(Current Issues in Linguistic Theory 194) Julie Coleman (Ed.), Christian J. Kay (Ed.) - Lexicology, Semantics and Lexicography - Selec

LEXICOLOGY, SEMANTICS AND LEXICOGRAPHY
AMSTERDAM STUDIES IN THE THEORY AND

HISTORY OF LINGUISTIC SCIENCE
General Editor
E. F. KONRAD KOERNER
(University of Ottawa)
Series IV CURRENT ISSUES IN LINGUISTIC THEORY
Advisory Editorial Board
Raimo Anttila (Los Angeles); Lyle Campbell (Christchurch, N.Z.)

Sheila Embleton (Toronto); John E. Joseph (Edinburgh)
Manfred Krifka (Austin, Tex.); Hans-Heinrich Lieb (Berlin)
E. Wyn Roberts (Vancouver, B.C.); Hans-Jrgen Sasse (Kln)
Volume 194
Julie Coleman and Christian J. Kay (eds)
Lexicology, Semantics and Lexicography

Selected papers from the Fourth G. L. Brook Symposium,
Manchester, August 1998
LEXICOLOGY, SEMANTICS
AND LEXICOGRAPHY
SELECTED PAPERS FROM THE FOURTH
G. L. BROOK SYMPOSIUM
Manchester, August 1998
Edited by
JULIE COLEMAN
University of Leicester
CHRISTIAN J. KAY
University of Glasgow
JOHN BENJAMINS PUBLISHING COMPANY

AMSTERDAM/PHILADELPHIA
TM The paper used in this publication meets the minimum requirements of American
8
National Standard for Information Sciences Permanence of Paper for Printed

Library Materials, ANSI Z39.48-1984.
Library of Congress Cataloging-in-Publication Data

G.L. Brook Symposium (4th : 1998 : Manchester, England)
Lexicology, semantics, and lexicography : selected papers from the fourth G.L. Brook Symposium,
Manchester, August 1998 / edited by Julie Coleman, Christian J. Kay.
p. cm. -- (Amsterdam studies in the theory and history of linguistic science. Series IV,
Current issues in linguistic theory, ISSN 0304-0763 ; v. 194)
Includes bibliographical references and index.
1. Lexicology--Congresses. 2. Semantics--Congresses. 3. Lexicography--Congresses. I. Coleman,
Julie. II. Kay, Christian. III. Title. IV. Series.
P326.G15 2000
413.028--dc21 99-088593
ISBN 90 272 3701 8 (Eur.) / 1 55619 972 4 (US) (Hb; alk. paper) CIP
2000 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 75577 1070 AN Amsterdam The Netherlands
John Benjamins North America P.O.Box 27519 Philadelphia PA 19118-0519 USA
Contents
Introduction vii
Brook Memoir by Alan Shelston ix
List of Contributors xiii
Andreas Fischer
Lexical Gaps, Cognition and Linguistic Change 1
Gabriella Rundblad and David B. Kronenfeld
Folk-Etymology: Haphazard Perversion or Shrewd Analogy? 19
Pivi Koivisto-Alanko
Mechanisms of Semantic Change in Nouns of Cognition: a General
Model? 35
Christian J.Kay
Historical Semantics and Historical Lexicography: will the twain
ever meet? 53
Julie Coleman
Strange Linguists: the Cant and Slang Dictionary Tradition 69
Maurizio Gotti
Lexical Choices in an Early Galilean Translation 87
Carole P. Biggam
Grund to Hrof: Aspects of the Old English Semantics of Building
and Architecture 103
Heli Tissari
Five Hundred Years of Love: a Prototype-Semantic Analysis 127
Louise Sylvester
The Vocabulary of Consent in Middle English 157
vi CONTENTS
Claire Cowie
The Discourse Motivations for Neologising: Action Nominalization
in the History of English 179
R.W. McConchie
The Vernacularization of the Negative Prefix Dis-in Early Modern
English 209
Brook Symposium Edited by Christian Kay 229
Author and Subject Index 241
Introduction
These papers were originally presented at the 10th International Conference

on English Historical Linguistics (10ICEHL), held at the University of
Manchester from August 2126, 1998. They formed part of the Fourth G.L.
Brook Symposium, which consisted of a day and a half of papers on Historical
Semantics, Lexicology and Lexicography, followed by a workshop on elec-
tronic resources.
The papers, we hope, show both the breadth and the depth of current
studies in the field. In keeping with the ICEHL tradition, they exemplify a
willingness to engage with relevant issues in theoretical linguistics, as well as
the detailed and meticulous scholarship characteristic of philology. Two
points stand out particularly. The first is the impact of prototype theory and
cognitive approaches generally in lexical studies. The second is the very
positive effects of the remarkable range of electronic resources now available
to historical linguists, notably corpora, dictionaries, bibliographies and the-
sauruses. These are important both quantitatively, in the amount of data they
make available, and qualitatively, in the versatility of their searches.
We have enjoyed editing this volume, and would like to thank everyone
who made it possible, including the ICEHL conference committee, and espe-
cially David Denison, for suggesting the Brook Symposium in the first place
and helping to bring it to fruition. We would also like to thank the Department
of English and American Studies at Manchester (and indeed the late G.L.
Brook himself) for financial support. Thanks are of course due to our contribu-
tors for their patience and co-operation, and to the anonymous reviewers who
refereed the papers. Our editor Anke de Looper helped to smooth the final
stages. And a special thank you is owed to Ian Hamilton at Glasgow Univer-
sity, who mastered the intricacies of the style-sheet and prepared the manu-
script.
Julie Coleman
Christian Kay
24/9/99
G L Brook: 19101987
Alan Shelston
University of Manchester
George Leslie Brook was Smith Professor of English Language at the Univer-
sity of Manchester from 1945 until his retirement in 1977, having previously
been a member of the English department since 1932. From 1951 the title of
his chair was extended to include Mediaeval Literature. A Yorkshireman from
Huddersfield, Brook completed his doctorate at Leeds but his forty-five years
of service to Manchester were uninterrupted, even by the war, since his poor
health disqualified him from active service. When I first encountered him at
my interview for an assistant lectureship in 1966 he already seemed, sitting at
the end of the table, a remote and slightly frightening presence, but it was he
who not only stayed behind to congratulate me on my appointment, but saw fit
to add a few words of advice on what would be expected of me. Throughout
my years as a colleague I continued to feel the remoteness, and sometimes the
threat, but I came to interpret the former as shyness, and the latter as a matter
of reputation as much as of reality.
But then I never worked directly for him. During his reign at Manchester
the department was very clearly divided into Language and Literature and
while assistant lecturers were appointed in English lest they think too readily
of themselves as specialists, the two paths in fact rarely crossed. Brook had a
strong sense of hierarchy, and it would be fair to say that those who came
directly under his command had little room for manoeuvre. Those of us on the
literature side however could afford to take a more detached view. When he
joked, as he did for example at an examiners meeting that he wanted to keep
my options open so that I can give him a third, or remarked of a Professor of
Education who had got into the news that I remember him well he failed
Intermediate English, we saw this as character acting: those who knew him
better would have argued that these were not jokes at all. He was in the
x ALAN SHELSTON
department every day, a short, rotund, slightly asthmatic and very myopic man
who would progress along the corridor, occasionally pausing to peer at a name
on a door, wondering perhaps whether some newcomer had arrived in the
night without his knowledge. In his own room which few actually entered
unless directly summoned into the presence was a large and very old
upholstered sofa, while around the walls were his books, many of them
volumes of considerable value.
For Brook, for all that he was Professor of Language, was also a literary
man. His publications were predominantly in the field of language notably
his English Dialects (1963), his editions of the Harley Lyrics (1968) and, with
R. F. Leslie, of LaZamons Brut (1963, 1975), and the surveys by which the
undergraduates came into contact with him, An Introduction to Old English
(1955) and A History of the English Language (1968) but he also published
Dickens as a Literary Craftsman (1966), The Language of Dickens (1970) and
The Language of Shakespeare (1976), and if these now have a very old-
fashioned ring the last two of them at least are still cited. He certainly knew his
literature the seventeenth century dramatists and religious writers and the
nineteenth century novelists in particular; furthermore he collected it. For
many years I had to resist the temptation of the three volumes of a first edition
of Felix Holt, the Radical which rested just by the half-open door; only the
thought that Brook himself might be behind it saved me from crime. At that
time in Manchester it was still possible to collect on an academic salary: first
editions of Dickens albeit shabby ones were left out on Eric Mortons
tables for less than five pounds, while for the more discriminating there were
Shaws and Gibbs, then still operating as antiquarian booksellers. Brook
invested well, and by the time he died his collection was worth a considerable
sum. The university became the beneficiary of his expertise when he left the
bulk of his collection, including rare first editions of the first part of Pilgrims
Progress and The Holy War, to the John Rylands University Library. The
leading part that he played in the Manchester Bibliographical Society was a
further reflection of these interests; he was also an active member of the
Manchester Mediaeval Society, the Philological Society, and the Lancashire
Dialect Society, whose journal he at one time edited.
I have suggested that Brook was a difficult man; I came to understand that
he was both a lonely and a shy man. Sadly his eyesight deteriorated to the
point where he began to learn braille, although as much as a precaution as for
immediate use. The loneliness was intensified after the death of his wife Stella
BROOK MEMOIR xi
Maguire, herself a scholar of the language of the English liturgy and intermit-
tently a teacher in the department. Just once did he call me into his room, after
a long department meeting. I cannot remember the exact pretext, but after we
had dealt with whatever it was he surprised me by moving into book talk and
reminiscence. The talk went on for some twenty minutes, until he equally
suddenly stopped and withdrew from this unusual moment of near intimacy. It
was as if he had said too much, and needed to retreat, but it was a revealing
moment to someone who had respected him, sometimes joked about him, but
never ever spoken more than a few nervous words to him directly. I had a
feeling that even that twenty minutes was more than most people had been
allowed.
What cannot be denied about Brook was his generosity to the university
and to the department that he had served for so long. As has been mentioned,
his collection of over four thousand items found its way to the John Rylands
University library where it is divided into separate Drama and Theology
collections. He founded undergraduate prizes one in his name alone, and
one in the name of Stella. There is also a family link to the Maguire Prize,
established in the name of Stellas father. In his will he left a sum of money to
his colleagues with the instruction that they enjoy themselves. More perma-
nently, he endowed a postgraduate prize to be awarded for research in lan-
guage studies and he provided the substantial investment which funds the
biannual Brook Symposium, the 1998 session of which this volume com-
memorates. It would be fair to say that these were bequests whose generosity
surprised as well as gratified some of their beneficiaries, but that perhaps
reflected more upon them than on him. Certainly they revealed the place that
English at Manchester had in Brooks priorities. As on so many previous
occasions it was he who had the last word.
Alan Shelston
Department of English Language and Literature
University of Manchester
Manchester M13 9PL
List of Contributors
Dr C.P.Biggam Prof Maurizio Gotti

High Garrett Universit degli Studi di Bergamo
17 Upper Glenfinlas Street Via Salvecchio 19
Helensburgh G84 7HD 24129 Bergamo
C.P.Biggam@btinternet.com ITALY
m.gotti@mediacom.it
Dr Julie Coleman
English Department Prof Christian J.Kay
University of Leicester Department of English Language
University Road University of Glasgow
Leicester LE1 7RH 12 University Gardens
jmc21@le.ac.uk Glasgow G12 8QQ
c.kay@englang.arts.gla.ac.uk
Dr Claire Cowie
Englische Sprachwissenschaft Dr Pivi Koivisto-Alanko
Philosophische Fakultt Tlnkatu 31 B 19
Technische Universitt Chemnitz FIN-00260 Helsinki
Reichenhainerstr. 39 FINLAND
D-09126 Chemnitz pukovis@cc.helsinki.fi
Germany
claire.cowie@phil.tu-chemnitz.de Dr R.W. McConchie
Department of English
Prof Andreas Fischer University of Helsinki
Englisches Seminar P.O.Box 4
Universitt Zrich Yliopistonkatu 3
Plattenstrasse 47 Fin-00014 Helsinki
CH-8032 Zrich FINLAND
SWITZERLAND roderick.mcconchie@helsinki.fi
afischer@es.unizh.ch
xiv LIST OF CONTRIBUTORS
Dr Gabriella Rundblad Penny Silva

School of Language, Linguistics Oxford English Dictionary
and Translation Studies Oxford University Press
University of East Anglia Great Clarendon St
Norwich NR4 7TJ Oxford 0X2 6DP
G.Rundblad@uea.ac.uk psilva@oup.co.uk
Dr Louise Sylvester Edmund Weiner

Dept of English Language and Oxford English Dictionary
Literature Oxford University Press
Kings College London Great Clarendon St
Strand Oxford 0X2 6DP
London WC2R 2LS eweiner@oup.co.uk
louise.sylvester@kcl.ac.uk
Prof Jane Roberts
Ms Heli Tissari Dept of English Language and
Department of English Literature
University of Helsinki Kings College London
JP 4 Strand
FIN-00014 Helsinki London WC2R 2LS
FINLAND jane.roberts@kcl.ac.uk
heli.tissari@helsinki.fi
Mrs Iren Wotherspoon
Brook Symposium Article, also: Department of English Language
Prof. Frances McSparran
12 University Gardens
Middle English Dictionary
Glasgow G12 8QQ
University of Michigan i.wotherspoon@englang.arts.gla.ac.uk
Ann Arbor, MI 48109,
United States of America
mcsparra@umich.edu
Lexical Gaps, Cognition and
Linguistic Change
Andreas Fischer
University of Zrich
Abstract
The paper first postulates that lexical gaps can be located by studying lexical
configurations of the type suggested in Cruses Lexical Semantics (1986). By
way of illustration it describes gaps in proportional series, hierarchies and sets
of complementaries. In a next step three cognitive factors are proposed for
explaining these gaps: (1) psychological salience or emotional involvement,
(2) perceptual salience, and (3) prototypicality. Finally, there are implications
for the study of linguistic change. From a diachronic point of view lexical gaps
may manifest themselves in four different ways: they may be permanent, that
is evident in a language from the earliest records to the present day; if not
permanent, they may have existed at an earlier stage and may then have been
filled, they may have opened up at some point in the history of a language,
or they may have opened up and then closed again.
Introduction
The cognitive turn in linguistics has led linguists to re-examine many

aspects of language in the light of cognitive science.1 Regarding the lexicon,
for example, important work has been done on categorization in general
(Taylor 1989, 2nd ed. 1995) and on categorization through metaphor (Lakoff
and Johnson 1980, Lakoff and Turner 1989). Language change in general and
lexical and semantic change in particular have received rather less attention,2
2 ANDREAS FISCHER
but some case-studies at least invite diachronic conclusions: Berlin and Kays
(1969) study of colour terms, for example, predicts lexical change (in the
sense of probable additions to the colour terminology), and Williamss (1976)
study of synaesthetic adjectives traces regularities of semantic change. Two
recent publications focus on prototypes: Geeraerts (1997) identifies character-
istics of prototypicality which result in four kinds of semasiological change,3
while Ungerer and Schmid (1996: Ch. 6.3 Lexical change and prototypical-
ity) suggest and illustrate different possible developments of prototypes.4
This paper will address a further, as yet largely unstudied, area in historical
lexicology where the cognitive approach may prove to be helpful: the problem
of lexical gaps.
Lexical gaps, to put it very simply, may be postulated when the lexical
structure of a language provides evidence that there is a concept which might
or should be lexicalized, but is not. To postulate such lexical gaps, therefore,
one needs a theory of lexical structure and a methodology to locate them.5 I
will address this problem first and, in the rest of the paper, look at the
cognitive and diachronic aspects of lexical gaps.
Lexical structure, in this paper, will be understood as emerging from
sense relations such as hyponymy, antonymy and synonymy. While these
sense relations have been known for a long time, they have only in recent
decades received systematic and detailed attention, beginning with John
Lyonss Semantics (1977) and culminating with D.A. Cruses Lexical Seman-
tics (1986). In this process a remarkable consensus has emerged,6 and as a
result there is now a good deal of agreement on the various kinds of sense
relations and their properties. Following Cruse (1986), sets of words con-
nected by sense relations will be called lexical configurations.7 Moreover,
sense relations are not simply the products of linguistic theorizing, for psycho-
linguistic research has shown that they have a basis in the meta-linguistic
knowledge of native speakers (Aitchison 1994: Ch. 8 Word-webs: Semantic
networks). In the following I will begin by looking at selected examples of
lexical gaps in English on the basis of three types of sense relations as
discussed by Cruse (1986), namely proportional series, hierarchies and oppo-
sites.8 After this structural part I will discuss cognitive and historical problems
arising from lexical gaps defined in this way. The evidence will come from
English, with an occasional glance at German.
LEXICAL GAPS, COGNITION AND LINGUISTIC CHANGE 3
Structure, or how to Locate Lexical Gaps
Proportional series, according to Cruse (1986:11819), consist of four or more

lexemes related to each other by analogy or, in his words, proportionality. He
explains:
The simplest proportional series consists of a single cell which has four
elements:
A B
| |
C D
The relations between the elements must be such that from any three of the
elements the fourth can be uniquely determined. The configuration is thus
structured by the following relations of proportionality:
A is to B as C is to D
B is to A as D is to C
A is to C as B is to D
C is to A as D is to B
Starting from a single cell, proportional series (especially the so-called open
ones) can be extended along both axes simultaneously (1986:120). In this way
a basic cell can be expanded to a much wider (and still open) proportional series.
Proportional series can be exemplified by the famous example introduced
by Chomsky in Aspects (1965:169, 231232). Briefly discussing accidental
gaps in the lexicon, Chomsky uses almost the same language as Cruse, that is,
he also sets up the minimal cell of a proportional series (1965:232), speaking
of the absence of a word that bears to plants the relation that corpse bears to
animals.9 This can be diagrammed as follows:
(1) a. animal corpse
plant ?
With only three elements this is not a complete minimal cell according to
Cruses criteria, and the example itself (the corpse of an animal?) is doubtful.
This may be the reason why the series is usually presented in the following,
modified and extended, form (Lehrer 1974:97, Ziegler 1984:67):
(1) b. human corpse
animal carcass
plant ?
4 ANDREAS FISCHER
Here human, corpse, animal and carcass form a complete minimal cell, and its
extension by plant reveals a lexical gap for dead plant. Note that this, like
any lexical gap, is only a structural and not a functional one. Paraphrases like
dead plant can always be used when there is an ad hoc need to express a
concept, and the term lexical gap simply indicates a structure point in a
lexical configuration which is not occupied by a lexicalized item.10
Cruses main example for a proportional series, the generic and gender-
specific terms for animals (1986:11829), provides a good starting point for
discussing some problems.
(2) generic adult adult young
(sg.) male female
horse stallion mare foal
sheep ram ewe lamb
pig boar sow piglet
? bull cow calf
?chicken cock/rooster hen chick/chicken
goose gander goose gosling
dog dog bitch puppy
bear bear bear (bear) cub
lion lion lioness (lion) cub
cat tom-cat tabby-cat11 kitten
goat billy-goat nanny-goat kid
etc.
An example like this one shows, first, that lexical configurations should not be
restricted to simple lexemes. It is of course worth noting where in a lexical
configuration patterns of derivation (for example, -ess for females, -let or -ling
for the young) or composition (for example tabby-cat and tom-cat) may be
found, but as long as they produce lexicalized items they do not affect such a
configuration from the point of view of semantics.12 Secondly, this series is a
good example of a fairly open lexical configuration. It is open vertically in that
the number of animals could be multiplied. On the horizontal axis, too, further
dimensions such as meat (horsemeat, mutton, pork) could be added, but the
possibilities are limited.13 Thirdly, the configuration highlights several cases
of underdifferentiation: bear, for example, is used not only as a generic term,
but also for the male and the female, and cub (when used without a premodify-
ing noun) may refer to the young of several animals. Note that cases of
underdifferentiation are not to be confused with lexical gaps: underdifferen-
tiation means that a word is used in two or more different senses (i.e. senses
for which separate words are used elsewhere in the same configuration), while
lexical gap points to a sense for which there is no word.14 This series,
therefore, reveals two lexical gaps (one certain, one less so; both indicated by
question marks). There is clearly no generic term in English for cow and/or
bull on the one hand, and chicken as a generic term for hen and cock/rooster is
not as well established as, say, sheep or dog: farmers may say that they keep
chickens, but when referring to a single animal it is more natural to speak of a
hen or a rooster than a chicken (while it is more natural to speak of a sheep
than a ewe or a ram).15 I will refer to chicken in the generic sense as a marginal
item.16
The following proportional series of kinship terms is comparable to the
one just dealt with.
(3) generic male female
(sg.)
?parent father mother
child son daughter
?sibling brother sister
? uncle aunt
? nephew niece
cousin cousin cousin
Here the generic terms corresponding to uncle/aunt and nephew/niece respec-
tively appear to be genuine lexical gaps (comparable to the missing generic
term for bull/cow), and the terms parent and sibling are marginal, though not
for the same reason as chicken: all other words in the series belong to the
everyday core vocabulary of English, but these two (especially sibling) are
formal and perhaps written rather than spoken. If we call parent and sibling
marginal on the grounds of register, two more gaps result. (Note that the
situation is similar in German: Elternteil parent is a complex item and
formal, Geschwister siblings is only plural).17
As a further example of a proportional series we may consider the
following (from Ziegler 1978:67):
6 ANDREAS FISCHER
(4) having a not having a

sense or faculty sense or faculty
(verb) (adjective)
see blind
hear deaf
smell ?
taste ?
feel18 ?numb
speak dumb/mute
walk lame
This series highlights the problem of consistency, which has some bearing on
the question of lexical gaps. Cruse (1986:122) points out the difference
between the animal series and series like the following:
(5) a. elevation b. body of water
mountain inland sea
hill lake
hillock pond
mound puddle
Series (5a) and (5b) are consistent because the relations holding between the
elements along the vertical axis are the same throughout: hillock can be
predicted from mountain and hill on the grounds of diminishing size or height,
and pond can be predicted from inland sea and lake by the same kind of
reasoning. Series (1) to (4), however, are not consistent in this way. In (4), for
example, the five senses seem to be a naturally limited series. Speaking and
walking are also human faculties, however, and their absence provides further
terms in the right hand column, supported by the collocational pairing of deaf
and dumb (German taubstumm). Is this enough to justify the series as we have
it here, and what prevents us from claiming that English lacks adjectives
expressing the inability to eat or to breathe? This remains an open question,
and I have to end my discussion of the proportional series with the proviso that
their (frequent or even usual) inconsistency presents a major theoretical and
practical problem.
Hierarchies, according to Cruse, can be taxonomies or meronomies (both
branching and non-branching), but here I will concentrate on taxonomies, that
is on hierarchies where subordinate terms stand in a kind of relationship to
their superordinate terms. Thus a spaniel is a kind of dog, a dog is a kind of
animal, an animal is a kind of creature, and so on. Cruse (1986:146) postu-

lates that in an ideal hierarchy all branches have nodes at each level but he
admits that in this respect natural [or: folk] taxonomies fall short of the ideal
[as against scientific taxonomies which do not]. Thus (6a) is an ideal (natural
or folk) taxonomy, while (6b) and (6c) are not.19
(6) a. creature
animal bird fish insect
dog elephant robin eagle cod trout ant butterfly
spaniel alsatian
b. creature
animal bird
dog cat
collie spaniel robin blackbird starling
c. creature
animal
dog cat bird
collie spaniel robin blackbird starling
According to Cruse (1986:118, cf. 146) (6b) or (6c) is the taxonomy of

many speakers of English [who] feel that the sub-classification of garden
birds into sparrows, robins, thrushes, blackbirds, etc. is comparable not with
the division of animals into dogs, cats, sheep, and so on, but with the sub-
classification of dogs into spaniels, poodles, alsatians, and the like.
8 ANDREAS FISCHER
As a result of this imbalance one might postulate a gap on the level of dog and
cat (in 6b) or on the level of animal (in 6c), but this solution is problematic for
at least two reasons. First of all it is difficult to conceive of a sense distinct
from bird that would fill the gap, and secondly either solution would violate
the principle that in an ideal hierarchy all branches have nodes at each level
(Cruse 1986:146). We conclude that folk taxonomies are often imperfect
(Cruse 1986:147 gives further examples), but that such imperfections do not
automatically allow us to speak of lexical gaps. However, Cruse also argues
(1986:147) that
[t]he lexical items in a taxonomic hierarchy may be considered to be labels for
nodes of a parallel hierarchy of conceptual categories. Now while the exist-
ence of a label without a corresponding conceptual category must be regarded
as highly unlikely, it is not impossible for what is intuitively recognised as a
conceptual category to be without a label.
Cruse calls them covert categories and gives an illuminating example from
among the verbs of locomotion for living creatures (1986:151):
(7) move
0 swim fly
run walk crawl hop jump
According to Cruses (1986:151) intuition there is a covert category in this

hierarchy: there is no superordinate term for the verbs denoting locomotion on
land []. Cruse here does not use the term lexical gap, but his own terms
conceptual category without a label or covert category are clearly synony-
mous with it.
By way of conclusion, and very briefly, I would like to look at one type of
opposites, namely complementaries. In Cruses words (1986:19899):
The essence of a pair of complementaries is that between them they exhaus-
tively divide some conceptual domain into two mutually exclusive compart-
ments, so that what does not fall into one of the compartments must
necessarily fall into the other. There is no no-mans-land, no neutral ground,
no possibility of a third term lying between them. Examples of complementa-
ries are: [adjectives like] true : false, dead : alive, open : shut, [or verbs like]
hit : miss (a target), pass : fail (an examination).
Now what about adjectives such as the ones discussed in (4) above,
namely blind, deaf, dumb or lame? People can either see or not see, they can
hear or not hear, so adjectives like blind or deaf appear to be members of a pair
of complementaries. Such pairs are incomplete, however, for the notions of
not blind / able to see and not ? / able to smell can be paraphrased with
expressions like able to see, able to smell, but are not, or not fully, lexicalized
(sighted,20 ?seeing?). This is partly due to the fact that the existing adjectives
in question express the absence of a quality and thus cannot be negated with
unlike at least some adjectives expressing the presence of a quality. This
holds true for adjectives in general: compare complementaries like true
untrue next to false, but false *unfalse, or antonyms like beautiful
unbeautiful next to ugly, but ugly *unugly. On this basis it seems reason-
able to call the missing complementaries of privative adjectives like deaf
lexical gaps.21
(8) having a having a not having a
sense or faculty sense or faculty sense or faculty
(verb) (adjective) (adjective)
see sighted, ?seeing blind
hear ?hearing deaf
smell ? ?
taste ? ?
feel ? ?numb
speak ? dumb/mute
walk ? lame
This discussion of gaps in lexical configurations such as proportional series,
taxonomies and complementaries lays no claim to completeness, but the
examples, in my opinion, are convincing enough to make the notion of lexical
gaps plausible, and to offer a discovery procedure to locate them.
Cognition, or how to Explain Lexical Gaps
Lexical gaps may thus be plausible structurally, but they pose new problems
when viewed psycholinguistically and diachronically. I shall address these
problems in the rest of the paper. Having established on structural grounds that
there are certain lexical gaps in a language at a given point in time, we may,
10 ANDREAS FISCHER
first, ask why they are where they are, and we may, second, study their history.
It will be seen that the two questions are interrelated.
How, then, can the existence of lexical gaps be explained? Looking at the
examples just presented, it seems to me that a number of ultimately cognitive
factors may be responsible. So far, and very tentatively, I have identified
three, partly overlapping, ones, which I will call psychological salience or
emotional involvement, perceptual salience, and prototypicality.
Psychological Salience or Emotional Involvement
Psychological salience may be illustrated with the kinship terms (3), where
parent and sibling are semi-technical and where there is no gender-neutral
term for uncle/aunt or nephew/niece. Mother and father, brother(s) and
sister(s), aunt(s) and uncle(s) etc. tend to be people one knows intimately as
individuals and refers to as individuals. As individuals, however, they are my
mother or my father, not my parent or one of my parents, and the absence of a
gender-neutral singular term can thus be explained psychologically.
Psychological salience may also explain cases of underdifferentiation,
which, as I have indicated, are separate from, but related to, lexical gaps. Why
does English have separate terms for the males, females and the young of
certain animals, but not of others (2)? The answer must be that the former tend
to be domestic animals and pets, i.e. animals that human beings are close to. A
farmer has to do with foals, lambs, calves etc. and many owners of pets are
familiar with puppies and kittens. Only zoologists and other specialists are
involved with the young of, say, lions, tigers, leopards, bears and foxes,
however, and so it is psychologically plausible that they are all called cubs
(Cruse 1986:127). With other animals even the distinction between males and
females becomes psychologically irrelevant for ordinary speakers.
Psychological salience may further explain the gap, pointed out by
Chomsky, for dead plant (1b). We are emotionally moved by dead human
beings, less so by dead animals, and certainly least by dead plants. (Note that
we always bury the first, that we sometimes bury the second, that we rarely
bury the third.) It is worth noting that new coinages, at least in German (cf.
Waldsterben dying of forests, Baumleiche corpse of tree, etc.) may be
explained as the result of increased human involvement with inanimate nature,
of the humanisation of nature, so to speak.
The gap among the verbs for locomotion (7) may also be due to psycho-
logical salience. A human beings natural way of locomotion is moving on land,

and this can be done in a great many ways, all captured by terms on the third
level such as run or walk. Swimming and flying, by contrast, are less common
forms of locomotion and thus psychologically more distant: the two general
terms on the second level are probably a reflection of this.22 Psychological
salience, finally, may account for the gaps among complementaries (8). The
vast majority of people can see and hear: seeing and hearing are thus normal,
while not seeing and not hearing are the psychologically salient exceptions.
Perceptual Salience
Perceptual salience may be less important as a factor, and it may go hand in

hand with psychological salience. I only suggest it as a possibility, therefore.
The verbs in (4) and the adjectives denoting their absence may serve as an
illustration. If a person is blind or lame, one can usually see the defect; if he or
she is deaf or dumb, this becomes apparent as soon as we try to communicate
through language. Not being able to smell or taste (for example when one has
a cold), however, is not detectable, except by the person affected by it. As just
pointed out, there is a potential overlap with psychological salience, for one
might maintain that blindness or deafness are felt to be more severe handicaps
than not being able to smell or taste.
Perceptual salience may also be responsible for the irregularity (for some
speakers) in the natural taxonomy for creatures noted by Cruse (1986:118,
14647) (6b+c). Anyone can see, one could maintain, the difference between
a dog and a cat, but not anyone can see the difference between a thrush and a
blackbird (or a robin and a titmouse, for that matter). There again the matter
may be psychological rather than perceptual; this at least seems to be sug-
gested by Cruse (1986:118) when he says that a taxonomy such as (6b) or (6c)
makes no biological sense, of course, but it has a certain psychological
validity, in that the significance to most members of our society of the
difference between, say, a thrush and a blackbird is roughly comparable to
that between a collie and a spaniel. (emphasis mine)
Prototypicality
Prototypicality is based on the notion of prototypes introduced by the psy-

chologist Eleanor Rosch. Roschs work (Cruse 1986:22) has shown that
12 ANDREAS FISCHER
informants judge some members of taxonomic categories to be better or

more central than others. The most central are called prototypes. Now it is
comparatively easy to imagine a prototypical cat, horse or duck, but what does
a prototypical bovine animal look like? Bulls have big horns, short necks, but
clearly no udder, whereas all cows have udders. By the same token one could
argue that it is difficult to describe the prototypical rooster-cum-hen, because
roosters are recognizable by their combs and dewlaps and by their distinctive
tails. To put it simply: there may be such a thing as a prototypical horse or
sheep, but there is no prototypical bull-cum-cow, no prototypical rooster-cum-
hen. I would like to suggest that it is the absence of a prototype for these two
animals which may be responsible for the gap among the generic terms.23
I do not claim to have solved the problem of causation with these
suggestions, but it seems to me that the three cognitive and perceptual factors
mentioned, namely psychological salience or emotional involvement, percep-
tual salience, and prototypicality, go some way toward explaining why we
find lexical gaps at all.
Diachrony, or what may Happen to Lexical Gaps
The cognitive explanation suggested in the previous section immediately

raises further questions. If some or all of the gaps discussed above are, indeed,
motivated cognitively, should we not expect to find them everywhere, that is,
throughout the whole history of a language and, cross-linguistically, in all or
many languages of the world? The answer, with some reservations, must be
yes. The reservations, of course, concern the universality of the three cogni-
tive factors mentioned above. While for human beings locomotion on land has
always been and presumably will remain of greater importance than swim-
ming or flying, different people in different cultures and/or at different times
will have different domestic animals, and this, in turn, may be expected to
influence the respective lexicons. Kinship, too, may be valued differently,
which can be expected to influence the kinship terms of a language.
From a diachronic point of view lexical gaps in a language may manifest
themselves in four different ways: they may be permanent, that is evident in a
language from the earliest records to the present day; if not permanent, they
may have existed at an earlier stage and may then have been filled, they may
have opened up at some point in the history of a language, or they may have
opened up and then closed again.

Permanent gaps are gaps which are motivated cognitively in the ways
outlined above. Given that this cognitive motivation has remained stable, one
would expect the gaps to have persisted throughout the history of a language.
Thus from Old English onwards the English language has not had gender-
neutral/common gender singular terms for father-or-mother, brother-or-
sister or uncle-or-aunt. The two terms parent and sibling which do exist are
technical in meaning and came into the language relatively late (parent pl.
c1450, sg. 1568; sibling is Old English with the meaning relative, kinsman,
while the modern meaning brother or sister is only attested from 1903
onwards). Parent, however, is becoming more common, reflecting perhaps a
change of the perception of parenthood: the high divorce-rate and the many
unmarried mothers in certain societies have made a concept like single parent-
hood much more acceptable (and much more talked about) than it used to be.
The case of parent, like that of Waldsterben/Baumleiche, may thus be seen as
a lexical gap being filled because of recent cognitive motivation. In this context
one might also mention the traditional, and I would claim motivated
absence of a common gender pronoun of the 3rd person singular, and the
problem this has created in the recent past. Generic he (If a baby cries, he may
be hungry.) is felt to be unsatisfactory nowadays because of the asymmetry it
represents: he is also, and more frequently, used to refer to males only and there
is no generic she. The missing common gender pronoun has clearly become
a motivated gap, but one that has proved difficult to fill. Pronouns are a more
or less closed set and changes in pronominal systems only happen very slowly,
but at present it looks as if singular they (see (9b) below) may be the most
acceptable of the many solutions offered.24 The common gender cow-or-bull
illustrates the opposite case: Old English had the term hrier, a cognate of
German Rind, but English lost it in the 17th century without replacing it (later
forms are dialectal; cf. OED s.v. rother). The evidence is slight, but I venture
the suggestion that this loss could happen because hrier/rother was cogni-
tively less motivated than other common gender animal terms.
Unlike the cognitively motivated permanent or lasting gaps, temporary
gaps come about by accident and are filled again by a therapeutic measure
(Aitchison 1991: Ch. 10 Repairing the patterns: Therapeutic changes). I can
offer two examples for this process. One concerns the 3rd person singular
personal pronouns, the other the terms for the seasons of the year.
The case of the 3rd person singular pronouns in Middle English (9a) is
14 ANDREAS FISCHER
well known (Samuels 1972:114116 and Lass 1992:116123) and needs little
exemplification. When the monophthongization of eo in early Middle English
led to the merger of Old English heo she with he he in large areas of
England, English in effect lost its 3rd person singular feminine pronoun. This
accident led to an unwelcome gap in the pronominal system, which was
filled almost immediately with the new pronoun she:
(9) a. generic male / female /
masculine feminine
OE ? he heo
early ME ? he ?
late ME ? he she
b. ModE ?they he she
My second and final example concerns the terms for the seasons of
the year (10) which exemplify one of Cruses minor lexical configurations,
namely a cycle or helix. It seems that Primitive Germanic only had two season
terms (summer and winter), while the various Germanic languages developed
four-season systems probably under the influence of the classical languages.
In Old English the seasonal helix was constituted by the four terms lencten,
sumer, hrfest and winter, but the old words winter and sumer were used
very differently from the newcomers lencten and hrfest. Anderson, who
has studied this situation, concludes that when it comes to the seasons, two
semantic systems co-exist in Old English: the earlier, Germanic system of two
seasons, winter and sumer, and the Latin system of four seasons, introduced
by means of interpretatio romana (1997:263). This difference is also evident
in the later history of English (Fischer 1994): while the terms for winter and
summer are characterised by remarkable lexical and semantic stability, the
terms for spring and autumn are highly unstable diachronically.25 Lencten
was already polysemous in Old English (meaning first season of the year as
well as time for fasting before Easter, Quadragesima). In the l3th to l5th
centuries Middle English lent(en) lost its seasonal meaning and became
restricted to its religious meaning (surviving as Modern English lent). This
meant the loss of a motivated term and the opening up of an accidental gap in
the seasonal helix, and it is fascinating to study the process of its replacement.
The first therapeutic measure was the extension of the meaning of sumer, but
this seems to have been felt as unsatisfactory, and from the 14th to the 16th
centuries a whole series of new terms was introduced from which spring
eventually (that is in the 16th century) emerged as the winner. This restructur-
ing also affected the terms for the third season: the loanword autumn and then
the new term fall were added to the lexicon and eventually replaced hervest/
harvest, which lost its seasonal meaning by the 18th century.
(10) spring summer autumn winter
OE lencten sumer hrfest winter
ME ? sumer hervest winter
ME ?sumer sumer hervest winter
ModE spring summer autumn/fall winter
Conclusion
Using examples from English I have discussed procedures for locating lexical
gaps in a language, I have suggested ways of explaining such gaps cognitively
and I have shown the diachronic implications of such explanations. The first
of these steps is not new, and my argument builds on and refines that of others.
The paper breaks new ground when it attempts to adduce cognitive factors to
explain lexical gaps and when it looks at the possible as well as the attested
diachronic developments in the lexical domains under investigation. The
findings invite cognitive and diachronic generalisations, but since this study is
based on a small body of evidence, further examples from English and, above
all, other languages will be required to substantiate them.26
Notes
1. For overviews see Radden (1992) and Ungerer and Schmid (1996).
2. See, however, Kellermann and Morrissey (1992).
3. He calls them modulations of core cases, radial sets through time, semantic polygen-
esis and semantic change from subsets, respectively (1997: Chapters 1.3 and 2). All his
examples are from Dutch.
4. The examples they give illustrate prototype stability (with lexical and semantic change
as well as metaphorical extensions), prototype shift and prototype split (1996:260
267), to which one might add prototype merger and prototype loss. Ungerer and
Schmids examples are taken from the history of English.
5. Note that while many linguists accept the idea of lexical gaps, others reject it entirely.
16 ANDREAS FISCHER
Lehrer (1974: Chapter 5), for example, takes the existence of gaps for granted, while
Ziegler (1984:7576) comes to a negative conclusion: Zum Begriff des Lexikons gehrt
zuallererst, dass es eine offene Liste darstellt, dass es grundstzlich und ohne jede
Einschrnkung vernderbar ist. Mit der Annahme von Lcken wird man diesem Begriff
nicht gerecht. Der Lckenbegriff unterstellt nmlich, dass das Lexikon einer Sprache
abschliessbar ist; Offenheit wre dann nur ein zuflliger Defekt des Lexikons, nicht
systematische Bedingung. Gesteht man aber zu, dass die Offenheit des Lexikons nur
systematische Offenheit sein kann, dann ist man gentigt, die Annahme fallen zu lassen,
jede Sprache habe ihre spezifische lexikalische Struktur. [It is part of the notion of the
lexicon that it is an open list, that it can be changed as a matter of principle and without
any restriction. By postulating gaps one does not do justice to this notion. The notion of
gaps presupposes that the lexicon of a language is finite; openness, then, would be a
coincidental defect, not a systematic requirement. If, however, one concedes that the
openness of the lexicon can only be a systematic openness, then one has to drop the idea
that each language has its own lexical structure.]
6. No such agreement has ever been reached on, for example, lexical/semantic fields or
semantic features/components.
7. In a note Cruse (1986:134) states that what he calls lexical configurations are often
referred to as lexical fields, or word fields. Field theorists tend to view whole configura-
tions as linguistic entities; for us, however, lexical configurations are merely by-
products, as it were, of particular sense relations. This is also the view I adopt here.
8. Lehrer (1974), who works with semantic features, sets up matrices which are very
similar to Cruses lexical configurations. She postulates matrix gaps which show up
when related lexical items are analyzed into semantic features and placed on a chart or
matrix (1974:97).
9. In the body of the text Chomsky, following Halle, uses phonology to illustrate the notions
of systematic and accidental gaps. His lexical examples are found in a footnote.
10. A number of items dealt with in the following may be more acceptable in technical
language/scientific terminology than in ordinary speech, or they may be thought of as
belonging to most peoples passive vocabulary only. Such words will be preceded by a
question mark and will be discussed.
11. The word is used to describe cats of a particular colour as well as she-cats. According to
the OED, the colour sense is attested from the 17th century onwards, the gender sense
since the 19th century.
12. Lehrer has a separate category of derivational gaps (1974:9697).
13. In proportional series the order of items along the two axes is largely random, but see
examples (5a) and (5b) below.
14. Lehrer (1974:9899) unfortunately does not distinguish between the two.
15. It seems to be true generally that generic terms are used more easily in the plural than in
the singular. This is particularly evident when one compares the plural parents (non-
technical) with the singular parent (technical, but becoming more general; see below).
Instead of plural generics one may have singular collective terms such as cattle or
poultry.
16. This is a point about which native speakers may disagree. It could be tested empirically
quite easily.
17. D.J. Allerton (forthcoming) neatly systematises the ways in which a language may mark
or not mark, gender differences. His typology complements my examples (2) and (3).
18. D.J. Allerton (personal communication) points out the parallelism of the verbs meaning
experience (with sense organ) (see, hear, smell, taste, feel) and the verbs meaning
attend to (with sense organ) (look, listen to, smell, taste, touch). There are two cases of
underdifferentiation here (smell and taste), which parallel the lexical gaps found in the
list of adjectives.
19. If (6a) was a scientific taxonomy one would find mammal instead of animal.
20. The OED attests sighted endowed with sight; able to see since 1836.
21. Alone is another such privative adjective. Ayto (1996:181) mentions the interesting case
of memorious which does plug a curious gap in English vocabulary: we have its
antonym forgetful, but the concept of having a good memory is not otherwise
lexicalized. The OED attests memorious having a good memory; mindful only in the
early 17th century (five quotations from between 1599 and 1656), but Ayto quotes an
example dated 1987.
22. For human beings swimming is natural, but not a primary means of locomotion, while
flying is unnatural.
23. However, generic terms do exist even when there are noticeable differences between the
males and females of animals (lion lion lioness, peacock peacock peahen).
Perhaps one could argue that psychological salience overrides prototypicality.
24. See Baron (1986: Ch. 10 The Word That Failed) for details.
25. Unlike the words for winter and summer they also show substantial regional variation.
26. I should like to thank D.J. Allerton, my anonymous reviewers and, of course, the editors
of this volume for helpful comments and suggestions.
References
Aitchison, Jean. 1991. Language Change: Progress or Decay? 2nd ed. (=Cambridge
Approaches to Linguistics). Cambridge: Cambridge University Press.
Aitchison, Jean. 1994. Words in the Mind: An Introduction to the Mental Lexicon. 2nd ed.
Oxford: Blackwell.
Allerton, D.J. forthcoming. Suffixes and other markers of gender in English.
Anderson, Earl R. 1997. The seasons of the year in Old English. Anglo-Saxon England
26. 23163.
Ayto, John. 1996. Lexical life expectancy a prognostic guide. Words: Proceedings of
an International Symposium, Lund, 2526 August 1995, organized under the auspices
of the Royal Academy of Letters, History and Antiquities and sponsored by the Founda-
tion Natur och Kultur, Publishers ed. by Jan Svartvik. (=Konferenser, 36). Stockholm:
18 ANDREAS FISCHER
Kungl. Vitterhets Historie och Antikvitets Akademien. 181188.

