or correcting
them unless there is a communication breakdown. She refers to this role as “monitor”.
The terms used by the teachers in this study to describe this role were “facilitator” and
“guide”. Five teachers described themselves as a facilitator. They saw their role as facili-
tating access to learning resources (on-line and in the classroom), as well as facilitating
learning and interactions in the classroom by eliciting and stimulating student output,
and by encouraging students to take greater initiative and risk when performing tasks.
Three of the same teachers and one other also described themselves as a “guide”. In
their view, this involves guiding students on how to use resources, clarifying task in-
structions, and providing some guidelines for the accomplishment of tasks. As with
Willis’ (2009) definition, the teachers did not mention error correction for these roles.
Illustration of teachers’ views relating to facilitator and guide roles
13. “But we also facilitate the student’s work and learning.” (T1)
14. “The role is a little bit like a guide...we give students some direction and support
[to accomplish tasks]” (T5)
The planning stage – teacher as adviser and mediator. Willis (2009, p. 232) attributes the
role of “language adviser” to teachers at the planning stage of the task cycle when stu-
dents are preparing oral or written presentations. She argues that the focus at this stage
is on helping students understand their objectives, providing advice on language ac-
cording to students’ needs, and encouraging learner independence. We can associate
two of the roles identified by teachers in this study with this stage: “adviser” and
“mediator”.
“Adviser” emerged as the most frequently cited role by teachers: 9 teachers en-
dorsed it and the subtheme comprises 28 CUs. It was particularly important for four
teachers (T1, T9, T12, and T14), whose comments account for almost two-thirds of
the total number of CUs. From the teachers’ perspective, this role involves listening to
students, providing advice on content and language, and also helping them to define
learning strategies that will favour language acquisition and greater learner autonomy.
The notion of “accompanying” students was strongly emphasised by teachers, which
includes being available outside of class time so students will not feel isolated and pro-
viding follow-up (sometimes psychological). Three teachers (T1, T12, and T4) saw this
role as being necessary in the French context, while two others (T8 and T14) high-
lighted the difficulties associated with the context in terms of having to advise large
numbers of students. A further 3 teachers (T4, T8, and T12) viewed this role as
 Chapter 14.  Task-based learning programme in a French University 

desirable in the sense that they perceived the exchange with students both enriching
and interesting.
Four teachers endorsed the mediator role which they said involves them helping
students to bridge the knowledge gap between what they actually know and what they
should know by drawing their attention to the gaps, providing scaffolding, and en-
couraging them to self-correct.
Illustration of teachers’ views relating to adviser and mediator roles
15. “So my role would be to show students how they can progress in English, give
them advice uh so they understand what they have to do uh to improve their mas-
tery of the language.” (T9)
16. “We have to help them identify their problems, uh, it’s something we have to give
them feedback on so they can self correct, therefore we have to target our com-
ments correctly. It’s not just something which is a sanction.” (T13)
The report stage – teacher as animator and tutor. According to Willis (2009, p. 233), at
this stage, when students present to the class or produce a written report, the teacher
acts as a “chairperson”. This role entails introducing the presentations, managing the
speakers and timing, summing up at the end, and providing feedback on content and
language. The roles identified by teachers in this study corresponding to the report
stage are “animator” and “tutor”. Two teachers endorsed the “animator” role and saw
it as involving managing the different groups of students and their interactions in
class when performing a task, keeping time, and summarising at the end. Five teach-
ers endorsed the “tutor” role which they said focuses on providing personalised cor-
rective feedback based on the students’ individual needs as highlighted during task
performance.
Illustration of teachers’ views relating to animator and tutor roles
17. “And then when we do group work we take on the role of an animator. We manage
the different groups and activities, we re-explain and reformulate etc.” (T1)
18. “We are also tutors insofar as we give students individual feedback based on their
productions.” (T1)
Teacher’s social role. One of the advantages of adopting a TBLT approach is that it com-
bines language teaching with a response to the very social needs of the learner. This
“social role” was traceable in the utterances of 7 teachers and incorporates both a so-
cio-affective and professional dimension. It involves stimulating learning through es-
tablishing a level of empathy between the teacher and student. In addition, teachers
perceived that they have a social responsibility to help prepare students for life and the
global community and also to prepare them for their future career by proposing real-
istic tasks and authentic business situations.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Illustration of teachers’ views relating to social role

19. “The way we work in this programme encourages interactions with the students
which are often continued outside the classroom and help them to feel more in-
volved. I believe we have a social role uh we have to build a social link with stu-
dents. That’s important.” (T8)
20. “It’s a big responsibility, I owe it to them to help them get a job, training and to
learn to reflect critically.” (T3)
Teacher as transmitter. As mentioned earlier, consistent with a TBLT approach, the
programme entails a shift from teacher-directed instruction to student-centred learn-
ing, with a transmission approach to teaching becoming ancillary rather than central.
Hence given the traditional importance of this mode of teaching, teachers’ thinking on
this aspect of their role is of particular interest.
As Table 4 shows, five teachers endorsed the “transmitter” role believing they
should transmit information, knowledge, and methodology to students. Of these five,
the comments of two in particular (T6 and T11) showed they strongly favoured the
adoption of language-centred and transmission approaches in the classroom. The claim
for centrality of this role would be inconsistent with assumptions underpinning the ap-
proach. The three others (T8, T9, and T14), however, emphasised a mix of roles in the
classroom (see Table 5). We can draw some parallels here between the teachers’ stance
on L2 acquisition (see Table 3) and the roles they said they adopted in the classroom
Illustration of teachers’ views relating to transmitter role
21. “Well, in any case I’ve always been convinced that we have to transmit uh work
methods and approaches...” (T8)
In sum, all 14 teachers accepted the types of roles intended by the TBLT programme
(see Table 5). Of these, 12 acknowledged the 6 roles we identified as corresponding to
the different stages of the task cycle (Willis, 2009) while 2 chose only to mention the
social role (T10 and T11). Twelve teachers referred to their role as a multiple one

Table 4.  Pedagogical Roles Endorsed by Teachers

Pedagogical roles Number of teachers Number of CUs in subtheme

1.  Adviser 9 28
2.  Social role 7 11
3.  Transmitter of knowledge 5 13
4.  Facilitator 5  7
5.  Tutor 5  5
6.  Guide 4 12
7.  Mediator 4  5
8.  Animator 2  2
 Chapter 14.  Task-based learning programme in a French University 

endorsing between two to four different roles, that is to say, all teachers with the excep-
tion of T4 and T10 (see Table 5). Their comments are in general evenly spread among
the roles, however, some teachers emphasised certain roles, for example T1
(adviser and guide), T9, T12, and T14 (adviser), and finally T6 (transmitter of
knowledge). It is noticeable that there is acceptance of a transmission orientation
amongst the various teaching roles.
As previously mentioned, we can again draw some parallels between the teachers’
stance on the SLA theories underpinning the programme (refer to Table 3 and the
summary that follows it) and the roles they said they adopted in the classroom. The 7
teachers (T1,T2,T3,T4,T5,T8, and T12) whose beliefs seemed consistently congruent
with the theoretical assumptions of the programme, and all who favoured interaction
and co-construction, endorsed roles such as facilitator and guide, which stress interac-
tions in the classroom, and mediator, which supports co-construction (see Table 5).
Three of the 4 teachers (T7,T9, and T14) who expressed mixed views relating to SLA
theories, and all who favoured a cognitive stance, endorsed roles such as adviser and
tutor which emphasise language feedback and a focus on form.
Illustration of teachers’ views relating to multiple roles
22.  “And uh so we have to take on several roles in one sense.” (T1)
23.  “I think as a teacher, we are a guide, a facilitator, coach and all those kind of things.
I see it more in terms of a multiple role.” (T2)

Table 5.  Summary of Roles Endorsed by Each Teacher

Teacher Roles (number of CUs)

T1 Adviser (5), facilitator (2), tutor (1), guide (4), animator (1)
T2 Adviser (2), facilitator (1), guide (1)
T3 Social role (1), facilitator (1), mediator (1)
T4 Adviser (3)
T5 Adviser (1), facilitator (2), guide (3)
T6 Adviser (1), transmitter of knowledge (5), tutor (1), guide (4)
T7 Social role (1), tutor (1)
T8 Adviser (3), social role (1), transmitter of knowledge (3), mediator (1)
T9 Adviser (4), transmitter of knowledge (2), mediator (2)
T10 Social role (2)
T11 Social role (2), transmitter of knowledge (2)
T12 Adviser (5), facilitator (1), tutor (1), animator (1)
T13 Social role (1), tutor (1), mediator (1)
T14 Adviser (4), transmitter of knowledge (1), social role (3)
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Teachers’ perceptions of the practical implications of the TBLT programme

In this final part of the study, we will explore the teachers’ perceptions of the practical
implications of the TBLT programme. What emerged from the data was that teachers
expressed their degree of satisfaction with the new programme, but also pointed at its
limitations. They specifically mentioned three themes relating to their practices, namely
group work, student and teacher investment, and autonomy. These themes provide fur-
ther elements of response to RQ2 as to whether teachers’ practices are in line with the
pedagogic requirements of the programme. There is some overlap, however, with RQ1
because teachers’ views on L2 acquisition will condition what they believe their prac-
tices should be, as acknowledged by Van den Branden (2009): “What language teachers
do in the classroom is inspired by what they know, believe, and think” (p. 403).
Group work. A major feature of the TBLT and sociocultural approaches, which under-
lies the new programme and differs from the teachers’ former classroom practices, is
the use of collaborative group work to enhance classroom communication. Group
work was a recurring theme in the teachers’ dialogue, with Alceste ranking it statisti-
cally as the second most important one. The content analysis revealed that teachers’
perceptions of group work were largely positive with more than three times as many
positive mentions of group work as compared to negative mentions (see Table 6). In
total, 13 teachers showed support for this approach and reported four key benefits as-
sociated with group work (see Table 6). However, it should be noted that 4 teachers’
comments account for just over half the total number of CUs (T8, T9, T10, and T12).
As summarised in Table 6, from the point of view of ten teachers, the main benefit
of working on tasks in tutorials with small groups of students was that students were
encouraged and, indeed, obliged to present and speak more in class. A key point that
emerged was that working in small groups gave students more confidence to “dare” to
speak in front of the class. Seven teachers believed that this active oral participation
coupled with greater student investment and mutual assistance contributed to stu-
dents’ progress.
Another advantage highlighted by 9 teachers was the ability to provide more
individualised feedback and support to students as a result of group work. Teachers
distinguished two types of individualisation: from an L2 acquisition development per-
spective, they were able to identify individual students’ difficulties more easily and
provide them personalised post-task feedback; from a socio-affective perspective,
teachers felt they established a more personalised relationship with students by getting
to know them individually. This, according to some teachers, was a radical change
from the traditional way of teaching at the university.
Despite these clear benefits, five teachers highlighted negative aspects associated
with group work (see Table 6). T12 stressed the difficulty of managing heterogeneity in
terms of students’ language and motivation levels. T9 highlighted a problem encoun-
tered with some students who perceived her feedback as negative criticism which led
 Chapter 14.  Task-based learning programme in a French University 