Baron, Dennis. 1986. Grammar and Gender. New Haven & London: Yale University
Press.
Berlin, Brent & Paul Kay. 1969. Basic Color Terms: Their Universality and Evolution.
Berkeley: University of California Press; repr. 1991.
Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge, Massachusetts: The
M.I.T. Press.
Cruse, D.A. 1986. Lexical Semantics. (=Cambridge Textbooks in Linguistics). Cambridge:
Cambridge University Press.
Fischer, Andreas. 1994. Sumer is icumen in: The seasons of the year in Middle English
and Early Modern English. Studies in Early Modern English ed. by Dieter Kastovsky.
(=Topics in English Linguistics, 13). Berlin & New York: Mouton de Gruyter. 7995.
Geeraerts, Dirk. 1997. Diachronic Prototype Semantics: A Contribution to Historical
Lexicology. (=Oxford Studies in Lexicography and Lexicology). Oxford: Clarendon
Press.
Kellermann, Gnter & Michael D. Morrissey, eds. 1992. Diachrony within Synchrony:
Language History and Cognition. Papers from the International Symposium at the
University of Duisburg, 2628 March 1990. (= Duisburger Arbeiten zur Sprach- und
Kulturwissenschaft/Duisburg Papers on Research in Language and Culture, 14).
Frankfurt am Main: Peter Lang.
Lakoff, George & Mark Johnson. 1980. Metaphors We Live By. Chicago & London:
University of Chicago Press.
Lakoff, George & Mark Turner. 1989. More than Cool Reason: A Field Guide to Poetic
Metaphor. Chicago & London: University of Chicago Press.
Lass, Roger. 1992. Phonology and Morphology. The Cambridge History of the English
Language, Vol. II: 10661476 ed. by Norman Blake. Cambridge: Cambridge Univer-
sity Press. 23155.
Lehrer, A[drienne]. 1974. Semantic Fields and Lexical Structure. (=North-Holland Lin-
guistic Series, 11). Amsterdam & London: North-Holland.
Lyons, John. 1977. Semantics. 2 vols. Cambridge: Cambridge University Press.
OED see Simpson, J.A. & E.S.C. Weiner.
Radden, Gnter. 1992. The cognitive approach to natural language. Thirty Years of
Linguistic Evolution: Studies in Honour of Ren Dirven on the Occasion of his Sixtieth
Birthday ed. by Martin Ptz. Philadelphia & Amsterdam: John Benjamins. 513541.
Samuels, Michael. 1972. Linguistic Evolution: With Special Reference to English. (=Cam-
bridge Studies in Linguistics, 5). Cambridge: Cambridge University Press.
Simpson, J.A. & E.S.C. Weiner, eds. 1989. The Oxford English Dictionary. 2nd ed. 20 vols.
Oxford: Oxford University Press.
Taylor, John R. 1995. Linguistic Categorization: Prototypes in Linguistic Theory. 2nd ed.
Oxford: Clarendon Press.
Ungerer, Friedrich and Hans-Jrg Schmid. 1996. An Introduction to Cognitive Linguistics.
(=Learning About Language). London & New York: Longman.
Williams, Joseph M. 1976. Synaesthetic adjectives: a possible law of semantic change.
Language 52. 461478.
Ziegler, Jrgen. 1984. Gibt es lexikalische Lcken? Linguistische Berichte 93. 6677.
Folk-Etymology:
Haphazard Perversion or Shrewd Analogy?
Gabriella Rundblad and David B. Kronenfeld

University of East Anglia and University of California
Abstract
Folk-etymology is a very popular phenomenon. Folk-etymologies often begin

as highly individual constructions, but yet seem to conform to some kind of
collective reality; because of this language users can accommodate to folk-
etymologised words with apparent ease. Using instances of folk-etymology
from various periods in the development of English, the semantic and cultural
nature of folk-etymology will be explored, the aim being a description,
identification and interpretation of the process of folk-etymology, the mecha-
nisms underlying it, and the insights it offers regarding the society of its
speakers.
Introduction
Folk-etymology has always been a popular activity and object of study, and
has received a lot of attention from both the speakers of languages and those
working professionally with the languages, such as writers, teachers and
linguists. But still several questions remain unanswered.
In books on language change, folk-etymology is normally discussed, if
discussed at all, under the heading analogy, where it is grouped together with
phenomena such as the tendency for irregular verbs to become regular (cf.
Hock 1991: 202203; McMahon 1994: 7576, 183184). One question that
arises is how folk-etymologies are similar to such (especially grammatical)
20 GABRIELLA RUNDBLAD AND DAVID B. KRONENFELD
analogies and how they differ. Analogies are usually treated primarily as
instances of formal innovation, but folk-etymologies seem to be a kind of
combination of formal and semantic innovation.
Using various instances of folk-etymology from several periods in the
development of the English language, we will discuss the semantic and
cultural nature of folk-etymologies. The majority of these instances were,
initially, extracted from Palmers dictionary of folk-etymologies (1890), but
have been scrutinised in the light of modern theories of semantics and lan-
guage change. However, other more frequently debated folk-etymologies
have also been included in our discussions.
An apparent and important aspect of folk-etymology is how easily folk-
etymologies are understood, accommodated to, adopted and spread further. As
in the case of all kinds of linguistic innovations, folk-etymologies begin as
highly individual constructions; nevertheless, these new, and derivationally
incorrect, constructions seem to adhere and conform to some kind of collective
reality. Language is a social phenomenon, and consequently successful changes
including the creation of new elements must conform to a shared pattern
of understanding. The historical inaccuracy of folk-etymologies helps undo
some of the opacity produced by borrowing. Folk-etymologies also facilitate the
elimination of the confusion and obscurity often introduced by the individual
and historically accidental processes used in the construction and usage of
words thereby making the cognitive elements that structure the lexicon of a
language (including the logic of word combinations and the salience of labelled
cultural entities and activities) particularly clear. Therefore, folk-etymologies
can be said to represent reasonably pure pictures of how the meaning relations
among terms in a language and the relation of these meanings to usage patterns
ought to be in the system of the given language.
The overall aim of this paper is, thus, to discuss the process of folk-
etymology and the semantic mechanisms underlying it, as well as to show
some of the insight folk-etymologies can offer into the cultures of those who
coin them and accept them.
Folk-Etymologies and Linguistic Analogy
Errors that people make in applying cognitive systems are rarely if ever
random or arbitrary. Instead, in the case of language, they usually unveil some
FOLK-ETYMOLOGY 21
underlying linguistic or cultural pattern. For example, the mistakes with

irregular inflections that children make when acquiring languages (such as
two mans, lots of deers, I singed a song, he hitted me) show that they are only
trying to form the plural forms or the past tense of the nouns and the verbs;
their attempts (and consequent mistakes) rest on the patterns of the plural form
and the past tense of the majority of the English words. This process is
commonly known as analogy. As pointed out by McMahon, analogy is
primarily concerned with the link between sound and meaning, which com-
bine to express particular morphemes or meaningful units. The task of analogy
is then to maintain this link by keeping sound structure, grammatical structure
and semantic structure in line, especially when sound change might have
made their relationship opaque (1994: 70).
As thinkers, and thus as speakers of a language in order to lighten our
memory load while maintaining our languages flexibility and adaptability
we keep trying to find productive patterns, patterns that make sense of the
otherwise opaque and confusing words often produced by the accidents of
history. But we also try to find cultural or social meaning in the patterning of
conceptual relations. Thus, analogies have a tendency to disclose the produc-
tive inflectional systems in that these are over-generalised.
Folk-etymology is usually regarded as one of the more noticeable sub-
categories of analogy. Just as analogies do, folk-etymologies reveal how
speakers regard linguistic and cultural matters.
A nice example of the sense in which cultural information can be re-
vealed by a folk-etymological process is one that Ardener (1971: 224225)
described in which members of a community in Wales folk-etymologised
asphalt as ashfelt. The mistake was occasioned by the fact that the actual
derivation of this word (Latin asphalton, -tum; Greek e, a variant of
e, of foreign origin (cf. OED)) was unknown to them and uncon-
nected with their cultural knowledge. Instead of seeking out the accidental
history (that is, the true etymological history) of the word, or leaving it opaque
and obscure, Ardeners Welsh speakers recognised that the process involved
in making an asphalt road was similar to a process familiar to them, namely
that of felting (that is, the pressing of wool or hair into a thick piece of cloth).
Thus, they seem to have presumed a meaning relationship between the two
processes (felt and asphalt) on the basis that asphalting or the laying of a
macadamised road was the pressing or felting of ash (ash used in the sense
powdery residue, composed chiefly of earthy or mineral particles, left after
the combustion of any substance).1

Reinterpretations of this kind can give us information about the culture of
that society information that we might not have otherwise known involv-
ing, in the Welsh case, both wool-working and road construction as well as the
(commonly) perceived similarity between these two processes.
Types of Folk-Etymology
Folk-etymology, as defined by Winer, is a popular but false hypothesis for a

word derivation, usually based on similarities of phonology or meaning
between two or more words [] or from similarity to results of a known
historical process (1992: 238239). Winers definition of folk-etymology
facilitates classification into two major groups of folk-etymology: we shall
refer to these as Class I and Class II. Class I contains folk-etymologies where
some change has occurred, either in meaning or form, or both. Folk-etymolo-
gies of the Class II type, on the other hand, do not usually change the meaning
or form of the word, but function mainly as some popular, though false,
etymological explanation of the word.
For the Class I folk-etymologies, the absence of a meaningful relation-
ship between the parts of the word and the word as a whole represents a source
of unease or a pressure for change or modification. Consequently, a back
and forth (feedback) process begins that involves, on the one hand (where
there exist suggestive possibilities), a search for meaning-sets for the parts
which would relate meaningfully to the whole. On the other hand (where there
is room for reinterpretation) the process involves a shift in the meaning of the
whole word in a direction suggested by the meanings ascribed to the parts. The
over-all intention guiding the Class I folk-etymologies is the de-obscuring of
opaque words or phrases.
The word chance-medley goes back to Anglo-French chance medle
mixed or mingled chance or casualty and the original sense of chance-
medley was accident or casualty not purely accidental, but of a mixed
character; haphazard or random action, into which chance largely enters.
However, due to the frequent use and influence of chance (and medley), the
expression has incorrectly been reinterpreted as a compound meaning acci-
dental (pure chance) medley or confusion.
Class I, which is the more common type of folk-etymology, can if needed
FOLK-ETYMOLOGY 23
also be divided up into sub-groups depending on what part of a word triggered

the change or what part was changed; that is, whether the form and/or meaning
triggered the change (as in quagmire becoming quakemire due to the phono-
logical similarity between quag and quake and the appropriate meaning of
the latter); or whether the change was triggered by some factor other than the
form and meaning, leading to a change of the form and/or meaning (for
example, silver type stemming from Elzevir type where the first element is the
name of the printer family that invented and used the type, but when that
connection was lost, this element became obscure, triggering a change).2
However, since such a classification would ultimately rely on linguists being
able to find out about each and every step of a folk-etymological process an
almost impossible event such a classification will not be attempted or used
here.
Other folk-etymologies only serve to explain the (etymological) history
of a word, possibly so as to be able to reinforce the word and its usage as well
as to ease the memory load. This group comprises Class II. In these cases, the
folk-etymologisers are very cautious regarding the form and meaning of the
words, in order to be able to hypothesise about the origin and creation of the
words. Thus, these creations have a tendency to preserve the form and
meaning of the words.
For example, it has been suggested that clipper a fast-sailing vessel is
actually a corruption of the name of one of Britains fastest ships, Cleopatra-
cum-Antonio, which in the mouths of the English was commonly referred to
as Clipater (Blackmore, in Palmer 1890: 66). Clipater was according to the
story changed into Clipper, hence clipper. Apart from the fact that this folk-
etymology is not true, it had no effect on the words form or meaning (to a
certain extent possibly also because the Cleopatra explanation was not com-
monly accepted), but mainly served as an attempt at an explanation (for some
people).
Similarly, the word forms key and quay (an artificial bank or landing-
place for boats and ships that is made of stone or other solid material) co-
occur in most varieties of English. Key, which was the first form to be used in
English, stems from Old French kay, cay sand-bank, bar. In the early 18th
century the spelling quay was introduced, modelled after French quai, but this
form did not become standard until the 19th century (OED). However, con-
cerned native speakers often regard key as some corruption of quay, as if the
presumed corruption was based on the notion that the key/quay shuts or locks
in vessels from the high sea. Even though people connect the two word forms
and it is quay that is regarded as good English (in Britain at least), key still
remains in use; and its pronunciation, [ki:], which is actually the pronuncia-
tion of both forms, has been left intact.
Occasionally, folk-etymologies belong to both classes at the same time;
that is, for one and the same word we can find two different folk-etymological
processes that are related to each other one way or another. This is the case for
herring-sue a heron, where a Class I folk-etymology has given rise to a
subsequent Class II folk-etymology. Herring-sue is a corruption of heronsew
(from Old French heronceau). In coastal dialects, heron- is commonly dis-
torted into herring, herrin, based, it is said, on the assumption that herons
pursue shoals of herrings. Additionally, in some English dialects, Old French
ceau came down as sue, and, therefore, heronsew was turned into herring-
sue. However, this last element has been further folk-etymologised as stem-
ming from the word pursue, based on the very same belief that herons pursue
herrings.
It also happens that one and the same word can give rise to several
separate and unrelated folk-etymologies sometimes to folk-etymologies of
both types. A freebooter is a person who goes about in search of plunder, a
pirate. The word freebooter is by many treated simply as a normal English
compound consisting of free+booter, a person (a pirate) who gathers booty
freely (that is, unencumbered by legal constraints etc.). However, the word
stems from Dutch vrijbutier (cf. German freibeuter), which makes the com-
pound interpretation nothing more than a Class I folk-etymology.3 It has also
been incorrectly assumed that freebooter is a corruption of Spanish filibustero
(cf. Middle English flibutor, flibustier, Modern English filibuster). But since it
is filibustero (and the equivalent English forms) that is the corrupted form,
stemming from the very same Dutch word, vrijbutier, this assumption must be
classified as a Class II folk-etymology.
A Collection of Folk-Etymologies
As mentioned earlier, the discussions and conclusions of this paper mainly

rely on a small collection of folk-etymologies. 100 alleged folk-etymologies
were extracted from Palmers dictionary of folk-etymology (the first entry on
every fourth page was picked out). The extracted words were investigated for
FOLK-ETYMOLOGY 25
how they were actually created (their true etymology). It was occasionally not
possible to establish the true origin of a presumed folk-etymology, as in the
case of thrush a species of sore-throat and tommy bread.
Thereafter, the words were classified according to which type of folk-
etymology they represent; that is, whether they belong to Class I or Class II.
Almost half of the words were not classifiable into the two classes: some
because they were not folk-etymologies at all, some because the folk-etymol-
ogy presented by, for example, Palmer could not be verified (that is, they
might prove to be idiosyncratic folk-etymologies), and some because it was
impossible to establish the true etymology of the words, which in many cases
made classification equally difficult and impossible. The results of the classi-
fication are illustrated in Table 1.
Table 1. The classification of the extracted words. NB. Two of the nouns contained two folk-
etymological processes each, hence the figures within parentheses
word class No of words Class I Class II Rest
noun 85 (+2) 41 5 (+2) 39
verb 6 4 0 2
adjective 6 2 1 3
adverb 1 0 0 1
interjection 1 0 0 1
noun/adjective 1 0 0 1
= 100 (+2) = 47 = 6 (+2) = 47
Of the two classes of folk-etymology, Class I is by far the largest. Almost half
of all the words and 85% of the classified words belong to this class. Another
conspicuous but not surprising find is that as many as 85 of the 100 words are
nouns.
Some of the folk-etymological processes seem to reveal interesting cul-
tural information, which, given our good historical and archaeological knowl-
edge of the culture in question, can hardly be said to be surprising. For
instance, the plant feverfew, Pyrethrum Parthenium (from Anglo-French
*fevrefeu), has leaves like feathers, something which shortly after its introduc-
tion gave rise to a change of its name from feverfew to featherfew. The change,
however, did not stop here, as the name probably was still considered too
obscure. Other names of the plant, still based on the feather-like appearance of
its leaves and its subsequent incorrect association with birds, are featherfold
and featherfowl.
Another example is bogbean, a common water plant (Menyanthes

trifoliata) that bears racemes of pinkish white flowers. Bogbean is clearly a
reinterpretation of buck bean (a translation of Flemish bocks boonen goats
beans). Since the plant mainly grows in bogs, the name bogbean seems more
adequate that the original buck bean. The folk-etymology was most likely
reinforced by the fact that many of the plants growing in bogs have acquired
names containing bog: bog asphodel, bog berry, bog moss, bog onions, bog
pink, bog nut and bog trefoil.4
Variations in Cultural Significance
Cultural knowledge and insight can similarly be provided by folk-etymologies

where the presence in language of apparent derivational relations can turn out
to be historically inaccurate (Class II). For instance, the word female appears
to represent the marked form of the unmarked male being formed by the
addition of some sort of prefix (*fe-).
Historically, female, from Middle English femelle (from Old French
femelle, a diminutive form of Latin femina woman/female), is unrelated
to male (Old French male/masle; Latin masculus (little) man/male); but
Middle English femelle was clearly remodelled into female based on the
association with male (approximately the 14th century) (OED). The remodel-
ling of female brought female and male into their current and apparently
sense-related and asymmetric relationship (one that many of us, now, are
going to some political lengths to unmake). What makes examples of this kind
interesting, of course, is the fact that the actual etymology has nothing to do
with such a semantic derivation. 5
On the surface of things, the word woman seems to have shared the
destiny of female. However, woman is, unlike female, truly a combination of
wo + man; woman stems from the compound wif-man(n), where man(n)
meant human being and wif (later wife) meant female. Wif-man(n) was
used to contrast with wp(n)man(n) male (originally armed) human being.6
Nevertheless, there is a problem; the word man in the construction referred
originally to a human without marked gender or age (cf. German Mann,
Mensch). The modern relationship, with its sense of a lexically unmarked
male vs. a lexically marked female, came through the historical ellipsis of
wp(n) in wp(n)man(n), which seems to have coalesced with an increasing
FOLK-ETYMOLOGY 27
use of man(n) in its specific sense, male, rather than its generic sense.7
Consequently, neither female nor woman were originally marked lexical
items contrasting with unmarked male equivalents (male and man). However,
the presence of a cultural pattern favouring maleness as unmarked has,
through the culturally based reinterpretation of Middle English femelle and
woman, where the former has changed to female by means of the process of
folk-etymology, resulted in the linguistically marked relationship between the
word pairs we find today.
In fact, the marked relation between woman and man has lately led to a
tendency to folk-etymologise woman (Class I). As shown by the quotes
below, which have been extracted from the internet, woman is occasionally
treated as stemming from *womb-man, especially in religious contexts. After
all, it is to speakers quite evident that -man in woman must come from man,
and what can wo- stem from but womb?
Our English word woman, means she-man, or female man and is a contrac-
tion from the Anglo-Saxon wombman, which means the man with a
womb.
(Two Covenants One for the Man, One for the Woman;
http://www.biblebelievers.org.au/bb940724.htm)
Male is man, female is womb-man, means a man with a womb -woman;
wombman has become woman. In God, woman is not without man and man is
not without woman.
(The Word of God, Jesus and Women;
http://www.tne.net.au/abdaacts/godjwo.html)
The hypothesis meets no phonological resistance since *womb-man [wu:m

mn] easily could become [wmn], the b in womb being conspicuously
silent. The only indication of womans true etymology, which is also the only
inconsistency in the womb-man theory, is the plural form of the word and its
pronunciation women [wmn].
The alteration of feverfew and buck bean (as well as the previously
mentioned asphalt) relied on the similarity between the words forms and
meanings, as well as on the similarity in their cultural connotations. However,
in some folk-etymologies, it is only the forms that contribute to the change.
This is often the case with loan words or foreign-sounding words where the
new form is as opaque as the old one, but sounds more English. This type of
folk-etymology can, for instance, be found in bagrag, wine produced at
Bacharach on the Rhine. The forms bagrag, backrag and backrack are all
phonetic distortions of Bacharach there is no connection between the
units of the folk-etymologised forms (that is, bag, rag, back and rack) and
Bacharach apart from the fact that the forms bagrag, backrag and backrack
sound far more English than Bacharach.
Deeper Folk Processes New Productive Derivational Affixes
Opaque words are, where the forms permit, remarkably often interpreted
as compounds or affixations consisting of two parts, as in gambone (for
gammon) or burster (Old English burstow).8 The tendency for this type of
reinterpretation can be found in other studies of analogy and folk-etymology;
Coates in his article on analogy was investigating the alterations found in
place-names. Looking at the examples of place names provided by Coates, we
can, for example, see that Crostwick has been reinterpreted as consisting of
crost + wick, rather than adhering to the etymologically correct interpretation
cross + thwaite (Old English cros, ON veit) (1987: 329). The reinterpretation
was, according to Coates, triggered by the name of a neighbouring village,
Postwick.
Similarly, names such as Watergate and Bikini have been reinterpreted in
a way that yields for the incorrectly presumed second element of the words,
that is, -gate and -kini, a (new) meaning; the new element -gate has come to
stand for an actual or alleged scandal, usually an attempted cover-up and
-kini denotes a minimal unit of a beach garment, but this element can only be
found in monokini (OED).
And that headlines about Monicagate will scream at us for months (1998,
http://www.flstarweb.com/columns/dyson/cd091298.htm)
Sunbathing.., she in a bikini.., he in a monokini (1964, OED). 9
More recent examples of this process, which we will here refer to as form
abstraction (based on the term abstracted form used by Barnhart (1980: 3)),
include the following new affixes quoted from Hargevik (1996: 230232):
-aholic (as in workaholic and leisureholic (from alcoholic)) meaning a
person who is (or appears to be) addicted to the object or activity specified in
the first element of the word
-nomics (as in Dukakinomics (from economics)) meaning the economic
FOLK-ETYMOLOGY 29
theory or policy of the person whose name constitutes the first part of the
word
-quel (as in prequel (possibly from sequel)) meaning an item (usually a book
or film) that is related chronologically, as specified by the first component of
the word, to some other already completed work
An investigation into the tendency for reinterpretation into meaningful

chunks or elements of the Class I and II folk-etymologies yielded the results
shown in Table 2. It should be noted that the folk-etymologies that were not
reinterpreted into elements were to a great extent monosyllabic or short
multisyllabic words, something that certainly would prevent that type of
reinterpretation.
Table 2. The number of folk-etymological processes where a tendency to structure words in
two or more elements can be found
word class Class I Class II
elements no elem. elements no elem.
noun 34 7 4 3
verb 3 1 0 0
adjective 2 0 1 0
= 39 =8 =5 =3
The overall number of Class II folk-etymologies is so small that conclusions

can hardly be drawn for that class except for the fact that the tendency for
reinterpretation into meaningful elements seems to be greater among the Class
I folk-etymologies. Among the Class I nouns almost 83% exhibit this ten-
dency; however, as in the case of the Class II words, the number of verbs and
adjectives in Class I is too small for any tendencies to be estimated.
In his place-name study, Coates often found a connection between the
altered name and the name of some neighbouring place. He concluded that the
analogical force causing the changes was shaped by the similarity between the
names, a similarity furthered by the change.
Yet, regardless of whether a neighbouring name or a similar sounding
word will trigger a reinterpretation, we still need to recognise that there is
certainly a tendency for humans to see multisyllabic words as being con-
structed as compounds or derivatives; and consequently when faced with an
obscure, complex-looking word, we tend to feel a need to find at least two
meaningful components in it. The usage of patterns and the search for patterns
in language is one of the most dominant forces governing language, including
the creation and reinterpretation of words.

When first encountering an opaque word that is multisyllabic, speakers
seek easily recognisable word parts. If no such already existing parts can be
found in the word, they will try to determine where and how to split up the
word into meaningful units. Similarly, if the word is found to contain one
established meaningful component, the other part of the word is automatically
treated as yet another, though still unknown, unit. That unit can, depending on
its appearance, be in the form of either an independent word or an affix.
Hence, the recognition of one meaningful unit gives life to a new cre-
ation. This is what has happened in the case of Hargeviks previously men-
tioned abstracted forms. Since the same treatment of opaque words can give
rise to new words or abstracted forms, we will treat the initial stage of the
creation of an abstracted form as part of a folk-etymological process. In a
sense, we can say that the continued usage of the abstracted forms is the best
possible confirmation of a successful folk-etymology.
As noted by Coates, many analogical acts illustrate the human mind in a
condition that is at the same time relatively free and relatively fettered (1987:
319). Among the most illustrative instances of this we can find folk-
etymologised words and phrases; speakers are fettered in that they adhere to
their shared need of patterns and units, but they still remain free in their
individual capability of creating new words and affixes, and especially ab-
stracted forms. However, folk-etymologies do not merely tell us about the way
the human mind is constructed and its craving for patterns, but also about the
way our minds structure the world around us, the way we perceive our culture
and society and also the way that culture and society are allowed to affect our
perception and language.
Underlying Linguistic and Cultural Forces
Language, as Saussure long ago noted, is a property of social groups of

speech communities rather than of individuals. The longevity of linguistic
entities and processes depends both on their ease of transmission and on their
usefulness in communication. To be successful, linguistic innovations have,
thus, to be easily learned and used by those other than their creators. The more
easily generalisable or productive they appear to be to some class of appro-
priate situations, and the more they serve to facilitate some facet of actual
FOLK-ETYMOLOGY 31
communication (whether content, attitude toward the content, or relevant

social facts), the more likely they appear to be to take hold. While we will not
explore the issue of collective reality in great depth in this paper, it is
important to our thinking (cf. Kronenfeld 1996).
Relevant to these considerations is also the invisible hand theory (cf.
Keller 1994; Nyman 1994), whether the invisible hand be seen as a direct
expression of some social constraint or indirectly as the effect of a set of
intuitions guided by some combination of our innate disposition and our
shared linguistic experience. The invisible hand theory speaks of the two
maxims of action that guide and influence our linguistic behaviour:
M1: Talk in such a way that you are not misunderstood
M2: Talk in such a way that you are understood
(Keller 1994: 94)
The two maxims are not equivalent, since being understood is not the opposite
of not being misunderstood. Simplifying matters, we can say that the two
maxims cause us, on the one hand, to avoid (that is, not to choose) one variant,
and, on the other hand, to choose another variant.
The reason why folk-etymologies occur is that people, for reasons of
clarity and efficiency, adhere to these maxims. In order to be certain that what
they are trying to communicate, which in the case of folk-etymology involves
the use of unfamiliar and confusing terms, will both be understood and not
misunderstood, speakers have two main alternatives: a) they can choose not to
use the opaque term or construction by choosing to use some other means of
expression (which might violate the need for brevity), or b) they can choose to
make sure that the opaqueness of the word is eased, which can mainly be
secured by the completion of some alteration (usually analogy or folk-etymol-
ogy) of the word.
The use of obscure and confusing words impairs language comprehen-
sion and clarity. According to the maxims, such usage should be avoided, not
promoted. Hence the folk-etymologising of words is not some kind of perver-
sion of the language as has often been claimed, but is the logical course of
action given the options at hand.
Cognitive Implications
In this paper, we have brought forth some of the often forgotten aspects of
folk-etymology in order to try to give a clearer picture of what folk-etymology
is, what triggers it, what makes it succeed or fail and whether it is such a bad
part of language. To summarise, the primary force that governs folk-etymol-
ogy (and similar processes, such as abstracted forms) is peoples constant
striving not only to make sense of whatever opaque forms they may encounter,
but also to ease the memory load in having to remember such opaque forms, a
task that would both ensure understanding and enable future usage.
Notes
1. It should be stressed that not only the pronunciation of the two words facilitated the
reinterpretation of asphalt as ashfelt, but it is likely the similarity in stress pattern also
enabled the reinterpretation; asphalt was at the time, according to the OED, stressed as
asphalt, thus, making it possible to perceive the word as a compound consisting of two
words ash and felt; today the word is stressed as asphalt.
2. On the distinction between triggers/causers of change, results of change and the difficul-
ties and necessity in making that distinction see Rundblad 1998: 2233.
3. It should be noted that Du vrijbuiter is actually a compound made up of the Dutch
equivalents of free + booter (we would like to thank Olga Fischer for pointing this out to
us). Due to this fact, freebooter could turn out to be a loan translation rather than a loan
word, especially since free and booty (or boot) already existed in English. However, the
Dutch dominated the naval scene in Europe during the 16th and early 17th century (later
both the Dutch and the English freebooted the Spanish), and it is generally accepted that
relevant terminology is usually borrowed together with the technological invention,
whatever form that invention may be in, which would clearly favour freebooter as a loan
word. There is also the fact that the word booter did not exist independently; it is
supposedly a shortening of freebooter, but according to the OED booter can only be
found in combination with free-. Additionally, the earliest form of freebooter was
frebetter, which shows no association with booter (OED). This would certainly favour
the word first being borrowed from Dutch, but later being reinforced, possibly, by an
additional loan translation of the very same Dutch word or by the possibility of creating
the word afresh using the already existing components free, booty/boot and er.
4. The last two names, bog nut and bog trefoil, are alternative names for buck bean.
5. In this case, the semantic and cultural information we glean is that, during the period in
which these words were taking on their modern form (especially gaining their apparent
marking relationship), there was a strong asymmetry in male/female relations, and that
women were in some public or political sense relatively insignificant. In the case of
female, we also have other kinds of historical information supportive of the semantic
FOLK-ETYMOLOGY 33
marking relations (though not of the presumptive derivational relationships) such as the
semantically similar but historically distinct relationship in the productive use until
recently in modern English of the ess ending to mark a female version of some
presumptively male category (as in tigress, lioness, poetess, actress, Negress, Jewess,
and so forth). The convergence in our example does not pertain to all sets of words that
contrast males and females (cf. father vs. mother) but to those which code sex differences
in the general public sphere.
6. The OED lists several instances of the pair, including the following early ones: Ne scride
nan wif hip mid wpmannes reafe, ne wpman mid wifmannes reafe (circa 1000);
Forbearn eall meast se burh of Lincolne & micel unperime folces wpmen & wimmen
forburnon (1123); Pif a lape weren nu, nalde na mon mis-don wi ore, ne wepmon ne
wifmon ne meiden (circa 1175). It should be noted that the first element in wifman(n),
that is wif, was also used for a female human being, and that wif was paired with Old
English wer, as in eos ylce wyrt pede t per pe wera pe wifa feax wexe (circa
1000). (OED)
7. On the issue of wif-man(n) and wp(n)man(n) and the process of ellipsis acting as a
trigger of semantic change see Rundblad 1998: 2227.
8. See also the previously discussed words bagrag, backrack etc.
9. Note that monokini is not frequently used any more, but seems to be more and more
restricted to use either in contrasting contexts with bikini or in swimsuit brand names.
References
Ardener, Edwin. 1971. Social Anthropology and the Historicity of Historical Linguistics.
Social Anthropology and Language ed. by Edwin Ardener, 209241. London, New
York: Tavistock Publications.
Barnhart, C.L. 1980. The second Barnhart dictionary of new English. Bronxville: Barnhart
Books.
Coates, Richard. 1987. Pragmatic sources of analogical reformation. Journal of Linguis-
tics 23. 319340.
Hargevik, Stieg. 1996. Trends In English Grammar Vocabulary & Pronunciation. Oxford:
Utbildningsstaden AB/UBS-Sweden.
Hock, Hans Henrich. 1991. Principles of Historical Linguistics. Berlin, New York: Mouton
de Gruyter.
Keller, Rudi. 1994. On Language Change. The invisible hand in language. London, New
York: Routledge.
Kronenfeld, David B. 1996. Plastic Glasses and Church Fathers. Semantic Extension from
the Ethnoscience Tradition. New York, Oxford: Oxford University Press.
McMahon, April M.S. 1994. Understanding Language Change. Cambridge: Cambridge
University Press.
Nyman, Martti. 1994. Language Change and the Invisible Hand. Diachronica 11: 2.
230258.
Oxford English Dictionary. 1992. 2nd edition on CD-ROM. Oxford: Oxford University
Press.
Palmer, A.Smythe. 1890. Folk-Etymology. A Dictionary of Verbal Corruptions or Words

Perverted in Form or Meaning, by False Derivation or Mistaken Analogy. New York:
Henry Holt and Company.
Rundblad, Gabriella. 1998. Shallow Brooks and Rivers Wide A Study of Lexical and
Semantic Change in English Nouns Denoting Watercourse. (Acta Universitatis
Stockholmiensis LXXXVIII). 2nd edition. Stockholm: Almqvist & Wiksell Interna-
tional.
Scholfield, Phil. 1988. Documenting Folk Etymological Change in Progress. English
Studies 69. 341347.
Scholfield, Phil. 1992. Folk Etymology and the Parallelism of Lexical Processes. Bangor
Research Papers in Linguistics. 4. 113.
Winer, Lise. 1992. Folk Etymology in Trinidad and Tobago Lexicography. Old English
and New. Studies in Language and Linguistics in Honor of Frederic G. Cassidy ed. by
J.H. Hall, N. Doane and D. Ringler, 238253. New York, London: Garland Publishing
Inc.
Mechanisms of Semantic Change in Nouns
of Cognition: a General Model?
Pivi Koivisto-Alanko
University of Helsinki
Abstract
This paper deals with subjectification in abstract nouns denoting cognition. The
focus is on the semantic change of the noun wit and its near synonyms in Late
Middle English and Early Modern English. The process of semantic change in
wit is analysed with prototype theory and then compared with studies of
subjectification and grammaticalization by Elizabeth Closs Traugott and Eve
Sweetser. The changes in the prototypical meaning structure of wit are triggered
by increased polysemy and an influx of overlapping loanwords. Both wit and
its near synonyms are shown to display signs of directionality towards in-
creased subjectivity in the process of semantic change.
Introduction
This article concerns the directions and mechanisms of semantic change in

abstract nouns denoting cognitive processes, with special reference to subjec-
tification. This vocabulary of cognition could also be termed the semantic
field of wit, the noun wit being my starting point, because it was a representa-
tive noun of cognition in Middle English (henceforth ME) but later developed
toward its modern meaning through a varied and intriguing process of change.
The polysemy of wit has been discussed earlier by linguists such as Charles
Barber (1976: 145147), William Empson (1979: 84100, 391396),
Manfred Grlach (1991: 197, 205207) and C.S. Lewis (1960: 86110). I
36 PIVI KOIVISTO-ALANKO
focus on Late Middle English (henceforth LME) and Early Modern English
(henceforth EModE) during which most of the changes took place, since I
intend to consider the fundamental changes in the structure of wit, and the
effect its increased polysemy and the influx of loanwords had on the whole
semantic or lexical field (for preliminary studies of wit and its near-synonyms,
see Koivisto-Alanko 1997 and forthcoming a).
The concept of a semantic field is very fuzzy and I have decided to limit
my study to the words with meanings synonymous or nearly synonymous with
wit at any point during the period studied. I use the term semantic field of wit to
discuss the diachronic semantic development of the noun wit by relating it to
words which have been partial synonyms of wit at some point of its existence.
There are stricter definitions of a semantic field (see Lipka 1990: 151152),
but since I have to operate both in diachrony and in synchrony, I have decided
to follow Adrienne Lehrers definition of a set of lexemes which cover a
certain conceptual field (Lehrer 1985: 283). Semantically related words have
been shown to undergo similar semantic changes (Lehrer 1985: 286) and my
purpose here is to consider reasons for such developments.
At the beginning of the period (in 1350) there were four or five important
words denoting cognition: wit, understanding, knowledge/knowing and mind.
Of these mind remained fairly stable, but the others were influenced by
polysemy and borrowing. The later arrivals were conscience, cunning,
conceit, intelligence, intellect, engine, ingeny, sense and ingenuity. Genius,
fantasy, fancy, imagination and invention also overlap to some extent with wit.
My study is partly corpus-based and the sources I have used are the Helsinki
Corpus of English Texts (HC), the Oxford English Dictionary (OED) and the
Michigan Early Modern English Materials corpus (http://www.hti.umich.
edu/dict/memem/). The examples are from the Helsinki Corpus unless other-
wise stated. The HC examples have a code which facilitates finding the
excerpt in the corpus. These codes can all be found in Merja Kyts Manual to
the Diachronic Part of the Helsinki Corpus (1996) which contains a list of
texts included in the corpus. I have, however, included also a general refer-
ence to the text (or the writer) to make reading easier.
The main objective of this study is to look at the changes in wit and in its
near-synonyms from a theoretical point of view. I have used prototypes to
analyse the meaning structure of wit at different points in time. The changes in
its prototypical structure exhibit some interesting tendencies which point to
increasing subjectification. Those changes seem to have some similarities
MECHANISMS OF SEMANTIC CHANGE 37
with the models of directionality used in the context of grammaticalization. A

lexical word such as wit does not become grammaticalized, but the changes
nevertheless have some striking resemblances to the grammaticalization pro-
cesses discussed by Elizabeth Closs Traugott and Eve Sweetser. When I
investigated other members of the semantic field, I discovered more similari-
ties and some interesting aberrations. I shall present here the case of the
directional change in wit, also offering some examples of its near-syn-
onyms. I shall go on to discuss the possibility of discerning some general
tendencies of semantic change, which could lead to a very general model
applicable to both lexical and grammatical change.
Prototypicality
Prototypes are fairly rarely used in historical linguistics, but I have found them
to be a very useful tool in categorising polysemous nouns. Dirk Geeraerts has
used them in his historical studies (see for example Geeraerts 1997), and
Xavier Dekeyser has discussed the loss of prototypical meaning in his 1998
article. What follows is the view of prototypes which has underpinned my
analysis of the semantic change in wit. For a good all-round account of
prototype theory, see Taylor (1989).
PROTOTYPES
B. ONE MEMBER GROWS
A. CORE AND PERIPHERAL MEMBERS
C. TWO OVERLAPPING CENTRES

(SAME WORD)
Figure 1. Prototypes formation of new prototypical centres

Figure 1 shows a word a category which has one prototypical

centre with a core meaning and several peripheral meanings. Some of the
peripheral meanings are close to the core and naturally it is sometimes
difficult to say whether a particular use is a core meaning or whether there is
something that separates it from the core (see A in Figure 1). Some meanings
can be far removed from the core. In B we have one peripheral meaning which
has begun to grow in importance and will perhaps further distance itself from
the core. At this point it is still categorised within the same prototypical centre.
In C we have the next stage in the development: two overlapping prototypical
centres, each of which has a core and (though not shown here) possibly several
peripheral members. They still belong to the same category, and the user can
perceive a connection between them, but may not always be able to tell under
which prototype a meaning should be categorised. There are, however, two
clearly distinguishable prototypical centres or different meanings of the same
word even though the boundaries of the categories are fuzzy.
Prototypes can be determined by several criteria, but the most significant
ones are frequency and generality (by generality I mean here the amount of
closely related meanings or meaning nuances one sense or prototypical centre
can be seen to cover). In historical semantics frequency can only be detected
from texts and editions or, more easily, from historical corpora. Generality is
determined mainly with historical dictionaries but naturally also by studying
the contexts of the word studied.
The Prototypical Structure Of Wit
The three central meaning clusters of wit may be schematised with prototypes.
The quantitative data for determining the prototypes has been collected from
the Helsinki Corpus (see Figure 2, below). The noun wit has two prototypical
centres at the beginning of the period studied (1350): I call these COGNI-
TION and PERCEPTION. PERCEPTION includes the five senses meaning
and COGNITION the mind, seat of consciousness types of meaning (see
examples 1 and 2).
(1) [PERCEPTION]
For, when any es tornede to delite of hys fyve wittes, alsonne
vnclennes entyrs in-to his saule. (Rolle RTREAT 16)
(2) [COGNITION]
And thenne they enteryde to the cytte of London as men that hadde
ben halfe be-sydde hyr wytte; (Gregorys Chronicle CHRLOND
11)
The PERCEPTION meaning disappears by EModE, being replaced by sense,
and a new centre is formed around the same time which I call EXPRESSION
(see example 3).
(3) [EXPRESSION]
Satyr and invective are the easiest kind of wit.
(Tillotsons Sermons TILLOTS II:ii 429)
This corresponds quite well to the most common modern meaning of wit.
EXPRESSION is formed out of COGNITION by a process in which the
COGNITION centre keeps growing and developing new meanings (denoting
superior intelligence etc.). At this time (Late Middle English) most of the
near-synonyms of wit enter the semantic field, and, most notably, the PER-
CEPTION meaning is replaced by sense. For quantitative data from the
Helsinki Corpus, see Table 1 and Figure 2. In EModE COGNITION becomes
less frequent and less polysemous and, at the same time, the EXPRESSION
centre grows.
Table 1. Quantitative distribution of the prototypical centres of wit, based on the Helsinki
Corpus
PERCEPTION COGNITION EXPRESSION
13501420 20 62
14201500 5 38
15001570 8 29 2
15701640 34 2
16401710 13 26
PERCEPTION
LME
COGNITION
EModE
EXPRESSION
Figure 2. Distribution of the prototypical centres of wit in LME and EModE

The three prototypical centres are the referents through which the abstract
noun wit is determined in diachrony. The semantic field of wit is formed by
words whose meanings overlap with any of the prototypical centres at any
point of the period studied.
Comparison with Models of Directionality
It is probable but impossible to prove, due to the scarcity of material that

the PERCEPTION meaning is the earliest one. Wit has its roots in the Proto-
Indo-European I have seen, which would indicate that the sensory meaning
is the oldest. The development would then have gone from PERCEPTION to
the coexistence of PERCEPTION and COGNITION and from there to the
coexistence of COGNITION and EXPRESSION (I have discussed this pro-
cess in detail in Koivisto-Alanko forthcoming b). This agrees with Xavier
Dekeysers view that loss of old core meanings is common, while peripheral
meanings tend to develop further (1998: 68). The change from PERCEPTION
to COGNITION is the basic semantic change from concrete to abstract. Eve
Sweetser argues that this is a metaphorical shift motivated by what she calls
the vision/intellection metaphor (seeing being the primary source of informa-
tion for humans) (1990: 3740). The PERCEPTION centre would, in
Sweetsers terminology, be rooted in the sociophysical world. In addition to
the five senses, PERCEPTION also signifies the five inner wits (common
wit, imagination, fantasy, estimation, memory) where the level is much more
abstract. In the OED, inwit means roughly the same as wit, without the
superior meanings, whereas outwit means faculty of observation or percep-
tion; an external sense (OED, s.v. outwit). The COGNITION centre clearly
exists on the abstract logical or epistemic level. The EXPRESSION centre
is certainly not very far from Sweetsers speech act level, although the term
speech act is overly restrictive. The inner logic the expression of proposi-
tions in the real world is, however, very close to what the EXPRESSION
meaning of wit really signifies. The result of this third (pragmatic) shift, from
COGNITION to EXPRESSION, is actually more concrete than the previous
one, from PERCEPTION to COGNITION.
These changes can also be fitted quite well into Elizabeth Closs Traugotts
models of directionality (Traugott 1989). The change from external described
situation I have seen to the internal (evaluative/perceptive/cognitive)
situation (Tendency I) corresponds with the development of the COGNITION

centre out of PERCEPTION. Traugotts next tendency (Tendency II) is the
change to a textual = cohesive (metalinguistic) situation (for example, mental
to speech act verbs). The EXPRESSION centre is formed gradually, new
meanings denoting quality (instead of the older ones denoting faculty or ability)
taking shape inside the COGNITION centre. When wit starts to signify the
ability to express those qualities (quickness of intellect, etc.), the EXPRES-
SION centre is formed. Here the word has a more active meaning than
previously, being concerned with the situation in which the quickness of
intellect of a person is exhibited. Of course a noun cannot be a cohesive factor
in a situation the same way a verb would be, but is there not a connection? More
broadly, this can also be seen as an evaluative and thus increasingly subjective
meaning. In her earlier work in particular (see for example Traugott 1989 and
1990), Traugott has given examples of tendencies leading to increasing subjec-
tification in changes not involving grammaticalization as well. The develop-
ment of wit, an abstract noun, conforms to this.
Some rarer meanings of wit seem to fall into this pattern as well (espe-
cially if the term metalinguistic is used). Wit can signify meaning, sense
(see example 4) (PERCEPTION) and opinion (see example 5) (COGNI-
TION). The EXPRESSION meaning of wit is not wholly without parallel,
either: conceit develops the sense a fanciful, ingenious or witty notion or
expression (OED 1513).
(4) [Th]e secounde witt is allegoryke. (c1380 Wyclif: Selected Works,
OED s.v. wit)
(5) What is Zoure wit? how thenke Zow? (c1400 Laud Troy Bk. 8135,
OED s.v. wit)
Further developments in the EXPRESSION centre lead to what C.S. Lewis in
Studies in Words called the Dangerous Sense (1960: 97) (see example 6),
where wit is interpreted negatively.
(6) To take men off from this impious and dangerous folly of prophane-
ness which by some is miscalled wit. (Tillotsons Sermons TILLOTS
II:ii 427)
Traugott speaks of a tendency for meanings to become increasingly
situated in the speakers subjective belief-state or attitude toward the situa-
tion (Tendency III). The negative overtones of wit (EXPRESSION) are
results of subjective interpretation, since it is, after all, often a matter of

opinion whether someone is being witty or insolent (pejorative/ameliorative).
It is fascinating to see that similar things happen with other words within the
same semantic field: conceit was a neutral term to begin with, its development
going from conception, idea, thought, through personal opinion, judge-
ment, fancy to an overweening opinion of oneself (= self-conceit) (see
example 7a-d).
(7) a. .. to replye to the seide answeris, and to the furste most specyally,
whiche is derke to my conceyte as yet; (Letters and Papers of
John Shillingford SHILLET 17)
b. .. preste [th]at I spoke to you of, and tolde hym my conceyte
howe he shal be demened in brekynge with my seid Mastres;
(Thomas Mull: The Stonor Letters and Papers TMULL I, 127)
c. .. the fancies and dreams of a company of melancholy men,
who were weary of the world and pleased themselves with vain
conceits of happiness and ease in another life. (Tillotsons
Sermons TILLOTS II:ii 430)
d. Allowing him a breath, a little scene,
To monarchize, be feared, and kill with looks,
Infusing him with self and vain conceit,
(Tragedy of Richard II 1595:1, lines 160162)
Cunning also is only known today in its negative sense, though in the
14thC the sense wit, wisdom, intelligence (OED 1340) was more common
(though artfulness, skill in deceiving (1325) was extant, too).
Subjectification in the Semantic Field of Wit
The development I here discuss is not grammaticalization as such, and al-

though pragmatization might be a slightly better term, it is not adequate either
in the case of content words. The direction from concrete or sensory to
abstract or cognitive to expressive and subjective is apparent, however. Sub-
jectification is probably the most descriptive term for the processes discussed
in this paper, but there is another important process affecting the semantic
field of wit. Metonymy also is very strongly present from the beginning and its
effect becomes more pronounced when we look at the whole semantic field. If
wit has its roots in a verb of seeing, what is the five senses meaning if not
metonymy? (see example 1). Personification, a form of metonymy, is com-
mon in both COGNITION and EXPRESSION, as in a person being called a
Wit (see example 8).
(8) Boileau among the French, and Cowley among the English Wits,
were those he admired most. (Gilbert Burnet: Some Passages of the
Life And Death of the Right Honourable John, Earl of Rochester
BURNETROC 8)
A witty story could also be called a wit. Here, it is the product of the
ability being described. A similar thing happens with other members of the
field: cunning has the sense skill, device, ingeny (a rare word, synonym of
wit, as in example 9) invention, clever contrivance.
(9) whiche Iniunction & comau[n]dement he accomplysshed & brought
to conclusion by his subtyl & industryous ingenye & wysedom
(Skelton Diodorus (tr.), bk. V, fol. 249r anted. royal soc. Canada;
c1485 University of Michigan Early Modern English Materials)
Engine is a case apart, since the earliest documented senses mean prod-
uct of ingenuity, machine and the later ones but not much later, about
eighty years (1300 - 1386) native talent, mother wit (see example 10).
Engine was borrowed from Old French (henceforth OF) where it already had
both abstract and concrete senses, so there is a clear pattern of development to
be seen in ME. On the other hand, the OF engin comes from the Latin
ingenium, which means natural talent and something close to mother wit;
the skill, invention and machine senses have developed in OF, thus
following the same pattern.
(10) But considere wel that I ne usurpe not to have founden this werk of
my labour or of myn engyn. (A Treatise on the Astrolabe ASTR
662.C2)
Ingenuity (example 11) has the sense ingenious device or contrivance;
artifice (1650); fancy also means inventive design; invention. Intelligence
means information, news, communication, knowledge means informa-
tion, notice, science craft, trade or occupation requiring trained skill.
Conceit also has a concrete, metonymous sense, developed from the fancy,
imagination senses: a fancy article, a fancy trifle for the table.
(11) Ingenuitie! I see his ignorance will not suffer him to slander her,
which he had done most notably, if he had said wit for ingenuitie!
(Ben Jonson; OED s.v. ingenuity)
Several of the near-synonyms also have a tendency to develop senses
signifying meaning (intellect (example 12), conceit, understanding, sense),
some also opinion (conceit, mind).
(12) Which verbe dothe accorde with the intellecte or significacyon &
not with the voyce. (Whitington Vulg. 1520; OED s.v. intellect)
The effect of the movement in the semantic field on the structure of wit
can be seen in Figure 3.
Figure 3 shows that during LME there was an influx of loanwords which
affected the structure of wit. The major changes take place directly after that,
at the beginning of the Early Modern period. It must be noted that during LME
there was also an increase in the meanings of wit, so that we can say that the
P CONSCIENCE (1325)
C
CUNNING (1340)
LME
CONCEIT (1374)
INTELLIGENCE (1380)
INTELLECT (1386), ENGINE (1386)
INGENY (1477)
E
SENSE (1526)
INGENUITY (1599)
EModE
Figure 3. Influx of loanwords and their effect on the prototypical centres of wit in LME and
EModE. The EXPRESSION centre is formed after the balance of the field has been
disturbed by loanwords
P = PERCEPTION, C = COGNITION, E = EXPRESSION
entire semantic field became really loaded. All the new arrivals, the loan-
words, have of course a narrower semantic range than wit and the other old
native words. It is rare for a word to be borrowed with several different
meanings (the way engine was; those meanings did not last very long).
Ingenuity and ingeny came later to replace the more abstract meanings of
engine. In addition to what is seen here, understanding, knowledge and mind
were basic and important members of the semantic field, and their position
was not much shaken by the arrival of new words, even though intelligence
and science did make inroads into understanding and knowledge.
Wit is the only one to undergo a substantial structural change. This kind of
semantic field has to have members both for basic cognitive functions (mind,
understanding) and for more specialised functions. Wit was stretched to cover
these new specialist functions, but the loanwords took over, affecting the
balance of the field more deeply. It is impossible to understand what happened
to wit without taking into account the loanwords and why they were needed
(the extralinguistic factors). Thus the semantic field actually shows the whole
development of wit (see Figure 4): the old meanings have not become redun-
dant even though they have disappeared from under wit (the five wits the
five senses). The division is based on the prototypical structure of wit; the
figure would naturally look different if it had been modelled on imagination,
for example.
LME
PERCEPTION COGNITION
MIND WIT UNDERSTANDING

WIT / INWIT /
OUTWIT KNOWLEDGE MEMORY
SCIENCE CUNNING
CONSCIENCE
CONCEIT
INTELLIGENCE EXPR
INTELLECT ENGINE
EMODE WIT
SENSE CONCEIT
INGENY
INGENUITY
FANCY
IMAGINATION
(INGENUITY) FANTASY
GENIUS
GENIUS
Figure 4. The semantic field of wit in LME and EModE

External factors always determine what words are needed; basic words
for the faculty of thinking are a constant, but in the late 20th century we need
more words to describe artificial intelligence, whereas words describing the
soul are not in such demand. A semantic field lives in real time, and changes in
a particular word never happen in a vacuum: there is always some trigger and
although the methods of change are those of the internal engine of language,
the reasons for starting the engine are external.
Models of Subjectification
Metonymy
So far we have seen that there is a certain directionality in semantic change,

even with content words. The next question is what happens when such words
reach the subjective level? More specifically, what is the importance of
metonymy? If the development went from the concrete I have seen via I
understand to understanding, wit, intelligence and continued towards more
subjective meanings (including, in the case of wit, the EXPRESSION sense),
what then is the next stage? From my evidence, it seems possible that it is the
return to the world of the concrete, personified meanings always being later
arrivals within prototypical centres and words such as engine apparently
having a tendency to develop thoroughly concrete meanings. In fact, if con-
crete things are used (through metaphor) to describe abstract things for which
there is no word (seeing equals understanding; compare Finnish tiet know
from tie road, way), then there is also a need to create words for the concrete
representations of an abstract ability or activity (such as thinking or creativ-
ity): a person who thinks, an intelligent person, a product of mental labour.
These are formed through metonymy. The metonymical change is similar to
those discussed by Beatrice Warren (1998: 308309): there is an obvious
representational relation between the ability and its product.
Towards a Model?
If we look at the development patterns in wit and its near-synonyms, we have

to wonder whether there is a model, some regularity, behind all this. As we
have seen, there are several valid ways of describing the process, the last one
being an attempt to combine all those above in order to describe abstract nouns
(Figure 5).
CONCRETE ABSTRACT
SOCIOPHYSICAL ABSTRACT LOGICAL EXPRESSIVE / PRAGMATIC
(Sweetser)
EXTERNAL DESCRIBED SITUATION INTERNAL (EVALUATIVE /
PERCEPTIVE / COGNITIVE) DESCRIBED SITUATION TEXTUAL /
METALINGUISTIC SUBJECTIVE ATTITUDE / BELIEF-STATE
(Traugott)
CONCRETE / SENSORY ABSTRACT (PERCEPTIVE / COGNITIVE)
METALINGUISTIC / EXPRESSIVE SUBJECTIVE (often
NEGATIVE) PRODUCT / PERSONIFIED ?
Figure 5. Models of semantic change
This is the general pattern, but it is not always followed through; often
only parts of it are realized, some stages may recur, and even if one prototypi-
cal centre of an abstract noun achieves the whole chain, things may go very
differently in another centre. The role of personification and other metony-
mous changes remains unclear. Extralinguistic factors affect the development
considerably, slowing it down in one word and accelerating it in another.
There may be movement backward and forward, again depending on the
situation. A word may, for example, lose its negative connotations when
social norms change. An example of this is the loss of the negative senses of
the noun enthusiasm (Hanks 1998: 152156). In fact, with nouns the thesis of
unidirectionality may not be as generally valid as with grammatical items,
because nouns are generally much more mobile (see also Hopper and Traugott
1993: 126129). Meanings can make little loops especially between varying
degrees of subjectivity and negativity (which often seems to occur in already
subjectified senses). However, on the scale of an entire semantic field, the
changes do follow this (uni)directional pattern towards increased subjectifica-
tion. Traces of the pattern are discernible also in the famous study by Jost Trier
of Middle High German terms of knowledge (although his starting point was
so different that drawing parallels is rather dangerous): list develops negative
(more subjective) meanings while wsheit loses most of its meanings and
develops religious and mystical, not altogether neutral, meanings (Ullmann
1962: 248249; Lehrer 1985 : 284).
External factors (on the scale of social and political changes in the
extralinguistic environment) cannot be overlooked. Their role is both a trig-
gering and a guiding one, and the changes they set in motion may have a
drastic effect on the structure of the semantic field. This effect cannot really be
called disturbance, since external factors are a natural and inevitable part of
semantic change. Context is naturally crucial for every individual change,
directing the changes on the micro-level. On the macro-level, the shifts in the
balance of an entire semantic field seem loosely to follow extralinguistic
changes (which play a role in setting the contexts in which individual words
are used). External factors do not, however, affect the actual mechanism of
change. The pattern of increased subjectification prevails at the level of the
semantic field, though the process may be stunted in the case of an individual
word (for example the arrival of a new loanword may render an old word
obsolete, but the meanings are carried on by the new word and continue to
develop new nuances with the mechanism already in use).
The pattern is also a little oversimplified. There are actually several levels
of semantic change at work simultaneously. If we think about wit, for ex-
ample, the formation of the EXPRESSION centre was only achieved through
an infinite number of smaller changes, some so tiny that they did not have a
lasting effect. The pattern toward more subjective meanings is the same, but it
is repeated at every level, every time a new meaning or meaning nuance is
born. The pattern which produced the EXPRESSION centre also produced
one miniature change which did not have a lasting effect on the big picture,
though it has been documented: Pope and his contemporaries called wit the
essential gift of the poet where we would say genius, but Popes interpreta-
tion was a tiny move in a more specialised direction (Lewis 1960: 106). I
would like to call this infiniteness of change at every level semantic fractals.
Figure 6 shows how, when one change is analysed, it is seen to have been
produced by a similar pattern of changes towards increasing subjectification.
1 2 3
1 2 3
Figure 6. Semantic fractals. The pattern of change is repeated on every level, inside a
semantic field, a lexeme and a prototypical centre.
Within the noun class the metonymous change (personification or the

product of an activity or an ability) seems to be the end of the line, like little
twigs in a big tree, in the sense that the change does not seem to go any further
within the noun. All the greater shifts happen by metaphorical means, moving
from concrete to abstract, from sensory (sociophysical) to perceptive/cogni-
tive/evaluative (especially evaluative consider the superior intelligence
meanings) to expressive and increasingly subjective. The metaphorical
change happens at every level; it could perhaps be said that whenever some-
body makes a novel use of a word, it is a more specific use this seems to be
the case at least with the vocabulary of cognition. But of course words such as
wit start out as very general, and thus probably cannot be generalised at
another (later) level, so that the pattern created somewhere in the distant past
with the first metaphorical shifts (I have seen I know, I understand) is
repeated at every level of the word (and the meanings are in constant move-
ment). Whenever there is a need to create a more concrete word out of an
abstract one, it is done by metonymy. This can happen at every stage of the
development (there are several metonymous meanings within each prototypi-
cal centre of wit) and this does not stop metaphorical change from taking place
(the EXPRESSION centre is very much later than the five wits). In this
sense, metonymic change is not the end result of the whole process of change,
but rather the big picture is something like Figure 7.
CONCRETE ABSTRACT SUBJECTIVE
M E T O N Y M Y
Figure 7. Direction of semantic change in abstract nouns with the effect of metonymy
Metonymy is, of course, especially relevant with abstract nouns. In the

case of wit, the changes can be compared with the general model as in Figure
8. Both Traugotts and Sweetsers models could be seen to fit in very nicely
(and indeed, they describe the same process).
COGNITION
VISION
PERCEPTION EXPRESSION
CONCRETE ABSTRACT SUBJECTIVE
M E T O N Y M Y
Figure 8. The semantic change in wit as evidence of subjectification
Conclusion
To conclude: could there be a common model for grammaticalization (or

pragmatization) and semantic and lexical change in general? I think that there
is a lot of evidence of directionality from less to more subjective, as we have
seen in the case of wit and its near-synonyms. It is also evident that the study
of semantic change needs tools such as prototypes and material such as
historical corpora to reliably document those directions. Even though wit is a
very pretty showcase word, studying just one word cannot tell us anything
reliable, since external factors often play havoc with regularity in individual
cases. And they are meant to: it is only the entire semantic field, however one
defines it and however fuzzy it remains, which can show us more general
tendencies. The change goes from the concrete towards the abstract, and from
the purely abstract towards the increasingly subjective evaluation. Basic
meanings do not disappear from the semantic field, though they may well be
lost in the case of individual words. The pattern of change, or, more often,
parts of it, is also repeated in the formation of new meanings inside the
prototypical centres (semantic fractality), so the picture looks the same,
whether we look at one meaning nuance with a microscope (micro-level) or
whether we look at the movement of the whole semantic field (macro-level).
For abstract nouns, metonymy is all-pervasive and extremely functional, but
the larger pattern of change uses metaphorical means.
References
Adamson, Sylvia, Vivien Law, Nigel Vincent & Susan Wright, eds.1990. Papers from the
5th International Conference on English Historical Linguistics. Amsterdam Studies in
the Theory and History of Linguistic Science. Series IV. Current Issues in Linguistic
Theory. Amsterdam & Philadelphia: John Benjamins.
Barber, Charles. 1976. Early Modern English. London: Deutsch.
Dekeyser, Xavier. 1998. Loss of prototypical meanings. Hogg & Van Bergen. 6371.
Empson, William. 1979. The Structure of Complex Words. London: Chatto and Windus.
Fisiak, Jacek, ed. 1985. Historical Semantics. Historical Word-formation. Trends in Lin-
guistics. Studies and Monographs; 29. Berlin: Mouton de Gruyter.
Fontenelle, Thierry, Philippe Hiligsmann, Archibald Michiels, Andr Moulin & Siegfried
Thiessen, eds. 1998. Actes Euralex 98 Proceedings Vol. I. Lige: Universit de Lige,
Dpartement danglais et de nerlandais.
Geeraerts, Dirk. 1997. Diachronic Prototype Semantics : A Contribution to Historical
Lexicology. Oxford: Clarendon Press.
Grlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: Cambridge
University Press.
Hanks, Patrick. 1998. Enthusiasm and condescension. Fontenelle et al.1998. 151166.
Hogg, Richard M. & Linda Van Bergen, eds. 1998. Historical Linguistics 1995. Selected
Papers from the 12th International Conference on Historical Linguistics, Manchester,
August 1995. Vol. 2: Germanic Linguistics. Amsterdam Studies in the Theory and
History of Linguistic Science. Series IV. Current Issues in Linguistic Theory.
Amsterdam & Philadelphia: John Benjamins.
Hopper, Paul J. & Elizabeth Closs Traugott. 1993. Grammaticalization. Cambridge: Cam-
bridge University Press.
The Illustrated Stratford Shakespeare. 1982. London: Chancellor Press.

Koivisto-Alanko, Pivi. 1997. The vocabulary of cognition in early English translations of
Boethius from Chaucer to Preston. Neuphilologische Mitteilungen. 1997:4. 397414.
Koivisto-Alanko, Pivi. (forthcoming a). Cognitive loanwords in Chaucer: is suprastan-
dard nonstandard?. Dimensions of Nonstandard writing. Symposium on Nonstandard
English.
Koivisto-Alanko, Pivi. (forthcoming b). Patterns of Semantic Change in Abstract Nouns:
the case of wit. Proceedings of the 2nd International Conference in Middle English.
(Helsinki May 29th June 1st 1997).
Kyt, Merja, ed. 1996. Manual to the Diachronic Part of the Helsinki Corpus of English
Texts: Coding Conventions and Lists of Source Texts. Helsinki: University of Helsinki.
Lehrer, Adrienne. 1985. The influence of semantic fields on semantic change. Fisiak
1985. 283296.
Lewis, C.S. 1960 (1996). Studies in Words. Cambridge: Cambridge University Press.
Lipka, Leonhard. 1990. An Outline of English Lexicology. Tbingen: Max Niemeyer.
Michigan Early Modern English Materials available at http://www.hti.umich.edu/dict/
memem/.
The Oxford English Dictionary. 1989. Oxford: Clarendon Press.
Sweetser, Eve. 1990. From Etymology to Pragmatics: Metaphorical and Cultural Aspects
of Semantic Change. Cambridge: Cambridge University Press.
Taylor, John. 1989. Linguistic Categorization: Prototypes in Linguistic Theory. Oxford:
Clarendon Press.
Traugott, Elizabeth Closs. 1989. On the rise of epistemic meanings in English: an example
of subjectification in semantic change. Language 65, number 1. 3155.
Traugott, Elizabeth Closs. 1990. The unidirectionality of semantic change. Adamson et
al.1990. 497517.
Ullmann, Stephen. 1962. Semantics. An Introduction to the Science of Meaning. Oxford:
Blackwell.
Warren, Beatrice. 1998. What is metonymy? Hogg & Van Bergen 1998. 301310.
Historical Semantics and Historical
Lexicography: will the twain ever meet?
Christian J.Kay
Abstract
Semantics and lexicography are often viewed as twin disciplines, both being
concerned with meaning and its expression, yet their relationship has gener-
ally been an uneasy one. Ironically, semantic theories which have had most to
offer lexicography, such as various types of lexical feature analysis, have been
least regarded within semantics itself. The development of the two subjects
will be briefly considered as a contributory factor in this situation. For histori-
cal semanticists, the pursuit of word meaning is often a primary concern, since
other kinds of meaning cannot be elucidated if word meaning remains elusive.
The development of cognitive semantics, with its emphasis on prototypicality
and meaning clusters, offers solutions to problems of lexical indeterminacy.
These are especially important in historical semantics, where indeterminacy
has both synchronic and diachronic impact, the natural vagueness of meaning
being compounded by its elusiveness through time.
Introduction
This paper is written from the point of view of a semanticist who is also a
working lexicographer, and one who has suffered frustration over the years
from a lack of connexion between the two activities. These two areas of
human endeavour have a natural affinity, yet the degree of cross-fertilisation
between them has been depressingly slight. To semanticists, lexicography
54 CHRISTIAN J. KAY
often appears largely and lamentably untheorised, uneasily poised between

the academic and commercial worlds. To lexicographers, on the other hand,
semantics may seem a remote, abstract and even frivolous discipline, with
little to contribute in the way of practical solutions. Dictionaries often merit
little more than a passing glance in handbooks of semantics, while semantic
theory is rarely mentioned in dictionary prefaces.
This situation can be related to both the history of lexicography and the
history of semantics. Lexicography is an ancient craft, but a relative new-
comer to the pantheon of academic disciplines. Despite the flourishing of
philology in the nineteenth and early twentieth centuries, semantics was
famously neglected in the narrow context of post-Bloomfieldian western
linguistics, as having data that were not amenable to scientific analysis. It
subsequently developed as a largely parasitic discipline, drawing on ideas
from subjects such as Philosophy, Mathematics, Anthropology and Psychol-
ogy, and contributing in turn to newer fields such as Pragmatics and Artificial
Intelligence. This situation has made the discipline receptive to new ideas, but
has hindered the development of a comprehensive and generally accepted
theory of linguistic or lexical semantics.
Recent developments within the framework of cognitive linguistics have
brought together some previous approaches and have been especially fruitful
in areas such as metaphor, and, increasingly, historical semantics. Geeraerts
(1988), for example, has shown how the assumptions and methodologies of
pre-structuralist historical-philological semantics can be reconciled with those
of the current cognitive semantics paradigm. As a practising historical lexi-
cographer, he has been interested in bringing the insights of lexicographers
working with large bodies of data to bear on theoretical work, writing:
It has long been my contention that theoretical lexicology could benefit
greatly from a closer acquaintance with the descriptive results and the meth-
odological problems of large-scale dictionaries such as the Woordenboek der
Nederlandsche Taal or the Oxford English Dictionary (1997: 5).
It is this contention that I would like to support from my own experience.
From Dictionary to Thesaurus
The primary concern of the two great dictionaries mentioned above is sema-
siological, that is they are concerned with the meaning, or more usually
HISTORICAL SEMANTICS AND HISTORICAL LEXICOGRAPHY 55
meanings, expressed by particular lexical items. My own concern is largely

onomasiological, that is an interest in the availability of lexical items to
express particular concepts.1 This interest arises from work on A Thesaurus of
Old English (TOE; Roberts and Kay 1995) and the Historical Thesaurus of
English (HTE; Kay, Wotherspoon, Samuels and Roberts, forthcoming).2 De-
scriptions of both projects can be found in van Dalen-Oskam 1997. A sample
of headings showing part of the taxonomy of the Historical Thesaurus is given
in the Appendix to this paper.
Both thesaurus projects use data compiled from dictionaries, which are
then re-organised into semantic fields, sub-fields, and lexical categories. TOE
is a synchronic snapshot of the extant Old English vocabulary, with no attempt
at further division into periods.3 HTE, which draws its data from the Oxford
English Dictionary (OED), with supplementation from the TOE materials, is
diachronic, adding dates of recorded use to each lexical entry.4 Thus, while
HTE is basically onomasiological, it is also semasiological, showing patterns
of semantic change and movement within the semantic field structures. Both
projects have provided those who work on them with a lengthy education in
problems of meaning and categorisation.
These projects date back to the 1960s, when the dominant structuralist
paradigm in semantics was being at least partially displaced by the transfor-
mational-generative approach, and truth-conditional semantics was yet to
make its mark (see, e.g., Geeraerts 1988, Saeed 1997). What Taylor (1995: 21
ff) calls the classical approach to categorisation dominated our thinking.
Even though our categories were not preconceived but were suggested by the
data, our objective was that every sense of every lexical item should fit in
neatly somewhere in the taxonomic structure, thus producing the ultimate
Saussurean network of sense relationships for English. As anyone who has
ever tried to classify anything, from Munsell colour chips to lexical meanings,
will know, such an objective is difficult, if not impossible, to achieve. There
will always be recalcitrant items which do not fit in neatly anywhere, or,
almost worse, fit equally well in several places. Because we were thinking
primarily of paper publication, economy of classification was also an issue;
electronic publication, with its infinite possibilities for repetition in the form
of cross-reference or multiple categorisation, may prove almost as liberating
as a new theory of semantics.
More encouraging during this early period was the development of
various types of semantic feature analysis, deriving mainly from anthropologi-
56 CHRISTIAN J. KAY
cal linguistics, and offering methodologies for determining the meanings of

words, and especially the differences between semantically close words, by
some form of lexical decomposition. Such work is perhaps best known under
the general heading of componential analysis.5 Its main drawback, at least in
its original form, was that it is a binary system: a meaning component is either
present in a word or it is not. Classical componential analysis thus has the
same drawbacks as the classical categorisation theory from which it derives,
in demanding either/or decisions of its users. In looser forms, where the
analysis attempts to go beyond the minimal disambiguation of limited subsets
such as kinship terminology, it has a tendency to become unruly, as ad hoc
components are added in order to make an ever-increasing number of seman-
tic distinctions.
Componential analysis has never achieved real respectability in theoreti-
cal semantics, and yet, acknowledged or not, it appears in many guises, in
work ranging from the transformational-generative struggles with the seman-
tic component of the grammar, as in Katz and Postal (1964), to recent work in
cognitive semantics. It also, I would like to suggest, forms part of the instinc-
tive mental toolkit of any lexicographer, whether engaged in definition or in
categorisation, and could thus claim the psychological salience required
within cognitive linguistics.
Dictionaries and Psychological Salience
This brings me back to dictionaries. Unsurprisingly perhaps, while semantic

theories have come and gone, dictionaries have continued to flourish. This
may just be because they fulfil a commercial need. However, I would like to
suggest that there could be a deeper reason and that in fact, far from doing
everything all wrong over the years, dictionaries have actually been getting
things more or less right. I have been greatly cheered recently to discover that
many of the procedures used in the HTE project can be incorporated within the
paradigm of cognitive semantics. What previously seemed like weaknesses or
ad hoc solutions in our approaches can now be justified in terms of their
psychological salience, that is, they reflect processes assumed, if not yet
proved, to be rooted in the general processes of cognition, which subsume
linguistic processes. Psychological salience, as thus defined, includes folk or
nave perceptions of how language works, and can be related to the apparently
untheorised activities of dictionary compilers and users.

Primary among these processes, and underlying many others, is that of
classification or categorisation, the area which has been most studied in
cognitive semantics. This process has two main aspects: firstly the semasio-
logical aspect of individual lexical items as clusters of meanings related to one
or more prototypical core meanings, and secondly the onomasiological aspect
of groups of lexical items forming a semantic category, again around a
prototypical core. Both of these aspects contribute to dictionary definitions,
are essential in the compiling of thesauri, and implicitly or explicitly draw on
lexical feature analysis.
Thus, when a researcher embarks on a new section of classification for
HTE, her first step is to sort the slips containing the data into initial categories,
using the dictionary definitions for guidance. A classic OED definition for a
phenomenon in the natural world might take the following form:
Hamster...A species of rodent (Cricetus frumentarius) allied to the mouse and
rat, found in parts of Europe and Asia; it is of a stout form, about 10 inches
long, and has cheek-pouches in which it carries the grain with which it stores
its burrows; it hibernates during the winter.
This definition relates the item in question to its superordinate category, and
adds a variety of other information, such as appearance, habitat and habits,
which form its stereotype and distinguish it from others of its kind. These
could be expressed either as a list of stereotypical characteristics or as compo-
nents such as [+HIBERNATE]. On HTE we may formulate such components
as a means of bringing a large number of closely-related meanings into order,
but even if we do not, they are subliminally present in our approach to
classification. In other words, units of meaning below word-level are neces-
sary to the process of categorisation. We use the dictionary definitions both for
their basic semasiological function and for their potential onomasiological
function of suggesting the categories to which the individual meanings may be
assigned. The semasiological is the source of the onomasiological.6
Dictionaries and Prototypes
Geeraerts (1997: 11) offers the following summary of the main characteristics
of prototypical categories:
58 CHRISTIAN J. KAY
(a) Prototypical categories exhibit degrees of typicality; not every member is

equally representative for a category.
(b) Prototypical categories exhibit a family resemblance structure, or more
generally, their semantic structure takes the form of a radial set of clustered
and overlapping readings.
(c) Prototypical categories are blurred at the edges.
(d) Prototypical categories cannot be defined by means of a single set of
criterial (necessary and sufficient) attributes.
The main points to note here are that not all members of a category are equally
central, and that consequently meanings overlap and the boundaries of catego-
ries are blurred. Although Geeraerts is mainly concerned with semasiological
analysis, his criteria will strike an instant chord with anyone who has had
dealings with an onomasiological thesaurus. Take, for example, a group of
definitions from the OED of words expressing the concept of ANGER:
Anger n 1. That which pains or afflicts, or the passive feeling which it
produces; trouble, affliction, vexation, sorrow. Obs. [Recorded from 1250 -
1475]
2. a. The active feeling provoked against the agent; passion, rage; wrath, ire,
hot displeasure. First recorded in 1375; still current.
3. Physical affliction or pain; inflammatory state of any part of the body. (Still
dial.) [Recorded from 1377 - 1698]
Passion, n 7. a. spec. An outburst of anger or bad temper. b. Without a:
Impassioned anger, angry feeling.
Rage, n 2. Violent anger, furious passion, usually as manifested in looks,
words or action; a fit or access of such anger; angry disposition.
Wrath, n 1. Vehement or violent anger; intense exasperation or resentment;
deep indignation.
Ire, n. Anger; wrath. Now chiefly poet. and rhet. [Only one sense.]
Displeasure n. 1. The fact or condition of being displeased or offended; a
feeling varying according to its intensity from dissatisfaction or disapproval to
anger and indignation provoked by a person or action.
With the exception of ire, which has only one meaning, the word-forms
individually constitute polysemic prototypical categories, as represented in
the OED by their numbered senses. These senses are categorised as polyse-
mous rather than homonymous on the basis of their semantic and etymological
relationship to one another, as in the links through the concept of pain in
Anger senses 1 and 3 (albeit apparently in the relatively unusual direction of

mental pain to physical pain7 ) and the metonymic link from the passive
sensation to the active expression in sense 2. In historical lexicography, such
relationships are usually etymologically determined, although this may not be
as clear-cut as it sounds where opinions about etymologies, or how far back to
trace them, varies. In synchronic lexicography, if such a thing can by defini-
tion exist, the appeal is increasingly to a folk perception of connectedness
rather than to strict etymological principle, presumably on the grounds that
most dictionary users are not historical linguists.8 Underlying this is an
assumption that core meanings in polysemous categories can be recognised by
speakers just as they can recognise prototypical cups or birds. Whitcut makes
a similar point about the psychological processes by which lexicographers are
presumed to select citations from their databanks:
Although citations are certainly objective as evidence of the existence and
use of a word or sense, when they are quoted as examples they must be
selected, perhaps from many other citations, as the most characteristic in-
stances of the word in use: they pass, in fact, through the lexicographers
brain, where some process of recognition must occur (1995: 255).9
The process of recognition produces the prototypical citation.
Definitions and Components
Collectively, the group of meanings discussed above forms part of an onoma-

siological sub-field of ANGER, which is in turn part of a much larger field of
EMOTIONS, as illustrated in the Appendix from HTE. The numbers displayed
there represent the taxonomic structure of the categories and sub-categories,
moving from the general to the particular. The example is incomplete, since
other forms of Emotion, such as Ardour or Excitement, fill the gaps preceding
02.02.21 Anger.
As described above, OED definitions and citations form much of the raw
material for our work on the HTE. Elements of the definitions are treated as
features of meaning and used as guides to categorisation, indicating whether a
meaning belongs to a particular category or to a subordinate or superordinate
one. The circularity of dictionary definitions, often a focus of criticism in
discussions of metalanguage, but inevitable when the defining language is the
same as the target language, works to our advantage in identifying the core
word of a category. Thus the fact that anger sense 2a contains several near-
60 CHRISTIAN J. KAY
synonyms in its definition, i.e. passion, rage, wrath, ire, and (hot) displeasure,
while these synonyms in turn are defined by the word anger, helps to identify
anger as the prototypical member of the set, at least for the 19th and 20th
century lexicographers who wrote the definitions. It thus forms the category
heading in category 02.02.21; as Geeraerts (1997: 43) points out, the choice of
category heading can in itself have psychological salience in indicating proto-
typicality.
Since the word anger is not recorded until around 1375, there must have
been earlier prototypes amongst the range of about twenty-five Old and Early
Middle English terms for the concept, possibly Old English ierre, Middle
English irre, last recorded around 1450 in the meaning anger, wrath. This
may have been lost through homonymic clash with, among others, French ire
and err, with the latter of which it sometimes coincided in form. Wrath is
another possible candidate, but seems always to have had a stronger meaning,
associated with OED 1 and 4a, Righteous indignation on the part of the
Deity.10
Anger, Rage and Dissatisfaction
As every lexicographer knows, assigning citations to senses is no easy busi-