to hostility. T7 and T14 found evaluating students problematic; they cited difficulties
in evaluating students’ individual contributions to collaborative written tasks (T7) and
in evaluating oral contributions when everyone spoke at the same time in class. Fi-
nally, three teachers (T3, T12, T14) mentioned the students’ lack of autonomy with
respect to organising their groups and working effectively in groups.
Illustration of teachers’ views relating to group work
Positive views
24. “They work more, they are more involved and in particular they are more active.
Uh the oral activities oblige them to speak in front of the whole class and that’s
something they don’t do during the first semester.” (T10)
25. “The most noticeable change is that we get to know them [the students] very
quickly whereas before it was impossible for me to know my groups. I didn’t know
them.” (T4)
Negative views
26. “One of the negative things about group work, written tasks in particular, is that
we don’t really know who does what...It is difficult to know what each student is
worth.” (T7)
Student and teacher investment. Robinson (2009, p. 197) reminds us that working on
meaningful real-world tasks can increase student interest and investment in classroom
pedagogic activities. Our previous analysis highlighted an association between group

Table 6.  Positive and Negative Views Reported for Group Work

Positive mentions about group work Negative mentions about group work
13 teachers (49 CUs) 5 teachers (13 CUs)

Increased oral participation (20 CUs) Students’ lack of autonomy (4 CUs)

10 teachers 3 teachers
Individualised follow-up/feedback (10 ECUs) Difficulty in evaluating students’ contributions
9 teachers to group work (3 CUs)
2 teachers
Greater student investment and active role Students’ negative reaction to feedback (2 CUs)
(11 CUs) 1 teacher
7 teachers
Co-construction (8 ECUs) Heterogeneity of language levels (2 CUs)
4 teachers 1 teacher
Students’ lack of motivation (1 CU)
1 teacher
Time consuming (1 CU)
1 teacher
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

work and increased student investment in terms of their work effort. Here we will
discuss teachers’ views regarding a possible link between tasks and student investment.
Ten teachers perceived tasks as having a positive impact on student investment: tasks
were seen to push students to work more and become more active and involved inside
and outside the classroom. Two teachers in particular appreciated this aspect: T1 and
T10, whose comments accounted for more than 40% of the total number of CUs
(15 out of 34). Eight of the teachers stressed the fact that the pre-task preparation re-
quirements of the programme and the relative frequency of scheduled tasks obliged
students to work. Nevertheless, according to 8 teachers, not all students demonstrated
the same level of investment and this was linked to individual differences and students’
lack of autonomy. Specifically, this lack of investment translated into unequal contri-
butions to group work and low usage of the on-line self-training centre by some stu-
dents. It was particularly problematic for T5, whose comments accounted for a quarter
of the total.
Illustration of teachers’ views relating to student investment
27. “They worked more because finally the tasks, uh, if they want to prepare them
well, they require a lot of time.” (T9)
28. “It’s a voluntary step to go and look at the resources and work on-line. I think that
could explain why some students progress and realise that they are progressing.”
(T10)
29. “[The students worked more due to] the fact that there are eight tasks to work on
during the semester including four oral tasks which oblige them to speak. Uh, it’s
a very sustained pace, perhaps too much sometimes.” (T14)
Illustration of teachers’ views relating to students’ lack of investment
30. “I noticed that a lot of students don’t use the self-training centre.” (T10)
31. “I think some students clearly put their names on the work handed in, but don’t
actually do any of the work. Although the majority do the work more or less cor-
rectly.” (T3)
In parallel, all teachers, with the exception of T12, noted that the new approach re-
quired greater personal investment from them in terms of time and additional effort
because of the increased workload (for example, preparing and evaluating tasks, pro-
viding personalised feedback post task, and communicating with students via email or
group forums etc.). They also mentioned that it required them to be more active and
attentive while monitoring student interactions in class and providing the right level of
scaffolding support, which they perceived as demanding. The increased workload was
a greater issue for two teachers (T11 and T13), whose comments account for just over
a third of the total number of CUs. Both teachers were contractual workers and have
since left the programme. Three teachers (T8, T9, and T11) explicitly stated that they
expected a return on their investment in terms of remuneration. Currently, teachers
 Chapter 14.  Task-based learning programme in a French University 

are remunerated for the extra work undertaken in the context of this programme, for
example: for task development, student follow-up, and attending meetings.
Illustration of teachers’ views relating to teacher investment
32. “There’s more correction [than the first semester] because we have tasks to evalu-
ate every week. In the first semester we only had four to correct, so we had more
time.” (T1)
33. “It takes up a lot of our time. I’m a little bit overwhelmed (laugh). But it’s satisfying
all the same.” (T5)
34. “This project demands a huge involvement from teachers and it’s right that they
should be remunerated at least in part for uh this uh extra effort which has been
the case. So that’s very good.” (T8)
Autonomy. Autonomy, in terms of students taking responsibility for their own learn-
ing, is built in to the new task-based blended learning programme because of the de-
mands of group work and of the self-directed context of the programme where stu-
dents work without teacher supervision. Table 7 shows that 9 teachers perceived these
features of the programme key to fostering greater learner autonomy. Ten teachers also
highlighted their own role in actively encouraging and supporting students to take
charge of their learning, for example, by guiding them to on-line resources, providing
feedback designed to encourage meta-reflection, and helping students to work effec-
tively in groups etc. However, only 6 teachers spoke positively about their students’
capacity for autonomy. This confirms Hurd’s assessment (2005) that we “cannot make
any assumptions or expectations about learners’ willingness or ability to become au-
tonomous learners” (p. 12). In contrast, 9 teachers, including the same 6 previously
mentioned, deplored the lack of autonomy of some students suggesting it is a key
problem in this context. It appeared to be particularly problematic for three teachers
(T5, T12, and T14), whose comments account for more than 60% of the total.
Seven teachers proposed reasons for this lack of autonomy: cultural differences
whereby French students are not used to taking responsibility for themselves (T2, T4,
T5, and T14), students’ lack of maturity (T2 and T13), and students’ representations
about the teacher’s directive role in the classroom (T1, T2, T12). Teachers viewed this
lack of autonomy as having a negative effect on student investment which, as we have
already seen, resulted in a weak involvement by some students in pre-task preparation
and in the use of online resources, as well as a greater passivity in class.
Two teachers specifically mentioned autonomy from the point of view of the
teacher’s autonomy. Little (2007) argues that learner autonomy and complementary
teacher autonomy are interdependent. The two teachers in question (T3 and T8) high-
lighted the importance they attached to maintaining their personal autonomy within
the programme by retaining their pedagogical freedom. Van den Branden (2009) af-
firms that for some teachers it is crucial to maintain control over what happens in the
classroom.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Table 7.  Teachers’ Views on Autonomy

Subtheme Number Number of CUs

of teachers in subtheme

Role of teacher in promoting autonomy 10 18

Role of TBLT programme in promoting autonomy  9 19
Students’ lack of autonomy  9 26
Students’ capacity for autonomy  6  8
Teacher’s autonomy  2  4

Illustration of teachers’ views relating to autonomy

35. “Distance learning means they have to be more autonomous, they have to prepare
their tasks and function in groups and all that is very good. It means they have to
become responsible.” (T5)
36. “It’s true that the French tend to want to be supported, we could say “spoon fed”,
it’s a bit like that all the same, uh, in our culture. I think it’s not surprising, but a
cultural revolution is necessary. The question is: will it happen?” (T4)

Discussion

The aim of the study was to explore teachers’ attitudes to the innovative TBLT pro-
gramme after two years’ experience working on it as part of a broader programme
evaluation for the stakeholders. Specifically, we wanted to investigate their beliefs
about L2 acquisition, and about their pedagogic roles and practices, so as to relate
them to the assumptions and intentions of the design team, with the broader view of
evaluating the future prospects of the programme. With respect to teachers’ beliefs
about SLA the results are mixed. It was postulated that representations that do not cor-
respond to the assumptions underlying the programme could hinder its successful
implementation. We wondered whether a common culture would emerge after two
years of active participation in the development of the programme. The study shows
that the thinking of at least half the teachers converges to that of the design team, that
two seem somewhat divergent, while a further four cannot be clearly positioned. As we
have seen, 11 of the 14 teachers interviewed, while not endorsing all the specific prin-
ciples drawn on in the design of the programme, nonetheless appear to be developing
a common culture. Significant here is the fact that only five of the teachers already had
prior knowledge of SLA theories, which may also help to account for some of the ap-
parent contradictions and incoherencies in some of the teachers’ utterances and for
their different interpretations of concepts. The findings suggest that sharing views and
teamwork in a non-threatening environment can contribute to the evolution of teach-
ers’ representations in the long run, although it is impossible to say how far this was
 Chapter 14.  Task-based learning programme in a French University 