ness, especially in historical lexicography, where there is no recourse to native
speakers or multi-million word corpora, but only to haphazardly surviving
written texts. The definition of displeasure above might be taken as a throw-
ing-in of the towel on the part of the lexicographer who wrote it. There is
surely a difference between mere dissatisfaction or disapproval on the one
hand and anger and indignation on the other. However, a look at the citations
reveals why the lexicographer may have erred on the side of caution:
Displeasure n 1
1484 Caxton Chivalry 81 Yre and dysplaysyre gyuen passion and payn to the
body and to the sowle.
1495 Act 11 Hen. VII, c. 57 Pream., All that that he hath doon to the
displeasure of your Highnes.
1489 Caxton Blanchardyn xxvi. 96 She brought thene in remembraunce how
swetly he had kyssed her, wherof she had take so grete a dyspleasure.
The first citation can, I think, clearly be assigned to anger, given its colloca-
tion with yre, passion and payn. The second is more ambiguous, since,
although there is no indication of strong feeling, the displeasure of a king

presumably had a strength of its own. The third is equally context-dependent,
though again does not suggest a very strong negative reaction if swetly is taken
into account. Taken together, the examples indicate the blurring that often
occurs at the edges of categories, where a word cannot be assigned with
confidence to either of the possible prototypes. In semasiological terms, we
have two possible solutions: we can seek out more evidence in the pursuit of
two separate senses or prototypes, or we can content ourselves with a broader,
more inclusive category. In historical lexicography, we may have no choice
but to adopt the latter course. However, this may also be the theoretically
preferable course if the ambiguity is inherent in the meaning. Prototype
theory, because of its tolerance of peripheral members of a category, would
support such an approach. In this context, it is interesting to note the following
statement in the preface to the New Oxford Dictionary of English:
Linguists, cognitive scientists, and others have been developing new tech-
niques for analysing usage and meaning Foremost among them is an
emphasis on identifying what is central and typical as opposed to the time-
honoured search for necessary conditions of meaning (i.e. a statement of the
conditions that would enable someone to pick out all and only the cases of the
term being defined). Past attempts to cover the meaning of all possible uses of
a word have tended to lead to a blurred, unfocused result, in which the core of
the meaning is obscured by many minor uses. In the New Oxford Dictionary
of English, meanings are linked to central norms of usage as observed in the
language. The result is fewer meanings, with sharper, crisper definitions
(Pearsall 1998: vii).
Where appropriate, meanings are divided into a core sense or senses and one
or more subsenses. Cognitive semantics and practical lexicography have
indeed come together.
For a thesaurus, there are various solutions to the problem of indetermi-
nacy of meaning. A meaning which is ambiguous or falls near a category
boundary can be repeated or cross-referred or can suggest the need for a more
general, higher-level category, as was the case with displeasure, which ap-
pears as a separate category 02.02.20.09. In such a case, there will usually be
support from other meanings of similar range. A new category may also be
formed at a lower level if sufficient words contain an additional meaning
component. From our set of ANGER words above, a strong case emerges for
a subordinate category with the extra component of extreme, violent, to
accommodate meanings such as those of rage and wrath, and this is indeed
62 CHRISTIAN J. KAY
what happened, as category 02.02.21.02 Furious Anger in the taxonomy of

headings illustrates.
Cognitive Semantics and Lexicography
The above examples seem to me to justify a good deal of traditional lexico-

graphical practice by suggesting that it has a sound psychological basis. While
the original OED editors did not think in terms of prototypes or meaning
components, any more than do the majority of their modern successors, they
nevertheless used techniques which can be retrieved in these terms. Their use
of approximate synonyms in definitions, often narrowed down by restrictive
terms such as extreme or violent, can be seen as an intuitive attempt to
decompose and relate meanings, which can then be reconstituted by those
seeking to differentiate and define words which are close in meaning. From an
onomasiological point of view, the relative frequency of such defining syn-
onyms, and the nature of the restrictive terms, can be used to establish
semantic categories. Geeraerts, indeed, claims that synonym dictionaries,
including those from the past, can represent the intuitive judgements of
speakers in identifying core meanings (1997: 171ff).
In historical lexicography there is the additional interest of potential
changes in the prototypical core. The chronologically earliest sense may
remain the core, as in many basic referring expressions such as mother or tree
(though even here perceptions may change). Alternatively, as a chain of
meanings develops, the original core may become peripheral or disappear
altogether, leaving meanings to cluster around a new core or multiple or
embedded cores (see, e.g., Taylor 1995: chapter 6). If some members of the
chain become obsolete, the links between the survivors may become unclear,
as in the series of metonymies leading from staff = stick to staff as in member
of staff.11 A diachronic study can help to retrieve the missing links and
uncover the underlying logic of the sequence. Thus an attractive and psycho-
logically convincing approach to change of meaning can be developed, where
successive meanings can be seen as part of a dynamic semantic system.
Two other advantages of a cognitive approach to lexical semantics are
worth mentioning. The first is the demise of synonymy as a theoretical
battleground. One no longer needs to argue about whether total synonymy is
possible, or to attempt to distinguish between degrees of synonymy (see, e.g.,
Lyons 1977: 242; Palmer 1981: 88ff). The following are among the words
listed under Category 891 Resentment, Anger, subsection Anger, in Rogets
Thesaurus of English Words and Phrases:
wrathfulness, irritation, exasperation, vexation, indignation; dudgeon, wrath,
ire, choler; rage, fury, passion; crossness, temper, tantrum, tizzy, paddy, fret,
fit, outburst, stew, paroxysm; rampage; shout, roar; fierceness, scowl; growl,
snarl, snappishness, asperity.
It is obvious that these words are by no stretching of the definition of the term
synonymous. Yet it is equally obvious that the words have semantic
connexions of one sort or another, whether expressed in terms of synonymy
and hyponymy or in terms of meaning clusters forming prototype categories
with core and peripheral members. One can assume that total synonymy is a
priori unlikely to exist: some synonyms are simply more synonymous than
others. Thus the problem of synonymy goes the same way as the problem of
the one-eyed man or the three-legged tiger.
The second advantage is the rejection of the dichotomy between lexical
and encyclopedic meaning which has so plagued academic discussions of
lexicology and reduced the value of lexicography in the eyes of semanticists,
particularly during the domination of formal semantics during the 1970s and
1980s. One textbook of the period contains a short and dispiriting chapter,
About Dictionaries, advising an individual known as the semanticist dictio-
nary-writer to strip the information in his dictionary down to the semantic
bone of formally described sense relations, and rigorously to exclude all
encyclopedic information as being irrelevant to the pursuit of semantics
(Hurford and Heasley 1981: chapter 16).
Such advice is hardly encouraging for the lexicographer confronting the
vast amounts of information contained in the lexicon, or indeed for the
computer scientist facing the same abundance when attempting to formalise
contextualised natural language. Language is concerned with organising our
perceptions of the world, and it is not therefore unreasonable that an appeal to
knowledge of the world should form part of the process of defining lexical
items. As Paul Deane says, in an article on polysemy, Human thought has
two complementary characteristics. While it displays flexible responses to
novel situations, it is also highly structured, incorporating detailed informa-
tion about the world (1988: 325). Thus in Rage, n 2 above, the information
that this emotion is usually manifested in looks, words or action might be
considered encyclopedic and not worthy of inclusion in a definition. Yet it is
64 CHRISTIAN J. KAY
often precisely this kind of information that enables the user to discriminate
among closely related meanings, or identify the referent, as in the Hamster
example. Backhouse, in an extended study of Japanese taste terms, points out
that
language is used in the world, and lexical items relate to aspects of this
world: in particular, lexical items are applied to extralingual categories of
entities, qualities, actions, events and states, and the relation between an item
and such categories is normally understood as constituting a central part of
its meaning (1994: 23).
Nida argues for an approach to lexical semantics where primacy is given to

context, both cultural and linguistic, and there is then a conceptual level
consisting of a focal term and the corresponding context (1997: 265). He
exemplifies this with an analysis of the focal verb run as used in a range of
sixteen contexts, such as printing, the electoral process, and knitted wear (ibid:
2724). Knowledge of the world is thus brought into the arena of definition.
In the light of these more flexible approaches, it is no coincidence that
semantics turns increasingly to other disciplines which attempt to understand
and categorise the external world, as in appeals to physiology and psychology
in Backhouse or in some of the articles on emotions in Athanasiadou (1998).
In historical lexicography, because of the limited evidence, we need all the
help we can get in elucidating the meanings of words, whether such help
comes from inside or outside the linguistic system. In support of this one
might also cite two studies in interdisciplinary semantics by Dr Carole
Biggam, Blue in Old English and Grey in Old English, where evidence from
fields such as botany, archaeology and early technologies is added to the
linguistic evidence in order to elucidate the semantics of the concepts in
question. This is, of course, a two-way process, with elucidation of word-
meaning sometimes shedding light on the societies of its speakers, as in such
socially-charged areas as words for the concept of Woman (see Kay 1997;
Kleparski 1997).
Conclusion
For all these reasons, and not least the use of the lexical item as the basic unit
of analysis, it seems to me that cognitive semantics offers a promising way
forward for semantics generally and for synchronic and diachronic lexicology
in particular. This approach will not solve all our problems, and we are still far
from achieving a comprehensive theory of lexical semantics, but it offers
enlightening ways of looking at some of them. In particular it suggests that
dictionaries have largely got things right: the apparent imperfections of their
methodologies in fact reflect the cognitive strategies of their compilers in
making the dictionaries, and of their users in understanding them.
Notes
1. For a discussion of the terms onomasiological and semasiological, see Hllen 1996.
2. The project is funded by the British Academy, the Carnegie Trust for the Universities of
Scotland, the Leverhulme Trust, and the University of Glasgow, to all of whom we would
like to express our gratitude. The work is about 75% complete. See also Wotherspoon in
this volume.
3. A better term might be achronic, since dates given in, for example, the OED, are not
incorporated.
4. The Old English entries are simply labelled OE.
5. An analysis of a set of key terms, using a method derived from Componential Analysis,
formed the basis of the original set of ordered categories for HTE (Kay and Samuels
1975). Although the internal structures of the semantic fields have been worked out by
individual classifiers, the original analysis has proved robust, and was used, with some
modifications necessitated by sociocultural factors and a much smaller corpus of data, for
TOE.
6. I am indebted to Prof. D. Geeraerts for a discussion of this point during a Symposium on
Linguistic Categories and Classification held at the Institute for the Historical Study of
Language, University of Glasgow, in September 1999.
7. One must, of course, bear in mind that such a link may represent different layers of
borrowing at different times rather than a straightforward chain of meaning in English.
See, for example, Considine 1997.
8. Check, for instance, whether your desk dictionary connects the expression to bale/bail
out (of an aeroplane) with the etymologically correct bale of cloth or the semantically
more attractive bailing out of water (from a boat), i.e. which meaning component is
selected as the basis for the semantic link. The latter explanation was preferred by both
staff and students in informal surveys carried out at Glasgow University, thus showing
that even historical linguists can succumb to folk etymology.
9. Whitcuts main concern, however, is to discuss how socio-cultural factors determine
lexicographical choice.
10. I am indebted to Dr Julie Coleman for the suggestion that the case for prototypical ierre
would be strengthened if ierre was the word commonly used to describe the third of the
66 CHRISTIAN J. KAY
seven deadly sins. I have not investigated this exhaustively, but ierre was certainly used
in that context.
11. An analysis of Staff, noun, from the OED suggests the following links:
I.1.a. A stick carried in the hand as an aid in walking or climbing.
6. Part of the insignia of the episcopal office, consisting of a rod or pole of wood, metal or
ivory The staff represents the possession of jurisdiction and was one of the insignia
connected with the investiture.
7. A rod or wand, of wood or ivory, borne as an ensign of office or authority; spec. as a
badge of certain chief officers of the Crown.
21. Mil.
a. A body of officers appointed to assist a general, or other commanding officer, in the
control of an army, brigade, regiment, etc., or in performing special duties (as the medical
staff).
[App. of continental Teut. origin. Cf. the like use of G. stab (also generalstab,
regimentsstab, etc.), Du. staf; prob. developed from the sense baton (= 7 above).]
E. 22. a. gen. A body of persons employed, under the direction of a manager or chief, in
the work of an establishment or the execution of some undertaking (e.g. a newspaper,
hospital, government survey, school, etc.).
References
Athanasiadou, Angeliki & Elzbieta Tabakowska, eds. 1998. Speaking of Emotions: Con-
ceptualisation and Expression. Berlin, New York: Mouton de Gruyter.
Backhouse, A.E. 1994. The Lexical Field of Taste: a semantic study of Japanese taste
terms. Cambridge: Cambridge University Press.
Biggam, Carole P. 1997. Blue in Old English. Amsterdam & Atlanta, Ga: Rodopi.
Biggam, Carole P. 1998. Grey in Old English. London: Runetree Press.
Considine, John. 1997. Etymology and the Oxford English Dictionary: a response.
International Journal of Lexicography 10, 3. 2346.
Deane, Paul D. 1988. Polysemy and Cognition. Lingua 75. 32561.
Geeraerts, Dirk. 1988. Cognitive Grammar and the History of Lexical Semantics. Topics
in Cognitive Linguistics ed. by B. Rudzka-Ostyn, 64777. Amsterdam: Benjamins.
Geeraerts, Dirk. 1997. Diachronic Prototype Semantics: a contribution to Historical
Lexicology. Oxford: Clarendon Press.
Hllen, Werner. 1996. An Onomasiological Version of the OED? Henry Sweet Society
Newsletter 27. 1516.
Hurford, James R. & Brendan Heasley. 1983. Semantics: a coursebook. Cambridge:
Katz, J.J. & P.M. Postal. 1964. An Integrated Theory of Linguistic Descriptions. Cambridge
MA: MIT Press.
Kay, Christian J. 1997. Historical Semantics and Material Culture. Experiencing Mate-
rial Culture in the Western World ed. by Susan M. Pearce, 4964. London and
Washington: Leicester University Press.
Kay, Christian J. & Jane Roberts. 1997. A Thesaurus of Old English. Van Dalen-Oskam.
3140.
Kay, Christian J. & M.L. Samuels. 1975. Componential Analysis in Semantics: its Validity
and Applications. Transactions of the Philological Society. 4981.
Kay, Christian J. & Iren Wotherspoon. 1997. Historical Thesaurus of English. Van
Dalen-Oskam. 4954.
Kleparski, Grzegorz. 1997. Theory and Practice of Historical Semantics: The Case of
Middle English and Early Modern English Synonyms of Girl/Young Woman. Lublin:
University Press of the Catholic University of Lublin.
Lloyd, Susan M., ed. 1982. Rogets Thesaurus of English Words and Phrases. London:
Longman.
Lyons, John. 1977. Semantics, Volume I. Cambridge: Cambridge University Press.
Nida, Eugene A. 1997. The Molecular Level of Lexical Semantics. International Journal
of Lexicography, 10, 4. 265-74.
Palmer, F.R. 1981. Semantics. 2nd ed. Cambridge: Cambridge University Press.
Pearsall, Judy, ed. 1998. The New Oxford Dictionary of English. Oxford: Clarendon Press.
Roberts, Jane & Christian Kay with Lynne Grundy. 1995. A Thesaurus of Old English.
(=Kings College London Medieval Studies XI). London: Kings College.
Saeed, John I. 1997. Semantics. Oxford: Blackwell.
Simpson, John & Edmund Weiner, eds. 1989. The Oxford English Dictionary (2nd ed.: 20
vols). Oxford: Oxford University Press.
Taylor, John. 1995. Linguistic Categorization: Prototypes in Linguistic Theory. Oxford:
Clarendon Press. 2nd edition.
Van Dalen-Oskam K.H., K.A.C. Depuydt, W.J.J. Pijnenburg & T.H. Schoonheim, eds.
1997. Dictionaries of Medieval Germanic Languages. (=International Medieval Re-
search 2). Turnhout: Brepols.
Whitcut, Janet. 1995. Taking It For Granted: Some Cultural Preconceptions in English
Dictionaries. Cultures, Ideologies and the Dictionary: Studies in Honour of Ladislav
Zgusta ed. by Braj B. Kachru & Henry Kahane, 2537. Tbingen: Max Niemeyer.
Appendix
Sample of categories from the Historical Thesaurus of English.

02.02. . . . . Emotion/feeling
02.02.01. . . . Seat of the emotions
02.02.02. . . . Emotional perception/consciousness
02.02.03. . . . Quality of affecting the emotions
02.02.10. . . . Absence of emotion
02.02.11. . . . Types of emotion/feeling
02.02.12. . . . Intense/deep emotions
02.02.13. . . . Sincere/earnest emotion
02.02.14. . . . Zeal/earnest enthusiasm
68 CHRISTIAN J. KAY
02.02.20. . . . Mental pain/suffering

02.02.20.01. . . (A) cause of mental pain/suffering
02.02.20.01.01. . Exacerbation of suffering
02.02.20.01.02. . Quality of being unendurable/intolerable
02.02.20.02. . . Mental anguish/torment
02.02.20.02.01. . Cause of mental anguish/torment
02.02.20.02.02. . Heart-brokenness/-strickenness
02.02.20.02.03. . Bitterness of heart
02.02.20.03. . . Sorrow/grief
02.02.20.03.01. . Sorrow caused by loss
02.02.20.03.02. . Lamentation/expression of grief
02.02.20.04. . . Regret
02.02.20.05. . . Misery
02.02.20.06. . . State of being upset/perturbed
02.02.20.06.01. . Worry
02.02.20.06.02. . State of being shocked
02.02.20.07. . . State of being harassed
02.02.20.07.01. . Condition of being oppressed
02.02.20.08. . . Dejection
02.02.20.08.01. . Nervous depression
02.02.20.08.02. . Melancholy
02.02.20.09. . . Displeasure
02.02.20.09.01. . Discontent/dissatisfaction
02.02.20.10. . . State of annoyance/vexation
02.02.21. . . . Anger
02.02.21.01. . . Manifestation of anger
02.02.21.02. . . Furious anger
02.02.21.03. . . Indignation/resentment
02.02.21.04. . . Irritation
02.02.21.05. . . Irascibility
02.02.21.05.01. . Touchiness
02.02.21.05.02. . Irritability
02.02.21.05.03. . Peevishness
02.02.21.05.04. . Ill-humour
02.02.21.05.05. . Ill-naturedness
University of Glasgow 1999

Strange Linguists: The Cant and Slang
Dictionary Tradition
Julie Coleman
University of Leicester
Abstract
This paper discusses the prefatory matter attached to cant and slang dictionar-
ies between the sixteenth and the nineteenth century, and argues that, despite
scholarly neglect, the dictionaries contain much that is worthy of attention.
They discuss social concerns and linguistic theories regarding universal gram-
mar and the origins of language. They also consider the problems of observa-
tional linguistics, though there is little evidence that the editors actually
undertook the fieldwork they claim to have done. In addition, it is only by
seeing these works as part of a long-running tradition that it is possible to
evaluate their contents.
Introduction
Early dictionaries of cant and slang have generally been regarded as peripheral
to the development of the mainstream English lexicographical tradition, as
indeed they are.1 Both their subject matter, the language of the criminal classes,
and their tendency towards ribaldry and outright obscenity rendered them unfit
even for antiquarian interest. As a result of this, most publications in this field
have been, and still are, aimed at a general rather than an academic audience.2
Where academics do write about these dictionaries, they tend to be literary
scholars who see them as sources for the study of a particular period, and not
lexicographers treating them as part of a long-running dictionary tradition.
70 JULIE COLEMAN
The English Cant Dictionary Tradition
The English cant dictionary tradition reaches back to the sixteenth century.3
This is well before Cawdreys Table Alphabetical of 1604, which is usually
considered the first monolingual dictionary of English. The origins of the
slang tradition lie in works like John Awdeleys Fraternity of Vagabonds
(1565), Thomas Harmans Caveat or Warening for Common Cursetors
(1567), Robert Greenes Groundwork of Cony-catching, The Second Part of
Cony-Catching, and The Third and Last Part of Cony-catching (15912), his
A Notable Discovery of Coosnage (1591), and his Disputation between a He
Cony-catcher, and a She Cony-catcher (1592). These works tell of the tricks
used to part unwary country folk or nave youths from their money. Lists of
words are included in some of these works, while the others were scanned for
cant vocabulary by the compilers of later dictionaries. Awdeley, for example,
lists types of rogue, but does not give a word-list as such. This list is copied
and added to for at least three centuries, being found under the entry for crew
in Groses dictionary and those derived from it. In Greenes Cony-catching
series there is a glossary rather than a dictionary proper, with the vocabulary
of each con-trick listed separately. Some, but by no means all, of Greenes
terms are found in the later dictionaries. Harman has a much more extensive
list of terms, over a hundred of them, this time with no apparent principle of
organization: they are grouped neither semantically nor alphabetically. It is
Harmans list that forms the basis of the cant dictionaries of the early seven-
teenth century.
The earliest seventeenth century cant dictionaries are found as part of or
appendices to larger works. In Thomas Dekkers Bellman of London series,
from 1608 onwards, the nocturnal vices of the capital are exposed. These lists
were incorporated into Richard Heads Life and Death of the English Rogue
(1665), which was the first of the word-lists to include citations, many of them
drawn from the cant dialogues and verses included in Dekkers works. Heads
English Rogue went through many editions in its own right, and spawned
accounts of the lives of rogues like Bampfylde-Moore Carew, Jeremy Sharp,
Jonathon Wild, and Meriton Latroon. These villains, some based on reality and
some entirely fictional, tended to have remarkably similar encounters with
bands of gypsies using the same vocabulary and singing the same songs across
a period of more than two centuries. This list developed into two separate
traditions, one alphabetized by the cant word and one by the standard English
CANT AND SLANG DICTIONARIES 71
term.4 Both of these word-lists survived in their own right, despite the
production of far more extensive dictionaries like B.E.s New Dictionary of the
Terms Ancient and Modern of the Canting Crew in c.1698,5 the anonymous
Scoundrels Dictionary of 1754, and the various editions of Groses Classical
Dictionary of the Vulgar Tongue, from 1785.6 After the publication of Elisha
Coles An English Dictionary in 1674, cant words also began to be included,
suitably stigmatized, in the ever-increasing flood of general dictionaries.7
The Dictionaries Prefatory Materials
The main focus of this paper is the cant dictionaries introductory materials:
the title pages, epistolary addresses, prefaces, forewords, and introductions.
These explain how and why the dictionaries were produced, and sometimes
include discussions of the nature of slang and cant. Some will have been
written by the editor of the dictionary, and some added by the printer or
publisher.8 What is important is that they represent the sales pitch: they are
designed to increase the dictionaries saleability, by appealing to as large a
section as possible of the potential market.
The dictionaries were not produced in isolation from contemporary con-
cerns about society and the economy. There are always more beggars now
than there used to be, and the prefaces often attempt to place what is seen as
the contemporary prevalence of sturdy vagabonds in its historical context.
Greene, for instance, looks back to the reign of Henry IV, when Richard
Whittington was Lord Mayor of London, and took on the task of ridding the
city of its sturdy beggars (Greene The Third and Last Part ). He prays that
another such will rise up and free his own time of the most hurtful and
dangerous enemies to the commonwealth (ibid). Dekker, however, suggests
that the beggars and thieves of his day are a throwback to the time of Jack
Cade, during the reign of Henry I (Dekker 1612). B.E., in c.1698, traces the
origins of begging in Britain to the dissolution of the monasteries, and ob-
serves, rather ominously, that no beggars are found in nations where slavery
persists.
Not only are there many more vagabonds than there used to be, which is
cause for concern in itself, but this mass of thieving deviants is also large
enough to threaten the very structure of society. They are lice (Dekker 1608),
locusts (Dekker 1612), drones (Greene The Second Part ), caterpillars
72 JULIE COLEMAN
(Greene ibid), and grasshoppers (Dekker 1608), all of which live off the
bodies or labours of others. They are hawks (Dekker 1612) and hounds
(Dekker 1608) which hunt after flesh, and dogs that return to their own vomit
(Greene The Third and Last Part ).9 In perhaps the most frightening
description of all, the beggars are a single monster whose neck grows so thick
as its body grows fat, that it becomes more difficult every day for the arm of
the law to strike off the head (Dekker 1608).
The Benefit to be Derived from Cant Dictionaries
As well as being objects for the curious, these dictionaries were to their
contemporaries what burglar alarms and steering-locks are to us: they were a
way to combat illegal activities. But before the reading public can be reassured
by their purchase, they must first be scared into making it.
In order to be encouraged to buy these dictionaries, readers have to be
convinced that beggars and vagabonds represent a threat, and that the cant
dictionaries are the antidote to it. The beggars and thieves are vipers that live on
poison, but from whom apothecaries can make valuable medicines (Greene The
Second Part ). They are gangrene, infecting the entire body if the diseased
limb is not amputated (Greene ibid), and pestilent sores that must be lanced for
the good of the Commonwealth (Dekker 1608).
What the early dictionaries imply by metaphor, the later works state
explicitly: these dictionaries are published for the benefit of the individual and
of the community as a whole. Coles justifies the inclusion of cant terms as
follows: Tis no disparagement to understand the Canting Terms. It may
chance to save your throat from being cut, or (at least) your Pocket from being
pickt (Coles 1676: To the Reader). The anonymous New Canting Diction-
ary of 1725 describes the plight of the Honest Man who is obliged to travel
the Road, and to frequent Inns and places of Publick Resort (whereby he is
often forced to mix with different Companies). Having bought the dictionary,
he may easily discover, by the Cant Terms and Dialects of the Persons, their
Profession and Intentions, and know how to secure himself from danger.
Similarly, Smiths Thieves New Canting Dictionary (1719) explains the
most mysterious Words, newest Terms, significant Phrases, and proper Idi-
oms, used at this present Time among our modern Villains; whereby Travel-
lers may oftentimes save both their Lives and Money. Parker (1789: 126)
makes a similar claim I have now only to say, that the man whose wish is to
avoid the stratagems and schemes that he is daily liable to fall into, I invite to
bear me company, whilst I describe and explain the following terms. The title
page of the Yokels Preceptor (?1855) states that it is Intended as a Warning
to the Inexperienced Teaching them how to Secure their Lives and Property
during an Excursion through London, and calculated to put the Gulpin always
upon his guard . In his glossary, published in 1728, Defoes advice is even
more straightforward:
And I shall give this by way of Advice before I begin, whenever any Person
hears such a Language, Speech, or Cant; or what you please to call it, let them
take Care of the Speaker; for they may depend ont they are certainly of the
Nimming Clan, and therefore to be avoided. (Defoe 1728: 29)
More enticing than any of this advice, are the claims made in the epistle to
Harmans Caveat or Warening for Common Cursetors (1567), which is
addressed to Elizabeth, Countess of Shrewsbury. Harmen praises Elizabeth
for her many virtues, including her charity, and explains that he describes the
tricks used by these rowsey, ragged rabblement of rakehelles so that she and
other householders can ensure that their alms go to the deserving poor, so that
justices will punish sturdy beggars more vigilantly, and so that bailiffs will
fulfil their obligations more circumspectly. The results, he promises, will be a
decrease in theft, safer travel to and from markets and fairs, an increase among
gentlemen in hospitality towards their neighbours and charity to the poor, and
a general decrease in sin. Even these peevish, perverse and pestilent people
benefit, because if it becomes more difficult for them to shorten their lives by
sinning, they may live long enough to repent. All of this will pacify Gods
wrath, with a resultant decrease in plagues and an improvement in the fortunes
of the nation. It is difficult to imagine how a dictionary could claim to do more.
In 1785, Grose argues that his dictionary will be a valuable tool for the
antiquarian, because of the customs and games it describes. He goes on to
explain that foreigners, and even native speakers from outside London, will
find his work very useful as a guide to the language used in common
conversation and periodical publications (Grose 1785: ii). Sometimes these
words pass into everyday use, in which case it is justifiable to record their
origins,10 and sometimes they fall from use altogether without leaving a trace
behind (ibid), which also justifies recording them.
As might be expected, the lexicographers also claim that they are merely
supplying their customers demands. Greene claims that The Second Part of
74 JULIE COLEMAN
Cony-Catching was published in response to public approval of A Notable

Discovery of Coosnage, but Judges (1930: 500) notes that it must have
reached the public within a few days of A Notable Discovery if not simulta-
neously with it, for both were registered in December, 1591 and both published
in 1591, though from different offices. Much later, Grose claims that the
favourable reception of his first edition justified the second (Grose 1788: iii).
The main justification, then, for the publication of these works, is the
benefit of the reader and of society at large. There are, of course, other
motivations, though few of the lexicographers are willing to admit them. In
1612, Dekker concedes that his intention in including cant in his work is to
procure delight to the Reader. The editor of the Bang-up Dictionary is
unusual in implying that his intention is not a moral one. He states that by
using his dictionary:
the whole tribe of second-rate Bang-ups may be initiated into all the
peculiarities of language by which the man of spirit is distinguished from the
man of worth. (Bang-up Dictionary 1812: v)
This is not the first preface to consider the possibility that the dictionaries
might lead to an increase in sin, however. Dekker (1608) concedes the
possibility that others that never before knew such evils, will be now in-
structed (by the book) to practise them, but reasons that the same argument
would justify suppressing the trials of traitors and murderers. He prefers,
instead, the notion that, just as seeing a common Drunkard acting his beastly
Scenes in the open street (ibid) is the best way to prevent a man turning to
drink, writing openly about vice will prevent its continuation, not least be-
cause it is difficult to cheat someone at dice if they know both that they are
being cheated, and also precisely how.
Obscenity
At the same time as justifying the publication of a dictionary of cant, some of

the later cant lexicographers acknowledged objections that might be raised,
and apologized in advance for causing offence. In his first edition, Grose
apologizes for the inclusion of offensive terms, and assures the reader that
when an indelicate or immodest word has obtruded itself for explanation, he
has endeavoured to get rid of it in the most decent manner possible; and none
have been admitted but such, as either could not be left out, without rendering
the work incomplete, or, in some measure, compensate by their wit, for the
trespass committed on decorum (Grose 1785: vii). He ends his preface by
evading responsibility for the words he has included. If anyone is offended, he
says, it is not his fault for recording the terms, but the fault of those who use
them. In the preface to the second edition of Grose, however, we find that
offensive words have been omitted or their definitions rewritten, with an
appeal to authorities such as Bailey, Mige, and Philips to justify the inclusion
of those that remain.11 Grose writes: it is hoped this work will now be found
as little offensive to delicacy as the nature of it would admit (1788: iii).
By 1865, Hotten is able to assure his readers that Filthy and obscene
words have been carefully excluded, and he provides a rhyming couplet
suitable for the occasion:
Immodest words admit of no defence,
For want of decency is want of sense. (Hotten 1865: xv)12
He confesses that, of those words remaining, he scarcely knew what to do

with some of the more repulsive (ibid), but in this he is referring to those that
describe theft rather than anything even more distressing: Their very exist-
ence is a lamentable fact; and the dry, unpoetic way they explain criminal
intentions and actions is miserable in the extreme (Hotten 1865: vii). This
squeamishness seems rather a product of social class than genuine delicacy of
mind: Slang is generally pithy and amusing, whereas Cant, like our lower
orders in their thoughts and actions, is unrelieved by any feeling approaching
to the poetic or the refined (ibid).
Somewhat disingenuously, the writer of the preface to the Bang-up
Dictionary (1812)13 argues that the production of such a work will actually
lead to a reduction in offence caused by obscenity, by providing the slang-user
with a secret language that the innocent will not understand:
We are very sure that the moral influence of the Lexicon Balatronicum will be
more certain and extensive than that of any methodist sermon that has ever
been delivered within the bills of mortality. We need not descant on the
dangerous impressions that are made on the female mind, by the remarks that
fall incidentally from the lips of the brothers or servants of a family; and we
have before observed, that improper topics can with our assistance be dis-
cussed, even before the ladies, without raising a blush on the cheeks of
modesty. It is impossible that a female should understand the meaning of
twiddle diddles, or rise from table at the mention of Buckingers boot.14
The introductory epistle to G.L.s dictionary of 1742 takes a much less

76 JULIE COLEMAN
conciliatory stance. If the reader is too foolish to appreciate what is set before
him, it is hardly the fault of the editor:
For my Part, since I am satisfied that what I here offer is Genuine, and not
spurious, I am not solicitous whether the READER likes it or not, for I am
sure it will commend itself to a Man of Understanding; and therefore he that
dislikes it must be otherwise; and I know no Reason I have to make Apologies
for the Imbecility of my Reader.
He continues, in mockery of the normal epistolary style:

But although I care not for to court my Reader, yet I love to be civil; and will
therefore assure him, That what he reads in the Title, he will find made good
in the Book: And if that will not satisfy him, he has no Reason to buy it; if it
will, he has no Reason to complain of being disappointed.
Authentication of the Material Included in the Dictionaries
As well as justifying and apologizing for the publication of their word-lists,

these early slang and cant lexicographers often tried to provide authentication.
If cant is the secret language of thieves, gypsies, and beggars, how is it
possible for a reputable editor or publisher to produce a dictionary of it?
For those word-lists that are included in a larger work, there is sometimes
an introduction, of sorts, to the glossary, but often it is nothing more than and
during my time with the gypsies/thieves/beggars, I learnt their language, and a
list of words is included at the end of this book. This is the most common
form of verification found, but where the dictionaries stand alone other
explanations have to be presented. In the second edition of The Bellman of
London, Dekker includes an account of an occasion in which he hid in a loft
and was able to spy on a gathering of vagabonds and rogues from around the
country. His unimpeded view of the ceremony by which a new rogue is
admitted to their ranks allows him a perfect opportunity to eavesdrop upon
their secret language, though the account is not explicitly provided to explain
the source of the cant vocabulary. In fact, the character of the Bellman, who
wanders around the capital at night, is already sufficient authentication of
Dekkers glossary.
In response to Dekkers work, S.R. produced Martin Markall, Beadle of
Bridewell in 1610.15 In this, the beggars are so enraged by the revelation of
their secret language that they gather together to find the culprit. Dekkers
Bellman is brought out to stand trial before them, but it soon emerges that
Dekker is guilty only of plagiarism, because the word-list had been published
forty years previously by Harman. As a further criticism of Dekker, S.R.
proceeds to annotate the word-list with corrections and new cant terms, a few
of which find their way into later editions of Dekkers own work.16 More
convincing still is Greene, who had uncertain thoughts in his younger years,
and delighted once to be wanton. In order to learn to avoid their snares, he
spent time with cheats and vagabonds, and what I saw in them to their
confusion, I can forewarn in others to my countrys commodity (Greene A
Notable Discovery ). His description of cheating at dice, however, suggests
that his involvement was as more than an observer: no man could better than
myself discover this lawe and his tearmes yet for some speciall reasons,
herein I will be silent (ibid).
Hottens dictionary of 1865 is preceded by a delightful account of the
deals he did with the chanters and patterers of Seven Dials to collect cant and
slang for him, which he checked with other chanters and tramps (Hotten 1865:
xiii). An intelligent printer in Seven Dials (ibid) put him in contact with
costermongers, pedlars, and hucksters, all of whom contributed to the produc-
tion of a dictionary that obviously owes much more than is admitted to
previous slang and cant lexicographers. Although Hottens account of his
data-collection is less than the whole truth, what it does demonstrate is that the
basics of dialectology were filtering through to cant lexicographers, along
with an understanding of social networks more characteristic of sociolinguis-
tics. All of this, however, is overlaid with the belief that even in this sub-
culture, there is a standard and correct form of language.
In producing his edition of the cant list, Head (1674) was troubled by the
lack of entries for cant terms after T in the alphabet. He too turned to cant
speakers for information on this matter, but this time by the more direct
method of going into Newgate and plying the rogues there with alcohol until
they revealed their secret language.17 They inform him that this language
frequently changes, and provide him with some new terms to add to his word-
list, but only those which have passed approbation of the Critical Canter. A
similar approach is taken by the anonymous author of Hell upon Earth (1703),
who writes, somewhat obliquely:
I have underneath set down the most necessary Words used by dishonest
persons; which I have with great Pains and Labour, as well as Charges,
collected from the Knowledge of the chief Professors of the Canting-Tongue,
now residing in this most noted Academy of Sin in London. (Hell 1703: 5)
78 JULIE COLEMAN
This language, he assures us, is daily more and more refind (ibid). Perhaps
even those who already possess a dictionary of cant ought to consider buying
an up-to-date edition after all the trouble he has gone to.
Sometimes, rather than take the dangerous action of venturing into pris-
ons or the underworld of thieves and beggars, lexicographers conduct their
dealings with canters on home-ground. Harmans dedicatory epistle, for ex-
ample, explains that he has been experiencing a period of ill-health, and has
had to stay at home. This has afforded him an opportunity to observe the wily
wanderers at work. By flattery, money, and good cheer, and by promising
them faithfully that neither their names nor anything they revealed would ever
go any further, he was able to prevail upon a few of them to explain their secret
language to him (Harman 1567: Epistle). Dekker similarly endeavours to
befriend a speaker of cant in order to study his language. In this case he takes
a sturdy big-limbed rogue into his service, and uses him kindly, until he
eventually admits that he does know cant, but had been afraid to reveal it
before for fear that the other beggars would kill him (Dekker 1612).
Awdeleys cant-informant is also afraid of the consequences of his co-
operation. Awdeley prefaces the Fraternity of Vagabonds with a poetic ac-
count of a rogue who was brought before a panel of magistrates:
Who promised if they would him spare,
And keep his name from knowledge then
He would as strange a thing declare
As ere they knew since they were men
But if my fellows should know (said he)
That thus I did, they would kill me.
The title page of the anonymous Scoundrels Dictionary of 1754 appeals both
to written authority and to a genuine contact with the underworld. It informs
the reader that the dictionary is:
printed from a Copy taken on one of their Gang, in the late Scuffle between
the Watchmen and a Party of them on Clerkenwell-Green; which Copy is now
in the Custody of one of the Constables of that Parish.
If this were to be believed, it would suggest that criminals themselves bought

copies of the cant dictionaries. Perhaps it would not be possible to communi-
cate with other thieves and beggars without learning this secret language.
Perhaps they just wanted to ensure that they sounded like authentic villains.
Perhaps the rogues were as concerned as their contemporaries with the cor-
rectness of their language.
Despite their editors protestations to the contrary, most of the slang and
cant dictionaries I have looked at so far are reprints of or compilations from
other slang dictionaries. It is probably Grose who is most honest about his
sources. He lists the dictionaries he has consulted for the 1785 edition, and
provides a further list used to find additions for the 1788 edition. He also
frequently cites other works in his definitions, and where he provides quota-
tions, he often gives their source. Unlike most of the other cant and slang
lexicographers, Grose makes a virtue of the unoriginality of his word-list: he is
self-consciously producing a work for antiquarians.
The Nature and Origins of Cant
Another matter often dealt with in the prefaces to these dictionaries is the
nature and origin of cant. It is, above all, characterized by its lack of rule:
Both the Father of this new kind of Learning, and the Children that study to
speak it after him, have been from the beginning and still are, the Breeders
and Nourishers of all base disorder, in their living and in their Manners: how
is it possible, they should observe any Method in their speech, and especially
in such a Language, as serves but only to bitter discourses of villanies (Dekker
1608).
The association made here, between ordered speaking and ordered living,
between linguistic and social rule-breaking, is usually considered more
characteristic of eighteenth than seventeenth-century linguistic attitudes (see
Mugglestone 1995: 28). But cant is not wholly unregulated, and Dekker goes
on to demonstrate ways in which compounds are formed, and to show that
some cant words are derived from Latin:
And yet (even out of all that Irregularity, unhandsomeness, and fountain of
Barbarism) do they draw a kind of form: and in some words (as well simple as
compounds) retain a certain salt, tasting of some wit, and some Learning. As
for example, they call a Cloake (in the Canting tongue) a Togeman, and in
Latin, Toga signifies a gown, or an upper garment. Cassan is Cheese, and is a
word barbarously coined out of the substantive Caseus which also signifies
Cheese (Dekker 1608).
The title of Smiths Thieves Grammar suggests that the language of villains
might be governed by rules that could be learnt along with the vocabulary, but
it proves to be a mock-grammar of the type that was to become popular a
80 JULIE COLEMAN
century later (see, for example, Reibel 1996). A brief quotation will illustrate
its level of humour:
Desperate Attempts ought to be performed in the Plural-Number; but if you
are Disposed not to put your Neck in the Power of such Comrades in Iniquity
as may turn Evidences against you, to save their own Lives upon Emergent
Occasions, you had then best to Rob in the Singular-Number; that is to say, by
your Self.
You have Five Declensions known by their several Terminations which end
your Lives in Misery, Disgrace, Poverty, Wickedness, and a Rope (Smith
1719).
As well as discussing the nature of cant, Dekker describes its origins. As

for other linguists of his time, his authority was the Bible. Dekker writes:
When all the World was but one Kingdom, all the People in that Kingdom
spake but one language, A man could travel in those days neither by Sea nor
land, but he met his Countrymen and none others. Two could not then stand
gabbling with strange tongues, and conspire together, (to his own face) how
to cut a third mans throat, but he might understand them (Dekker 1608).
He traces the threat that lies in cant to the origins of the diversity of language,
at the tower of Babel, and in a reversal of Harmans argument (see above), that
cant leads to sin, Dekker demonstrates that sin led to cant, through the
messenger sent to Babel by an angry God:
This strange Linguist, stepping to every Artificer that was there at work,
whispered in his ear: whose looks were thereupon (presently) filled with a
strange distraction: and on a sudden whilst every man was speaking to his
fellow, his language altered, and no man could understand what his fellow
spake. (Dekker 1608).
Saman, conversely, states that cant was not spoken at Babel:

Thus, Reader, I have given you a Light in to this New created Language,
which was never known to our Fore-fathers; nor heard of at the Confusion of
Babel (Saman 1725: 156).
Unlike all other languages, then, cant cannot trace its origin to the actions of
God, even in his rage at mans insubordination. It is much worse than that:
cant is a purely human product.
Harman gives a much more precise and recent date for the origin of cant.
He writes that cant began but within these xxx years little above, and that the
first inventor thereof was hanged, all save the head, for that is the final end of
them all, or else to die of some filthy and horrible diseases (Harman 1567:
Epistle). Shirley gives this individual a name:
the Foundation of which Gibberish was laid on one Rugosa, a sturdy Wanderer,
who first prescribed Rules and Orders for the wandering Tribe, and became
their Head or Superior; but not long enjoyd his ragged Dignity, before he fell
sick of a filching Fever, for which the Doctor of the Triple Tree applied the
powerful Cordial of Hemp to his Jugular Vein (Shirley 1724: 143).
This attribution of the development of cant to the inventiveness of a single

individual is repeated in later dictionaries, including the 1785 edition of
Grose. Grose also cites Mr. Harrisons Description of England prefixed to
Hollingsheads Chronicle, in which the development of cant, a language
without all order or reason, is considered to be the work of beggars and
gypsies collectively (Grose 1785: iii). A depraved individual who devises a
secret language is threatening enough, but Harrison speaks of a body of ten
thousand vagabonds throughout England united by their common, but secret,
language.
The earliest cant glossary published in America, containing some of the
same words as the British dictionaries, is the first that I have seen that admits
that cant is not a language in itself:
From this sample it may appear, that nouns and principal verbs, as being the
more important words in a sentence, are generally flashified; while pronouns,
auxiliary verbs and abbreviations retain their English uniform; so that the
flash tongue is nothing else than a mixture of English, with other words,
fabricated designedly for the purposes of deception; it can be useful to rogues
and sharpers only. I once acquired such a facility in this dialect, as to converse
in it with much the same ease as in plain English, although now I have lost its
familiar use. But no more of this futile language; may it return to Europe,
where it received its misshapen birth (Tufts 1807: 294).
Tufts also assumes that cant was invented with a deliberate intention to
deceive, but he blames Europe as a whole rather than individual rogues or
even an individual nation. In agreement with this is Winstanleys assertion, in
his discussion of European languages, that cant is used generally in most
Countrys of Europe (1669: 130). Shirley agrees that CANT is found to be
the peculiar Language of no Nation (Shirley 1724: 143). It may be that they
are guilty of a blurring of distinctions found in many of these works: between
vagabonds in general and gypsies proper. Dekker is not alone among the cant-
lexicographers in denying that gypsies are anything other than vagabonds who
paint their faces:
82 JULIE COLEMAN
A man that sees them would swear they had all the yellow jaundice, or that
they were Tawny Moors bastards, for no Reddaker man carries a face of a
more filthy complexion, yet are they not born so, neither has the Sun burnt
them so, but they are painted so, yet they are not good painters neither; for
they do not make faces, but mar faces (Dekker 1608).
Conclusions
The scholarly and lexicographical neglect of cant and slang dictionaries is

undeserved. The linguistic theories underlying the dictionaries demonstrate
the extent to which informed thought on the matter had filtered down to the
non-scholarly public from which the cant lexicographers were largely drawn.
Although the contents change relatively slowly, the introductions and prefaces
address the concerns of contemporary readerships about the insecurity of
society and the nature of obscenity. The editors of these dictionaries may not
have undertaken the fieldwork they claim to have done, but their claims
foresee some of the fieldwork techniques associated with modern linguistic
studies.
It is true that large numbers of derivative dictionaries were produced,
which merely copied from their predecessors, but the same carefree plagia-
rism is found in the general dictionary tradition. It is also true that we
sometimes have no evidence from outside the dictionaries to demonstrate that
their lexis was ever actually used, but a careful comparison between them and
the earlier dictionaries can at least identify which terms are slavishly copied
and which might represent genuine contemporary cant or slang. If nothing
else, the dictionaries demonstrate contemporary linguistic theories and prac-
tices, and attest to an enduring fascination with the idea of a secret and deviant
language used to undermine the fabric of society.
Notes
1. Although the dictionaries usually contain the same basic word-list, the earlier ones tend
to attribute it to gypsies, beggars and thieves, which suggests that it is cant, while the later
dictionaries often include, or at least claim to include, the language of fashionable young
men about town, which implies that the words are slang. I have tended to use the two
terms together here in order to emphasize that both types of dictionary are included.
2. Modern editions of, for example, Groses Classical Dictionary and Matsells Rogues
Lexicon tend to be produced by non-specialist publishers and priced for a general
audience. Examples include Harris (1980), Lovric (1997), and Matsell (1997).
3. From the mid-sixteenth to the mid-nineteenth century, there were over 200 editions of
English cant and slang dictionaries, more than half of them published in the eighteenth
century. They range from cheaply-produced pamphlets to expensively-bound books.
4. Heads alphabetization by the standard English term is not entirely successful. The word
list begins: A Curious wench, An Apron, An instrument to break a door, A part or share,
An hole (Head 1674).
5. B.E. lists not only cant, but also jargon, drawn especially from the fields of science (e.g.
Amphibious creatures of a doubtful kind, or of a double element; as a Bat is between a
Bird and a Beast; and Otter between a Beast and a Fish, and a Puffin with the rest of the
Sea-Fowl, between Fowl and Fish) and hunting (e.g. Entries where the Deer have lately
passed the thickets).
6. In his title, Grose puns on many difference senses of the word vulgar: vernacular (OED
sense 3), of or pertaining to the common people (OED sense 8), Commonly current or
prevalent, generally or widely disseminated, as a matter of knowledge, assertion, or
opinion (OED sense 6), unrefined, obscene (OED sense 13). His title puns on the usual
opposition between classical and vulgar as historical periods (OED vulgar sense 1), but
also on the use of classical in the sense Of the first rank or authority; constituting a
standard or model; especially in literature (OED sense 1).
7. See Gotti (1996).
8. This observation was made by Geoffrey Hughes at the G.L. Brook Symposium, Manches-
ter, 1998. I refer here to the introductory materials by means of the dictionary they are
attached to, without necessarily suggesting that the editor of one is also the author of the
other. Where page numbers are omitted, it is because the pages are not numbered.
9. See II Peter 2.22.
10. This argument is also put forward by Hotten (1865: xv-xvi).
11. For example, bagpipe a lascivious practice too indecent for explanation is omitted after
1785. Emendations are not always in the direction of decency, however. For instance,
ballum-rancum A hop or dance, where the women are all prostitutes, is expanded after
1788 for the benefit of those with little imagination N.B. The company dance in their
birthday suits.
12. It may be that Hottens choice of verse is a joke for the benefit of those who know what
follows it:
What modrate Fop would rake the Park, or Stews,
Who among Troops of faultless Nymphs may chuse?
Variety of such is to be found;
Take then a Subject, proper to expound:
But Moral, Great, and worth a Poets Voice,
For Men of sense despise a trivial Choice:
And such Applause it must expect to meet,
84 JULIE COLEMAN
As woud some Painter, busie in a Street,

To Copy Bulls and Bears, and evry Sign
That calls the staring Sots to nasty Wine.
(Earl of Roscommon, An Essay on Translated Verse, 1685)
but it seems unlikely, in any case, that the editor of a slang dictionary could
wholeheartedly approve of Roscommons views.
13. The preface promises that this will be a dictionary of the slang of the fashionable world,
but it is actually just another edition of Grose.
14. Buckingers Boot The monosyllable. Matthew Buckinger was born without hands and
legs; notwithstanding which he drew coats of arms very neatly, and could write the Lords
Prayer within the compass of a shilling; he was married to a tall handsome woman, and
traversed the country, shewing himself for money. (Grose 1788, and subsequent edi-
tions). Twiddle-diddles Testicles (Grose 1796, and subsequent editions). Bang-up
probably derives both from Grose 1811.
15. Originally attributed to Samuel Rowlands, but now to Samuel Rid (see Judges 1930:
514).
16. For example, Dekker (1608) has ruff-beck bacon, but by 1612 the form is ruff-peck, as
in S.R.
17. A similar request, but this time made to a prisoner in Pennsylvania Penitentiary, was
made by the author of an article called The Flash Language in the Ladies Repository in
1848. Alcoholic inducements were not involved on this occasion.
References
Awdeley, John. 1565. The Fraternitie of Vacabondes. London: John Awdeley.

Bang-up Dictionary. 1812. London: M. Jones.
Carew, Bampfylde-Moore. 1745. The Life and Adventures of Bampfylde-Moore Carew.
London: Printed by the Farleys for Joseph Drew.
Cawdrey, Robert. 1604. A Table Alphabeticall. London: I.R. for Edmund Weaver.
Coles, Elisha. 1676. An English Dictionary. London: Samuel Crouch.
Defoe, Daniel. 1728. Street Robberies Considered: The Reason Of their being so Frequent,
with Probable Means to Prevent em Written by a Converted Thief. To which is
prefixd some Memoirs of his LIFE. London: J. Roberts.
Dekker, Thomas. 1608. The Belman of London. London: Nathaniel Butter.
Dekker, Thomas. 1612. O per se O. London: John Busby.
E., B. c.1698. A New Dictionary of the Terms Ancient and Modern of the Canting Crew.
London: W. Hawes.
The Flash Language. The Ladies Repository. Oct. 1848. 3157.
Gotti, Maurizio. 1996. The canting terms of Coles dictionary. Linguistica e Filologia
3.231 252.
Greene, Robert. 1591. A Notable Discovery of Coosnage. London: John Wolfe for T.N.
Greene, Robert. 1591. The Second Part of Cony-Catching. London: John Wolfe.
Greene, Robert. 1592. Groundworke of Cony-catching. London: John Danter for William
Barley.
Greene, Robert. 1592. The Third and Last Part of Cony-catching. London: Thomas Scarlet
for Cutberd Burbie.
Greene, Robert. 1592. A Disputation between a He Cony-catcher, and a She Cony-catcher.
London: A.I. for T.G.
Grose, Francis. 1785. A Classical Dictionary of the Vulgar Tongue. London: S. Hooper.
Grose, Francis. 1788. A Classical Dictionary of the Vulgar Tongue. London: S. Hooper.
Grose, Francis. 1796. A Classical Dictionary of the Vulgar Tongue. London: Hooper & Co.
Grose, Francis. 1811. Lexicon Balatronicum. London: C. Chapel.
Grose, Francis. see also Harris (1980), a reprint of Grose (1811), and Lovric (1997), a
compilation from the first three editions of Groses Classical Dictionary.
Harman, Thomas. 1567. Caveat or Warening for Common Cursetors. London: William
Griffith.
Harris, Max. 1890. 1811 Dictionary of the Vulgar Tongue. A Dictionary of Buckish Slang,
University Wit, and Pickpocket Eloquence. Adelaide: Bibliophile Books.
Head, Richard. 1665. The English Rogue. London: Henry Marsh.
Head, Richard. 1674. The Canting Academy. London: F. Leach for Mat. Drew.
Hell upon Earth: or the most Pleasant and Delectable History of Whittingtons Colledge,
Otherwise (vulgarly) called Newgate. 1703, London: no publishers details.
Hotten, John Camden. 1865. The Slang Dictionary. London: John Camden Hotten.
Judges, A.V. 1930. The Elizabethan Underworld. A collection of Tudor and early Stuart
tracts and ballads telling of the lives and misdoing of vagabonds, thieves, rogues and
cozeners, and giving some account of the operation of the criminal law. London:
Routledge.
L., G. 1742. The Amorous Gallants Tongue. London: C. Hitch, L.Hawes, T. King et al.
Latroon, Meriton. 1671. The English Rogue. London: Frans. Kirkman.
Lovric, Michelle. 1997. The Scoundrels Dictionary. A Copious and Complete Compen-
dium of 18th-Century Slang. Oxford: Past Times.
Matsell, George W. 1859. Vocabulum. New York: George W. Matsell & Co.
Matsell, George W. 1997. reprinted as The Secret Language of Crime. The Rogues Lexicon
Springfield, Illinois: Templegate.
Mugglestone, Lynda. 1995. Talking Proper. The Rise of Accent as Social Symbol. Oxford:
Clarendon.
Murray, J.A.H. et al., eds. 1989. The Oxford English Dictionary. 2nd ed. Oxford: Clarendon.
New Canting Dictionary. 1725. London: The Booksellers.
Parker, George. 1789. Lifes Painter of Variegated Characters. London: J. Ridgway.
Reibel, David, ed. 1996. Lindley Murrays Grammar in Caricature: Four Parodies.
London: Routledge.
R[id], S[amuel]. 1610. Martin Markall. London: for John Budge and Richard Bonian.
Roscommon (Wentworth Dillon). 1685. An Essay on Translated Verse. 2nd ed. English
Poetry, software version 4.0, Chadwyck-Healey Ltd, 1995.
Saman. 1725. The Golden Cabinet of Secrets. London: A. Bettesworth, C. Hitch, J. Osborn
et al.
86 JULIE COLEMAN
The Scoundrels Dictionary. 1754. London: J. Brownell.

Sharp, Jeremy. 1741. The Life of an English Rogue. London: T. Read.
Shirley, John. 1724. The Triumph of Wit, or ingenuity displayd in its perfection. 8th ed.
London: A. W. for J. Clarke.
Smith, Alexander. 1719. The Thieves New Canting Dictionary. London: Sam Briscoe.
Smith, Alexander. 1719. The Thieves Grammar. London: Sam Briscoe.
Tufts, Henry. 1807. A Narrative of the Life, Adventures, Travels and Sufferings of Henry
Tufts. Dover, New Hampshire: Samuel Bragg, jr (from Edmund Pearson, ed. 1931. The
Autobiography of a Criminal. Henry Tufts. London: Jarrolds).
Wild, Jonathon. 1725. History of the Lives and Actions of Jonathon Wild. London: Edw.
Midwinter.
Winstanley, William. 1669. The New Help to Discourse: or, wit, mirth, and jollity intermixt
with more serious matters. London: T. J.
Yokels Preceptor ?1855. London: H. Smith.
Lexical Choices in
an Early Galilean Translation
Maurizio Gotti
Universit di Bergamo
Abstract
The paper takes into consideration Thomas Salusburys rendering of Galileos

main scientific works (1661 and 1665), focusing in particular on his lexical
choices, in order to point out the criteria followed in the selection of appropri-
ate equivalents and in case of lack of suitable English words in the
creation of the new terminology required. In order to evaluate the translation
methodology followed by Salusbury and in particular the criteria adopted in
his rendering of Galileos terms, a few passages of the Dialogue on the Great
World Systems have been analysed. This work, which can be considered
fundamental for the growth of scientific thought in this period and which has
greatly influenced the evolution of several disciplines not only in England but
all over Europe, introduced many new terms, which were soon included in the
specialized terminologies of the various European languages.
The seventeenth century was a period of great increase in the English vocabu-
lary and certainly one of the most productive in the whole history of that
language (cf. Finkenstdt et al. 1970). A great contribution to this lexical
growth was provided by translations of foreign texts, particularly those pertain-
ing to the various scientific branches, in which important and rapid innovations
were continuously brought about all over Europe. The great epistemological
and methodological evolutions taking place in that period determined the need
for corresponding changes both in the ways of communicating the new
discoveries attained by means of innovative procedures and apparatus, and in
88 MAURIZIO GOTTI
the expressive tools to be used to describe and argue about the new phenomena
observed and analysed.
This need for linguistic systematization and improvement caused a
heated debate in 17th century England about the coining of new terminology
and the style to be adopted in writing specialized texts.1 Specific accusations
were made against English, particularly because of the limited vocabulary
present in that language. The field in which English proved particularly
inadequate was that of termes of art, an expression commonly used in that
period to refer to the technical words that make up the basic lexis of a subject.
The criterion most frequently adopted to overcome this handicap was the
borrowing of terms from other languages, particularly from Latin.2 The choice
of a loan rather than the specialization of an existing word or the coinage of a
new term was often suggested by the fact that the concept to be referred to was
already expressed in a foreign language by an existing term. The availability
of such a term was particularly evident in the case of translation of texts
written in a foreign language. In such a case when the translator came across a
word with no equivalent in English, he found it easier to use the original word,
thus enriching the lexical load of the receiving tongue. Moreover, once a loan-
word had been introduced, it was commonly used as a root from which further
words could be formed by means of affixation.
The present paper will take into consideration one of these translations
that is, Thomas Salusburys rendering of Galileos main scientific works
focusing in particular on his lexical choices, in order to point out the criteria
followed in the selection of appropriate equivalents and in case of lack of
suitable English words in the creation of the new terminology required.
The Text and the Author
Not much is known about Thomas Salusbury. Indeed, the few fairly reliable
details about his life have become available only with the recent discovery of
a series of letters written by him (cf. Zeitlin 1959). The information deduced
from them has suggested a date of birth in the decade 162030, his belonging
to the royalist movement, and a period of several years spent in Italy and
France before the return of Charles II to England. His stay in those countries
enabled him to perfect his knowledge of the local languages and particu-
larly in Italy to cultivate his interest in astronomical and physical matters.
LEXICAL CHOICES IN TRANSLATION 89
On those subjects he translated several works in particular those written by

Galileo besides writing a treatise of his own on the comparative gravity of
bodies in air and water.
Thomas Salusburys main body of translations is to be found in his
Mathematical Collections and Translations, which consists of two volumes,
the first published in 1661 and the second in 1665. Of the latter, only eight
copies exist. These are incomplete, as they contain only the first part of the
second volume; the second part seems to have been destroyed by the Great
Fire of London, just one month after Salusburys death, which took place in
August 1666. Complete copies of the second volume, however, are very likely
to have existed, as the presence of one of them in the library of the Earl of
Macclesfield housed at Shirburn Castle has been reported by various readers;
the latest testimony is dated 1829 a few passages from Salusburys second
part of the second volume were quoted directly by John Elliot Drinkwater-
Bethune but since that date the book has disappeared.
The importance of these Mathematical Collections and Translations
must have been great, particularly for Salusburys contemporaries, as they
contained the first English version of several works written by Galileo,3 and
remained the only version translated into that language for almost three
centuries.4 The reason that mainly prompted Salusbury to provide an English
version of some of the most important innovative specialist works written in a
foreign language is of a pedagogic nature, as can be deduced from his own
testimony:5
Mathematical learning (to speak nothing touching the necessity and delight
thereof) has been so sparingly imparted to our countrymen in their native
English, especially the nobler and sublimer part, that in compliance with the
solicitations of several of my noble and learned friends and the inclinations of
such as are mathematically disposed, more especially those who either want
time or patience to look into the vulgar and unstudied languages, I did
adventure upon this work of collecting and translating from among the
excellent pieces that are so abounding in the Italian and French tongues some
of those that from my own observation and from the intimation of friends
were most useful and desired and, withal, most wanting in their own.
(Salusbury 1661: foreword)
Only recently have these translations been commented on by scholars, and

they have received contrasting evaluations. Giorgio de Santillana, who edited
a reprint of Salusburys translation of Galileos Dialogue on the Great World
Systems, although admitting some merits on the part of the translator,6 finds
90 MAURIZIO GOTTI
that his translating is miserably unreliable. [] One wonders what the

translator must have thought on rereading the nonsense that he had written
down. There is a suspicion that he never did. (Santillana 1953: XXXII)
This very negative comment may be due to the fact that the volume of the
Mathematical Collections and Translations that Santillana was working on
did not include the leaf of errata (copies of which were inserted in other
volumes) in which the author pointed out the many misspellings, alterations of
punctuation and other mistakes that the printer of his work had made.
A very appreciative comment comes, in contrast, from the second editor
of Salusburys translation of the Dialogo:
In his own edition Salusbury was meticulously faithful to his original, and
never edited out or deliberately omitted a word of Galileos text (or for that
matter of the almost equally important marginal notes and glosses). []
Salusburys edition of the Dialogue, even neglecting the errata, ranks among
the good scientific translations (Drake 1958: 2627).
Apart from these general statements, no specific evaluation has been

made of Salusburys rendering of Galileos specific terms and specialized
concepts.
Salusburys Approach to Translation
In order to evaluate the translation methodology followed by Salusbury, and

in particular the criteria adopted in his rendering of Galileos terms from the
many texts contained in the Mathematical Collections and Translations, I
have chosen to analyze a few pages of the Dialogue on the Great World
Systems, a work which can be considered fundamental for the development of
scientific thought in this period and which has greatly influenced the evolution
of several scientific disciplines not only in England but all over Europe. The
text, which presents a well-reasoned defence of the Copernican innovative
views opposed to the traditional Ptolemaic system, introduced several new
terms, which were soon included in the specialized terminologies of the
various European languages.
The corpus analysed consists of two passages Galileo (1632/1982:
393407 and 460470) and Salusbury (1661: 302308 and 352361) taken
from the third day, in which Galileo discusses the annual movement of the
earth and reports on a series of experiments and observations carried out by
means of a telescope. In these passages many specialized terms are made use
of, several of which have been coined for the first time by Galileo.
The first feature that stands out from this analysis is the faithfulness with
which Salusbury translates Galileos dialogue, not only in rendering the single
lexemes, but in reproducing the whole of the original text. Indeed, the sentences
are usually translated in a very similar form, and only minor alterations in word
order or syntactic construction are made in order to satisfy specific rules which
diverge in the two languages. This is the case, for example, with the nominal-
ization of verbal forms, which is commonly rendered in Italian with the
infinitive of the verb preceded by the definite article, a construction which is
not typical of the English language. To translate such forms, Salusbury resorts
to different means. At times he uses the -ing form, which enables him to leave
the original construction of the sentence unaltered:
(1) Il non mutar figura in Venere. (p. 399)
The not changing figure in Venus. (p. 302)
(2) Dal non si inclinare o elevar gi mai. (p. 467)
From its not inclining or elevating at all. (p. 358)
(3) Circa lalzarsi e abbassarsi. (p. 468)
About its elevating or declining. (p. 358)
On other occasions he uses a noun, a device which again enables him not
to alter the word order of the original sentence:
(4) Lo strumento stesso del vedere. (p. 401)
The very instrument of sight. (p. 303)
In other cases Salusbury manages to make use of an infinitive himself,

although not preceded by the article:
(5) Il far che tutti i pianeti, insieme con la Terra, si muovano intorno al
Sole. (p. 399)
To make the Planets, together with the Earth, to move about the Sun. (p.
302)
(6) E qui rimetto al vostro parere il giudicare quello che abbia pi del
verisimile. (p. 470)
And here I refer it to your judgement to determine which of the two is
the most probable. (p. 361)
Salusburys faithfulness to the form of the original text leads him at times
to use constructions which are typical of Italian but rare in English, such as the
placing of an adjective after a noun, as the following example testifies:
92 MAURIZIO GOTTI
(7) Nelle parti inferiori. (p. 406)

In the parts inferiour. (p. 307)
It must be said, however, that on the whole the form of Salusburys

translation follows the standard rules of the language and although it
succeeds in maintaining the characteristics of Galileos prose it reads in a
very natural English style. There are times, however, when Salusbury has to
make slight alterations to the word order of the original sentence, or modifica-
tions to the syntactic function of some of its elements, so as to make them
sound appropriate in English. Here are a few examples:
(8) Della poco variata grandezza di lei. (p. 400)
Of her small variation of Magnitude. (p.302)
(9) La fama della sublimit del suo ingegno. (p. 405)
The fame of his sublime wit. (p. 306)
(10) Per esser egli cos vicino al Sole. (p. 406)
By reason of its vicinity to the Sun. (p. 307)
Salusburys Translation of Specialized Terms
In his choice of terms too, Salusbury generally makes use of common words
which are the direct equivalent of Galileos. This direct transposition is made
easier by the fact that a large amount of specialized terminology is very similar
in the two languages. Below, for example, is a list of words drawn from the
two texts which are clearly very much alike (the dates appearing after the
English terms are those of their first quotations reported by the OED):
congiunzione conjunction (1374)
emisferio hemisphere (1374)
opposizione opposition (1386)
risplendente resplendent (1448)
splendor splendour (1450)
umidit humidity (1450)
sferico spherical (1523)
apparizione apparition (152530)
riflettersi reflect (1530)
superficie superficies (1530)
conversione conversion (15401)
punti cardinali cardinal points (1549)
digressione digression (1552)
illuminazione illumination (1563)
pupilla pupil (1567)

cilindro cylinder (1570)
semidiurno semidiurnal (1594)
seminotturno seminocturnal (1594)
telescopio telescope (1619)
opaco opacous (1621)
eclissato eclipsed (1633)
meridiano meridian (1633).
Similarity of form can be found also in many non-technical words em-

ployed by Salusbury in his translation. These, however, have not been in-
vented by him following Galileos example, but were commonly used
especially in the written language in English texts, which show a frequent
use of Latinate forms.7 Here is a short sample of such words with the
indication of their first quotation in the OED:
terminato terminate (1432)
vivacit vivacity (1432)
mattutino matutine (1445)
avvertimento advertisement (1475)
amplo ample (1485)
vespertino vespertine (1508)
situato situate (1523)
assurdit absurdity (1528)
cospicuo conspicuous (1545)
stupendo stupendious (1547)
perspicacit perspicacity (1548)
insensibilmente insensibly (1584)
eminente eminent (1588)
avventizio adventitious (1603)
vivace vivacious (1645).
In inventing new technical terms, Galileo generally adopts words used in

everyday language or present in the basic terminology of the various scientific
disciplines and gives them his own specific meaning. This same process of
specialization is also followed by Salusbury, who finds such a process facili-
tated by the fact that the words to be given a specialized meaning have a form
which is similar to the Italian one. A few examples of these terms deriving
from a process of specialization are cono/cone (in the expression cone of
shade), esatto/exact (applied to very precise instruments), irradiazione/irra-
diation (with its optical sense), momento/moment (with the meaning of mo-
mentum, force), occultato/occulted (as used in astronomy), orbo/orb (with its
astronomical meaning), parallelo/parallel (in its astronomical sense), sistema/
94 MAURIZIO GOTTI
systeme (with its astronomical value) and trasmettere/transmit (in its physical
sense).
In order to maintain a form similar to that of the Italian version, Salusbury
sometimes adopts an existing term and assigns a new meaning to it. This
process of semantic innovation is followed, for instance, in rendering the
adjective gioviale, used by Galileo to refer to the moons of the planet Jupiter;
in this case the English translator uses the existing adjective jovial, up to then
commonly used with reference to the god Jove.
In Salusburys use of the term moons, one can see a recourse to the
process of calquing. Indeed, this term was commonly used in English to refer
to the earths satellite and was adopted by Salusbury in his translation to refer
to the satellites of Jupiter. In doing this, Salusbury imitates the word-forma-
tion process followed by Galileo in using the term lune for the same purpose.
Another instance of imitation of form can be found in Salusburys render-
ing of Galileos cristalli referring to the lenses8 of the telescope (a clear case of
metonymy), for which Salusbury makes use of the existing word chrystals, a
word which was mainly used to indicate ice or minerals as well as glass, of
which the telescope lenses are made.
On other occasions Salusbury makes use of the process of conversion, by
means of which an existing word is inserted in a text with a syntactic function
which is different from the usual one. An example of this can be seen in the
translation of the Italian adjective reflesso, for which Salusbury makes use of
the term reflex (which had only been used as a noun up to then) to perform the
same adjectival function.
A form similar to the Italian original is also adopted by Salusbury in the
creation of those neologisms which were needed to render terms which had no
equivalent in the English language. At times, the form of these loans remains
the same, facilitated by the presence not only of the same root but also of the
same suffix in the two languages. This is the case with terminator, which,
however, appears in Salusburys text with only a nominal value, while the
same word is used by Galileo in the double function of adjective and noun.
In other cases the Italian suffixes are rendered with the English ones that
are most similar to them: examples can be seen in the loans Mediceo/Medican
and capellizio/capellitious or capillitious.9 As regards the latter, the addition of
a suffix commonly used for adjectives causes Salusbury to use this neologism
only in adjectival position, even when Galileo uses it as a noun, as in the case
of capellizio radioso (p. 401) rendered as capillitious rayes (p. 303).
In the case of Galileos term disco, Salusbury adopts the form discus, a
word generally found in works written in Latin and which was starting to be
used also in English, although not in its astronomic sense, as can be seen from
its earliest quotation in the OED taken from Cowleys Pindarique Odes:
The chief Exercises there were Running, Leaping, Wrestling, the Discus,
which was the casting of a great round Stone, or Ball, made of Iron or Brass.
(Cowley 1656, iii: note)
As regards units of measurement, these are not translated literally, but

rendered with the nearest equivalent. Thus Galileos dito (literally finger,
roughly equivalent to one inch) is translated by Salusbury into inch, while
braccio (literally arm, corresponding approximately to 21 inches cf.
Drake 1960: XXV) is translated as yard.
Salusburys Rendering of Metaphorical Language
In translating the Italian text, Salusbury takes great care to preserve the
figurative language of the original. Galileo uses a colloquial form of Italian,
particularly suitable for the expressive genre that he has chosen, that is, that of
the dialogue. Moreover, his style is enriched by the use of metaphors, which
render his text admirable also from the literary point of view. Some of the
metaphors that Galileo makes use of are dead metaphors, that is, figures of
speech that have become common in the language and are therefore perceived
as unmarked by interlocutors. One of these is ghirlanda, used in the figurative
connotation of ring or circular band; to render this metaphor, Salusbury uses
the word garland, which had been in use for at least one century10 with that
semantic value.
It is interesting to observe, however, that in translating Galileos meta-
phorical language, Salusbury avoids the inventing of neologisms when figura-
tive terms already exist. This can be seen, for example, in the rendering of
inghirlandato, for which Salusbury rather than coining the derived form
garlanded from the existing verb to garland makes use of the adjective
fringed, already popular with that figurative value. Another English figure of
speech common at that time is shining locks, which is adopted by Salusbury to
render Galileos metaphorical expression crini risplendenti to refer to the
phenomenon of the irradiation of the sun.
96 MAURIZIO GOTTI
In certain cases Salusbury provides two equivalents for the same figura-
tive expression. For example, the Italian adjective falcata, used by Galileo to
refer to the shape of the moon as similar to that of a sickle (falce in Italian) is
rendered by Salusbury by means of the two adjectives forked or horned, both
of which were part of contemporary common language.11 Another double
rendering of a metaphor is used in the case of crini, an Italian term denoting
hair that Galileo uses with reference to the suns rays, for which the English
author provides the double translation hair or fringe.
Sometimes the second equivalent is added as a sort of explanation for the
first term, which is of a more technical nature. This can be seen, for instance,
in Salusburys rendering of irragiarsi as irradiate or beam forth rayes, of
apparenze as Phnomena or appearances and of vertice as apex or top.
Indeed, the three technical terms given as first equivalents were quite recent
borrowings (the first quotation in the OED for irradiate is dated 1617, for
phnomena 1605 and for apex 1610) and the English translator was thus
careful to accompany them with more familiar synonymic expressions.
The explanatory purpose of these binary renderings is indicative of the
care for clarity and precision that characterizes also other parts of Salusburys
translation, and which at times leads him to use a paraphrase rather than a
direct equivalent. An example of this lexical choice can be seen in Salusburys
use of the paraphrase the observators standing in Jupiter (p. 308) instead of
the nominalized form of the existing adjective Jovial to translate Galileos
innovative use of Gioviali (p. 406). In another case, the risk of ambiguity
which might derive from using the word mistress endowed with so many
meanings, not all of a positive nature in the translation of Galileos phrase
la vera maestra ci insegna (p. 403, literally the true mistress teaches us, in
which mistress has the semantic value of female teacher) leads Salusbury to
accompany that term with appropriate specifications and to provide the fol-
lowing equivalent text: the true Mistris of Astronomy, Experience, teacheth us
(p. 305).
A detailed analysis of the Italian and the English versions shows that on
the whole there are fewer metaphorical expressions in Salusburys text than in
the original, since some figures of speech are rendered in a more technical way
by the English translator. Examples of this kind of lexical choice occur in the
rendering of Galileos capellatura (p. 404, literally hair) as irradiation (p.
305), and of corna (p. 405, horns [of the moon]) as crescents (p. 306).12 In
the case of non-figurative language too, Salusbury often renders Galileos
expressions with more specific terms, as can be seen in the translation of mezo
cerchio (p. 406, literally half circle) as Semicircle (p. 307) or of punto
ugualmente lontano (p. 463, equally distant point) as point equidistant (p.
354).
Salusburys Influence on the English Language of Science
As the present analysis has shown, Salusburys faithful rendering of Galileos

work led him to the coining of new terms by means of the processes of
borrowing and calquing or to the adoption of existing lexical elements en-
dowed with new meanings, thus making use of the semantic redefinition
processes of specialization, semantic innovation, metonymy and conversion.
Moreover, the new terms were at times accompanied by synonyms or para-
phrases so as to be made more easily understandable.
The influence of Salusburys Galilean translations on the English lexicon
is difficult to assess. Although his works are rarely mentioned as providing the
earliest instances of neological formations in the OED,13 the great majority of
his lexical innovations are to be found in that dictionary,14 which means that
they were quickly adopted by contemporary English scientists and included in
their texts. It is also possible that Salusburys terms could have been indepen-
dently coined by other writers at the same time rather than their re-use being
an indication of his influence. To assess the validity of these two hypotheses
the testimony of the great men of science of the period is unfortunately not of
great assistance to us, as Newton is the only leading figure who clearly
mentions his reading of Salusburys translation of Galileos Dialogo (cf.
Turnbull 1963, III: 52).
Salusburys influence on British scientists of his time may have been
limited both by the habit of many eminent members of the Royal Society of
reading the works of the most important foreign scientific innovators in their
original language, or more frequently in their Latin versions,15 and by
the too limited number of copies available of his translations, due to the almost
complete loss caused by the Great Fire of London of 1666.
Certainly other figures have been more influential in the development of
English scientific language. Robert Boyle, for example, is quoted 632 times in
the OED for words which either appear for the first time in his works or are
employed by him with a meaning different from those previously recorded in
98 MAURIZIO GOTTI
that dictionary. (For an analysis of the new words in Boyles texts cf. Gotti
1996: Chapter 3.) This higher degree of influence may be due to the greater
number of works written by Robert Boyle, the wider range of disciplines dealt
with, and his particular importance in the scientific world of his time, also
testified to by his leading role in the early life of the Royal Society.
However, even analyses of minor works of this age, such as that de-
scribed above, may prove to be very interesting, as they can provide important
details to support and integrate the main views commonly held on the forma-
tion of modern scientific language. Indeed, in the corpus taken into consider-
ation here we have found a confirmation of the great linguistic creativity
characteristic of this period and of the main word-formation processes
adopted to facilitate the rapid lexical growth that was so greatly needed by
English men of science in the 17th century.
Notes
1. For an analysis of various opinions expressed in this heated debate on the language of
science in 17th century England cf. Gotti 1996: Chapter 1.
2. As Barbers analysis established (1976: 166195), many of the new words that became
part of the language in the 17th century were Latin loan-words, and many of them
belonged to specialized fields of discourse.
3. A previous translation into English (attributed to Joseph Webbe) of Galileos Dialogue
had been made (a mention of it can be found in a letter written by Thomas Hobbes to
William Cavendish, Earl of Newcastle, in January 1634 (Drake 1967: 2), but was never
published. The manuscript, which can be found in the British Museum, had no influence
on Salusbury, as this English translation is quite different in style from Salusburys and
was doubtless unknown to him (Drake 1967: 3).
4. As regards Galileos Dialogo, Salusburys first printed translation into English was not
followed until almost three centuries later by that published by Stillman Drake in 1953.
5. It is interesting to note that, in providing a translation into the vernacular, Salusbury
claims a motivation similar to that which had led Galileo to choose Italian rather than
Latin for the composition of many of his books, and made explicit in a letter written to
Paolo Gualdo:
I am induced to do this by seeing how young men are sent through the
universities at random to be made physicians, philosophers, and so on; thus
many of them are committed to professions for which they are unsuited,
while other men who would be fitted for these are taken up by family cares
and other occupations remote from literature. [] Now I want them to see
that just as nature has given to them, as well as philosophers, eyes to see her
works, so she has also given them brains capable of penetrating and under-
standing them (Translated by Drake 1957: 84).
6. These are the positive evaluations that Giorgio de Santillana expresses:
The Salusbury translation, with all its faults, remains better (once corrected)
than any modern one could be. Although we had to modernize it and shorten
the sentences, the text preserves a measure of the original spirit. It has in it
the leisurely unrolling of seventeenth-century thought, with that peculiar
ingenuous quality that would be lost in any imitation (Santillana 1953:
XXXII).
7. A confirmation of this behaviour can be found in the following words by Walter
Charleton, who clearly identifies its motivation in the limitedness of the existing English
lexis:
If my Stile shall sound somewhat harsh and ungrateful many times to Ears
unaccustomed to any but their Mother tongue, as coming too near to the
Latin; I intreat you to consider, this is either no indecency in this place, or
such a one at worst, which I could not otherwise avoid, than by involving
my sense in the obscurity of words less proper and significant; the nature
and quality of Subjects treated of, being such, as cannot be fully expressed
in our yet imperfect Language (Charleton 1680: Sig.E3v).
8. The term lens is first reported in the OED in a quotation dated 1693.
9. This can be considered a nonce word, as it appears not to have been used by other people.
Indeed, it is not even listed in the OED.
10. The OED reports the following as the earliest quotation of this metaphorical use of
garland: Round about the edge of the urine there appeareth a garland, circle, or ring
(1548, Recorde, Urin. Physick x. (1651) 81).
11. The Italianate form falcated is first reported in Harris Lexicon Technicum (1704) with
the following explanation:
The Astronomers say the Moon, or any Planet appear falcated, when the
enlightned part appears in the Form of a Sickle, or Reaping-hook; which is
while she is moving from the Conjunction to the Opposition, or from New
Moon to Full; but from Full to a New again, the enlightened part appears
Gibbous, and the dark Falcated.
12. In behaving thus, Salusbury seems to be following the dictates of the Royal Society to
reject all the amplifications, digressions, and swellings of style: to return back to the
primitive purity, and shortness, when men deliverd so many things, almost in an equal
number of words (Sprat 1667/1959: 113).
13. Salusburys works are quoted only twice in the OED as sources of neologisms, namely
equijacent and the physical meaning of force. It should be remembered, however, that the
method by which the OED was compiled meant that assistants read texts and noted usages
that struck them as novel or unusual, a practice that would certainly favour works by more
famous writers.
100 MAURIZIO GOTTI
14. Indeed, of all the neologisms found in our corpus, only the term capellitious/capillitious
is not reported in the OED.
15. A confirmation of this practice can be found in the following words by Hall: We may be
sure that for every Englishman who read Galileo in Italian or Descartes in French there
were ten who read the alternative Latin editions (Hall 1961: 26). Latin versions of
Galileos works were available in most European countries, including England, where
some of these versions were published by local printers.
References
Barber, Charles. 1976. Early Modern English. London: Andre Deutsch.