due to the particularities of the innovation and the way it was conducted, or how far it
reflected their readiness for change. In contrast, the more transmission-oriented views
of two teachers (T6 and T11) on L2 acquisition appear to be to some degree inconsis-
tent with the broad assumptions of the design team, though it does need to be ac-
knowledged that there are potential inconsistencies even within the data of individual
teachers. For example, it is worth noting that 13 of the 14 teachers endorsed the value
of group work, and it is of course quite possible for a teacher both to believe in the
value of an element of transmission-oriented teaching and at the same time in the
importance of ensuring an element of student-centred or group-based work. It is per-
haps significant however that the two teachers concerned both chose not to participate
in team meetings nor in the development of the programme. The direction of causality
cannot of course be inferred from this.
Concerning the pedagogic implications of the programme, the thinking of all
teachers seemed convergent with the types of roles anticipated in the design of the
programme, with 12 specifically endorsing the types of roles identified by Willis (2009)
in her TBLT framework. It is also clear that the overwhelming majority of the teachers
(12) see themselves as taking on a multiplicity of roles in this type of programme, in-
cluding adviser, facilitator, guide, and tutor. Five of the teachers continue to espouse
the use of a transmitter role. This was particularly important for T6 and T11, the same
two teachers whose beliefs about L2 acquisition were at odds with the theoretical as-
sumptions used by the design team. We were able to draw some further parallels be-
tween the teachers’ beliefs about L2 acquisition and the roles they said they adopted in
the classroom. Specifically, those teachers that saw a significant role for interaction and
co-construction also endorsed roles such as facilitator, guide, or mediator, which are
congruent with the use of pedagogic activities which emphasise interaction and co-
construction in the classroom. On the other hand, teachers that favoured cognitive
approaches tended to endorse roles such as adviser and tutor which focus on providing
language feedback. This suggests some degree of convergence in their views on L2 ac-
quisition and classroom roles. Further research is planned to examine how the teach-
ers perform the roles in the classroom by analysing classroom interactions and then
comparing what teachers say about their practices to what they actually do in class.
The programme gained support from the teachers at the start because they were
interested in the potential benefit for them to deal with smaller groups of students in
face-to-face tutorials. In that respect, the teachers clearly recognised the benefits of the
TBLT practice of groupwork in terms of greater student involvement in the tasks and
increased oral participation in class, which they saw as promoting language acquisi-
tion. Working with smaller groups also enabled them to provide a more individualised
form of teaching. These perceived benefits may have contributed to the teachers’ en-
dorsement of the theoretical underpinnings of the programme. It is possible that the
more they perceived the approach to pay off, the more they endorsed it. However, the
increased personal investment required is problematic for thirteen of the fourteen
teachers interviewed and some teachers may not remain committed to the programme
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

because of this additional workload. The lack of student autonomy in this context re-
quires teachers to provide increased support therefore adding to their workload. It
appears then that what was gained on the one hand was counterbalanced on the other.
Clearly, the researchers will have to consider the issue and study what solutions could
be proposed. The subject is on the agenda for future negotiations with the regional
government authority and the university governing body in the coming months.

Conclusion

Teachers’ representations and beliefs should be taken into account when innovation is
implemented. As this study shows, innovative TBLT programmes can represent a rad-
ical departure from prevailing teaching practices and the process of change requires
time, pragmatism, and continued efforts in communicating the benefits to all involved.
We have seen how the role of the teacher has been redefined in the TBLT environment
and results indicate that the majority of teachers seem to have accepted this changing
role. Despite the fact that we identified converging views on L2 acquisition, apparent
incoherencies and disparities indicate that group discussion, seminars, and teamwork
should be encouraged and become part of the programme – and not be limited to the
experimental phase – as they could play a key role in developing a common culture
between teachers in terms of language learning approaches.
Teachers’ recognition of the key benefits of the programme may have contributed
to their endorsement of its underlying principles. In this case, the TBLT approach was
seen as promoting the implementation of smaller class sizes and a focus on group
work, which led to greater student motivation and involvement and enabled teachers
to provide more personalised support. On the other hand, the study also highlights
key issues for them in terms of increased workload and students’ autonomy which are
linked. With regard to student autonomy, it remains an issue in this context and we
cannot assume that the task-based blended learning programme will automatically
lead to learner autonomy. Indeed, the idea of autonomy does not necessarily fit easily
with the French culture, where it is often assumed by students (at least in language
programmes) that the teacher has sole responsibility for directing learning, defining
objectives, evaluating work, and assessing progress. Thus, teachers have a key role to
play in preparing students for autonomy by helping them to develop their self-man-
agement and critical reflection skills through dialogue and feedback which could in
turn contribute to reducing their workload. Nevertheless, the issue of workload is not
limited to student’s lack of autonomy and the successful implementation and extension
of the programme could be under threat if this problem is not resolved. The team are
working to address this issue by examining solutions grounded in theory which could
require institutional adaptation and approval.
 Chapter 14.  Task-based learning programme in a French University 

References

Anzieu, D., & Martin, J-Y. (2001). La dynamique des groupes restreints. Paris: Presses Universi-
taires de France.
Bardin, L. (2007). L’analyse de contenu. Paris: Presses Universitaires de France.
Bareil, C. (2004). Gérer le volet humain du changement. Montréal: Les Éditions Transcontinental.
Bertin, J-C., Gravé, P., & Narcy-Combes, J-P. (2010). Second language distance learning and
teaching. New York, NY: Hershey.
Blanchet, A., & Gotman, A. (2007). L’enquête et ses méthodes, l’entretien. Paris: Armand Colin.
Bruner, J. (2000). Culture et modes de pensée: l’esprit humain dans ses œuvres. Paris: Retz.
Buck, J. & McAllister, J. (2011). Mise en place d’un dispositif d’apprentissage hybride à l’université.
Les Cahiers de l’Apliut, 30(1), 83–101.
Bygate, M., Skehan, P., & Swain, M. (Eds). (2001). Researching pedagogic tasks: Second language
learning, teaching and testing. Harlow, UK: Longman.
Bygate, M. (2009). Effects of task repetition on the structure and control of oral language. In K.
Van den Branden, M. Bygate, & J.M. Norris (Eds.), Task-based language teaching: A reader
(pp. 249–274). Amsterdam: John Benjamins.
Chapelle, C.A. (2003). English language learning and technology. Amsterdam: John Benjamins.
Crozier, M. (1963). Le phénomène bureaucratique. Paris: Le Seuil.
Donato, R. (2000). Sociocultural contributions to understanding the foreign and second lan-
guage classroom. In J. Lantolf (Ed.), Sociocultural theory and second language learning (pp.
27–50). Oxford: Oxford University Press.
Dörnyei, Z. (2005). The psychology of the language learner: Individual differences in second lan-
guage acquisition. Mahwah NJ: Lawrence Erlbaum Associates.
Ehrenberg, A. (2010). La société du malaise. Paris: Odile Jacob.
Ellis, R. (2003a). Task-based language learning and teaching. Oxford: Oxford University Press.
Ellis, R. (2003b). Second language acquisition. Oxford: Oxford University Press.
Ellis, R. (2004). Individual differences in second language learning. In A. Davies & C. Elder
(Eds.), The handbook of applied linguistics (pp. 525–551). Oxford: Blackwell.
Fullan, M. (2007). The new meaning of educational change 4th edition. New York, NY: Teachers
College Press.
Goguelin, P. (1989). Le management psychologique des organisations. Paris: Editions Sociales
Françaises.
Hampel, R. (2006). Rethinking task design for the digital age: A framework for language teach-
ing and learning in a synchronous online environment. ReCALL, 18(1), 105–121.
Holec, H. (1981). Autonomy and foreign language learning. Oxford: Pergamon Press for the
Council of Europe.
Hurd, S. (2005). Autonomy and the distance language learner. In B. Holmberg, M. Shelley, & C.
White (Eds.), Distance education and languages: evolution and change (pp. 1–19). Clevedon:
Multilingual Matters.
Joffe, L.S. (2000). Getting connected: Online learning for the EFL professional. In Crossroads of
the New Millennium. Proceedings of the Technological Education and National Develop-
ment (TEND) Conference, (2nd, April 8 – 10, 2000, United Arab Emirates).
Kintsch, W. (2009). Learning and constructivism. In S. Tobias & T. M. Duffy (Eds.) Constructiv-
ist instruction: Success or failure? (pp. 223–241). London: Routledge.
Lamy, M.-N., & Goodfellow, R. (1999). Supporting language students’ interactions in Web-based
conferencing. Computer Assisted Language Learning, 12(5), 457–477.
 Julie McAllister, Marie-Françoise Narcy-Combes and Rebecca Starkey-Perret

Lantolf, J. (2000). Sociocultural theory and second language learning. Oxford: Oxford University
Press.
Little, D. (2007). Language learner autonomy: some fundamental considerations revisited. In-
novation in Language Learning and Teaching, 1(1), 14–29.
Narcy-Combes, J.-P. (2012). Tutoring at a distance, modeling as a tool to control chaos. In J.C.
Bertin & J.-P. Narcy-Combes (Eds), Computer-Assisted Language Learning 25(2), special
issue Tutoring at a Distance (pp. 111–127). London: Routledge.
Narcy-Combes, M.-F. (2008). Conflits de représentations et adaptation des dispositifs
d’enseignement/apprentissage. Les Cahiers de l’APLIUT, 27(1), 33–50.
Narcy-Combes, M.-F. & McAllister, J. (2011). Evaluation of a blended language learning envi-
ronment in a French university and its effects on second language acquisition. Asp, 59,
115–138.
Robinson, P. (2001). Cognition and second language instruction. Cambridge: Cambridge Univer-
sity Press.
Robinson, P. (2002). Individual differences and instructed language learning. Amsterdam: John
Benjamins.
Robinson, P. (2009). Task complexity, cognitive resources, and syllabus design: A triadic frame-
work for examining task influences on SLA. In K. Van den Branden, M. Bygate, & J.M.
Norris (Eds.), Task-based language teaching: A reader (pp. 193–236). Amsterdam: John
Benjamins.
Rogers, E.M. (2003). Diffusion of Innovations 5th edition. New York: Free Press.
Roots-Buck, J. (2005). Le scénario comme démarche d’apprentissage et mode d’évaluation. Univer-
sity of Nantes. Unpublished doctoral dissertation.
Samuda, V. (2009). Guiding relationships between form and meaning during task performance:
the role of the teacher. In K. Van den Branden, M. Bygate, & J.M. Norris (Eds.), Task-based
language teaching: A reader (pp. 379–400). Amsterdam: John Benjamins.
Schmidt, R. (Ed.) (1995). Attention and awareness in foreign language learning. Hawai’i: Univer-
sity of Hawai’i Press.
Skehan, P. (1991). Individual differences in second language learning. Studies in Second Lan-
guage Acquisition, 13(2), 275–298.
Soubrié, T. (2008). La difficile articulation du présentiel et de la distance dans le cadre d’un cours
hybride en master. Alsic, 11 (2), 105–127.
Swain, M. (2000). The output hypothesis and beyond: Mediating acquisition through collabora-
tive dialogue. In Lantolf, J. (Ed.), Sociocultural theory and second language learning
(pp. 97–114). Oxford: Oxford University Press.
Van den Branden, K. (Ed.) (2006). Task-based language education: From theory to practice. Cam-
bridge: Cambridge University Press.
Van den Branden, K. (2009). Training teachers: Task-based as well? In K. Van den Branden, M.
Bygate, & J.M. Norris (Eds.), Task-based language teaching: A reader (pp. 401–429).
Amsterdam: John Benjamins.
Van Patten, B. (1994). Evaluating the role of consciousness in SLA terms, linguistic features, and
research methodology. AILA Review, 11, 27–36.
Willis, D., & Willis, J. (2007) Doing task-based teaching. Oxford: Oxford University Press.
Willis, J. (2009). The TBL framework: The task cycle. In K. Van den Branden, M. Bygate, & J.M.
Norris (Eds.), Task-based language teaching: A reader (pp. 227–242). Amsterdam: John
Benjamins.
epilogue

What is next for task-based language

teaching?
 chapter 15

TBLT in EFL settings

Looking back and moving forward

David Carless
University of Hong Kong, Hong Kong

This final chapter builds on issues which have been discussed implicitly or
explicitly within the volume. I address five themes: research methodology;
contextual adaptations; TBLT in Chinese contexts; assessment; and teacher
education. I conclude by speculating on some possible future directions for
TBLT and some avenues for further research.