Charleton, Walter. 1680. Enquiries into Human Nature, VI.Anatomic Praelections. Lon-
don.
Cowley, Abraham. 1656. Pindarique Odes. London.
Drake, Stillman. 1957. Discoveries and Opinions of Galileo. New York.
Drake, Stillman. 1958. Galileo Gleanings II A Kind Word for Salusbury. Isis IL:
155.2633.
Drake, Stillman. 1960. Introduction to Galileo Galileis Discourse on Bodies in Water.
Translation by Thomas Salusbury edited by Stillman Drake, IX-XXVI. Urbana: Univer-
sity of Illinois Press.
Drake, Stillman. 1967. Introduction to Thomas Salusburys Mathematical Collections and
Translations. Facsimile edition, London: Dawson of Pall Mall, and Los Angeles: Zeitlin
& Ver Brugge, 127.
Elliott Drinkwater-Bethune, John. 1829. Life of Galileo Galilei with Illustrations of the
Advancement of Experimental Philosophy, London.
Finkenstdt, Thomas, Ernst Leisi and Dieter Wolff, eds. 1970. A Chronological English
Dictionary, Heidelberg: Carl Winter Universittsverlag.
Galilei, Galileo. 1982. Dialogo sopra i due Massimi Sistemi del Mondo, Edited by L. Sosio.
Turin: Einaudi. (First edition: Florence, 1632.)
Gotti, Maurizio. 1996. Robert Boyle and the Language of Science, Milan: Guerini.
Hall, A. Rupert. 1961. English Scientific Literature in the Seventeenth Century, Scientific
Literature in Sixteenth & Seventeenth Century England ed. by C. Donald OMalley and
A. Rupert Hall, 2346. Los Angeles: William Andrews Clark Memorial Library,
University of California.
Oxford English Dictionary on Historical Principles, 2nd edition, Oxford: Clarendon Press
1989.
Salusbury, Thomas. 1661 and 1665. Mathematical Collections and Translations, London:
William Leybourn.
de Santillana, Giorgio. 1953. Historical Introduction to Galileo Galileis Dialogue on the
Great World Systems (in the Salusbury Translation), XI-LVIII. Chicago: University of
Chicago Press.
Sprat, Thomas. 1667. A History of the Royal Society of London for the Improving of Natural
Science. London: Martyn; rpt. London: Routledge & Kegan Paul 1959.
Turnbull, H., ed. 1963. Correspondence of Isaac Newton, Cambridge: Cambridge Univer-
sity Press.
Zeitlin, Jacob. 1959. Thomas Salusbury Discovered. Isis L:162.455458.
Grund to Hrof: Aspects of the Old English
Semantics of Building and Architecture1
C.P. Biggam
Abstract
The paper builds two Anglo-Saxon virtual buildings, a stone church and a
timber house or small hall, using Old English vocabulary. At each phase of the
construction, certain of the vocabulary is discussed, drawing on evidence from
archaeology and surviving buildings, from etymology, and from contemporary
translation from Latin. The aim is to investigate the potential of interdiscipli-
nary studies in the field of Anglo-Saxon architecture and building processes.
Introduction
On first consideration, it would seem highly productive to combine the forces

of archaeological and semantic research to elucidate the subject of Anglo-
Saxon buildings (Grundy & Roberts 1997). More detailed investigation, how-
ever, often leads to disappointment, as Old English lexemes, usually occurring
without elaborate explanation, are difficult to link to excavated structural
remains, which are not, of course, conveniently labelled with Old English
technical building terms. Nevertheless, it has been found profitable to link
evidence from the two disciplines in certain cases, and the aim of this article is
to explore the possibilities for interdisciplinary research in this area.
There is plenty of evidence to consider. A Thesaurus of Old English
(TOE) lists more than 630 words connected with building, building materials,
types of building, parts of buildings, and rooms within buildings (Roberts &
104 C.P. BIGGAM
Kay 1995: 1: 23141), while the archaeological evidence includes over 260
Anglo-Saxon stone churches, surviving in whole or in part above the ground
surface (Taylor & Taylor 196578), and much more evidence from excavated
stone and timber structures. In this article, two virtual buildings will be
constructed from foundations to roof, one in stone and one in timber, and
various semantic points will be considered in the process.
Foundations
Wisely starting with foundations, it can be seen that the lexeme grund has an
important role as both a simplex and as one element of compound terms. It can
mean foundation, or earth, the latter like ModE ground, and it can denote
the bottom or basal part of anything, including the bed of the sea and the
depths of hell.2 In the context of buildings, it could indicate the ground surface
under the structure, or the foundations below ground, both invisible after the
structure has been built, or the lowest visible part of the building, which forms
the apparent base of the structure once the foundations have been covered.
These possibilities will now be explored.
Grund can certainly indicate the ground surface, as it does in the poem
Andreas (line 776)3 in which reference is made to grene grundas, green
plains (Bradley 1982: 130), but it can also refer to below-ground features.
When Christ likened those who follow his teachings to a man who builds his
house on rock (Luke 6.48), the Old English of the Lindisfarne Gospels reads
delf on heanise [and] gesette a grundas of[er] carr [ve]l stan, [who]
dug in depth and set the grundas on rock or stone (Skeat 187187 (Luke):
71). The Anglo-Saxon interpretation of this passage would, presumably,
envisage the builder digging foundation trenches down to the bed-rock. The
grundas, in this case, are clearly below ground level. There are no clear cases
in Old English texts of grundas referring to the visible base of a wall, and,
although it would be unwise to exclude this possibility, the situation may be
understandable in view of the suggestion in the next paragraph.
Grund occurs in the compound term, grundweall, which is also defined
as foundation. Is there a subtle difference in meaning from that of grund as a
simplex? A crucial passage occurs in King Alfreds translation of Gregorys
Cura Pastoralis, in which instructions are given that the irresolute and weak-
willed of the Christian flock are to be encouraged to be strong. The metaphor
GRUND TO HROF 105
is presented of a strong wall needing to be built on firm foundations: y

mon r gehawige t se grund fs sie, r mon one grundweall
onlecgge, where the grund has been seen to be firm, there the grundweall
is laid (Gregory I 1871: 2: 308; ch. 42, line 934). This passage suggests an
elementary difference between the two words, in that grund refers to horizon-
tal foundations, that is, the surface on which the building is raised, whether it
is at ground level or at the bottom of a trench, while grundweall refers to
vertical foundations, namely, the lowest part of the walls, whether they are
above or below the ground. This suggestion is confirmed by the meaning of
the simplex weall, wall, which has a core meaning of a vertical, or near
vertical, face, being applicable to a wall but also to an earthen embankment,
such as a rampart or dam, or to a natural cliff.
Clearly, vertical foundations below ground could have been grund-
weallas, and there is a text by lfric which confirms this. It is concerned with
the dedication of a church, and the reader is told dylf one grundweall
swye deopne [and] leg hine mid stane, he digs the grundweall very deep
and lays it with stone (Brotanek 1913: 14, lines 89). In this context, grund-
weall appears to indicate base-wall (trench), but it is clearly a vertical
feature, and it is below ground level.
To suggest appropriate features at ground surface level which could have
been called grundweallas, it is necessary to turn to the archaeological and
architectural evidence. First of all, buildings often have a distinctive lower
section of walling, which may be for extra strength and stability, protection
against damp and/or passing traffic, or for aesthetic effect. At the Saxon
church at Bradford-on-Avon, for example, the lower courses of the exterior
walls project beyond the surface of the wall above (Backhouse, Turner &
Webster 1984: pl. XXIX). It may have been possible to refer to a feature like
this as a grundweall, meaning something like a base-wall.4
The translation of grundweall as base-wall means it could also refer to
the plinth occurring at the base of certain Anglo-Saxon stone buildings. A
particularly massive example occurs at Repton, Derbyshire (Taylor 1971:
36569).
Another possibility for the meaning of grundweall is basal revetment
wall. A revetment is a retaining structure, designed to prevent the collapse or
slip of another structure or a natural feature, such as an earthen bank. While
some revetments can stand as high as the feature they are strengthening, others
are part of the internal construction of a wall or bank, or are lower structures
106 C.P. BIGGAM
built externally at ground level to contain slippage. Such revetments could

have been seen as basal features of the main structure. An example is the rear
wall of the Late Saxon Rampart F at South Cadbury (Alcock 1995: 46; illus.
2.1 & 3.1).
In conclusion, it would appear that the grund of a building is its horizontal
base, either below or on the ground surface, and that the grundweall is its
vertical base, which can certainly be below ground, but which may also be an
above-ground feature.
Another compound with grund is grundstan, which is defined as founda-
tion stone. This seems like a very reasonable interpretation of a compound
word meaning ground (or base) stone, especially when the basal stones of
Anglo-Saxon arches and doorways are not infrequently larger and more
noticeable than the other stones (Taylor & Taylor 196578: 3: 806). It is
suggested here, however, that this interpretation of grundstan is mistaken. In
the Microfiche Concordance to Old English (MCOE), grundstan only ever
occurs in the plural and in glossaries, where it translates the Latin word
cementum mortar (Healey & Venezky 1980; Latham 1965). It looks like an
Anglo-Saxon translation error, causing a word meaning mortar to emerge
with the meaning of foundation stone, but this is not the case.
The Classical Latin word caementum means small stones, rubble, used
in making a kind of concrete and it is based on the verb caedo, in this case,
meaning to break or crush up, referring to the rubble (Oxford Latin Diction-
ary: OLD). As the earlier caementum developed into the Mediaeval Latin
word cementum it gradually shifted its semantics away from the little stones in
a type of concrete, to any type of concrete or mortar, with or without rubble.
That shift appears not to have happened by the time of at least one of the Latin-
Old English glossary entries, since, in the Harley Glossary, grundstanas
translates cementa i petre mortars, that is, stones (Oliphant 1966: 70, line
794).
Cement with a rubble content is most suitable for foundations, so it seems
that OE grundstanas probably indicates foundations consisting of cement and
rubble, not a single foundation stone, as implied in dictionary definitions. This
semantic conclusion is well supported by archaeology. Taylor refers to foun-
dations at Repton as well-mortared rubble, and to most of the foundations at
Deerhurst as coursed rubble laid in good mortar (Taylor & Taylor 1965
78: 3: 761). Rodwell explains that, in lowland Britain, Anglo-Saxon church
foundations have been found to consist of alternate layers of coarse and
GRUND TO HROF 107
granular material like gravel, chalk or stone rubble, and finer material like
clay, sand or crushed mortar from demolished buildings (Rodwell 1986: 157).
It seems likely that grundstanas would refer only to the former type of
material.
Timber buildings were far more common in Anglo-Saxon England than
stone-built structures (Addyman 1972). They are found on excavated sites as
marks of different coloured soil, most usually as patterns of post-holes, and/or
as long foundation trenches (Rahtz 1976: 82). Substantial vertical timbers
were placed in each post-hole, or in the foundation trench, and firmly packed
around with stones, rubble and soil (Greene 1995: 67, fig. 3.12). The TOE
includes several words for lengths of timber, and they may have had specific
meanings which are no longer apparent (Roberts & Kay 1995: 2: 233). The
combination of archaeological and etymological evidence in one case, how-
ever, suggests a more precise definition is possible.
Old English staca, which gives ModE stake, can also mean pin in Old
English, and the main semantic force of the word is a length of material with a
pointed end. When post-holes are vertically sectioned, it is often possible to
see the shape of the post itself (the post-pipe) as an area of a different colour
from that of the post-hole fill. The actual wood of the post has most often
rotted away, although certain conditions, such as pertain in waterlogged areas,
can preserve the wood intact. Some of these former or surviving timbers can
be seen to have had pointed ends, and it can be presumed that they have been
pile-driven into less resistant ground, for example sand, or marsh (Milne &
Milne 1992: 45). Bearing in mind the etymology of staca, it seems likely that
the word could have been applied by Anglo-Saxon builders to any such
sharpened stake, but not to a blunt-ended post.
Walls
Having, hopefully, acquired firm foundations, the walls must now be con-
structed. Considering ground-level features first, it is noted that, in various
Anglo-Saxon churches, decorative pilasters, and strip moulding around arches
and doorways often spring from a larger basal stone (Taylor & Taylor 1965
78: 2: 570). There is good evidence that these could be referred to as fotstanas
foot-stones. The word occurs in Byrhtferths Manual, and the text refers
specifically to an illustration of arches in the manuscript, in which the
108 C.P. BIGGAM
fotstanas are clearly the bases of columns (Byrhtferth 1929: 90, lines 2931;
91, fig.). This is a rare example of an architectural term which can be firmly
and objectively linked to an architectural feature.
As the wall is built up, it may be decided that substantial quoin-stones, or
angle-stones, are desirable, at the meeting point of two differently aligned
walls. These quoin-stones are called hyrnstanas, the first element representing
OE horn, which denotes not just horn but any other projection. In architec-
ture, it seems to include the restricted sense of an angle, which, of course,
also projects. The Anglo-Saxons used several types of quoining, and good
examples of these types can be seen in surviving churches (Kerr & Kerr 1983:
38, fig. 21). Probably the most common is known as long and short work, in
which the angle-stones are alternately laid flat and stood upright (Taylor &
Taylor 196578: 3: 94143). Another style of quoin construction is side-
alternate quoining, where the long sides of the stones are set at alternate sides
of the wall angle (Taylor & Taylor 196578: 3: 94041).
In a homily on the Epiphany, lfric writes that Christ is se hyrnstan e
gefeg a twegen weallas togdere, the quoin-stone which joins the two
walls together, the two walls representing the Jews and the Gentiles (lfric
18446: 1: 106). Looking at an example of this often impressive feature of
Anglo-Saxon buildings, it is easy to understand the image of Christ that this
would have conjured up in the Anglo-Saxon mind. The hyrnstan is bigger,
stronger, and more impressive than the other stones, and it bonds together two
parts of a single structure, lending it stability, strength, and distinction (Kerr &
Kerr 1983: 39).
What of the main fabric of the wall? Some Saxon stone structures re-used
Roman stone, bricks and/or tiles from the increasingly ruinous remains of
Roman domestic and military buildings in the area. An example of the re-use of
Roman stone occurs at Escomb church, County Durham (Pocock & Wheeler
1971: plate IV). There are various general terms for stone building in Old
English, and their meanings appear to be very similar. Walls are examples of
stangefog stone-joining, and stangetimbre stone-building, and, of course,
they are stnen made of stone.
The use of Roman building stone at Escomb lends a neatness and regular-
ity to the Anglo-Saxon walling, which is, more usually, absent. Convenient
building stone has always been expensive, scarce or non-existent, depending
on the area, and it usually had to be gathered together from all quarters. The
geological mix in a single building is often amazing, not to mention the wide
GRUND TO HROF 109
variety of sizes, colours and textures of stones used. There is now an academic
specialism, ecclesiastical petrology, which is devoted to elucidating the
sources of building stone in churches, and a project at the Anglo-Saxon church
of Brixworth is a good example of the challenge offered by buildings of this
date (Sutherland & Parsons 1985).
It is in this context that the word stanlesung should be considered. Clark
Hall defines it as building with loose stones (Clark Hall 1960), and Bosworth
and Toller say more precisely without cement (Bosworth & Toller 1898
1921), so the term is understood to mean a dry-stone wall, as can be seen in the
countryside to this day. It occurs only once in the extant texts, as a gloss to
lithologia, a Greek borrowing into Latin which indicates a gathering together
of stones. The Old English compound, formed from stan stone and the verb
lesan, meaning to collect, pick, select, gather, glean, is a precise translation of
the Greek, and there is a possibility that it was specifically coined to define
lithologia. Even if it were never in common usage, however, it would have
evoked certain local images for the Anglo-Saxon who invented it. It is
suggested here that stanlesung need not have exclusively indicated dry-stone
walling, but it could also have denoted the eclectic nature and jumbled
appearance of much of the fabric of Anglo-Saxon stone buildings, for which a
term meaning a collection of stones is particularly apt. It is interesting to note
in this context that Rodwell and Rodwell say of the masons at St Peters, Barton-
upon-Humber, that they tended to build in heaps, giving rise to wild
irregularities in their work (Rodwell & Rodwell 1982: 297).
Returning to timber construction, our building was left with foundations
and a skeleton of upright timbers, so the next consideration is to infill the
spaces between the timbers.5 This can be done with several materials, one
example of which is planks, each plank being a bord, bred, or el. The
completed wall can then be termed breden made of planks, and the structure
thus built is treowen made of wood.
Planking may be vertical, and evidence for this was found at Thirlings,
Northumberland (Webster & Cherry 1975: 22627), or horizontal, as was
found at basement level in some of the Anglo-Scandinavian buildings at York
(Hall 1994: 60; pl. 8). In some of these buildings in York, the upright timbers
were not set into post-holes but rested on a horizontal sill- or foundation-beam
below the ground surface. In other cases, the uprights were found to be simply
resting on the earthen floor surface without a beam (Hall 1994: 60; figs. 37 &
39; pl. 8).
110 C.P. BIGGAM
Apart from being breden made of planks, timber buildings could also be
stoccen made of logs, and this is best illustrated by Greensted church in
Essex, the only Anglo-Saxon timber building to survive above ground
(Christie, Olsen & Taylor 1979).6 The timbers are split in two lengthwise so
the outside walls of the church are clearly of log construction, but the inside
walls are flat. The entire nave of the church survives, and the timbers were
originally set in post-holes in the ground, although they are now resting on a
concrete platform to preserve them. Greensted offers a vivid illustration of the
meaning of the place-name Stokenchurch in Buckinghamshire, which is re-
corded as stockenechurch in c.1200 (Mawer & Stenton 1925: 194). Stave
construction like this was also used for high status secular buildings in Anglo-
Saxon England, as in, for example, the ninth century long-hall at Goltho,
Lincolnshire (Beresford 1982: figs. 6.2 & 6.3).
Another means of filling the gaps between the timber uprights is by
hurdles, or panels of wattle, as is still used for fencing. The interpretation of
the evidence from the Anglo-Saxon phase at Cowderys Down in Hampshire
allows for several ways of incorporating wattling into a structure (Millett
1984: 22834, figs. 5963). The actual remains of early tenth century wattling
were recovered at York (Hall 1994: 567). Appropriate to this construction
method are the Old English words hyrdel, windung and watel, referring to a
hurdle, that is, a frame of wattle-work.
Exterior wattle walls usually serve as a core for some form of daub or
plaster, so the building can be weather-tight. Dictionaries tend not to use
daub as a definition of any Old English word, but OE clam is used in Anglo-
Saxon glossaries to translate the Latin word litura plaster (Healey &
Venezky 1980), so it is likely that it would also have been used of daub. The
remains of daub were found, for example, at Yeavering in Northumberland,
with the impressed marks of wattling (Hope-Taylor 1977: 197, fig. 93).
Anglo-Saxons may well have described this as the marks of the hyrdel on the
clam.
The word feorstuu a buttress, support literally means a far-support.
Studu is translated as column, pillar, post, buttress, which is applicable to
both stone and wooden buildings. Stone buttresses were added to the south
walls of Building A, interpreted as a refectory, and others, at the monastery of
Jarrow, presumably to support the buildings where the ground slopes down to
the river (Cramp 1971: 48). This sort of structure would fit the meaning of
far-support very well, since it projects far beyond the wall of the building,
GRUND TO HROF 111
and supports it. Another massive type of feorstuu can be seen in some early
Kentish churches like Reculver, where certain angles of the building are
supported by two buttresses, one against each wall (Taylor & Taylor 196578:
2: 507).
Timber buildings also sometimes have supports which could be described
as far-supports. At Cowderys Down, evidence for slanting timbers was found
in the angles of the post-pipes in post-holes situated beyond the outer walls of
buildings. This enabled the excavators to reconstruct the angles of these raking
timbers, and they surmised that their purpose was to make contact with the
underside of the wall-plate and prevent it rotating off the top of the wall under
the pressure exerted by the weight of the roof (Millett 1984: 24243; figs. 57,
64 & 68). Since the word feorstuu is used to translate Latin obstipus meaning
slanting, inclined (Healey & Venezky 1980), this meaning fits the archaeo-
logical evidence very well. It is, furthermore, illustrated by Bedes story of the
death of Aidan, Bishop of Lindisfarne, who died whilst leaning against an
inclined prop fixed against the outer wall of the church (adclinis destinae,
quae extrinsecus ecclesiae pro munimine erat adposita) (Bede Hist. Eccles.
3: 17; Bede 1896: 160).
Doorways
The walls have now been completed, but, of course, doors and windows have
not been forgotten. Dictionary definitions suggest that Old English did not
make the modern distinction between door, which opens into an interior
space, and gate, which gives access through an exterior boundary, such as a
garden wall, a city wall, or a fence. In the MCOE, however, the entries for geat
include nearly three pages of references to land charter boundaries, while duru
has no such references (Healey & Venezky 1980). This suggests that the
modern distinction, while perhaps not invariable, was already well developed.
It is suggested, therefore, that the many surviving Anglo-Saxon church door-
ways which can be inspected today (Kerr & Kerr 1983: 42, fig. 25) would
have been described as a dor or duru rather than a geat.
The words durstodl, durustod and gedyre refer to door-posts. It is a
characteristic feature of Anglo-Saxon stone-built doorways that they are
edged with through-stones, that is, single stones which are large enough to
extend from front to back of the doorway (Brown 1925: 27, figs. 910). The
112 C.P. BIGGAM
door would have been hung on iron hinges, or heorras, set into the stone, or
into timber frames called doorcases, constructed in the stone doorway. Al-
though doorcases are a well-known means of hanging doors in stone build-
ings, for many years no evidence for Anglo-Saxon examples could be found.
Then, in the 1970s, evidence was found, first at Hadstock, and then at Barton-
on-Humber churches, in the form of deep post-holes against the inner faces of
stone-built doorways (Rodwell 1986: 16667). This is clearly what the Anglo-
Saxons meant by the terms durstodl and durustod, since the word stod means
a post. In Exodus, at the Passover, when the first-born of Egypt died, the
first-born of the Hebrews were spared because the people had marked their
lintels and door-posts with blood. In the Old English prose Exodus, each door-
post is a gedyre (Crawford 1922: 245; Exod. 12.22).
Door-posts also occur in wooden buildings, of course. At Chalton, the
doors were hung inside the building, closing against the back of the door
frame (Addyman, Leigh & Hughes 1972: 23). This probably gave greater
security than doors closing against the outside of the frame. Excavated timber
buildings show evidence for post-holes at either side of the doorways, as at
Yeavering, for example (Hope-Taylor 1977: 64, fig. 22), and a timber which
rose from a post-hole in such a position was, probably, also called a durstodl.
Two Old English words mean a lintel, which is a horizontal beam or
stone at the top of a window-space or doorway. The words are oferdyre and
oferslege, and they were probably a little more specific than the modern lintel.
Oferdyre contains the word for door and is, therefore, likely to have been
used specifically of a door lintel. Nothing in the etymology, however, suggests
it would have been specific to either wood or stone. If oferdyre is specific to a
door, was oferslege used exclusively of window lintels? The answer is in the
negative, since, in the story of the Passover in the Old English prose Exodus, it
is the lintel of the door which is referred to as an oferslege (Crawford 1922:
245; Exod. 12.22).
Slege is cognate with Old Norse sl a beam of wood, so is likely to have
been specific to a timber building. There appears, therefore, to be a lexical gap
in Old English for the denotation of a stone window lintel. There may have
been an appropriate lexeme which has not survived, but it should also be
remembered that building in stone did not have a long history for the Anglo-
Saxons. They must have had a large building vocabulary for wooden struc-
tures, but, from the seventh century, they had to expand this vocabulary, or,
alternatively, expand the semantics of existing words to cope with the con-
GRUND TO HROF 113
cepts of building in stone. It is possible that oferslege could have been applied
to stone lintels by analogy, but this cannot be proved or disproved.
Windows
Anglo-Saxon windows, many of which can be seen today in churches across

the country, were very small by modern standards (Taylor & Taylor 196578:
3: 83668). They would have been referred to as a yrel, which means
anything with a hole through it, or, using compound terms, as an eagyrel an
eye-hole or an eagduru an eye-door, indicating holes through which the
eyes can see. In rare cases, as in the church at West Barsham, Norfolk, round
windows were inserted which actually resemble eyes. The glazed area at the
centre of each of these windows, moreover, resembles the pupil of an eye
(Taylor & Taylor 196578: 2: pl. 377).
Among the Old English words for window is fenester, an obvious
borrowing from the Latin fenestra. Why was this borrowing necessary, when
at least four native Old English words for window were available (Roberts &
Kay 1995: 1: 236)? It is possible that the word implied a window-form which
the Anglo-Saxons regarded as Roman and/or Continental. The word could
have been applied to any stone-built window, since these were usually small,
round-headed, and with single- or double-splayed construction, as occurred in
Roman buildings and post-Roman churches on the Continent. Such windows
can be seen in many surviving Anglo-Saxon churches, for example at Escomb
(Brown 1925: 141, figs. 65A & 65B). The presence of glazing in a stone-built
window must have added to the exotic impression, particularly if it consisted
of a design in stained glass. Evidence for such windows has now been found
on several Anglo-Saxon sites, including Jarrow monastery (Cramp 1975). The
founder of the monastery, Benedict Biscop, is recorded as bringing glass-
workers from Gaul to adorn his buildings (Bede Hist. Abbat. 5; Bede 1896:
368). Such exotic windows may well have triggered the use of fenester rather
than the more homely eagyrel. It is interesting to note that the two occur-
rences of fenester in Old English occur in Gregorys Dialogues, in connection
with a church in Rome (Hecht 1900: 220, lines 15 & 22).
Another Old English word which occurs in connection with windows is
crypelas. It occurs only once in extant texts, as a translation of the Latin word
cancelli, denoting an iron lattice or grille forming a barrier. The Latin text is a
114 C.P. BIGGAM
Biblical passage, from the Book of Proverbs, in which the speaker is looking
out of a window, through cancelli (Prov. 7.6): de fenestra enim domus meae
per cancellos prospexi. This context would have made it clear to the Anglo-
Saxon translator that the referent was some form of barrier across a window,
which did not obscure the view.
Crypelas is defined as a lattice because of the Latin word it translates,
but it is cognate with the verb cryppan to bend, crook, contract (DOE), and
also with Modern English cripple. The core sense is crookedness, and
crypelas should be literally interpreted as crooked things, in this case across
a window. Such a meaning would be suitable for a Roman window-grille, or
for several other means of creating a barrier across a window-space, such as
loosely woven wattle-work or a pattern of lead cames holding panes of glass.
The changes of angle in a lattice or other pattern, or the bending of lines in a
woven structure, could all be described as crooked, but the Anglo-Saxon
translator would probably have interpreted this particular Biblical context in
terms of Roman or contemporary Continental architecture, which suggests a
window-grille or a lattice of glass panes. The remains of a Roman window-
grille were found at Duston, Northamptonshire (Harden 1961: pl. V), and
some of these could have survived into Middle Saxon England. In conclusion,
it may be too specific to define crypelas as a lattice, since it could also have
indicated a grille or leaded panes of a different pattern.
Towers
Before progressing to roofs, it is interesting to consider some optional extras.

Several surviving stone churches of this period have towers, including a group
of unusual round towers in East Anglia (Kerr & Kerr 1983: 268, figs. 1012;
Fisher 1969). Towers are clearly denoted by the words torr, tur and stantorr, in
which the torr/tur element is borrowed from the Latin turris, a tower or turret,
although tur entered English via Old French. This may indicate a similar case
to that of fenester, namely, the use of a Latin loanword for the technology of
building in stone. The use of the compound stantorr, however, suggests that a
torr could also be of wood, and that meaning is, in theory, permissable for
another word for tower stipel, cognate with steap high, lofty.
It is also possible to illustrate the word windelstan winding stone [stair-
case] with a very few extant buildings. Two of the best preserved spiral
GRUND TO HROF 115
staircases are at Broughton and Hough-on-the-Hill, both in Lincolnshire

(Taylor 1959: 15255; pls. I-II). Surviving Anglo-Saxon stairways are not
internal to the buildings, but are housed in a round stair tower attached to the
outside of the normal square church tower. There are small openings at
intervals up the tower to light the staircase (Kerr & Kerr 1983: 25, fig. 9).
Vestibules and Porches
One of the most noticeable things about the architectural vocabulary in the
TOE is the large number of words under the headings concerned with porches
and vestibules (Roberts & Kay 1995: 1: 236). Once again, a Latin loan-word,
portic, is included in the list, and the MCOE shows that it is used of church
buildings, and of temples in the Holy Land, so another loan-word appears to
be linked with the exotic concept of building in stone.
Although portic appears to indicate a vestibule in a few cases, this is not
its usual meaning. A brief consideration of its source word, the Classical Latin
porticus, is helpful in this case. Porticus denoted a portico or colonnade, that
is, a covered walk in which the roof is supported by columns (OLD). Many
early Christian churches were built on the basilican plan, copied from Roman
public buildings, and such churches consisted of a nave divided from two side
aisles by colonnades. This Roman architectural feature, along with its Latin
name, was, therefore, adopted into early Christian culture. Over the centuries,
there was a tendency to divide up the side aisles of basilican churches into
separate chapels or sanctuaries, each entered through an opening in the
colonnade, and, in some cases, the side aisles were completely blocked by the
chapel walls which extended from the exterior wall to the columns. As chapels
became an increasingly valued feature, churches were built with chapels
extending beyond the outer walls, and many existing buildings were enlarged
by similar additions. By the time the Anglo-Saxons encountered the word
porticus it appears that its semantic connection with colonnades had been
subordinated to the concept of the side-chapel or sanctuary. It entered Old
English in the form portic, and appears most often in the context of the burial
of important persons in this type of structure. Its meaning in Old English is,
primarily, therefore, chapel, sanctuary (Cherry 1976: 16369). In Modern
English, porticus is a technical term used of a chamber entered from the nave
of a church.
116 C.P. BIGGAM
Early churches could have any number of porticuses, but it was common
for small churches to have two, one on the north side, and one on the south
side of the nave, amounting to rudimentary transepts. This can be seen, for
example, in the remains of the Saxon church, St Pancras, in Canterbury
(Taylor & Taylor 196578: 1: 147, fig. 64). These structures clearly illustrate
the Old English terms suportic and norportic.
Old English portic had a further meaning, however, since it is used to
gloss Latin absida an apse. A porticus, in the form of a structure extending
beyond the outer wall of the church, could have an apsidal, rather than square
or rectangular plan, so the extended meaning is understandable. The majority
of apsidal structures in surviving Anglo-Saxon churches occur at the east end,
as, for example, at Deerhurst in Gloucestershire (Rahtz & Watts 1997: 180,
fig. 106). In this connection, it is interesting to note Bradys suggestion that
porticus might sometimes indicate a choir or chancel area in early mediaeval
Irish churches. His theories are based on a new interpretation of some vocabu-
lary in De oratoria, a poem in the Hisperica famina (Brady 1997: 33033;
Herren 1974: 1089).
Any practical number of extra side- and end-chambers could be built onto
a church in this way, and, in the Blickling Homily on Holy Thursday, the
church on the Mount of Olives is described as having ry porticas emb a
ciricean utan geworhte, three chambers around the church built on the
outside (Morris 1967: 125). Morriss translation refers to three porches and
he omits built on the outside (Morris 1967: 124). Since ModE porch indi-
cates a covered approach to the entrance of a building, and since porticuses
were usually only accessible from inside the church, this translation is mis-
leading.
At the west end of the great abbey church at Sherborne, Dorset, there was
a true porch (Cherry 1976: 190, fig. 4.11), and it is necessary to turn to the Old
English vocabulary for porches and vestibules to consider which lexemes could
have been applied to this feature. It is immediately evident that this is an area
requiring more research, although certain compound words, such as foreduru
a before-the-door, superficially suggest their appropriateness. There are
particular difficulties with Old English words which survive only as glosses to
Latin words, since the huge differences between Roman or Roman-style
buildings and traditional Anglo-Saxon buildings, and the suspicion that some
of these words may have been coined specifically for translation purposes, all
create problems for accepting apparently obvious interpretations.
GRUND TO HROF 117
Roofs
Finally, the structures must be roofed. This is the most difficult part of an
architectural reconstruction, because it is the part of the building furthest
removed from the excavated evidence. The evidence nearly always permits
various possible reconstructions, as was clearly shown in the report on the
excavated Structure C12, a hall, at Cowderys Down, Hampshire (Millett
1984: 24546, figs. 70 & 71). The ground-plan, the buttresses, the position of
the doorway, and the construction of the walls are all based on hard evidence,
but the nature and shape of the roof rests largely on speculation. The options
are, of course, limited by certain small hints from the excavated evidence.
Thus, the size of the timbers in the post-holes and the angle of the buttressing
suggest the height of the building (Millett 1984: 23640), but there is little
more to help in this case.
There is a considerable Old English vocabulary for roofing, and some of
it will now be discussed. A roof is the ofergeweorc of a building, that is, the
over-work, and the process of roofing is hrefan or oferhrefan, to roof or
over-roof. A roof is hrof or c, the etymology of both words indicating a
covering. The hrof may be an internal or external covering, that is, in modern
usage, a roof or a ceiling, and its cognates and descendants include words for
the hard palate of the mouth in Modern English, a coffin lid in Dutch, and a
boat shed in Old Icelandic (OED). Old English c is cognate with Modern
German dach roof, and refers to any type of roof-covering, not just thatch.
Turning to more specific terms, it should, first of all, come as no surprise
that Anglo-Saxon roofs were sloping rather than flat. Water damages build-
ings, so rainwater needs to be removed as quickly as possible. In the English
climate, therefore, sloping roofs could be expected, and the Old English
vocabulary shows that this was the norm. Words like first ridge-pole and
hrycg roof-top, the latter with an etymology in the concept of a backbone,
are incompatible with flat roofs.
Another word which is associated with roofs is horn, which occurs as a
simplex, and also in compound constructions. Bosworth and Toller (1898
1921) interpret horn, in architectural contexts, as the horn-shaped projection
on the gable-end of a house, a pinnacle, horngeap as having a wide extent
between the horns, horngestreon as an abundance of pinnacles, and
hornreced as a house having horns or pinnacles. The horns envisaged in
these definitions refer to the crossed ends of the beams at the highest point of
118 C.P. BIGGAM
the gable-ends of a timber built house. This feature can be seen in two of the
suggested reconstructions of Structure C12 at Cowderys Down (Millett 1984:
24555, figs. 7071). Clark Hall (1960) agrees that horn denotes a feature
which extends beyond the roof-line, or is added on top of it, and he defines
horn, in an architectural context, as projection, pinnacle. However, when he
turns to interpreting the compound terms, he appears to change his definition
of horn to gable. He defines horngeap as broad between the gables, and
hornreced as hall with gables, although horngestreon remains as wealth of
pinnacles (on a house).
The question, therefore, is whether horn-words refer to a gable or an
ornamental addition to the roof. A gable is the triangular, upper portion of a
wall which occurs at the ends of a ridged roof. At its highest point, at either
end of the building, it has been common, at most periods of history, to add
embellishments such as finials, animal figures, crossed beams, and many
others. The point of the gable and an ornamental finial of some kind are,
therefore, in close proximity, and may have been semantically undifferenti-
ated. Can surviving texts cast any light on this problem?
First of all, there is good evidence that horn, in one context at least,
referred to an addition to the roof, and not the gable. The word involved in this
context is hornpic, in which pic, as a simplex, means point. Hornpic appears
in the Biblical context of the Temptation of Christ, during which the Devil
took Him to the top of the Temple (Luke 4.9). It is clear from Bedes De
templo that he fully understood the roof of the Temple in Jerusalem was not
gabled like traditional Anglo-Saxon buildings. He writes The temple did not
have a ridge on top any more than the tabernacle did, but was flat, as was the
prevailing fashion with all who built houses in Palestine and Egypt (Bede
1995: 34). If Bede understood this, it would be unwise to assume that other
Anglo-Saxon writers did not, so, in this context of a flat roof, the word horn
must indicate a turret or some other ornamental addition to the roof, and
hornpic obviously indicates the tip of the turret, not of an imagined gable.7
It has not proved possible to present equally strong evidence that horn
means gable, although there is an indication. Horn occurred earlier in this
paper in connection with hyrnstanas, or quoin-stones, denoting the angle at
which two walls meet, and this shows that horn could indicate a projection in
the form of a simple angle, in addition to that of a typical horn-shape like, for
example, a bulls horn. Since the point of a gable is the angle at which the two
slopes of the roof meet, it could indicate a second example of horn denoting an
GRUND TO HROF 119
angle, but there is no conclusive example to prove this, since it would require
a statement that the building had no turrets or finials. Furthermore, it is
suggested that the definitions by Clark Hall (1960) of hornreced as hall with
gables and hornsele as house with gables, appear to contain redundant
semantic elements, since the vast majority of halls and houses had gables.
Translations such as hall with finials or house with horn-like cross-beams
would have much more semantic significance. So far, it is suggested that horn
in architecture signifies finial, turret, cross-beams, or other addition to a roof-
top, and it may, in addition, signify gable-point although the evidence for
the latter is poor. It is now proposed to examine the same Biblical context in
the Latin text.
The Vulgate text refers to the place where Christ and the Devil stood as
the pinna or pinnaculum of the Temple, indicating a projecting part, and these
words, in Medieval Latin, are usually translated as spire or steeple, al-
though Lathams examples are all post-Conquest (Latham 1965).8 Classical
Latin pinna meant a feather, wing of a bird, fin, but it also had an architec-
tural meaning, namely, a raised part of an embattled parapet, a merlon
(OLD). The basic semantic concept is the same as that in OE horn, that is,
something which projects. Just as wings and fins project from birds and
fishes bodies, the merlons of a battlement are the upstanding portions of the
upper wall between the embrasures, or openings. It is suggested here that
pinna, indicating a projection, came to mean any vertical extension beyond the
roof-line of a building, and that it underwent a semantic extension to spire or
steeple, an extreme form of projection, in Anglo-Norman times. Brady
suggests that pinna indicates the wings or horns discussed above, namely,
the cross-beams at the gable-ends. This may well be correct, although I would
argue that the semantics of pinna should encompass any other type of exten-
sion or addition to the roof. It seems much more likely that the later mediaeval
sense of spire would develop from an earlier meaning such as turret or
pinnacle, than from a specific sense of cross-beams.9 The suffix -culus,
-culum in Classical Latin indicates a diminutive (OLD), so pinnaculum means,
literally, a little projection. It is possible that this word specifically indicates
a finial or other small ornamental flourish on the roof.
The image of Christ and the Devil in this text was intended to convey the
idea of a dramatic mental struggle at the highest point of the Temple, and,
while some were content to envisage a turret, others, it seems, preferred to
indicate the finial on the top of the turret. In conclusion, horn and its com-
120 C.P. BIGGAM
pounds are most likely to indicate additions or extensions to the roof-line, and
rather less likely to denote a gable-point.
Old English vocabulary shows that various types of roof-covering were
known. Thatch, in its modern sense, is represented by the word fennc,
indicating roofing materials from a fen, which must mean rushes or reeds. The
word riscen made of rushes also occurs. The availability of this material
would, obviously, have geographical limits.
Shingles were also used for roofing, and examples, dating to the tenth or
eleventh century, have been recovered from excavations at Winchester, some
with the nail-holes still visible (Keene 1990: 323, fig. 73). Shingles are,
essentially, tiles of wood, usually made of oak, which may be square or oval,
and they are probably illustrated on the Bayeux Tapestry, on the roof, for
example, of Harold Godwinsons house (Wilson 1985: pl. 34). The words
scindel and scingal refer to these oak tiles, and both words are thought to be
descended from the Latin scandula, also meaning shingle.10 The hrycg, or
roof-ridge, may not always have been a substantial structure, but a means of
sealing the ridge of a shingled or tiled roof (Beresford 1982: 119).
Stone or ceramic roof-coverings are represented by words like hrofstan
roof-stone, tigel tile and ctigel roof-covering tile. Hrofstan, in theory,
should be specific to the splittable stone which occurs in certain parts of
England, and which is traditionally used for roofing in those areas. A limestone
roofing-stone was found, for example, at Monkwearmouth, Northumberland
(Cramp 1971: fig. 24.4). Tigel refers to a fired ceramic tile, and comes from the
Latin word tegula. ctigil refers specifically to a roof-tile, since there were
also Anglo-Saxon floor-tiles (Backhouse, Turner & Webster 1984: 13637).
Excavations at Jarrow produced further evidence in the form of large roofing-
nails, and clips for lead roofing-sheets (Cramp 1971: 56, fig. 24.23). The latter
are evidence for Bedes plumbi lamminae sheets of lead, used to re-roof the
first church at Lindisfarne which had previously been thatched with reeds
(Bede Hist. Eccles. 3.25; Bede 1896: 181).
Conclusion
Having constructed virtual buildings in stone and timber from the Old English
vocabulary, it is evident that, while there are certainly difficulties in linking
semantic and archaeological evidence, there are also some fascinating ex-
GRUND TO HROF 121
amples of the compatibility of the two forms of evidence. Obvious avenues of

future research are revealed, for example the nature of Latin words in the Old
English building vocabulary. Some may be loan-words from Continental
Latin, or even from Romano-British Latin. James, Marshall and Millett (1985)
discuss the distinctive structural types found in lowland England in the sixth to
eighth centuries, which have some affinities with north German buildings and
others with Romano-British buildings. Whatever this may mean in ethnic
terms, it would appear that Anglo-Saxons and Britons influenced each other in
their building methods (Dixon 1982), and, must, therefore, have discussed
such matters. It would not be surprising if the semi-technical building vocabu-
lary of the Britons was partly Latin, or if the Anglo-Saxons borrowed certain
of those words into Old English. Further research should also consider the
extent to which some Old English terms may have been coined specifically to
translate Latin words, and surveys of building terms in other Germanic
languages, in Middle English and dialectal Modern English would, no doubt,
also be enlightening. The particularly exciting aspect of research in this
subject is that archaeology is always producing more evidence of the struc-
tures that the Anglo-Saxons must have linguistically encoded in every detail.
Notes
1. An earlier version of this paper was read to the School of English, Adam Mickiewicz
University, Pozna, at the kind invitation of Prof. Jacek Fisiak.
2. Dictionary definitions in this paper are taken from Clark Hall 1960, except for those
which appear in the published fascicles of the Dictionary of Old English (DOE) which is
currently in process of publication by the Centre for Medieval Studies of the University
of Toronto. Where the DOE has been used, this abbreviation will appear after the
definition.
3. Line references to Old English poetry throughout this article are taken from Krapp &
Dobbie 193142.
4. This argument refers to the walls appearance from Late Saxon times to the present. It is
not suggested that the base-wall is of separate build from the upper part, nor is any
opinion expressed here on the controversial date and history of this and other features of
this churchs external walls.
5. The virtual timber building in this paper is a house or hall. The type of building known
variously as a grubenhaus, (grub-)hut, or sunken-featured building, has not been specifi-
cally considered, although some of the discussion is also relevant to this type. The
grubenhaus has a floor excavated below the ground surface, although, in some cases, this
122 C.P. BIGGAM
may simply have provided a ventilation space beneath a wooden floor, and its roof rests
on the ground, or very nearly so, thus obviating the need for vertical walls of any great
height except at the two gable ends. West (1985: 1: 11321) discusses the evidence for
grubenhuser from West Stow, and this is provided in summary form in Welch (1992:
216).
6. The surviving stave construction at Greensted may not be pre-Conquest in date, although
it is Anglo-Saxon in type. According to recent dendrochronological dating, the timbers
were felled in 1063, and the church is estimated to have been built between then and the
end of the century (Tyers 1996).
7. The illustration of the Temptation of Christ in the Book of Kells shows a gabled Temple
(Dublin, Trinity College Library, MS 58, folio 202v). This may mean that the style of
Near Eastern architecture was not universally understood within early mediaeval insular
Christianity, or it may simply denote artistic licence.
8. In spite of speculation, there is no evidence that spires were constructed in Anglo-Saxon
England (Gem 1995: 50).
9. I do not understand Bradys equating of pinnacle with flagpole: Brownes choice of
pinnacles is equally unsatisfactory, since the thought of a flagpole rising from the roof
of any Irish building in this early period is quite unimaginable (Brady 1997: 333). The
OED interprets pinnacle as a small ornamental turret.
10. The evidence from manuscript illustrations for shingles and tiles is classified by Carver
(1986: 121 & 125; fig. 2).
References
Addyman, P.V. 1972. The Anglo-Saxon House: a New Review. Anglo-Saxon England
1.273307.
Addyman, P.V., D. Leigh & M.J. Hughes. 1972. Anglo-Saxon Houses at Chalton,
Hampshire. Medieval Archaeology 16.1331.
lfric. 18446. The Sermones Catholici, or Homilies of lfric in the Original Anglo-
Saxon with an English Version ed. by Benjamin Thorpe. 2 vols. London: lfric Society.
Alcock, Leslie, in collaboration with S.J. Stevenson & C.R. Musson. 1995. Cadbury Castle,
Somerset: the Early Medieval Archaeology. Cardiff: University of Wales Press.
Backhouse, Janet, D.H. Turner & Leslie Webster, eds. 1984. The Golden Age of Anglo-
Saxon Art 9661066. London: British Museum.
Bede. 1896. Venerabilis Baedae historiam ecclesiasticam gentis Anglorum, historiam
abbatum, epistolam ad Ecgberctum una cum historia abbatum auctore anonymo ed. by
Charles Plummer. Oxford: Clarendon Press.
Bede. 1995. Bede: On the Temple trans. with notes by Sen Connolly, with an introduction
by Jennifer OReilly. Translated Texts for Historians 21. Liverpool: Liverpool Univer-
sity Press.
Beresford, Guy. 1982. The Reconstruction of Some Saxon Buildings at Goltho,
Lincolnshire. Structural Reconstruction: Approaches to the Interpretation of the Exca-
GRUND TO HROF 123
vated Remains of Buildings ed. by P.J. Drury, 11323. British Archaeological Reports,
British Series 110. Oxford: B.A.R.
Bosworth, Joseph & T. Northcote Toller with Alistair Campbell. 18981921. An Anglo-
Saxon Dictionary Based on the Manuscript Collections of the Late Joseph Bosworth,
ed. and enlarged by T. Northcote Toller, with revised and enlarged addenda by Alistair
Campbell. 2 vols. Oxford: Oxford University Press.
Bradley, S.A.J. 1982. Anglo-Saxon Poetry: an Anthology of Old English Poems in Prose
Translation with Introduction and Headnotes. London: Dent.
Brady, Niall 1997. De oratorio: Hisperica famina and Church Building. Peritia 11.327
35.
Brotanek, Rudolf, ed. 1913. Texte und Untersuchungen zur altenglischen Literatur und
Kirchengeschichte. Halle: Niemeyer.
Brown, G. Baldwin. 1925. The Arts in Early England: Anglo-Saxon Architecture. 2nd ed.
London: John Murray.
Byrhtferth. 1929. Byrhtferths Manual (A.D. 1011) Now Edited for the First Time from MS.
Ashmole 328 in the Bodleian Library by S.J. Crawford. Early English Text Society,
Original Series 177. London: Oxford University Press.
Carver, M.O.H. 1986. Contemporary Artefacts Illustrated in Late Saxon Manuscripts.
Archaeologia 108.11745.
Centre for Medieval Studies, University of Toronto. 1986 . Dictionary of Old English.
Toronto: Centre for Medieval Studies, University of Toronto.
Cherry, Bridget. 1976. Ecclesiastical Architecture. The Archaeology of Anglo-Saxon
England ed. by David M. Wilson, 151200. Cambridge: Cambridge University Press.
Christie, Hkon, Olaf Olsen & H.M. Taylor. 1979. The Wooden Church of St Andrew at
Greensted, Essex. Antiquaries Journal 59:1.92112.
Clark Hall, J.R. 1960. A Concise Anglo-Saxon Dictionary. 4th ed. Cambridge: Cambridge
University Press.
Cramp, Rosemary. 1971. Excavations at the Saxon Monastic Sites of Wearmouth and
Jarrow, Co. Durham: an Interim Report. Medieval Archaeology 13.2166 (1971 for
1969).
Cramp, Rosemary. 1975. Window Glass from the Monastic Site of Jarrow. Journal of
Glass Studies 17.8896.
Crawford, S.J., ed. 1922. The Old English Version of the Heptateuch, lfrics Treatise on
the Old and New Testament, and his Preface to Genesis. Early English Text Society,
Original Series 160. London: Oxford University Press.
Dixon, Philip. 1982. How Saxon is the Saxon House? Structural Reconstruction: Ap-
proaches to the Interpretation of the Excavated Remains of Buildings ed. by P.J. Drury,
27587. British Archaeological Reports, British Series 110. Oxford: B.A.R.
DOE see Centre for Medieval Studies, University of Toronto.
Fisher, E.A. 1969. Anglo-Saxon Towers: an Architectural and Historical Study. Newton
Abbot: David & Charles.
Gem, Richard. 1995. Staged Timber Spires in Carolingian North-East France and Late
Anglo-Saxon England. Journal of the British Archaeological Association 148.2954.
Glare, P.G.W., ed. 1982. Oxford Latin Dictionary. Oxford: Oxford University Press.
Greene, Kevin. 1995. Archaeology, an Introduction: the History, Principles and Methods
of Modern Archaeology. 3rd. ed. London: Routledge.
124 C.P. BIGGAM
Gregory I. 1871. King Alfreds West Saxon Version of Gregorys Pastoral Care ed. by
Henry Sweet. 2 vols. Early English Text Society, Original Series 45 & 50. London:
Oxford University Press.
Grundy, Lynne & Jane Roberts. 1997. Shapes in the Landscape: Some Words. Names,
Places and People: an Onomastic Miscellany in Memory of John McNeal Dodgson ed.
by Alexander R. Rumble and A.D. Mills, 96110. Stamford: Paul Watkins.
Hall, Richard. 1994. English Heritage Book of Viking Age York. London: Batsford.
Harden, D.B. 1961. Domestic Window Glass: Roman, Saxon and Medieval. Studies in
Building History: Essays in Recognition of the Work of B.H. St.J. ONeil ed. by E.M.
Jope, 3963. London: Odhams.
Healey, Antonette diPaolo & Richard L. Venezky. 1980. A Microfiche Concordance to Old
English. Toronto: Centre for Medieval Studies, University of Toronto.
Hecht, Hans, ed. 1900. Bischofs Wrferth von Worcester bersetzung der Dialoge Gregors
des Grossen ber das Leben und die wunderthaten italienischer Vter und ber die
Unsterblichkeit der Seelen. Bibliothek der angelschsischen Prosa 5. Leipzig: Georg H.
Wigand.
Herren, Michael W. 1974. The Hisperica famina, I: the A-Text: a New Critical Edition with
English Translation and Philological Commentary. Pontifical Institute of Mediaeval
Studies, Studies and Texts 31. Toronto: Pontifical Institute of Mediaeval Studies.
Hope-Taylor, Brian. 1977. Yeavering: an Anglo-British Centre of Early Northumbria.
Department of the Environment Archaeological Reports 7. London: HMSO.
James, Simon, Anne Marshall & Martin Millett. 1985. An Early Medieval Building
Tradition. Archaeological Journal 141.182215 (1985 for 1984).
Keene, Derek. 1990. Shingles. Object and Economy in Medieval Winchester, ed. by
Martin Biddle, 1: 32026. Winchester Studies 7.ii. 2 vols. Oxford: Clarendon Press.
Kerr, Mary & Nigel Kerr. 1983. Anglo-Saxon Architecture. Princes Risborough: Shire.
Krapp, George Philip & Elliott van Kirk Dobbie, eds. 193142. The Anglo-Saxon Poetic
Records: a Collective Edition. 6 vols. New York: Columbia University Press.
Latham, R.E. 1965. Revised Medieval Latin Word-List from British and Irish Sources.
London: Oxford University Press.
MCOE see Healey, Antonette diPaolo & Richard L. Venezky.
Mawer, A. & F.M. Stenton. 1925. The Place-Names of Buckinghamshire. English Place-
Name Society 2. Cambridge: Cambridge University Press.
Millett, Martin with Simon James. 1984. Excavations at Cowderys Down, Basingstoke,
Hampshire, 197881. Archaeological Journal 140.151279 (1984 for 1983).
Milne, Gustav & Chrissie Milne. 1992. Catalogue of Waterfront Installations. Timber
Building Techniques in London c.9001400: an Archaeological Study of Waterfront
Installations and Related Material ed. by Gustav Milne, 2377. London and Middlesex
Archaeological Society Special Paper 15. London: London and Middlesex Archaeo-
logical Society.
Morris, R., ed. 1967. The Blickling Homilies with a Translation and Index of Words
Together with the Blickling Glosses. Early English Text Society, Original Series 58, 63
& 73. London: Oxford University Press.
OED see Simpson, J.A. & E.S.C. Weiner.
OLD see Glare, P.G.W.
Oliphant, R.T., ed. 1966. The Harley Latin-Old English Glossary. Janua Linguarum, Series
GRUND TO HROF 125
Practica 20. The Hague: Mouton.

Pocock, Michael & Hazel Wheeler. 1971. Excavations at Escomb Church, County
Durham, 1968. Journal of the British Archaeological Association 34.1129.
Rahtz, Philip. 1976. Buildings and Rural Settlement. The Archaeology of Anglo-Saxon
England ed. by David M. Wilson, 4998. Cambridge: Cambridge University Press.
Rahtz, Philip & Lorna Watts, with the late Harold Taylor & Lawrence Butler. 1997. St
Marys Church, Deerhurst, Gloucestershire: Fieldwork, Excavations and Structural
Analysis 19711984. Reports of the Research Committee of the Society of Antiquaries
of London 55. Woodbridge: Boydell Press.
Roberts, Jane, Christian Kay with Lynne Grundy. 1995. A Thesaurus of Old English in Two
Volumes. London: Centre for Late Antique and Medieval Studies, Kings College
London.
Rodwell, Warwick. 1986. Anglo-Saxon Church Building: Aspects of Design and Con-
struction. The Anglo-Saxon Church: Papers on History, Architecture, and Archaeol-
ogy in Honour of Dr H.M. Taylor ed. by L.A.S. Butler & R.K. Morris, 15675. CBA
Research Report 60. London: Council for British Archaeology.
Rodwell, Warwick & Kirsty Rodwell. 1982. St Peters Church, Barton-Upon-Humber:
Excavation and Structural Study, 197881. Antiquaries Journal 62:2.283315.
Simpson, J.A. & E.S.C. Weiner, ed. 1989. The Oxford English Dictionary. 2nd ed. 20 vols.
Skeat, Walter W., ed. 187187. The Holy Gospels in Anglo-Saxon, Northumbrian, and Old
Mercian Versions Synoptically Arranged, with Collations Exhibiting all the Readings of
all the MSS, Together with the Early Latin Version as Contained in the Lindisfarne MS,
Collated with the Latin Version in the Rushworth MS. Cambridge: Cambridge Univer-
sity Press.
Sutherland, Diana S. & David Parsons. 1985. The Petrological Contribution to the Survey
of All Saints Church Brixworth, Northamptonshire: an Interim Study. Journal of the
British Archaeological Association 137.4564 (1985 for 1984).
TOE see Roberts, Jane & Christian Kay.
Taylor, H.M. 1959. Some Little-Known Aspects of English Pre-Conquest Churches. The
Anglo-Saxons: Studies in Some Aspects of their History and Culture Presented to Bruce
Dickins, ed. by Peter Clemoes, 13758. London: Bowes & Bowes.
Taylor, H.M. 1971. Repton Reconsidered: a Study in Structural Criticism. England
Before the Conquest: Studies in Primary Sources Presented to Dorothy Whitelock ed. by
Peter Clemoes and Kathleen Hughes, 35189. Cambridge: Cambridge University Press.
Taylor, H.M. & Joan Taylor. 196578. Anglo-Saxon Architecture. 3 vols. Cambridge:
Tyers, I. 1996. Tree-Ring Analysis of Timbers from the Stave Church at Greensted,
Essex. Ancient Monuments Laboratory Reports 20, no. 14/96.
Webster, Leslie E. & John Cherry. 1975. Medieval Britain in 1974. Medieval Archaeol-
ogy 19.22060.
Welch, Martin. 1992. English Heritage Book of Anglo-Saxon England. London: Batsford.
West, Stanley. 1985. West Stow: the Anglo-Saxon Village. East Anglian Archaeology
Report 24. Ipswich: Suffolk County Planning Department.
Wilson, David. 1985. The Bayeux Tapestry: the Complete Tapestry in Colour. London:
Thames & Hudson.
Five Hundred Years of Love:
a Prototype-Semantic Analysis
Heli Tissari
Abstract
This article describes the semantic microfield formed by the English lexeme
LOVE. By relating the meaning of LOVE to the conceptually-oriented proto-
type theory, it makes a corpus-based diachronic comparison between Early
Modern and Present-Day English. It also creates a five-fold division of love
into family love, friendship, sexual love, religious love and love of
things. This division appears in C. S. Lewiss The Four Loves but also
corresponds fairly well with dictionary data. To describe the changes within
the field more fully, the study also considers text types and societal reasons for
change.
Introduction
This article suggests a model which helps to explain changes in the meaning of
LOVE. It discusses this lexeme in a broad sense which includes both the noun
and the verb.1
The data which I have been exploring includes a pertinent quotation from
Bacons Twoo Bookes of the Proficience and Advancement of Learning (19r):
for wordes are but the Images of matter, and except they haue life of reason
and inuention: to fall in loue with them, is all one, as to fall in loue with a
picture. The main topic here is the image, or signifiant. The matter, or
signifi (concept), also plays an important role, but is no longer entirely
128 HELI TISSARI
approachable by linguistic means. We should also discuss the reality behind

the name and the concept, but here even more interdisciplinary knowledge
would be required to draw a full picture.2
Another distinction which created unresolvable tension was whether love
is basically an emotion or a disposition. It is characteristic of love as an
emotion that it is considered more or less uncontrollable and that it can be
fairly short-lived (see stman 1989). This is especially the case in romantic
love, while a disposition is likely to be of a more lasting nature.3
Despite the ambiguities and tensions, it can be assumed that there is a
certain degree of collective subjectivity, a way of thinking and feeling
shared by most if not all speakers of a language. The present analysis tests a
model which is concerned with how this is reflected in their linguistic output.
Simultaneously, it tells us how language deals with abstract concepts.
The main idea is that there are five central or prototypical loves which
occur in the domains of family, friendship, sexuality, religion and the non-
human world (Taylor 1991: 8387). The main hypothesis is that their (rela-
tive) frequency is likely to vary with time. I suggest that such variance is not
random but coincides with text types and even the development of society in
general. I consider the single lexeme LOVE as a microfield of such changes.
I am going to describe these changes between the Early Modern period
and the present day, using three corpora and Shakespeares prose as data for
this purpose. The corpora are (the Early Modern period of) The Helsinki
Corpus (HC), The Brown Corpus (Brown), and The Lancaster-Oslo/Bergen
Corpus (LOB).4 Shakespeares prose is included in order to complement the
otherwise smaller amount of data from the Early Modern period.5
The Roots of the Model
It is clear that no lexeme exists in a vacuum, and field theorists have tradition-
ally described a host of related lexemes which form a network of meanings
(see e.g. Kjellmer 1973, Schwyter 1996). However, an approach which begins
from the meaning of a single lexeme is not infrequent either (see e.g. Grlach
1978: 179, Koivisto-Alanko 1999). Here I want to build a solid basis for the
description not only of LOVE but also of related lexemes. However, since I
wish to treat these matters in more detail in subsequent publications, I deliber-
ately concentrate on what there is to say about LOVE. The approach is
FIVE HUNDRED YEARS OF LOVE 129
prototype-semantic because it allows me to focus on a restricted set of relevant

information, while I can overlook a myriad of features accompanying a
lexeme which is so familiar, clichd, surprising, multifaceted, and emotionally
charged as LOVE.
My first decision was to consider primarily who is loving whom, which
means that I am not even attempting to decode LOVE as a set of semantic
constituents. Instead, I want to look at the participants in the situation where
LOVE occurs, and more generally, the situation itself. I was dissatisfied with a
pilot analysis which divided the semantic subjects (hereafter called experienc-
ers) and objects (hereafter causes or causers) of LOVE into human and non-
human (cf. Diller 1994). This only indicated that the subjects were almost
always human, and even the objects were more likely to be human than non-
human.
Eventually, I found a new point of view in Lewiss (1960) The Four
Loves. Lewis sees LOVE alternatively as
(1) storge (affection or family love),
(2) philia (friendship),
(3) eros (sexual love), or
(4) agape (religious love).
He also briefly treats likings and loves for the sub-human, although he is of
the opinion that people can only really love animate objects. Nevertheless, a
linguist cannot ignore those cases where the semantic object of LOVE really is
an inanimate object, such as wine, dice, or even pleasure, which all occur in
my data.
Such cases are not a real problem because it is obvious that if one
disregards Lewiss evaluation of the fifth category, he actually provides five
clearly differentiated categories of love, which can be defined via their partici-
pants in the following manner:
(1) The participants in family love are family members.
(2) The participants in friendship are friends.
(3) The participants in sexual love are lovers.
(4) In order for love to be religious, God has to be a participant.
(5) If one of the participants is non-human, love is situated in the fifth
category, which will be called love of things.
The main problems of this approach are:
130 HELI TISSARI
1. Problems of definition arise because although it is rather easy to define

family love, even quite specifically in terms of the family members included,
it is quite impossible to say exactly what friendship is or what differentiates
spouses as family members from lovers as participants in sexual love.
2. The second question concerns the problems of fuzziness which appear
between the categories. In this study I arrived at a context-based solution in
most of the cases, but in order to describe this fuzziness adequately one would
need more space and tools.
3. The context-based approach is closely related to the concept of love. In
other words, my analysis is not based purely on participant criteria, but also on
situational knowledge. Consequently, an exact identification of the partici-
pants becomes unnecessary. This agrees well with the prototype theory of
concepts and word meaning discussed in further detail below.
4. The deductive nature of the present approach means that it does not show
what would happen if the data were analysed without any preconceived
categories. However, I consider this as a choice-of-method problem to which
there is no single ideal solution.
The advantages of the present approach include at least the following:
1. The advantage of abstraction means that it is possible to categorize the items
without precise knowledge about the participants. LOVE often appears in
contexts where an exact identification of the participants is not relevant. My
example comes from LOB where the writer is talking about sexual love:
(1) If I am to speak the whole truth about my knowledge of love, I will have to
stop trying to emulate the transcendant nightingale (Belles Lettres, G33: 22).
2. The model corresponds fairly closely to dictionary entries. The dictionaries

which I consulted for this study, and which I will discuss below, are The
Oxford English Dictionary (OED) and Dr. Johnsons eighteenth-century Dic-
tionary of the English Language (DEL).
3. The model follows the prototype theory of semantics, which suggests that
meaning is situated in domains and organized in clusters, and that some
meanings are more typical than others. Prototype theory also predicts that
there will be no clear lines between categories but that they will blend in a
fuzzy border area. (See e.g. Taylor 1991: 5455, Geeraerts 1985: 136139.)
4. The basic simplicity of the approach helps to form a coherent complete view
of the data. It also supports the assumption that although the categories have
not been tested on informants and/or evaluated by a psychologist they may be
close to some reality in our minds.
The Dictionaries
I do not claim a complete correspondence between the five categories and the
entries for LOVE in OED and DEL, but there seems to be an affinity between
them. The affinity appears to best advantage when one looks at Johnsons
definition of the verb LOVE. Johnsons five senses correspond exactly with
the five categories, as can be seen in his definitions:
(1) To regard with passionate affection, as that of one sex to the other =
sexual love.
(2) To regard with the affection of a friend = friendship.
(3) To regard with parental tenderness = family love (although the
example comes from the Bible and I would rather situate it in religious
love).
(4) To be pleased with = love of things.
(5) To regard with reverent unwillingness to offend = religious love.
If one thinks of the five loves as conceptual categories, it is clear that

Johnson here says little about them, but what he says is relevant. His example
of the sense To be pleased with is worth quoting in full because it sheds
more light on the love of things category:
(2) Fish used to salt water delight more in fresh: we see that salmons and
smelts love to get into rivers, though against the stream. Bacons Nat. Hist.
N. 703.
Besides the fact that salmon and smelt satisfy the criterion one of the
participants is non-human, what they love is not human either, but here
LOVE occurs with another verb. In analysing my data, I have considered such
verbs as non-human participants in LOVE.
Tables 1 and 2 summarise other correspondences between the five catego-
ries and the noun LOVE in each dictionary. Table 1 shows that Johnson seems
to associate parental care chiefly with God. His definition of LOVE conveys the
idea of God as a loving parent whom we must revere. Although family relations
remain explicitly absent, we may assume a reciprocal relationship between
family love and religious love, where each is defined in terms of the other.
Such fuzziness is thus inherent even in Johnsons definition.
It is also clear that good-will is a more general attitude than friendship
and that courtship or lewdness are specified aspects of sexual love. I have
considered my categories as rather loose but containing a variety of selectional
features or attributes which I cannot analyse in detail here.
132 HELI TISSARI
Table 1. The correspondences between the five categories and the noun LOVE in DEL
Category Johnsons definitions Example
and sense numbers
Family love 4. Tenderness; parental care. No religion that ever, was [sic] so fully
represents the goodness of God, and his
tender love to mankind, which is the
most powerful argument to the love of
God. Tillotsons Sermons.
Friendship 2. Kindness; good-will; God brought Daniel into favour and
friendship. tender love with the prince. Dan. i.9.
9. Fondness; concord.
Sexual love 1. The passion between the My tales of love were wont to weary
sexes. you; I know you joy not in a love dis-
3. Courtship. course.
7. Lewdness. Shakespeare.
Religious love 13. Due reverence to God. I know that you have not the love of
God in you. John.
Love of things 5. Liking; inclination to (as, the
love of ones country).
Table 2. The correspondences between the five categories and senses of the noun LOVE in
OED. (The definitions have been shortened for this purpose.)
Category OED definitions Example
and sense numbers
Family love 1. Disposition or feeling 1818 CRUISE Digest (ed. 2) II.
towards a person. 346 The natural love which
Thomas Kirby bore to his brother.
Friendship 1. Disposition or feeling 1535 COVERDALE 2. Sam. i. 26.
towards a person. Thy loue hath bene more speciall
vnto me, [ then the loue of wemen. ]
Sexual love 4. Feeling of attachment based 1975 D. BAGLEY Snow Tiger
upon difference of sex. xvi. 138 Dont you believe in
6. The animal instinct between love at first sight?
the sexes, and its gratification.
Religious love 2. a) Gods affection and 1526 TINDALE I John v. 3 This is
benevolence to His children. the love of god, that we kepe his
b) The affectionate devotion commaundementes.
due to God from His creatures.
c) Affection of one created
being to another by the sense
of their common relationship to
God.
Love of things 3. Strong predilection, liking or 1422 tr. Secreta Secret., Priv.
fondness for, a devotion to. Priv. 218 Philosophie is no more
but loue of witte and cvnnynge.
OED distinguishes fairly well between sexual love, religious love,

love of things and other love, but family love and friendship merge in
its entries into this category of other. I see no reason why this should be so.
Family love and friendship form fuzzy boundaries in my data, but so do
religious love and sexual love. I regard the OED editors decision first and
foremost as a choice between levels of abstraction, especially because the
examples given fit all five categories.
The level of abstraction is high in the definition of the verb, where OED
only distinguishes between loving a personal object and loving a thing. It is
interesting to note the editors comment: In the U.S. a frequent vulgarism for
like; and the example:
(3) 1859 BARTLETT Dict. Amer., To Love, for to like. Do you love
pumpkin pie?
Although OED suggests a twofold distinction between different loves,
it is not wholly consistent because it presents a rather analytical definition of
the specific category religious love, as Table 2 shows. This definition
suggested to me that I should consider religious love not only in terms of
participants but also in terms of the larger domain where it occurs, i.e. a
religious context in which the motive for love is derived from God, even if He
is not directly a participant.
The dictionaries also list other senses, some of which are simply hom-
onyms (such as a kind of thin silk stuff). Mainly relevant to our present
purposes is the sense a beloved person (OED), which I first saw as a
separate category, but since the beloved person is most often a lover, I decided
to situate these instances in the domain where they occur, in this case the
domain of sexual love.
On Prototype Theory
Several aspects of prototype theory have already come to the fore in this
article, but I will now summarize its main implications for this study. Here my
134 HELI TISSARI
main sources are Aitchison (1994: 3972), Geeraerts (1997), and Taylor
(1991).
1. The whole procedure is based on the way that the mind probably deals with
concepts, through (proto)typical representations and contextual information.
The approach minimizes the importance of word-class, as it is assumed that
an identical conceptual core underlies both linguistic forms (Szagun 1983:
160, with reference to Miller and Johnson-Laird 1976), here the noun and verb
LOVE.
2. LOVE is considered a large conceptual cluster, which consists of five
principal smaller clusters: family love, friendship, sexual love, religious
love, and love of things. These clusters have a prototypical core but also
include more peripheral senses. Koivisto-Alanko (1999) suggests that the
totality of a prototype semantic structure resembles fractals in that it is
infinitely detailed.
3. The five major categories which I use are defined both in terms of their
participants (section 2 above) and in terms of the domains where they occur
(following the method in Taylor 1991: 8387):
(1) Family love occurs within the family.
(2) Friendship occurs in the domain of the world at large, between people
whose mutual (or even one-sided) love is not based on a family situation
or sexual relationship.
(3) Sexual love occurs within any range of romantic, sexual and erotic
relationships and encounters.
(4) Religious love occurs in a world defined in terms of Gods dominance.
(5) Love of things contains the remaining instances.
This way of defining lacks refinement but works rather well in the actual
analysis, where a strict participant analysis often does not yield results.
4. Prototype theory implies degrees of typicality (Geeraerts 1997: 1113, 23
24, 3243), which means that it is likely that loving a personal object is more
typical than loving a thing; or that when one looks at the five categories, one of
them is likely to be more typical than the others; and even within the five
categories themselves a certain core or general sense is likely to be more
typical than a host of peripheral senses adjoining it. This is actually merely an
enlargement of implication 2. above, although it must be noted that the
relationship between typicality and frequency is problematic because it is by
no means straightforward (Geeraerts 1988).
5. To elaborate further, prototypical categories tend to have a family resem-

blance structure (Geeraerts 1997: 11, 1516, 2324, 61). In the present case it
is self-evident that while the differences between the five loves can be
considered conjectural, it is only common sense that there is normally a very
real difference between loving ones spouse and loving ones grandmother,
not to mention loving ones sandwich. While there may simultaneously be
elements common to all these loves, such as pleasure, they still belong to
quite different domains in ones life.
6. Prototype theory takes up elements of former theories and moulds a new
whole out of them. The existence of fuzziness has been noted by linguists, but
prototype theory places fuzziness on a par with the other phenomena which it
discusses. The existence of so-called blurred edges is a natural continuation of
the assumption that concepts and/or words have a core and a periphery which
may overlap with some other periphery (Aitchison 1994: 3950, Geeraerts
1985: 136139 and 1997: 11, Taylor 1991: 5455). Consequently, sexual
love and religious love may overlap where Gods love is described symboli-
cally as Christs betrothal to the Church, and so on.
7. Prototype theory acquires a flexible and functional character through
propounding the unsuitability of definition by a set of criterial attributes
(Geeraerts 1997: 11). In the present case this means relief from attempting to
collect all the possible features of LOVE before analysing it.
8. Geeraerts (1997: 8, 19, 2526, 6879) also emphasizes the encyclopaedic
nature of linguistic changes, which means that he wants to relate them to
issues beyond the purely linguistic world, as would I, but my task proves very
problematic, because it involves a vast array of relevant questions. What I do
here is to limit myself to a very brief note on these topics.
A Brief Note on the Extralinguistic Context of the Model
Three major issues suggest themselves to me at this point. Firstly, there are what
a linguist normally calls synchronic and diachronic points of view inherent in
my approach, which concerns two periods separately and together. What one
can do is to look at peoples world views and how they could have changed.
Secondly, it is not only a society which develops and changes but also
every individual. While studying each of them is impossible, an ontogenetic
model corresponds to how love is actually processed in the mind. It would
136 HELI TISSARI
probably be only natural for children to learn to understand love in terms of

certain situations before they are able to recognize the emotion in themselves
and others, and talk about it. The prototype-semantic approach may be a kind
of return to a simpler view of the world, so to say. (See e.g. Murphy 1991,
Nelson 1983, Szagun 1983.)
Thirdly, a topic which relates closely to the latter is the question of how
much of the subject-matter here concerns intercultural or universal phenom-
ena and how much is typical of the English language. Lewis (1960) himself
took his terms from the Greek language, and the dictionary definitions cannot
have been entirely immune to outlandish influence. As a matter of fact, even
Tables 1 and 2 suggest otherwise, since they present us with quotations from
the Bible. Within this overall framework, it should be self-evident that what I
can do here is merely scratch the surface of the subject.
Since Tillyards (1958) The Elizabethan World Picture was one source of
inspiration for constructing my model, I visualized the relationships between
the participants in love as situated within a chain of being where God sits on
the highest throne and everything has its proper place below a superior. This
affected my analysis of the category love of things, into which I, defying
Lewis and some very common sentiments, put everything that could have
been considered to exist below humans. Consequently, all animals became
things, although people may of course (rightfully) claim to love at least some
animals more than some of their fellow humans. Nevertheless, it is probable
that peoples sentiments in this respect have undergone a significant change
(Thomas 1983). In the hypothesis section below, which will follow a descrip-
tion of the data, I will point out some other societal changes which could have
influenced the development of LOVE.
Data
It is important to realize that using corpus data, like any other procedure, has
both advantages and restrictions for my study. The main advantages are that I
can easily access a large and representative collection of texts, that one of
these collections contains evidence of a language (Early Modern English) the
speakers of which can no longer be interviewed, and that the partially overlap-
ping structure of the corpora facilitates a comparison between Early Modern
and Present-Day English.
But although corpora are very useful tools for linguistic research, they are
by no means an unproblematic source of knowledge. To mention a couple of
general problems, one has to decide whether to treat texts and genres individu-
ally or in groups, and to realize that size does not necessarily mean quality
(Rissanen 1993: 7677, Svartvik 1992: 10). Nevertheless, instead of planning
better corpora, it is probably more reasonable to explain why I have chosen to
use exactly these corpora and not, for example, the hundred-million-word
British National Corpus.
It is important to note that the purpose of this study was to test the model
within reasonable limits, which at the time were reached when I had collected
a little over two thousand words representing the verb and noun LOVE and
some related adjectives and adverbs. The adjectives and adverbs were dis-
carded from further study when I decided to concentrate on the noun and verb.
The Early Modern English period of The Helsinki Corpus, Shakespeares
prose, Brown and LOB provided a fairly balanced amount of data from both
periods without resorting to any further selection criteria. Shakespeares prose
was assumed to be more similar to the data gained from HC than his verse,
which is also likely to be less straightforward and therefore more difficult to
interpret.
Table 3 and Figure 1 describe this initial collection of LOVE words from
the four corpora.6 It is worth mentioning that I also planned to include The
London-Lund Corpus of spoken English in the data, but it only contained 35
LOVE words and was therefore discarded.
Table 3. LOVE in the sources. Note that this initial collection was not as restricted as the
final selection for the analyses proper, which explains the differences between
the figures in this and other tables (see note 6)
Total of N/
Corpus Total of LOVE 1000 Nouns Verbs Adjectives Adverbs
words words for
LOVE
HC 551,000 498 0.9 223 260 11 4

Shaksp. 203,892 562 2.8 301 257 4 -
Brown 1,014,000 418 0.4 219 148 50 1
LOB 1,013,737 526 0.5 278 176 67 5
138 HELI TISSARI
3
2.5
2
HC
1.5
SH
1 BROWN
0.5 LOB
0
ADV
ADJ
VERB
NOUN
TOTAL
Figure 1. LOVE in the sources. The columns show the frequency (N / 1000) of words in
each of the sources and word classes respectively. The last columns show the
total for LOVE in each of the sources.
Among other things, Table 3 shows that LOVE is a much more frequent
item in the Early Modern English sources than in the Present-Day English
sources. That it is especially frequent in Shakespeare can be seen even better
in Figure 1. Theoretically, it is even possible to speculate that LOVE appears
to be slightly more frequent in Present-Day British English (LOB) than in
Present-Day American English (Brown), but a proper study of transatlantic
variation would require more evidence. It is also interesting to note that HC is
the only source in which the verb outnumbers the noun.
All the corpora are divided into text categories. The Early Modern
English period of The Helsinki Corpus includes the following text types: BIA
(autobiography), BIBLE, BIO (biography, other), COME (drama, comedy),
CORO (non-private correspondence), CORP (private correspondence), DI-
ARY, EDUC (educational treatise), FICT (fiction), HANDO (handbook other
than astronomy or medicine), HIST (history), LAW, PHILO (philosophy),
SCIM (science, medicine), SCIO (science, other than medicine or astronomy),
SERM (sermon), TRAV (travelogue), and TRI (proceeding, trial). Kyt and
Rissanen (1993: 1213) further divide these into oral and literate text types.
The oral text types contain drama and private letters, some parts of fiction,
sermons and trial records. Note that Shakespeares prose would belong there if
it were included in the corpus.
The division into oral and literate resembles that of Brown and LOB into
informative and imaginative prose. The informative prose ranges from catego-
ries A to J, and the imaginative comprehends the rest. The categories are: A.
Press: reportage, B. Press: editorial, C. Press: reviews, D. Religion, E. Skills,
trades and hobbies, F. Popular lore, G. Belles lettres, biography, essays, H.
Miscellaneous, J. Learned and scientific writings, K. General fiction, L.
Mystery and detective fiction, M. Science fiction, N. Adventure and western
fiction, P. Romance and love story, and R. Humour (see e.g. Francis and
Kucera 1982: 45). It is noteworthy that all the corpora consist of samples
representing these text types, not of whole texts.7
Expectations
The main hypothesis concerns changes in the meaning of LOVE and espe-
cially the five prototypical loves (family love, friendship, sexual love,
religious love, and love of things) between two different periods, the Early
Modern English period and the Present-Day English period. The Early Mod-
ern English period of The Helsinki Corpus covers the years 15001710 (Kyt
1996: 2; cf. Barber 1976: 13, Grlach 1978: 22). The Present-Day English
corpora were compiled in the 1960s and should represent the then current
usage.
The primary inquiry is what could have happened between these two
periods. My first question was whether LOVE had become any less powerful
and absolute a word, whether its meaning had faded, so to say. One way to
measure this would be to see if love of things had become more common
towards the present remember that OED criticises the American vulgarism
that people use LOVE instead of LIKE. A parallel situation can be found in
the French language which uses aimer for both LOVE and LIKE.8
But then other things could have occurred as well. Here I suggest three
areas where a change in peoples world view might have influenced the usage
of LOVE. Firstly, because of the secularisation of Western culture, religious
love could have become less frequent towards the present. Secondly, Early
Modern society was undoubtedly more patriarchal and stricter about family
hierarchy (Stone 1979: 93148). The importance of the family could show in
more frequent occurrences of family love. Thirdly, the data could confirm
the reality of a sexual revolution or at least a growing courage and willing-
140 HELI TISSARI
ness to talk about sexual love in the 1960s, and show if such usage might
differ significantly from our Early Modern ancestors habits.
These are rather rough and ready suggestions, and in order to comple-
ment the description of my data, I also decided to see what had happened to
LOVE in the various text types. Expressions of affect tend to be most frequent
in personal letters, face-to-face conversations, telephone conversations, letters
of recommendation and romance fiction (Biber and Finegan 1989: 103, 105
108). If LOVE behaved in the same way, we should expect it to be most
frequent in personal correspondence in HC and in the category P. Romance
and love story in the two other corpora. If we think of the content of these
texts, we can expect to find family love in the personal correspondence and
sexual love in the love stories. More generally, LOVE should be more
frequent in the Early Modern oral text types than in the literate, and more
frequent in Present-Day imaginative than informative prose.
Love in the Texts
The results of the text type analysis confirmed the expectations, as Tables 46
show. LOVE was most frequent in the private letters in HC and in the category
P. Romance and love story in Brown and LOB. Indeed, all the highest
figures in HC are in the oral group. After private letters come sermon, fiction,
and comedy. Similarly, LOVE is more frequent in the imaginative than in the
informative prose in the Present-Day English corpora. However, the differ-
ence between these two categories is fairly small in Brown. What surprised me
was that in HC the category BIBLE contained very little LOVE, but then HC
only includes samples of the Bible.9
In the Present-Day English corpora it was surprising that there were
categories of informative prose where LOVE was almost as frequent as in
imaginative prose. For example, the press includes many book reviews which
deal with emotion. The line between the two categories thus appears to be
fairly fluid, at least when one considers it from a LOVE point of view.
At their face value, the content of the categories also suggests that some
prototypical loves might be more frequent in the corpora than others. We
have already noted that private correspondence is likely to involve family
love, if not also sexual love in love letters. The problem is that the Present-
Day English corpora do not contain any such correspondence and therefore
are not absolutely comparable to HC.