Introduction

The aim of this chapter is to identify some themes arising in this book, to include some
related commentary, and to chart some issues for further exploration. As the title sug-
gests, the chapter looks back at this volume and also at what might come next. The
chapters have already been admirably summarized in Shehadeh’s introduction, so one
of my aims is to indicate some common themes. My choice of issues is necessarily
something of a subjective enterprise, and I should first acknowledge that the realities
of my role as a teacher educator and researcher in the Faculty of Education, University
of Hong Kong bias me towards issues in Chinese (and wider East Asian) contexts and
those related to the implementation of TBLT in state school systems.
The contributors to this book all researched TBLT in EFL contexts, and bringing
together the chapters in a single volume is a highly valuable extension of the existing
literature. Part of the editors’ rationale for the volume is that it provides alternative
perspectives to the current knowledge base on TBLT which, at least in its first decades,
has been dominated by studies conducted in Anglophone settings. Given that context
is an issue I discuss later in the chapter, the geographical spread of the contributions is
worth some comment. The majority of the chapters come from Asian settings, a fitting
echo to the seminal early work on task-based teaching in the Bangalore Communica-
tional Teaching Project (Prabhu, 1987). The multiple contributions from Japan are also
a noteworthy feature. Europe is comparatively under-represented in the collection
with just three chapters from Turkey, Spain, and France. This balance of contributions
provokes some mild surprise in view of the solid body of work on TBLT emanating
 David Carless

from Western Europe. Examples include studies of task complexity (Kuiken & Vedder,
2007) from scholars based in Amsterdam; Ramon Ribé’s account of project work in
Catalonia (e.g., Ribé, 2000); and the edited collection compiled by Garcia Mayo (2007),
to name but a few.
Not surprisingly, the majority of the contributions focus on TBLT with young
adults, mainly in university settings. This is the dominant trend, presumably for the
pragmatic reason that many of us in the higher education sector are involved in
advancing the English language skills of students of this type and age-group. The lit-
erature on TBLT in relation to schooling remains comparatively modest. A notable
exception is Kris Van den Branden’s careful and insightful analysis of the teaching of
Dutch as a second language in Flanders (Van den Branden, 2006a). It is probably fair
to say that the implementation of TBLT is even more complex with school-age students
than adults, in view of challenges, such as large class sizes, classroom management,
limited resources, the needs of school examination systems, and the teacher factors of
attitudes, understanding, and capacities. Three noteworthy chapters in the collection
focus on the school sector: Chapter 2 by Sasayama and Izumi using data from Japanese
high school students; Chapter 9 by Chan focused on primary schooling in Hong Kong;
and Chapter 10 by Park based on data from a Korean secondary school.
In what follows, I discuss five themes which occur in the collection and on which
I feel qualified to comment. The first, research methodology, has already been high-
lighted by Shehadeh in Chapter 1. Secondly, context has been signaled as an issue
above in my discussion of the geographical spread in the collection and under this
theme I address the need for contextual adaptations to TBLT. Thirdly, I discuss prog-
ress in implementing TBLT in Chinese settings. The fourth theme of assessment is
addressed because it is often a powerful force impacting on what teachers and students
perceive as being important in the teaching and learning process. The fifth issue I ad-
dress is teacher education, a key issue for the development of TBLT in EFL settings.
Finally, I sketch some possible future developments.

Research methodologies used

The knowledge base on TBLT is obviously most effectively advanced by rigorous stud-
ies which are a good fit for the issues they are exploring. A distinctive feature of the
current collection is the range of carefully designed research methods in use. Many of
the chapters use quantitative methods, including: questionnaire surveys; various pro-
cedures involving the counting and classification of classroom interaction data or the
analysis of oral production; and the use of various statistical means to present data, for
example, Weaver’s use of multi-faceted Rasch analysis to explore student peer assess-
ment of a PowerPoint presentation task.
Qualitative approaches are also featured, but not as frequently as quantitative ones.
Chan (Chapter 9) analyzed 20 lessons in Hong Kong primary schools facilitated by
 Chapter 15.  TBLT in EFL settings 

multiple data sources: classroom observations; interviews with teachers; interviews

with students; and documentary analysis of artifacts, such as lesson plans, teaching
materials, and student work. Chacón (Chapter 11) studied issues related to a series of
film tasks through focus group interviews, reflective student diaries, and audio-­
recordings of student presentations.
Mixed method approaches are often thought to be a sensible way of combining
quantitative and qualitative approaches and thus exploiting assets of both. A difficulty
lies in a single researcher possessing approximately equivalent expertise in both ap-
proaches. Jackson (Chapter 12) negotiates these issues well in his analysis of a course
on a teacher education program in Japan. His chosen methods are written retrospec-
tive comments on student perceptions of gains from a task; analysis of classroom dis-
course; and a questionnaire gauging student attitudes towards TBLT in comparison
with a parallel control group.
This brief review of research methods used in the collection amply illustrates the
range of approaches deployed. The chapters demonstrate both rigor and carefully de-
signed research which meets the aims of a particular study. The detailed discussions of
research methods serve to distinguish this collection from an earlier one by Edwards
and Willis (2005) and a more recent one by Shehadeh and Coombe (2010), which
mainly place emphasis on valuable accounts of practice, with data collection being less
at the forefront. My own methodological bias is towards qualitative research involving
naturalistic classroom observations and associated interviews with participants. More
studies of this nature would, in my view, advance research on TBLT in EFL settings,
but it is heartening to see in this volume researchers delving into some of the com-
plexities of practice that are in need of empirical attention.

Contextual adaptations to TBLT

The need to identify teaching approaches which are grounded in local needs and val-
ues has been well-established over the last couple of decades. Influential within this
theme is research on the interplay between methodology and social context (Holliday,
1994), and the notion of context-sensitivity (Bax, 2003) in relation to communicative
language teaching (CLT). Resonating with this line of thinking, an important issue lies
in considering how TBLT might be adapted to suit the EFL settings in which it is being
implemented, or vice versa, the extent to which educational traditions may need to
change in order for effective language learning to occur. A possible repercussion is that
adaptations of TBLT may involve some form of merging the global with localized
methodologies (Littlewood, 2011). Implicit in such perspectives is the need for inclu-
sive non-doctrinaire approaches to TBLT. Within such an orientation, it would be
useful to identify key features central to all forms of TBLT, and explore further those
aspects amenable to contextual adaptation.
 David Carless

In a number of chapters, authors refer to contextual features of their setting at

macro and/or micro levels. Iwashita and Li (Chapter 7), for example, review some of
the challenges for the implementation of CLT or TBLT in China, including a view of
teaching predicated on a process of transmitting knowledge, information, or skill from
the teacher to the student. Park (Chapter 10) notes issues, such as heavy focus on
preparation for college entrance exams, rigid national curricula, and conventional
teaching methods, which act as challenges for the implementation of TBLT in the
Korean context.
A theme which arises in relation to Chinese settings is how TBLT can be adapted
to make it fit with the exigencies of the prevailing educational culture. Hu (2002, 2005),
for example, has written insightfully about the interplay between cultural norms in
Chinese settings, which prescribe well-established expected roles of classroom partici-
pants and their potentially different roles under CLT. Whilst caution has to be applied
to avoid cultural stereotyping, TBLT in many EFL settings requires changes to conven-
tional roles of the teacher and possibly the student.
For example, TBLT approaches the acquisition of grammatical form in a different
way to the more explicit teacher-fronted explanation practiced by many teachers. The
chapters in Section 1 of this volume all examine how different variables impact on
task-based language processing. This presages an aspect of contextual adaptations in
relation to how grammar is perceived and how it is normally treated in a specific locale.
This aspect may involve some move away from conventional presentation-­practice-
production (P-P-P) routines, containing limitations which have been well-articulated
by scholars (e.g., Ellis, 2003; Lewis, 1996), if not always accepted by classroom teach-
ers. However, there may still be potential for developing productive versions of P-P-P
(Carless, 2009), possibly through some form of reconciliation with task-supported
teaching (Ellis, 2003; Samuda & Bygate, 2008). This may include further exploration of
the complementary functions of analytic and experiential strategies, as highlighted by
Littlewood (2011).
The role of context in the implementation of TBLT is an issue I addressed in
Carless (2007) when, on the basis of interview data from teachers and teacher educa-
tors, I proposed three dimensions of what I referred to as situated task-based ap-
proaches. These comprised: the issue (noted above) of clarifying or enhancing the
role of explicit grammar instruction; integrating tasks with the needs of examinations
(a theme I return to later in the chapter); and a need for balance in the modes of task
interaction. For example, the latter could entail careful planning of balance between
reading and writing tasks, as opposed to oral ones. The current collection seems to
contain further indications that studies of TBLT tend to be dominated by a focus on
oral production, although a notable exception is Chapter 5 in which Horiba and
Fukaya focus on the processing of written texts, whilst in Chapter 4 Genc explores
both written and oral tasks.
 Chapter 15.  TBLT in EFL settings 

TBLT in Chinese contexts

Two of the chapters in the volume are set within Chinese settings. Within this theme,
I make reference to developments in Hong Kong over the last two decades and more
recent ones in the People’s Republic of China.

Hong Kong

The story of TBLT in Hong Kong is set against the backdrop of a predominantly con-
ventional teaching culture in which grammar has generally been taught through
explicit explanation and controlled practice (Andrews, 2007). My own work has par-
ticularly focused on the implementation of TBLT in Hong Kong schools, and I address
selected issues within this theme, in the belief that they carry messages for other con-
texts. A task-based curriculum was proposed for implementation in primary schools
from the mid-1990s onwards and in secondary schools from 1999. In an analysis of the
Hong Kong guidelines for TBLT, Candlin (2001) put forward the case that the docu-
mentation of the educational authorities in Hong Kong was exemplary. The reality at
the chalk-face revealed, however, different issues to the more idealized picture pre-
sented in curriculum guidelines.
My doctoral research, conducted in the late 1990s, involved detailed case studies
of teachers’ attempting to implement some version of TBLT in primary schools. The
ensuing publications (Carless, 2002, 2004) highlighted three main challenges in the
classrooms under discussion. These features were student use (or overuse) of the
mother tongue; tensions between teacher desire for an orderly, well-disciplined envi-
ronment and a ‘noisier’ activity-based classroom; and concerns about the quality and
quantity of student English language production. Chan (this volume), also basing her
study in Hong Kong primary schools, discusses pedagogic issues which illustrate some
of the complexities of teacher enactment of tasks with young learners. She includes a
detailed analysis of aspects, such as strategic use of visual input, scaffolding, and creat-
ing conditions for noticing salient features of form.
TBLT was introduced in Hong Kong secondary schools from 1999 onwards. Sec-
ondary school case studies (Adamson & Tong, 2008) indicated that the version of
TBLT being implemented in schools was less ‘strong’ (cf. Skehan, 1996) than that
envisaged by the official government documentation and guidelines. In current man-
ifestations of the Hong Kong school curriculum, TBLT is embedded within a “New
Senior Secondary Curriculum” which involves changes to the structure and content
of the curriculum. The structural element involves a move from a four-year to a three-
year course of study with students entering university one year earlier than previ-
ously. In terms of content, TBLT is supposed to be embedded within a core focused
mainly on grammar, communicative functions, vocabulary and text-types; and elec-
tives, such as learning English through drama, short stories, popular culture, or social
issues. A further feature is school-based assessment (SBA), explored in more detail
 David Carless

later in the chapter, whereby teacher marks awarded for work done in schools count
towards the overall grade in the key high-stakes examination. Part of the rationale for
SBA rests on an assumption that reforming assessment is an effective way of stimulat-
ing pedagogic change. The interplay between TBLT and SBA under the new Hong
Kong curriculum at senior secondary level is an issue for ongoing attention, with the
possibility that the newer examination-focused SBA will attract more resources and
attention than TBLT.

China

In the last few years, China has designed and promoted a so-called New Curriculum
Standards (Wang, 2007) in which TBLT forms a part of an agenda for change across
the school curriculum. In their contribution to this volume, Iwashita and Li investigate
patterns of interaction in a task-based oral English class at the university level in China.
A useful addition to our understanding of TBLT in China would be further studies in
the school and university sectors.
Recent research on TBLT in China reinforces some of the challenges of introduc-
ing TBLT in a mass school system. Zhang (2007), for example, found that whilst teach-
ers claimed to be carrying out tasks, what was going on in the classrooms she re-
searched did not contain features of what one would normally consider as TBLT. More
recent case studies conducted by Deng (2011) on the implementation of TBLT in pri-
mary schools in Guangdong province used Littlewood’s (2004) communicative con-
tinuum to gauge the extent of task-based activities in the classrooms of four teachers
in two contrasting schools. A significant (although hardly surprising) finding was that
in the lessons observed, there were more activities which focused on forms rather than
on meaning. Such activities were perceived by teacher participants as easier to manage
and more effective in preparing students for examinations (Deng & Carless, 2009).
One of the case study teachers (pseudonym, Jane) was, however, found to possess con-
fidence in the effectiveness of communicative language teaching and her lessons evi-
denced more focus on meaning than in the lessons of the other three teachers. Jane
seems to share some similar characteristics with Debbie (Iwashita & Li, this volume)
who was able to promote an interactive classroom, despite contextual limitations such
as a large class size. A facilitating factor, common to both the cases of Jane and Debbie,
is a belief in and well-developed understanding of the potentials of communicative
interaction for both language development and student motivation.
Examinations are often seen as a barrier to the implementation in schools of such
as communicative or task-based approaches (Deng & Carless, 2010; Littlewood, 2007).
This is clearly the case in China (Iwashita & Li, this volume), although the issues are
complex and interwoven with contextual factors, such as those related to teachers’
values. An analysis of changes to the writing section of the NMET (the National
Matriculation English Test) used for university entrance indicated that teachers may
utilise their own preferred approaches, whilst side-stepping the more communicative
 Chapter 15.  TBLT in EFL settings 

aspirations of the test developers (Qi, 2007). It seems that teachers’ values and beliefs
about pedagogy are even more influential than the examinations, and there is consid-
erable variation in how teachers respond to the tension between a communicative
curriculum and a more traditional examination-system (Deng & Carless, 2010). The
relationship between test preparation and communicative language teaching remains
an important issue for further research. This leads me to consider, in the next sub-
section, assessment in relation to TBLT.

Assessment and TBLT

Only one chapter in the collection focuses explicitly on the important dimension of
assessment in TBLT (Weaver, Chapter 13). We know that in formal mainstream edu-
cational situations where certification is at stake, assessment is what most powerfully
captures the minds, or at least the study behaviors, of students. This is largely the case
everywhere, but particularly so in China given its long history of examinations, dating
back to the Han dynasty. This history of large-scale public testing in China is noted by
Iwashita and Li (Chapter 7) as being a contextual barrier to the implementation of
communicative and task-based approaches.
Weaver (Chapter 13), researching Japanese business students, provides an account
of the incorporation of a formative assessment cycle into a TBLT curriculum in rela-
tion to a task with a strong sense of authenticity: a PowerPoint presentation aiming to
persuade peers to invest in a particular company. A positive aspect of the study in-
volves student engagement with exemplars of performance in association with the
application of explicit assessment criteria. Student engagement with exemplars and
criteria is an important step in developing self-evaluative capacities which lie at the
heart of formative assessment.
An issue bubbling under the surface of the chapter, but not addressed in much
detail, is the extent to which the processes described by Weaver actually acted forma-
tively. In other words, to what extent did the cycle support the development of wider
student learning capacities or more narrowly help them to improve their PowerPoint
presentations? As the seminal work of Black and Wiliam (1998, 2003) has argued, an
assessment process can only be said to have acted formatively if it advances student
learning. It may be that a way forward for formative assessment in relation to TBLT is
to engage more with the expanding educational literature on formative assessment
(e.g., Andrade & Cizek, 2010) and developments in dynamic assessment (e.g., Poeh-
ner, 2008). Indeed, along these lines and within the TBLT literature, Norris (2009) has
highlighted the formative as well as summative and other uses for task-based language
assessment in his overview of task-based teaching and testing, and a handful of ex-
amples (e.g., Byrnes, 2002; Byrnes, Maxim, & Norris, 2010) point to the potential of
task-based assessment as a core element of curriculum renewal initiatives.
 David Carless

A lesson from innovation theory (e.g., Barnes, Clarke, & Stephens. 2000) is that
attempts to reform pedagogy often have limited impact, unless there is also change to
how students are assessed. Again I wish to return to developments in Hong Kong. To
stimulate the implementation of CLT and TBLT, high-stakes examinations have over
the last twenty years increased the weighting awarded to oral performance, and the
examinations have become increasingly task-based. The high-profile introduction of
SBA involves students carrying out oral tasks within the school which are graded by
the teacher and count for 15% of the English mark in the high-stakes examination at
the end of secondary schooling (Davison, 2007; Davison & Hamp-Lyons, 2010). The
oral tasks are either group discussions or oral presentations, in both cases responses to
a print or non-print ‘text’, such as a book or movie.
The use of teacher assessment in a traditionally examination-oriented system is
both a powerful impetus for change and a challenge to teachers’ workloads and mind-
sets. In such a setting, there is a likelihood that test-takers (often with the assistance of
the ubiquitous after-hours tutorial sector) may seek to subvert the aims of the test de-
velopers. A careful analysis of the discourse of test interaction (Luk, 2010) shows how
students colluded in producing utterances aimed at creating the impression of being
effective interlocutors for the purpose of scoring marks rather than for authentic com-
munication. In contrast, Gan, Davison, and Hamp-Lyons (2009) analyzed the dis-
course derived from a task which in their particular case was based on choosing a gift
for the main character in the movie Forrest Gump. They found that peer group discus-
sion as an oral assessment format has the potential to provide opportunities for stu-
dents to demonstrate ‘real-life’ spoken interactional abilities. Discourse analysis of oral
production in SBA tasks looks like an emerging research interest in Hong Kong.
Some form of synergy, or at least peaceful co-existence, between the needs of as-
sessment in specific locales and communicative pedagogy is an important ongoing
dilemma for TBLT in EFL settings. Some possible ways forward relevant to task-based
assessment include: further development of assessment literacy amongst stakeholders,
including Education Ministry officials, school and university managers, teachers, and
teacher educators; greater alignment of high-stakes assessment with curriculum and
pedagogy; and a focus on productive student learning to be made explicit in all forms
of assessment, including high-stakes examinations (Carless, 2011).

Teacher education and TBLT

For teachers to be able to develop the full potential from TBLT they are likely to need
a variety of opportunities to learn about and engage with communicative and task-
based teaching. Two chapters in this volume are indicative of some of the potentials
and challenges for task-based teacher education. In the study by Jackson (Chapter 12),
Japanese student teachers expressed positive responses to TBLT concepts, although
the extent to which they would use these notions in their future teaching was beyond
 Chapter 15.  TBLT in EFL settings 

the scope of the study. Chacón’s contribution (Chapter 11), set in Venezuela, indicated
positive participant perceptions of working in pairs or groups. The Venezuelan trainee
teachers in the study reported that they would be willing to implement TBLT with
their own students in the future. Two key issues arise from these cases, both of them in
relation to support in the school context. First, to what extent would the prevailing
school cultures in the two settings encourage the adoption of task-based activities?
Second, what kind of quality and depth of support would novice teachers receive in
implementing TBLT?
A related theme in many EFL contexts is the broader background to innovation,
and in particular the difficulty of providing sufficiently in-depth teacher education to
illuminate the key issues in implementing TBLT successfully. This challenge is exacer-
bated, for example, in a setting as vast and widely dispersed as China. A further issue
for China is the disparity between the mainly advanced metropolitan and coastal areas
versus the rural hinterland (Hu, 2003; Iwashita & Li, this volume). In such situations,
mass centralized short-term training programs are unlikely to do more than provide a
basic introduction to some of the issues in TBLT. These limitations can be compound-
ed by the perceived complexity of TBLT, an issue I have discussed elsewhere (Carless,
2009). The mode of teacher training also merits further consideration. The argument
that TBLT teacher education should itself be experiential and task-based is a good one
(see Van den Branden, 2006b), but this might be difficult to achieve on the scale re-
quired by China in view of the enormous numbers of teachers and sites involved. That
is not to say that smaller-scale pockets of exploratory developmental work could not
be attempted. For example, the ‘lesson study’ or ‘research lesson’ is a professional de-
velopment method practiced extensively in Japan (see, for example, Fernandez, 2002)
through which a team of colleagues analyze a taught lesson and seek ways to improve
it. This could be a contextually grounded source of professional development in col-
lectivist East Asian societies, such as China (Carless, 2011).

Future developments in TBLT

The literature on educational change indicates that many innovations last only a rela-
tively short time because they are replaced by other innovations (Waks, 2007); condi-
tions are not receptive to the proposed change (Datnow, 2002); or they are reinvented
and rebranded (Snyder, Acker-Hocevar, & Snyder, 2008). Reinvention can have posi-
tive connotations when it encourages customization of the innovation to fit it more
appropriately to local conditions (Rogers, 2003), or more negative ones when it leads
to dilution of its underlying principles.
TBLT seems to have stood the test of time, and as Shehadeh indicates in the intro-
duction, it possesses many markers of a well-established field of study. Its vitality, rich-
ly illustrated in this volume, is indicative of enduring interest in different facets of
TBLT. In terms of future research agendas, a particularly useful and comprehensive
 David Carless

discussion has already been compiled by Samuda and Bygate (2008) in the final chap-
ter of their book. Here I comment on selected issues which resonate with the discus-
sion in this chapter.
Although it is notoriously problematic to try to demonstrate the superiority of one
teaching approach over another, it might be worth attempting a quasi-experimental
research design in which experimental groups taught through TBLT are compared
with control groups taught by the existing approaches being used in that setting. In a
discussion of the well-known Pennsylvania project, Lynch (1996) discusses some of
the myriad challenges of this kind of research: the optimum number and characteris-
tics of teacher participants; unwanted variability in teacher implementation of the ap-
proach they are supposed to be using; and developing achievement tests that do not
favor one group or another. Nonetheless, studies of this kind could help to explore dif-
ferences both in process and outcome, and with due sensitivity to local context could
yield valuable information. This is a research challenge awaiting a team brave enough
to take it on, although clearly the lessons learned from evaluations of the Bangalore
Project (e.g., Beretta, 1992) could be utilized as a starting point for such an endeavor.
The acquisition of grammatical form within meaning-focused activities is at the
heart of TBLT. A classic example of the interweaving of form within meaning inten-
tions is Samuda (2001). As Samuda notes, the role of the teacher in mediating task-
based language development is well-worth further exploration (see also Van Avermaet,
Colpin, Van Gorp, Bogaert, & Van den Branden, 2006). I have already alluded to the
chapters in Section 1 which investigated a number of variables impacting on task-
based performance. Although Moore (Chapter 8) touches on awareness-raising
activities, the current volume does not include a detailed discussion of the role of con-
sciousness-raising (CR) tasks as a productive way of encouraging a focus on language
form. This line of research within a TBLT framework does not seem to have advanced
as much as one might have anticipated. CR could represent the task itself or it could be
an option to encourage language analysis in the post-task stage. The post-task is a
critical stage of the task cycle and seems particularly under-researched within school
settings.
Students’ perceptions of tasks are a further area which merits future inquiry. A
general question arises: to what extent do students in different geographical settings
appreciate communicative tasks as much as some proponents of TBLT appear to do?
For example, a study of South African secondary school students (Barkhuizen, 1998)
asked respondents to evaluate a list of fifteen language learning activities; a surprising
finding for the teachers concerned was that students preferred mechanical written ac-
tivities rather than oral communicative tasks. A sample of students of different profi-
ciency levels in a private Brazilian language school, surveyed by Garrett and Shortall
(2002), indicated that students perceived peer work in groups to be enjoyable but that
teacher-fronted work was thought to be more likely to enhance their learning. This
study also reinforces the value of attending to the voice of the learner. More on what
university and/or school students think about TBLT and their perceptions of tasks
 Chapter 15.  TBLT in EFL settings 

would be valuable extensions of the existing literature. Particularly valuable would be

longitudinal studies, as there is some evidence (e.g. McDonough & Chaikitmongkol,
2007) that students become more favorably disposed towards TBLT, after they have
some time to engage with its practice.
Both Chacón (Chapter 11) and the Hong Kong SBA tasks use film as a stimulus for
task-based teaching. Notable features of Chacón’s study include: co-operative work in
groups of four; peer and self-assessment; and careful design of a post-task (here a Film
Guide containing some of the features often found in reviews of movies). Film as a
stimulus for tasks could be an area for further research. The use of audio-visual input
could be extended in various ways to build on new literacies and the development of
technologies associated with Web 2.0. As a final pointer to the future, a new generation
of tasks may well be inspired by analysis of discourse from YouTube or other digital
media in line with changes in authentic language use.
In conclusion, this volume has made useful steps in developing the knowledge
base on TBLT in EFL settings. The points I have made suggest the need for the follow-
ing: more reports on the implementation of TBLT from different EFL settings; detailed
qualitative accounts of what is really taking place in classrooms in which the teacher is
trying to implement some version of TBLT; further research on contextual adaptations
to TBLT; continued scrutiny of the interface between assessment and TBLT; and a
search for appropriate forms of teacher education and support for the implementation
of TBLT.

References

Adamson, B., & Tong, A.S.Y. (2008). Leadership and collaboration in implementing curriculum
change in Hong Kong secondary schools. Asia Pacific Education Review, 9(2), 180–189.
Andrade, H., & Cizek, G. (2010). (Eds.), Handbook of formative assessment. New York, NY:
Routledge.
Andrews, S. (2007). Teacher language awareness. Cambridge: Cambridge University Press.
Barkhuizen, G. (1998). Discovering learners’ perceptions of ESL classroom teaching/learning
activities in a South African context. TESOL Quarterly, 32, 85–108.
Barnes, M., Clarke, D., & Stephens, M. (2000). Assessment: The engine of systemic curricular
reform? Journal of Curriculum Studies, 32(5), 623–650.
Bax, S. (2003). The end of CLT: A context approach to language learning. ELT Journal, 57,
278–287.
Beretta, A. (1992). What can be learned from the Bangalore evaluation? In J.C. Alderson & A.
Beretta (Eds.), Evaluating second language education (pp. 250–273). Cambridge: Cambridge
University Press.
Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education,
5(1), 7–74.
Black, P., & Wiliam, D. (2003). In praise of educational research: Formative assessment. British
Educational Research Journal, 29 (5), 623–637.
 David Carless

Byrnes, H. (2002). The role of task and task-based assessment in a content-oriented collegiate
foreign language curriculum. Language Testing, 19(4), 419–437.
Byrnes, H., Maxim, H., & Norris, J. M. (2010). Realizing advanced FL writing development in
collegiate education: Curricular design, pedagogy, assessment. Modern Language Journal,
Monograph. Cambridge: Blackwell.
Candlin, C. N. (2001). Afterword: Taking the curriculum to task. In M. Bygate, P. Skehan & M.
Swain (Eds.), Researching pedagogic tasks: Second language learning, teaching and testing
(pp. 229–243). Harlow, UK: Longman.
Carless, D. (2002). Implementing task-based learning with young learners. English Language
Teaching Journal, 56(4), 389–396.
Carless, D. (2004). Issues in teachers’ re-interpretation of a task-based innovation in primary
schools. TESOL Quarterly, 38(4), 639–662.
Carless, D. (2007). The suitability of task-based approaches for secondary schools: Perspectives
from Hong Kong. System, 35(4), 595–608.
Carless, D. (2009). Revisiting the TBLT versus P-P-P debate: Voices from Hong Kong. Asian
Journal of English Language teaching, 19, 49–66.
Carless, D. (2011). From Testing to Productive Student Learning: Implementing Formative Assess-
ment in Confucian-heritage Settings. New York, NY: Routledge.
Datnow, A. (2002). Can we transplant educational reform, and does it last? Journal of educa-
tional change, 3(3–4), 215–239.
Davison, C. (2007). Views from the chalkface: English language school-based assessment in
Hong Kong. Language Assessment Quarterly, 4 (1), 37–68.
Davison, C., & Hamp-Lyons, L. (2010). The Hong Kong certificate of education: School-based
assessment reform in Hong Kong English language education. In L.Y. Cheng & A. Curtis
(Eds.), English Language Assessment and the Chinese Learner (pp. 248–266). New York, NY:
Routledge.
Deng, C.R. (2011). Communicativeness of activities in EFL primary school classrooms in Guang-
dong, China: Teachers’ interpretations of task-based language teaching. Unpublished doc-
toral dissertation, University of Hong Kong.
Deng, C.R., & Carless, D. (2009). The Communicativeness of Activities in a Task-based Innova-
tion in Guangdong, China. Asian Journal of English Language teaching, 19, 113–134.
Deng, C.R., & Carless, D. (2010). Examination Preparation or Effective Teaching: Conflicting
Priorities in the Implementation of a Pedagogic Innovation. Language Assessment Quar-
terly, 7(4), 285–302.
Edwards, C., & Willis, J. (Eds.). (2005). Teachers exploring tasks in English language teaching.
Basingstoke, UK: Palgrave Macmillan.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.
Fernandez, C. (2002). Learning from Japanese approaches to professional development: The
case of lesson study. Journal of Teacher Education, 53(5), 393–405.
Gan, Z., Davison, C., & Hamp-Lyons, L. (2009). Topic negotiation in peer group oral assessment
situations: A conversation analytic approach. Applied Linguistics, 30(3), 315–334.
Garcia Mayo, M.P. (Ed.) (2007). Investigating tasks in formal language learning. Clevedon, UK:
Multilingual Matters.
Garrett, P., & Shortall, T. (2002). Learners’ evaluations of teacher-fronted and student-centred
classroom activities. Language Teaching Research, 6(1), 25–57.
Holliday, A. (1994). Appropriate methodology and social context. Cambridge: Cambridge Uni-
versity Press.
 Chapter 15.  TBLT in EFL settings 

Hu, G. W. (2002). Potential cultural resistance to pedagogical imports: The case of communica-
tive language teaching in China. Language, culture and curriculum, 15, 93–105.
Hu, G.W. (2003). English language teaching in China: Regional differences and contributing
factors. Journal of Multilingual and Multicultural Development, 24(4), 290–318.
Hu, G. W. (2005). Contextual influences on instructional practices: A Chinese case for an eco-
logical approach to ELT. TESOL Quarterly, 39, 635–660.
Kuiken, F., & Vedder, I. (2007) Task complexity and measures of linguistic performance in L2 writ-
ing. International Review of Applied Linguistics in Language Teaching (IRAL), 45, 261–284.
Lewis, M. (1996). Implications of a lexical view of language. In J. Willis & D. Willis (Eds.), Chal-
lenge and change in language teaching (pp. 10–16). Oxford: Heinemann.
Littlewood, W. (2004). The task-based approach: Some questions and suggestions. English Lan-
guage Teaching Journal, 58(4), 319–326.
Littlewood, W. (2007). Communicative and task-based language teaching in East Asian class-
rooms. Language Teaching 40(3), 241–249.
Littlewood, W. (2011). Communicative language teaching: An expanding concept for a chang-
ing world. In E. Hinkel (Ed.), Handbook of research in second language teaching and learning
(pp. 541–557). New York, UK: Routledge.
Luk, J. (2010). Talking to score: Impression management in L2 Oral Assessment and the Co-
construction of a Test Discourse Genre. Language Assessment Quarterly, 7(1), 25–53.
Lynch, B. (1996). Language program evaluation: Theory and practice. Cambridge: Cambridge
University Press.
McDonough, K., & Chaikitmongkol, W. (2007). Teachers' and learners' reactions to a task-based
EFL course in Thailand. TESOL Quarterly, 41(1), 107–132.
Norris, J. M. (2009). Task-based teaching and testing. In M. Long & C. Doughty (Eds.), Hand-
book of language teaching (pp. 578–594). Cambridge: Blackwell.
Poehner, M. (2008). Dynamic Assessment: A Vygotskian Approach to Understanding and Promot-
ing L2 Development. Berlin: Springer.
Prabhu, N. (1987). Second language pedagogy. Oxford: Oxford University Press.
Qi, L. (2007). Is testing an efficient agent for pedagogical change? Examining the intended wash-
back of the writing task in a high-stakes English test in China. Assessment in Education,
14(1), 51–74.
Ribé, R. (2000). Introducing negotiation processes: An experiment with creative project work.
In M. Breen & A. Littlejohn (Eds.), Classroom decision-making (pp. 63–82). Cambridge:
Cambridge University Press.
Rogers, E. (2003). The diffusion of innovations (5th edition). New York, NY: Free Press.
Samuda, V. (2001). Guiding relationships between form and meaning during task performance:
The role of the teacher. In M. Bygate, P. Skehan, & M. Swain (Eds.), Researching pedagogic
tasks: Second language learning, teaching and testing (pp. 119–140). Harlow, UK: Longman.
Samuda, V., & Bygate, M. (2008). Tasks in second language learning. Basingstoke, UK: Palgrave,
MacMillan.
Shehadeh, A., & Coombe, C. (2010) Applications of task-based learning in TESOL. Alexandria,
VA: TESOL Inc.
Skehan, P. (1996). A framework for the implementation of task-based instruction. Applied Lin-
guistics, 17, 38–62.
Snyder, K. J., Acker-Hocevar, M., & Snyder, K. M. (2008). Living on the edge of chaos: Leading
schools into the global age. Milwaukee, WI: ASQ Quality Press.
 David Carless

Van Avermaet, P., Colpin, M., Van Gorp, K., Bogaert, N., & Van den Branden, K. (2006). The role
of the teacher in task-based language teaching. In K. Van den Branden, (Ed.), Task-based
language education: From theory to practice. (pp. 175–196). Cambridge: Cambridge Univer-
sity Press.
Van den Branden, K. (Ed.). (2006a). Task-based language education: From theory to practice.
Cambridge: Cambridge University Press.
Van den Branden, K. (2006b). Training teachers: Task-based as well? In K. Van den Branden,
(Ed.), Task-based language education: From theory to practice. (pp. 217–248). Cambridge:
Cambridge University Press.
Waks, L. J. (2007). The concept of fundamental educational change. Educational Theory, 57(3),
277–295.
Wang, Q. (2007). The National Curriculum changes and their effects on English language teach-
ing in the People’s Republic of China. In J. Cummins & C. Davison (Eds.), International
handbook of English language teaching (pp. 87–105). Boston, MA: Springer Science &
Business Media.
Zhang, Y.F.E. (2007). TBLT-Innovation in Primary school English language teaching in main-
land China. In K. Van den Branden, K. Van Gorp, & M. Verhelst (Eds.), Tasks in action:
Task-based language education from a classroom-based perspective (p. 68–91). Newcastle,
UK: Cambridge Scholars Press.
 About the contributors

David Carless is Associate Professor and Head of the Division of English Language
Education, Faculty of Education, University of Hong Kong. His main research interest
is in how assessment can be reconfigured to stimulate productive student learning. His
latest book is entitled: From testing to productive student learning: implementing forma-
tive assessment in Confucian-heritage settings published by Routledge in 2011.
Carmen Chacón is a retired English professor of the Department of Modern Lan-
guages in the Faculty of Education at the University of Los Andes, Venezuela. She
holds a PhD from The Ohio State University and worked as an EFL teacher and teach-
er educator for more than twenty five years. She has published several ELT articles and
book chapters in Venezuela and in the USA. Her research interests include non-native
English speaking teachers’ (NNESTs’) issues, TESOL pedagogy, and teacher efficacy.
Shirley Chan is the Director of Education and School Services of the Whole Person
Development Institute (WPDI). Before joining WPDI, she was the Assistant Professor
of the Hong Kong Institute of Education. Her research interests are English language
teaching and learning for young learners, language assessment, curriculum change
and innovation and teacher education. Currently, she works with a team of psycholo-
gists and psychiatrists to promote mind-wellness of the general public and school chil-
dren by integrating education, psychology and psychiatry.
Christine Coombe has a PhD in Foreign/Second Language Education from The Ohio
State University. Christine is co-editor/author of numerous publications including:
Language Teacher Research in the Middle East (2007, TESOL Publications), Leadership
in English Language Teaching and Learning (2008, UMP) Applications of Task-based
Learning in TESOL (2010, TESOL Publications), The Cambridge Guide to Second Lan-
guage Assessment (2012, Cambridge University Press) and Reigniting, Retooling and
Retiring in English Language Teaching (2012, University of Michigan Press). She was
TESOL Convention Chair for Tampa 2006. She is currently TESOL President
(2011–2012).
Keiko Fukaya is a researcher at the Institute for Research in Language and Culture,
Tsuda College. She worked as associate professor of English at St. Luke’s College of
Nursing. She holds an MA in TESOL from College of New Rochelle, NY. Her research
interests include second language reading, vocabulary acquisition and English for
specific purposes. Her publications include a chapter on extensive reading in an
edited book.
 Task-Based Language Teaching in Foreign Language Contexts: Research and implementation

Zubeyde Sinem Genc is Associate Professor in the Faculty of Education, Uludag

University in Bursa, Turkey. She received her PhD from the Department of English,
Indiana University of Pennsylvania (IUP), USA, specializing in TESOL and SLA. She
taught EAP at IUP, and both graduate and undergraduate courses at Southern Illinois
University. Her research interests include second language teacher education, SLA,
task-based language teaching, and curriculum development.
James Hobbs is from Britain and is an Associate Professor in the Department of
Foreign Languages at Iwate Medical University, Japan. He has been living and teaching
in northern Japan since 1991 and has an MSc in TESOL from Aston University. He has
taught in a wide range of contexts and his research interests include language teaching
methodology, medical English education, the analysis of task-generated discourse,
and Japanese-English translation.
Yukie Horiba is Professor of Language Education/Applied Linguistics at the Graduate
School of Language Sciences, Kanda University of International Studies, Japan. She
holds a PhD in second languages/cultures education from University of Minnesota.
Her research interests include second language text comprehension, vocabulary acqui-
sition, reading-writing connection, language testing, instructional task and teacher
development. Her work has appeared in Discourse Processes, Language Learning,
Modern Language Journal, and Studies in Second Language Acquisition, and some
edited books.
Noriko Iwashita is a senior lecturer in Applied Linguistics at The University of
Queensland. Her research interests include the role of classroom interactions (teacher-
students and student-student) and tasks in SLA, the interfaces of language assessment
and SLA, task-based assessment, and cross-linguistic investigation of four major lan-
guage traits. Her work has appeared in Language Learning, Language Testing, Applied
Linguistics, and Studies in Second Language Acquisition.
Shinichi Izumi is a professor in the faculty of foreign studies at Sophia University. He
earned his Ph.D. in Applied Linguistics from Georgetown University, USA. His prin-
cipal research interest is in second language acquisition and English teaching. His
previous research appears in his books on focus on form and CLIL and in many inter-
national journals. He currently serves on the editorial (advisory) boards for several
international journals including Studies in Second Language Acquisition.
Daniel O. Jackson (M.S. in Education/TESOL, University of Pennsylvania) is a PhD
student in the Department of Second Language Studies at the University of Hawai‘i at
Mānoa. His primary research interests are in task-based instruction and second lan-
guage acquisition and he currently serves as managing editor for Language Learning &
Technology. Before returning to the US to complete his studies, he taught and admin-
istered university EFL courses in Japan for over eight years.
 About the contributors 

Mayya Levkina received a Bachelor degree in French and Spanish Linguistics from the
Department of French Language from the Moscow State Linguistic University with
cum laude. She then received an M.A. in Applied Linguistics from the University of
Barcelona. At present she is carrying out her PhD project with the focus on task se-
quencing and individual differences (e.g. Working Memory Capacity and Attention)
in L2 development and L2 acquisition.
Huifang (Lydia) Li obtained her PhD degree in the field of Applied Linguistics from
The University of Queensland, Australia in 2010. She has lectured at Anyang Normal
University, Henan Province in the People’s Republic of China, The University of
Queensland and Griffith University, Queensland Australia. Her particular field of in-
terest is in facilitating roles of tasks and conversational interaction in second language
learning and teaching.
Aleksandra Malicka is a PhD candidate at the University of Barcelona, Spain. She
obtained her Bachelor’s degree in Ethnolinguistics from Adam Mickiewicz University
in Poznan, Poland. She then completed a Master’s degree in Applied Linguistics and
Language Acquisition at the University of Barcelona. She is currently working on her
PhD dissertation on task complexity, task sequencing and language performance. Her
research interests include task-based language teaching, bilingual teaching, and lan-
guage policy and planning.
Julie McAllister is a PhD student in language teaching methodology at the University
of Nantes under the supervision of Professor M.-F. Narcy-Combes. Her research
focuses on the evaluation of learning environments in a context where ICT plays an
increasing role. She has been teaching business English and Marketing in higher edu-
cation since 2003. Prior to this, she held several management positions in multina-
tional companies within the IT and Telecoms sector.
Paul Moore is a lecturer (language specialist) in Learning Development at the Univer-
sity of Wollongong. His current research interests involve several aspects of task-based
language learning and teaching, including influences of interaction on performance
and development, form-focused instruction, and use of the first language in the EFL
classroom. He is also researching intercultural team work in higher education, and has
been involved in consultations regarding the implementation of Japanese-as-a heri-
tage-language courses in high school curricula.
Marie-Françoise Narcy-Combes is a full professor at the University of Nantes where
she is involved in the coordination of the Applied Languages Department. Her main
teaching fields include business English within this department as well as pre-service
and in-service language teacher training. She originated and conducted the imple-
mentation of the Task-Based Learning and Teaching (TBLT) programme at the
University of Nantes. Her research fields concern Second Language Acquisition and
Language Teaching Methodology.
 Task-Based Language Teaching in Foreign Language Contexts: Research and implementation

Moonyoung Park is an ESL lecturer, and research assistant while working towards a
PhD in Applied Linguistics and Technology at Iowa State University. He has taught
secondary and college level EFL/ESL courses in Korea, Thailand, and the US. His re-
search interests include applications of TBLT and technologies in language learning
and language testing. He is currently researching task-based aviation English training
and testing and the use of automatic writing evaluation software in college ESL writing
courses.
Late Teresa Pica was Professor and TESOL program me director at the University of
Pennsylvania. She was an academic advisor to universities in Korea, Spain, India, and
Japan. She served on the editorial boards of many of the leading journals of language
study, and had held guest lectureships at the US Department of Defense Language Insti-
tute; Temple University, Japan; the University of Sydney; the Federal University of Cuiaba,
Brazil, the University of the Basque Country, and the TESOL Summer Institute.
Shoko Sasayama is currently a PhD student in the Department of Second Language
Studies at the University of Hawaii at Manoa. Her primary academic interests include
second language acquisition and language pedagogy. Currently, her research is focused
on how cognition influences language processing, performance, and learning. Peda-
gogically, she advocates the use of tasks in language classrooms for the development of
learners’ communicative as well as linguistic competence.
Ali Shehadeh is Associate Professor in and past chair of Department of Linguistics at
the UAE University. His research papers have appeared in Language Learning, TESOL
Quarterly, System, Journal of Applied Linguistics, and ELT Journal. He has served on the
Editorial Boards of a number of international journals. Currently he is the co-editor of
Brief Reports and Summaries of TESOL Quarterly and co-editor of Asian Journal of
English Language Teaching. His research interests are SLA, task-based language learn-
ing and teaching, and L2 writing research and pedagogy.
Rebecca Starkey-Perret is a research designer for the task-based blended language
learning programme for Business English students at the University of Nantes and a
PhD student in Second Language Acquisition and Classroom Oriented Research, su-
pervised by Professor Marie-Françoise Narcy-Combes. Her professional focus is on
designing materials for the programme as well as providing technical and pedagogical
support for the teachers involved. Her research focuses on language teachers’ repre-
sentations in French secondary schools.
Christopher Weaver belongs to the Department of Business Administration at Toyo
University, Tokyo, Japan. His research interests involve inquiries into pre-task plan-
ning and applications of the Rasch measurement model in task-based assessment.
 Index
A E
Affective variable question-
naire  50, 51, 55, 58, 66
AS units  157
Attentional capacity  24, 38, 59
C Estimation Time Judgment
Carless, David  3, 6, 7, 13,
15–17, 189, 208, 272, 284, 345,
348–353, 356, 359
Chacón, Carmen Teresa  6, 7,
13–15, 17, 241, 244, 257, 271,
281, 282, 347, 353, 355, 359
Chan, Sui Ping (Shirley)  5, 7, 12,
13, 187, 189, 208, 346, 349, 359
China  4–7, 11, 137–139, 348–351,
356–358, 361
CLAN mode of CHILDES  52
Clarification request  146
Code complexity  25, 44, 190
Cognition hypothesis;
Robinson’s Cognition
Hypothesis  9, 28, 29, 38, 40,
43–45, 59
Cognitive complexity  9, 25, 40,
47–51, 59, 61, 190, 191, 201
Communicative language teach-
ing (CLT)  138, 217, 267, 347,
350, 351, 357
Communicative stress  25, 44,
190, 193
Computer-assisted language
learning (CALL)  216, 342
Computer-assisted task-
based language teaching
(CATBLT)  215, 220
Computer-mediated communi-
cation (CMC)  318
Conversation analysis  164, 294,
306
Coombe, Christine  xi, xv, 3, 8,
19, 20, 347, 357, 359
Corrective feedback  xvi, 137,
164,, 168, 181, 331
Corrective precasts  111
Elicitation  146, 147, 149–155
English as a foreign language
(EFL)  xi, 1, 3, 5, 90
ESOL (English to Speakers of
Other Languages)  5
Task  54
Evidence-driven practice  8
Explicit correction  146, 149,
150, 152
Explicit grammar-teaching
focus  5
F intersubjectivity  167, 168, 170
Feedback  11, 110, 112, 120, 121,
123, 125, 137, 138, 140–143, 222,
270, 271, 326, 331
Focus on form (FONF)  xi, 12,
141, 163–166, 217, 319360
Foreign language contexts  xi,
xv, xviii, 1, 4, 19, 166
Form-focused episodes
(FFEs)  164, 165
France  4, 7, 15, 16, 313–316, 345
Fukaya, Keiko  10, 89, 348, 359
G
Genc, Zubeyde Sinem  10, 39, 67,
101, 348, 360
Global accuracy measures  38
glossing  94, 101, 103
Guiraud’s Index of Lexical
richness  52
H (LREs)  12, 168, 173, 177
Hobbs, James  11, 109, 360
Hong Kong  xv, 5–7, 12, 13, 17,
159, 187–190, 208, 283, 284,
345, 346, 349, 350, 352, 355,
359
Hong Kong Target Oriented
Curriculum  5
Horiba, Yukie  10, 89, 103, 348,
360
I
Incidental vocabulary
acquisition  10, 91, 93, 101
Information gap task/
activity  203
Initiation-Response-Feedback
(IRF)  120, 123
Interactive lexical phrases  110,
112
Interactive moves  112, 116, 118,
127
International Consortium
on Task-based Language
Teaching (ICTBLT)  16
Iwashita, Noriko  xvi, xviii,
6, 7, 11, 30, 137, 348, 350, 351,
353, 360
Izumi, Shinichi  9, 23, 80, 82, 91,
92, 101, 269, 346, 360
J
Jackson, Daniel O.  6, 14, 15, 267,
347, 352, 360
Japan  xv, 4, 6–8, 14, 16, 23, 109,
163, 267–269, 287, 345, 347,
353, 360, 362
K
Korea  6, 7, 142, 152, 216, 218,
362
L
L2 task performance  24, 28
Language-related episodes
Learner orientation  68
Learner’s attentional
allocation  39
Levkina, Mayya  9, 43, 52, 361
Lexical chunks  112, 123
Lexical variety  31, 33, 38
Li, Huifang (Lydia)  6, 7, 11,
137–139, 158, 217, 348, 350, 351,
353, 361
 Task-Based Language Teaching in Foreign Language Contexts: Research and implementation

Limited capacity hypothesis  9, Resource-directing Task organizers  118–120, 123

23, 40, 80 dimensions  25–28, 45, 46 Task performance  3, 10–12,
Rote learning  7 14, 24–29, 67, 109–111, 123,
M 181–183, 287–290, 300, 303
Malicka, Aleksandra  9, 43, 361 S Task structure  9, 11, 109–112, 127
McAllister, Julie  7, 15, 313, 314, Sasayama, Shoko  9, 23, 80, 82, Task-supported language
361 91, 92, 101, 346, 362 teaching  8
modality  10, 67, 191 Scaffolding  12, 167, 195, 325, 349 TBLT implementation  3, 109, 124
Monologic narrative tasks  27, 30 Second language (SL) Teacher agency  139, 140
Moore, Paul J.  5, 12, 163, 354, 361 contexts  1, 3 Teacher-centered instruction  5
MSTTR  31, 33 Shehadeh, Ali  xi, xv, xvii, 1, 3, Teacher recasts  149, 150
8, 14, 155, 166, 243, 345–347, Text processing  9–11, 89–93,
N 353, 362 101–103
Narcy-Combes, Situation models  91, 106 Time judgment task  54, 66
Marie-Francoise  7, 313, 314, Skehan’s trade-off hypothesis; Topic familiarity  91, 191
318, 321, 361 trade-off hypothesis  9, 43, Transmission-based
Native-speaker task 44, 59 approach  15, 313
interaction  109 Starkey-Perret, Rebecca  7, 313, Triadic Componential
362 Framework  24
O Strategic planning  9, 10, 67–70, T-unit  31–33, 73
Opinion-exchange task  125 72, 73, 87 Turkey  xv, 16, 67, 345, 360
Opinion-gap tasks  122, 123 Student recasts  149, 150
Oxford Placement Test Surface code  90 U
(OPT)  50 Syntactic complexity  9, 23, Uptake  11, 142, 154, 164
26–28, 38, 40
P V
Park, Moonyoung  13, 215, 217, T Venezuela  xv, 4, 6, 13, 14, 16,
346, 348, 362 Target-Oriented Curriculum 241, 255–257, 283, 353, 359
Pica, Teresa  xv, xvi, xvii, xix, (TOC)  188, 189 Verbatim repetitions  32, 39
3, 141, 156, 165, 191, 362 Task-based interaction  xii, Vocabulary Levels Test  94
Presentation-Practice- xviii, 11, 12, 137, 140–142, 151,
­Production (PPP)  110, 348 156, 163, 167, 180 W
Pre-task planning  9, 23–28, Task-based language assessment Weaver, Christopher  14, 15, 287,
38–40, 67, 68, 70, 253, 362 (TBLA)  11, 14, 229, 287, 351 288, 303, 346, 351, 362
Propositional textbase  90 Task-based syllabus  2, 14, 287 Wechsler Adult Intelligence
Task complexity  9–11, 23–41, Scale  30
R 55–60, 92, 109, 189, 190, 346,
Reasoning-gap tasks/ 357, 361 X
activities  114, 117 Task difficulty  9, 24, 45, 49, 50, X-Lex  46, 49, 50, 61
recasts  19, 111, 141, 142, 146, 157, 53, 57, 58
165, 166, 271 Task implementation  xv, 39, 67, Y
Resource-depleting 189, 270 Y-Lex  46, 49, 50, 61
dimensions  25–28, 39 Task manipulation  9, 23