Sexual love should be an issue in romances and love stories, although it
is actually surprising that the Early Modern fiction and comedies contain more
LOVE. In addition, it seems that there is so little religion in the Present-Day
English corpora that religious love cannot be very frequent in them. How-
ever, it is quite difficult to say anything conclusive about the five loves
without looking at the LOVE words themselves in their actual contexts.
Table 4. The frequencies of the love words per text type in the Early Modern English
period of the Helsinki Corpus.
The items included: loud, loue, loued, loue-letters, louer, loues, louest, loueth, loufyng,
louing, louinge, louyd, louyng, louynge, lovd, love, loves, loved, loveing, loveinge, lover,
lovers, loves, lovest, loveth, loving, lovinge, lovingest, lovyng, lovynge, lovg.
Text type Number of words Number of N/1000
love words
Law 36,750 1 0.03
Handbooks, other 33,660 10 0.31
Science, medicine 12,800 1 0.08
Science, other 24,400 0 0.00
Educational treatises 32,980 27 0.82
Philosophy 25,590 7 0.27
Sermons 32,240 86 2.67
Proceedings, trials 43,960 13 0.30
History 32,820 24 0.73
Travelogue 39,350 4 0.10
Diaries 36,790 4 0.11
Biography, autobiography 15,420 12 0.78
Biography, other 16,420 14 0.85
Fiction 36,080 79 2.19
Drama, comedies 35,120 71 2.02
Letters, private 35,370 97 2.74
Letters, non-private 17,830 14 0.79
Bible 43,420 20 0.46
TOTAL CORPUS 551,000 484 0.88
142 HELI TISSARI
Table 5. The frequencies of love words per genre (text category) in the Brown Corpus
The items included: love, love-, loves, loves, loved, love-in-action, love-making, lover,
lover-, lovers, lovers, lovering, loveways, lovin, loving, loving-.
Genre Number of words Number of N/1000
love words
A. Press: reportage 88,690 8 0.09
B. Press: editorial 54,505 18 0.33
C. Press: reviews 35,346 22 0.62
D. Religion 34,590 21 0.61
E. Skills and hobbies 72,590 9 0.12
F. Popular lore 97,223 24 0.25
G. Belles lettres, etc. 152,064 115 0.76
H. Miscellaneous 62,477 2 0.03
J. Learned 162,211 15 0.09
SUBTOTAL: Informative 759,696 234 0.31
K. General fiction 58,380 30 0.51
L. Mystery and detective 48,208 11 0.23
M. Science fiction 12,042 6 0.50
N. Adventure and western 58,416 17 0.29
P. Romance and love story 58,625 57 0.97
R. Humor 18,277 8 0.44
SUBTOTAL: Imaginative 253,948 128 0.50
TOTAL CORPUS 1,013,644 363 0.36
Table 6. The number of love words per text category in the Lancaster-Oslo/Bergen
Corpus
The items included: love, loved, love-affair, love-letters, love-making, love-play, lover,
lovers, lovers, lovers, loves, loves, love-stories, lovin, loving.
Category Number of words Number of N/1000
love words
A. Press: reportage 89,138 12 0.13
B. Press: editorial 54,447 1 0.02
C. Press: reviews 34,321 53 1.54
D. Religion 34,387 8 0.23
E. Skills, trades, and hobbies 76,913 19 0.25
F. Popular lore 89,090 32 0.36
G. Belles lettres etc. 155,336 71 0.46
H. Miscellaneous 60,761 0 0.00
J. Learned and scientific 161,900 11 0.07
SUBTOTAL: Informative 756,293 207 0.27
K. General fiction 59,204 91 1.54
L. Mystery and detective 49,145 17 0.35
M. Science fiction 12,119 3 0.25
N. Adventure and western 59,391 23 0.39

P. Romance and love story 59,382 107 1.80
R. Humour 18,203 5 0.27
SUBTOTAL: Imaginative 257,444 246 0.96
TOTAL CORPUS 1,013,737 453 0.45
The Prototypes
To familiarize the reader with the kind of data I analysed, I will first discuss
the five loves separately and give examples of each kind. I would like to
stress again that alongside the participant analysis, I relied on contextual
information concerning the domain of love. By so doing I was able to
differentiate between the five loves in most cases.
At times I was aware of different possible readings, but it was only
afterwards that I fully realized their value as indicators of fuzzy areas. There
seems to be a tension between the flexibility of the model and the idea that it
would help to deal with fuzziness. In this study at least I tended to situate each
item in one of the five categories because they were not strict enough to
challenge my readings. I will discuss these problems separately before pre-
senting the numerical results.
Family Love
Family love was more frequent in the Early Modern English texts than in the
Present-Day English period, but a comparison between the two periods does
not reveal all, because it was fairly rare in Shakespeare. We have already seen
that LOVE was especially frequent in private letters in HC, which contributed
to this result.
This is problematic in two ways: firstly, because one could consider the
LOVE words in letters quite separately as formulaic phrases; and secondly
because Brown and LOB do not contain any such correspondence. However,
if we consider examples like (4) merely formulaic, we seem to challenge the
sincerity of the writers even if this was the expected way to begin letters:
(4) To my very lovinge sonne, Mr. Henry Oxinden, at Corpus Christi Collidge
in Oxfoord, give this. (HC: Oxinden, Letters [ROXINDEN] 26)
144 HELI TISSARI
The least that one can say is that the Early Modern speakers were quite aware
of their family duties.10
The second result was that in both periods it was chiefly the nuclear
family which was talked about. In other words, the participants in family
love were mostly the closest relatives: parents, children, or spouses, as in (5)
and (6).
(5) [Between spouses:] Love will account that to be well said, which it may be
was not so intended. (HC: Taylor, The Marriage Ring [JETAYLOR] 28)
(6) If his mother loves him, he clings to that love as a ballast. (Brown: Press:
Editorial B13: 102)
Sometimes the family member is a grandparent or a cousin:

(7) beleeve me, if I say he looses non of his grandfather loue (HC:
Harley, Letters [HARLEY] 2)
(8) His younger brother Charlie held up his chubby arms with love towards
cousin Nelly. (LOB: P: Romance, love story P18:5)
It is also possible that there is no mention of a specific relationship

between the participants:
(9) I left behind me brave men, whom captivity had robbed of all hope. They
too loved their families (Brown: Belles Lettres, G27: 68)
Friendship
Friendship was more common than family love in both periods, but even
more frequent in the Early Modern English period. If the central sense of
friendship is a mutual and free relationship between people who simply
enjoy each others company, then the more peripheral senses were much more
frequent in this cluster. For example, when Falstaff exclaims to Prince Hal:
(10) Thou owest me thy love (The First Part of King Henry the Fourth
3.03.137),
he is talking about money.

Friendship, or perhaps it would be better to say friendliness or loy-
alty, often appeared in contexts where it had to do with ones duty towards a
subject or a superior, or quite simply neighbourly love, or universal benevo-
lence (cf. Wrightson 1993: 5157). The power relationships between people
appear more clearly in the Early Modern English period, and the Early
Modern English data also reflect a more male-oriented view of friendship in
society.
The master-servant relationships of the Early Modern data could be of
various kinds:
(11) [the] motive that [{led to{] my stirr with my Lo. was my love to my
Lord (HC: The Trial of the Earl of Essex [ESSEX] 17)
(12) such entier loue he bare vnto king Edward and his children (HC:
More, The History of King Richard III [MORERIC] 78)
Hospitality is also typical of the Early Modern data:

(13) where with all loue I was entertained with much good cheere
(Taylor, The Pennyles Pilgrimage [JOTAYLOR] 140.C2)
In the Present-Day English data friendship becomes a general principle:

(14) there is in the universe of persons a moral law, the law of love, which is
a natural law in the same sense as is the physical law. (Brown: Belles Lettres,
G55: 36)
It also includes example (15), which discusses a teacher of German:

(15) She was a sweet, kind creature and we all loved her. (LOB: G: Belles
lettres, biog G22: 82)
In terms of power relationships, it is of course difficult to consider this

friendship, but the item suggests that her way of dealing with power was
very friendly. All in all, it seems that people are more likely to use the word
LOVE when they are not talking about their friends in the strictest sense.
Sexual Love
It was to be expected that sexual love should be quite frequent because

LOVE was most frequent in the Present-Day English category P. Romance
and love story. Sexual love was indeed the most frequent category in both
periods, although it was even more dominant in the Present-Day English data.
It is noteworthy that the Early Modern English data did not contain any pure
love letters.
The whole process of falling in love and getting married is described by
Kvecses (1986), who sketches both an ideal and a typical model of romantic
love. In the ideal model the lovers reach a stage where loves intensity is
146 HELI TISSARI
maximum and will not be disappointed, while in the typical model everyday
reality intervenes and even if the partners get happily married, love cools to
affection (Kvecses 1986: 9596, 103104). The ideal model is behind many
of the Present-Day English items:
(16) You swept her off her feet, and she was all prepared to love you for ever.
(LOB: P: Romance, love story P11: 45)
(17) No price is too high when true love is at stake. (Brown: Romance P22:
114)
Both models are probably summarized in:

(18) Ive dreamed about falling in love and getting married. (LOB: P: Ro-
mance, love story P25: 203)
but the ideal remains an ideal. People offer advice on how to cope with the
reality, where some hopes are lost:
(19) [the] marriage bed is a place of unity and harmony. The partners each
bring to it unselfish love, and each takes away an equal share of pleasure and
joy. (Brown: Popular Lore F08: 2)
(20) our sympathies are usually enlisted on the side of the man whose love
is not returned, and we condemn the woman as a coquette (Brown: Belles
Lettres, G31: 4)
Older conceptions of love underlie the Early Modern data, which pro-
vides us with violence, sighs and tears:
(21) This love will undo us all. (Troilus and Cressida 3.01.110)
(22) my violent love, my trade in arms, and all my vast desire of glory, avail
me nothing (HC: Behn, Oroonoko [BEHN] 160)
(23) I can love to all the Tenderness of Wishes, Sighs and Tears (HC:
Farquhar, The Beaux Stratagem [FARQUHAR] 59)
Shakespeare provides a multitude of examples, among them

(24) I will not be sworn but love may transform me to an oyster. (Much Ado
About Nothing 2.03.24)
(25) Sir, the maid loves you, and all shall be well. (The Merry Wives of
Windsor 1.04.120)
The peripheral cases include (26) where dogs behave according to

Kvecsess model:
(26) It is love at first sight, marriage at first opportunity, and soon fifteen
beautiful puppies. (LOB: C: Press: reviews C16: 30)
not to mention:
(27) the nymphomaniac trip-hammer operator who falls hopelessly in love
with a middle-aged steam shovel. (Brown: Humor R08: 44)
Religious Love
As against sexual love, it was to be expected from the beginning that

religious love would not be the most frequent prototype. Shakespeare pro-
vides hardly any examples, but HC contains more religious love than the
Present-Day English corpora. In HC the LOVE words usually appear in Bible
citations or in explanations of Biblical passages, while the Present-Day Eng-
lish data what little there is is slightly more varied.
Obviously, most of the items appear in the biblical texts and sermons in
HC and in the category D. Religion in the Present-Day English corpora.
Example (28) from LOB is an exception:
(28) God may hold us and our Roman Catholic brethren firmly in his love. (G:
Belles lettres, biog G66: 43)
There God is the source of love but we may also expect it to mean that the
brethren will love one another. In the second, biblical, example God is the sole
lover:
(29) Who shall separate us from the love of Christ? (Brown: Religion D07: 96)
Considering the Christian world-view, one could say that religious love
in its primary sense is Gods love for people, but God can also inspire love in
and among the created. Therefore in its secondary sense religious love is a
divinely inspired love between people. Both aspects are summarized in ex-
ample (30) from HC. This sermon again quotes the Bible:
(30) Loue God aboue all things, and thy neighbour as thy selfe. (HC: Smith,
Two Sermons Of Usurie [SMITH] B4R)11
Love of Things
Love of things is somewhat more frequent in the Present-Day English data

than in the Early Modern, where it is about as frequent as family love and
148 HELI TISSARI
friendship. The hypothesis about its becoming more frequent towards the
present is hardly proved, but it does become more dominant with respect to the
other loves.
In comparison to the other categories, the range of the participants is
rather large. To get a better idea of their nature, I conducted an analysis of the
complementation of the verb in this category. The analysis showed that the
verb was mostly complemented by a direct object. Infinitives and participles
were rarer. However, love of things was typically a verb category, while the
other loves were more often represented by the noun.
The following items can be considered typical because they attest
complementation through a direct object:
(31) If you love an addle egg as well as you love an idle head (Troilus and
Cressida 1.02.133)
(32) Some likewise have had pictures and letters printed in this manner on the
back side of a pack of cards, to entice children, that naturally love that sport, to
the love of learning their books. (HC: Hoole, A New Discovery of the Old Art
of Teaching Schoole [HOOLE] 8)
(33) The thing is that these bees love a fine-grained soil that is moist
(Brown: Learned J10: 89)
The less typical complementational patterns appear in:

(34) I love not to be crossed. (Loves Labours Lost 1.02.32)
(35) I will bar no honest man my house, nor no cheater, but I do not love
swaggering (The Second Part of King Henry the Fourth 2.04.103)
(36) I am no kill-joy, but would love to see this period of the year treated with
more reverence. (LOB: E: Skills, hobbies E12: 73)
Lastly, a couple of examples which feature the noun LOVE:

(37) I was in love with my bed. (Two Gentlemen of Verona 2.01.81)
(38) Of course his love was expressed in intelligent care. He planted the pansy
seeds himself, buying them from a pansy specialist. (LOB: EO2: 7)
Fuzziness
The blurred edges between the categories are actually quite as interesting as
the relative frequency of the categories themselves. I will here briefly describe
seven areas of fuzziness, beginning with four which could come under the title
problems of participant analysis.

The idea of the participant analysis was that it would be possible to label
each instance of LOVE by identifying its participants. Problems arise not so
much when the participants are not exactly identifiable but rather when they
could be assigned to several groups simultaneously. In the present data these
questions arose in four situations:
1. When the participants could simultaneously be considered lovers and
spouses. It seems absolutely impossible to say that marriage would mean that
a couple cease to be lovers, although they do become spouses. However,
Kvecsess model suggested that when people get married their love cools to
affection, which is one way of describing something which people tend to
experience (Kvecses 1986: 103104). Lewis gives affection a positive
flavour: As for erotic love, I can imagine nothing more disagreeable than to
experience it for more than a short time without this homespun clothing of
affection (1960: 36).
There are basically two ways of dealing with this problem short of
changing the present approach. One must either separate marital love as a
category of its own or label the items by some other criteria. In this study I
decided to use contextual criteria for distinguishing between family love and
sexual love in these cases. Therefore items such as (5), which describe the
spouses attitude towards each other, became family love, while items such
as (19), where sexuality was clearly at issue, became sexual love.12
2. It was not clear what to do with texts discussing incest:
(39) We are also struck by the fact that this story of a boys love for his mother
does not offend, while the incestuous love of the man, Paul Morel, sometimes
repels. (Brown: Belles Lettres G26: 35)
Here the participants are a son and a mother, but if the love has a repellent
quality, it would seem slightly out of place to call it family love. I therefore
categorized such items as sexual love.
3. Homosexuality was not a problem if it was explicitly stated:
(40) the poems that deal with love (always homosexual love) (LOB: C:
Press: reviews C12: 40)
In (40), we need not even know the participants in order to know that they are
lovers and that the item belongs in the category sexual love. But it was
impossible to conduct any deep analyses of cases which on the surface seemed
to belong in the category friendship because they discussed love between
150 HELI TISSARI
two (fe)males. I decided not to read homosexuality between the lines.

4. Metonymy was typical of Early Modern love discourse:
(41) I can be charmd with Sapphos singing without falling in Love with
her Face (HC: Farquhar, The Beaux Stratagem [FARQUHAR] 7)
When part of the beloveds body represents the beloved, a literal interpretation
might also situate the item in the category love of things. However, it seemed
fittest to situate such items in sexual love. This decision was not only based
on the participant analysis, reading the whole for the part, but also on what I
knew about the context or domain where such love occurred.
The remaining three areas of fuzziness could probably be called domain
problems. Here the problem was whether to look at the context or the explicit
reference:
5. One of the cases which I had to deal with was Smiths sermon on usury:
(42) the Vsurer loueth the borrower to grow rich by him. (HC: Smith, Two
Sermons Of Usurie [SMITH] C1R)
This is only one of the several LOVE items in the sermon. The question was
whether to consider them religious love because they appeared in a sermon,
or friendship because they dealt with the relationship between two people
who were not family members or lovers, and because Smith advocated a
concern for our fellow humans. I decided to treat each item separately.
The love which Smith discusses in (42) does not seem to have anything to
do with religion, rather the contrary. I thus labelled it (negative) friendship
unfortunately there was no way to categorise (the) false love. However, if Smith
explicitly talked about Gods commandments the label was religious love:
(43) For Christ sayd to his Disciples, Loue one another, as I have loued you.
(HC: Smith, Two Sermons Of Usurie [SMITH] C1R)
6. Another problem also occurred in a sermon. This time it was Taylors

sermon giving practical advice on how to behave towards ones spouse (see
example 5); he also talks explicitly about sexuality, and just a little about the
relationship between Christ and the Church in very religious terms. Having
realised that the Present-Day English data contained similar advice to married
couples, I decided to give Taylors sermon special treatment and consider it
first and foremost a treatment of marriage (family love or sexual love).
7. The last note concerns my example (26) about the dogs falling in love.
Strictly speaking, my analysis regards dogs as things, so that the example
ought to be situated in love of things. However, since the dogs behaviour

was described from an anthropomorphic perspective, I decided that the item
belonged rather to sexual love.
The dogs romance is an example of the family resemblances which
people can see between two different loves. I consider friendship as against
homosexuality (sexual love) another area where the family resemblance
structure can be so strong that two people may see things in two quite different
ways and be quite convinced of their own opinions.
According to prototype theory, fuzziness appears in the categorical pe-
ripheries where edges meet and blur. However, it is not necessarily always so.
In the case of married love the overlap between family love and sexual
love seems to be quite close to the centre of both categories. Of course one
can say that married love is a prototypical category of its own and thus
resolve the problem. Otherwise my problems with the present data could
probably be considered peripheral, but they are simultaneously a very
relevant part of the whole.
Numerical Results
The numerical results are based on the methodological decisions that I have
just discussed. In other words, they give a general picture of the microfield(s)
of prototypical meanings but mostly overlook fuzziness by approximating
each LOVE item to a category. This is a simplified but probably quite reliable
way to measure change between the two periods. A discussion of the blurred
edges would have to be more complex.
The results can be viewed in two ways. One can either compare the
absolute frequencies of the five loves or the proportions of each love in the
total. If one compares the absolute frequencies, then each prototype becomes
less frequent in the Present-Day English data because it contains less LOVE.
It is only when one compares HC with the Present-Day English corpora that
an exception appears: sexual love is more frequent in LOB (0.31/1000
words) than in HC (0.21/1000 words).
However, the proportions of the prototypical loves in the period totals
tell a slightly different story. The proportions of sexual love and love of
things increase, while the proportions of friendship and family love de-
crease. Surprisingly, nothing seems to happen to religious love. Figure 2
summarizes the trends.
152 HELI TISSARI
Even if they must remain tentative, the results indicate changes between
the relative proportions of the prototypical meanings. These changes are partly
analogous and partly contrary to expectations. Love of things has become
more dominant because family love and friendship have lost some of their
status, but it has not become much more frequent if it is considered on its own.
Sexual love has clearly gained more ground in the Present-Day English data,
but religious love has remained almost as it was.
In other words, the data mainly behaved in accordance with the sexual
revolution and loss of patriarchal society hypotheses which predicted an
increase in sexual love and a decrease in family love. This interesting result
raises new questions, above all two:
1. Were the results really reliable? First and foremost, it is possible that if
the Present-Day English corpora had contained family letters the results
would have looked quite different. There is no exact correspondence between
the text types in other respects either, for example because the Present-Day
English data contain no drama. It is also important to note that the data do not
contain any spoken English, which is likely to be different from written
English. A study of Present-Day spoken English could be compared with the
Early Modern oral text types.
2. If the result indicates real changes, what are the reasons behind them?
This is an even more complicated question and beyond a linguist alone to
answer.
If the data reflect sexual liberation, it is impossible to say whether it was a
short-term or a long-term process, ongoing at the time of the compilation of
Brown and LOB or a development which began a long time ago. It is likely, for
example, that literature has increasingly affected peoples minds. There are
text types which did not even exist in the Early Modern English period as we
know them (the novel, book reviews), not to mention the fact that the influ-
ence of literature is associated with the growth of literacy.
The development of the family and family values since the Renaissance is
also a big question, although it seems certain that peoples attitudes have
changed a great deal, for example in the area of what are considered family
duties. Duty also relates to the concept of friendship which, if defined as
broadly as here, concerns the nature of the whole society and its real or imagined
development (from hierarchical to democratic, from Christian to pluralistic).
Storge Storge
Philia Philia
Eros Eros
Agape Agape
Things Things
The Early Modern English Data The Present-Day English Data
Figure 2. The proportions of the prototypes in the two periods

Family love decreases from 14% to 5% and friendship from 16% to 6%, while sexual
love increases from 49% to 64%. Religious love stays the same (7% and 6% respec-
tively), while love of things increases from 14% to 19%. The figure excludes some
unanalysable items which formed 3% of the data.
Conclusion
The present analysis suggests that it is possible to view LOVE in prototype-

semantic terms, as consisting of five major clusters or senses which are
situated in different conceptual domains. These senses are: family love,
friendship, sexual love, religious love, and love of things.
There are strong family resemblances between these loves, which means
that a look at dictionary definitions will show that their edges meet and blur. A
study of LOVE in The Early Modern English period of The Helsinki Corpus,
Shakespeares prose, The Brown Corpus and The Lancaster-Oslo/Bergen
Corpus suggests that it is possible to categorise real data quite satisfactorily by
looking at the participants and context (domain) of these five loves. However,
sometimes the participant analysis can simultaneously suggest two categories,
and the contextual information may also differ from the participant analysis.
Statistically, the study indicated that sexual love is dominant in both
periods, although it becomes even more frequent in Present-Day English. The
numerical analysis suggested that the relative frequencies of the five loves
had changed between the two periods. Family love and friendship had
become less frequent, while the share of love of things and religious love
had remained almost the same. The content of the data sources has influenced
these results to an extent, but their indicative value should not be overlooked.
154 HELI TISSARI
The article further suggests that LOVE could be considered at several

levels of abstraction. For example, one could add a category of married love
to the five prototypes. On the other hand one could reduce the categories to
two by noting only whether the cause(r)s of love are personal or not. A closer
view (catching the semantic detail) would require another, auxiliary model.
Clearly, this is a fairly rough picture of a simultaneously very small and
very rich semantic field. Nevertheless, the study offers a new point of view on
the history of LOVE.
Notes
1. Throughout this article, I will use LOVE in capital letters to refer to the lexeme in a broad
sense which comprises all its tokens. Italics on the other hand will emphasize the type (cf.
Lyons 1977: 1825).
2. For the triangle of word meaning see e.g. Smith (1996: 115) or Ullmann (1981: 55).
3. For example in the OED definition of LOVE, disposition and feeling appear side by side.
4. I used the WordCruncher program to search for the relevant tokens.
5. I used Spevack (1968 and 1970) for the relevant information.
6. To begin with the nouns, I not only looked at the basic noun love but also at lover and
twelve compounds: love-affair, love-broker, love-cause, love-god, love-in-action, love-
letter, love-making, love-play, love-prate, love-song, love-story and loveway(s). The
initial count which is presented here also included loveliness, which was discarded from
the analyses proper. The verb category mainly consisted of the basic (to) love in its
various inflections. All the instances of the type loving were counted as verbs. In addition,
there was a single instance of the rare lovering. There were eight adjectives: lovable,
love-alluring, loveless, lovelorn, lovely, lovely-ugly, love-mad and love-shakd. These
and the only adverb, lovingly, were later discarded.
7. For more specific information on the texts in HC see the manual (Kyt 1996: 1417, 165
230). I have not found a similar list of texts in Brown and LOB, but Francis and Kucera
(1982: 46) gives some idea of what they contain.
8. However, a French etymological dictionary does not suggest a meaning development
where aimer would have bleached. Instead, it simply gives the word two senses
(Gamillscheg 1969: 22).
9. For a discussion of LOVE vs. CHARITY, see Tissari, in prep.
10. See Abbott (1996: 180189) for an example of The duties of husbands and wives,
parents and children, masters and servants.
11. This is necessarily a simplification of religious love.
12. I use the other approach in Tissari (1999), where I look at marital love separately.
References
Abbott, Mary. 1996. Life Cycles in England 15601720. Cradle to Grave. London:
Routledge.
Aitchison, Jean. 1994. Words in the Mind: an Introduction to the Mental Lexicon. 2 ed.
Oxford: Blackwell.
Barber, Charles. 1976. Early Modern English. London: Andre Deutsch.
Biber, Douglas & Edward Finegan. 1989. Styles of Stance in English: Lexical and
Grammatical Marking of Evidentiality and Affect. Text Vol. 91. 93124.
Diller, Hans-Jrgen. 1994. Emotions in the English lexicon: a historical study of a lexical
field. English Historical Linguistics 1992. Papers from the 7th International Confer-
ence of English Historical Linguistics ed. by Fernndez, Fuster & Calvo, 219234.
Amsterdam: Benjamins.
Francis, W. Nelson & Henry Kucera. 1982. Frequency Analysis of English Usage: Lexicon
and Grammar. Boston: Houghton Mifflin.
Gamillscheg, Ernst. 1969. Etymologisches Wrterbuch der Franzsischen Sprache.
Heidelberg: Carl Winter.
Geeraerts, Dirk. 1985. Cognitive restrictions in the structure of semantic change. Histori-
cal Semantics. Historical Word-Formation ed. by Jacek Fisiak, 127153. Amsterdam:
Mouton.
Geeraerts, Dirk. 1988. Where Does Prototypicality Come From. Topics in Cognitive
Linguistics ed. by Brygida Rudzka-Ostyn, 207-229. Amsterdam: Benjamins.
Geeraerts, Dirk. 1997. Diachronic Prototype Semantics. A Contribution to Historical
Lexicology. Oxford: Clarendon.
Grlach, Manfred. 1978. Einfhrung ins Frhneuenglische. Heidelberg: Quelle & Meyer.
Johansson, Stig & Knut Hofland. 1989. Frequency Analysis of English Vocabulary and
Grammar. Vols. 12. Oxford: Clarendon Press.
Johnson, Samuel. 1755/1979. A Dictionary of the English Language. London: Times.
Kjellmer, Gran. 1973. Middle English Words for People. (=Gothenburg Studies in
English 27). Stockholm: Almqvist & Wiksell.
Koivisto-Alanko, Pivi. 1999. Mechanisms of Semantic Change in Nouns of Cognition: A
General Model? In this volume.
Kvecses, Z. 1986. Metaphors of anger, Pride and Love: a Lexical Approach to the
Structure of Concepts. Amsterdam: Benjamins.
Kyt, Merja & Matti Rissanen. 1993. General introduction. Early English in the Com-
puter Age. Explorations through the Helsinki Corpus ed. by Merja Kyt, Minna
Palander-Collin & Matti Rissanen, 117. Berlin: Mouton de Gruyter.
Kyt, Merja. 1996. Manual to the Diachronic Part of The Helsinki Corpus of English Texts.
Coding Conventions and Lists of Source Texts. 3rd ed. Helsinki: Department of English,
University of Helsinki.
Lewis, C.S. 1960. The Four Loves. London: Collins.
Lyons, John. 1977. Semantics. Vol. I. Cambridge: Cambridge University Press.
Miller, G. A. & P.N. Johnson-Laird. 1976. Language and Perception. Cambridge: Cam-
bridge University Press.
156 HELI TISSARI
Murphy, Gregory L. 1991. Meaning and Concepts. The Psychology of Word Meanings
ed. by Paula J. Schwanenflugel, 1135. Hillsdale, NJ: Lawrence Erlbaum Associates.
Nelson, K. 1983. The Conceptual Basis for Language. Concept Development and the
Development of Word Meaning ed. by Th. B. Seiler & W. Wannenmacher, 173188.
Berlin: Springer.
OED. 1989. The Oxford English Dictionary. 2nd edition. Oxford: Oxford University Press.
stman, Jan-Ola. 1989. English love: temporary or permanent? Journal of Pragmatics
13. 10191022.
Rissanen, Matti. 1993. The Helsinki Corpus of English Texts. Corpora Across the
Centuries. Proceedings of the First International Colloquium on English Diachronic
Corpora ed. by Merja Kyt, Matti Rissanen & Susan Wright, 7379. Amsterdam:
Rodopi.
Schwyter, J.R. 1996. Old English Legal Language. The Lexical Field of Theft. Odense:
Odense University Press.
Smith, Jeremy. 1996. An Historical Study of English. Function, form and change. London:
Routledge.
Spevack, Marvin. 1968. A Complete and Systematic Concordance to the Works of
Shakespeare. Vol. 1. Hildesheim: Georg Olms.
Spevack, Marvin. 1970. A Complete and Systematic Concordance to the Works of
Shakespeare. Vol. 5. Hildesheim: Georg Olms.
Stone, Lawrence. 1979. The Family, Sex and Marriage in England 15001800.
Harmondsworth: Penguin.
Svartvik, Jan. 1992. Corpus Linguistics Comes of Age. Directions in Corpus Linguistics.
Proceedings of Nobel Symposium 82 Stockholm, 48 August 1991 ed. by Jan Svartvik,
713. Berlin: Mouton de Gruyter.
Szagun, G. 1983. From Outer to Inner Reality: On the Meaning of Some Moral
Words. Concept Development and the Development of Word Meaning, ed. by Th. B.
Seiler & W. Wannenmacher, 157172. Berlin: Springer.
Taylor, John R. 1991. Linguistic Categorization. Prototypes in Linguistic Theory. Oxford:
Clarendon.
Thomas, Keith. 1983. Man and the Natural World: changing attitudes in England 1500
1800. New York: Pantheon Books.
Tillyard, E.M.W. 1958. The Elizabethan World Picture. London: Chatto & Windus.
(Reprint.)
Tissari, Heli. 1999. LOVE Shakes the Spheres: Five Prototypical Meanings in
Shakespeares Plays. The Proceedings of the Conference of the Nordic Association of
English Studies 1998. (= Anglicana Turkuensia No. 20) ed. by Sanna-Kaisa Tanskanen
& Brita Wrvik, 175-191. Turku: Turun Yliopisto.
Tissari, Heli. in prep. AFFECTION, FRIENDSHIP, PASSION and CHARITY. A history
of four love lexemes since the fifteenth century. To be published in Neuphilologische
Mitteilungen.
Ullmann, Stephen. 1981. Semantics. An Introduction to the Science of Meaning. Oxford:
Blackwell. (Reprint.)
Wrightson, Keith. 1993. English Society 15801680. London: Routledge.
The Vocabulary of CONSENT
in Middle English
Louise Sylvester
Kings College, London
Abstract
Against the background of the difficulties of tracing out satisfactory defini-

tions of the crime of rape from the available medieval sources, this paper
considers the idea of CONSENT in Middle English. The definitions of the term
consent in the major dictionaries of English are examined, as is the vocabulary
which may be seen as synonymous or partially synonymous with the term
consent around the time of its adoption into English. The essay then considers
the lexicalization of the concept in Old English and developments in the
lexical field which led to the borrowing of the term consent in Middle English.
Introduction
The origins of this paper lie in my interest in the idea of CONSENT as a legal
concept in the medieval period. Scholars examining the law of rape in the
Middle Ages are agreed on the difficulty of teasing out definitions from the
statutes, eyre rolls and legal treatises of the period (Hanawalt 1979: 104;
Carter 1985: 35). In the earliest English legal treatise, known as Glanvill,
raptus is defined in terms of rape, and this is elaborated by Henri de Bracton in
his thirteenth century legal treatise. The statutes of Westminster I (1275) and
II (1285) both make explicit mention of the womans consent, and ban the
possibility of redemption by marriage (Post 1978: 158). However, in these
treatises the distinctions between rape and abduction (a crime against those
158 LOUISE SYLVESTER
under whose authority the woman lived) are blurred. It has been suggested that
often a woman would allow herself to be abducted by her lover, but at the
same time seek to show that the abduction was carried out against her will
(Bellamy 1973: 58). The issue of consent is crucial to legal definitions of rape
and to the manner of its prosecution in the medieval period (Carter 1985: 45).
It is crucial also to the way in which we frame our constructions of gender and
sexuality (there is a growing bibliography on this topic, but see especially
MacKinnon 1989: 175). These issues led me to consider the vocabulary
available to express CONSENT in Middle English.
The thirteenth century has been portrayed as the period in which Anglo-
Norman became established as a distinct language and English was exposed to
a new wave of French influence of a more central and cultivated kind (Strang
1970: 217); more recent research on French borrowings in texts of this period
seeks to question the extent to which this vocabulary may be classified as
borrowed into English, rather than as self-consciously French usages (Dor
1992: 484485). The focus of this paper is in the main semasiological: I shall
examine the definitions of the term consent in the major dictionaries of
English, including consideration of the ordering of the senses and the use of
citations. I shall also look at vocabulary which may be seen as synonymous or
partially synonymous with the term consent around the time of its adoption
into Middle English, and I shall consider the lexicalization of the concept of
CONSENT in Old English and developments in the lexical field which led to the
borrowing of the term in the Middle English period. A particular concern is
the way in which the language of CONSENT is treated in the Oxford English
Dictionary (OED) and the Middle English Dictionary (MED), but I shall look
also at the evidence provided by the Thesaurus of Old English (TOE). The
paper is in part an onomasiological study, reflecting on the ways in which the
concept of CONSENT developed and on the competing possibilities for the
lexical expression of CONSENT in the medieval period.
Consent in the Dictionaries
The Oxford English Dictionary and the Middle English Dictionary both note
the verb form of the word consent as the earliest recorded English form of this
word shape. Both suggest that the word was borrowed alongside the term
consense, with which it shares some of its meaning in English (see below).
THE VOCABULARY OF CONSENT IN MIDDLE ENGLISH 159
The earliest use of the verb consent in English is defined and illustrated in the
major dictionaries as follows:
OED:
6. Voluntarily to accede to or acquiesce in what another proposes or desires;
to agree, comply, yield a1225 Ancr. R. 272 So sone so me biginne
kunsenten to sunne.
MED:
3 (c) to yield (to a sinful impulse, to temptation, to a seducer), surrender;
c1230(?a1200) *Ancr. 74a: is heteward [reisun] li to slepen sone se me
biginni consenti to sunne.
It is interesting to note the different extrapolations made by the dictionaries

editors: the OED, which offers citations up to 1848 to illustrate this sense of
the term, offers a more generalized definition, which seems close to the
meaning the term carries today. The MED offers a definition that seems more
closely tied to the specific context of the citation.
As consent is a loan word, we need to consider its meaning in the source
language from which it was borrowed. Godefroy (1883) lists two verb forms:
consentir and consentre.1 Without a dictionary which indicates which senses
were in play at a particular historical moment, it is difficult to be sure about the
definitions of these terms, but the first appears to mean to reconcile, bring into
harmony; to grant, concede or admit a proposition, to grant a favour, etc.; to
approve of something, to consent, agree to, sanction something; to be in
agreement, while the second carries the part of the meaning to approve of
something, to consent to, sanction something.
Consent in the OED
The OED notes that consent v. is an adoption from OF, ultimately from Latin
consentire to feel together, agree, accord, harmonize, and adds that [t]he
sense consent to a thing being done was a subsequent development, but
occurs in 12th c. in Fr., and is app[arently] the earliest recorded in Eng. The
first (but later) group of senses is headed: I To agree together; and five senses
are listed within it: 2
1. To agree together, or with another, in opinion or statement, to be of the
same mind, a1300+15351865
2. To agree in sentiment, to be in accord, to be at one, 1382+16111651
3. To come to an agreement upon a matter or as to a course of action,

1375+1546+1668
4. Of things: to agree, be in harmony, 15401794
The fifth sense, To act or be affected in sympathy, is not recorded for the
medieval period. Despite its being the earliest recorded sense in English, the
OED lists consent to a thing being done in the second sense branch of the
verb consent:
II. To agree to a proposal, request etc.
6. Voluntarily to accede to or acquiesce in what another proposes or desires;
to agree, comply, yield, a1225
7 to be consented: to be agreed; to be an accessory or consenting party (to
something), c13861485
8. To allow, agree to, consent to, 13861534.
The OEDs ordering of the senses of the verb consent emphasizes the
etymological sense of the term, the Latin senses derived from con together
and sentire to feel, think, judge etc. The list of senses for the noun consent is
not subdivided, but from examination of the pattern and order of the senses
listed for the verb it is clear that in the case of the noun the order is reversed:
1. Voluntary agreement to or acquiescence in what another proposes or
desires; compliance, concurrence, permission, a13001440+15901875
2. Agreement by a number of persons as to a course of action; concert, 1382
1650+1780
3. Agreement or unity of opinion, consensus, unanimity, 15291658+1785
1879
4. Agreement in feeling, sympathy; also, more generally, harmony, accord,
agreement, 1382+15791870
Interestingly, given the order of definition terms in the OED, sense 1 is first
recorded in what appears from the citation to be the sense permission:
Cursor M. 4955 (Cott.) Yee sald him an gain mi consent. The other
citations from Middle English texts tend much more towards the other defin-
ing words offered in the definition, however. This first sense is followed by
three sub-divisions, none of which is recorded for the medieval period. Of the
remaining six senses, we may note that four begin with the word agreement.
The OED lists other consent forms used in the medieval period:
consentant, a., Consenting, c1385;

consenter, One who consents, or is a party to anything, 1303c1380;
consenting, vbl. sb., The action of the verb consent, the giving of consent,
c1380+1482;
consenting, ppl. adj. Agreeing or giving consent (to a proposal or course of
action; formerly, also, to an opinion, a13001382;
consentment, The action of consenting, consent, 1340+1491+1525; also,
noted as irreg. formed on consent after assent, assentation: consentation
a1529, cited once, from Skelton.
It is perhaps possible to infer from these forms that from the end of the
thirteenth century the term consent had become naturalized in the language as
a productive base for word-formation, and that, within the medieval period,
the sense of granting permission, of agreeing to something or to someones
proposition, was gaining ascendency over the etymological sense of agree-
ment.
Consent in the MED
The MED uses Middle English spellings as headwords. Thus the noun form of
consent appears first, with the inflected verb form consenten following. MED
offers four sense divisions for the noun consent, and, as in the OED, these are
not subdivided into sense-branches. At first sight it seems as if the first sense,
recorded from the S[outhern] Leg[endary] c1300, was in use a century earlier
than the three other senses cited, which are from the Cursor Mundi (senses 2
and 3) a1400, and Wycl.Church (sense 4), a1425. If the suggested composition
dates for these texts, a1325 and ?1384 respectively, are accepted, however, the
noun form of consent would seem to have been used in English in all four of
the senses listed at roughly the same time. The MED thus notes:
Consent n.
1 (a) Agreement in sentiment; harmony (of hearts or sounds); in o ~, of one
mind, in agreement; c1300+c1450/51+1532 rev. (c1385);3
(b) resemblance (between things), c1384;
(c) of ~, (persons) of like mind or purpose, c1400
2. Consent, approval; gain mi ~, against my wishes, a1400(a1325)-1464
3. Inclination or yielding (to sinful desire, etc.), surrender; yeven fleshli ~, to

yield to lust, a1400(a1325)-a1450(c1412);
4. Connivance, abetting, acquiescence, a1425(?1384)-a1500(?c1378)
Interestingly, MED editors, unlike those of the OED, do not think that it is
possible to assign the origins of the English term solely to Old French, and
they suggest that the verb form of consent was coming into English from Latin
at the same time. The MED does not offer notes on meaning, and so, although
the Latin and OF forms consentire and consentir are noted in MED (the same
OF verb form is given in OED), no etymological discussion is offered. Four
main senses are listed. The senses for the verb form of consent are not divided
into overall groups, but the first two senses are both divided into sense-
branches, each with its own subdivisions of meaning.
1a. (a) To agree mutually, ~ togeder: ben consent(ed, be in agreement,
a1400(c1303)-c1450;
(b) ~at, of til, to on (ende), to agree unanimously, a1400(c1303)-c1450;
1b. To be consistent (with sth.), be adequate, (c1390)
2a. (a) To give ones consent (at, til, to, sth.), assent; ben consent(ed, be
agreeable (to sth.), accede (to), c1300(a1464);
(b) to approve (of sth.), express approval, c1300(c1384)+1450a1500
(a1415) [citations listed as (c), but are presumably meant for (b).];
2b. (a) To grant or allow (that sth. is true, that sth. be done, etc.), (c1395)-
a1450(c1410);
(b) to concede (sth.), c1400(?a1387)
3. (a) To comply, give in, (c1395)-a1425(?a1400);
(b) to comply (with a rule, truthfulness, a suggestion), conform; yield (to an
emotion, a false doctrine), (c1384)+14501475(a1400);
(c) to yield (to a sinful impulse, to temptation, to a seducer), surrender,
c1230(a1200)+1340(c1390);
(d) to submit or give in (to sb.), c1390(?1325)-c1475(?c1400)
4. (a) To connive or acquiesce (in a crime, a sin, etc.); ben consenting,
consented; give tacit encouragement or support, tolerate, c1325(c1300)-
(a1456)
(b) to connive (with a traitor, thief, etc.), be an adherent (of), give encourage-
ment (to), a1400(c1303)-(a1456).
Other items follow on in the MED from consent, including

consenta(u)nt adj & n., ppl of OF consentir, 1. (a) Agreeing in sentiment; (b)
inclined, willing; (c) acquiescent; accessory (to a crime);
2. As noun: an abettor (in) or accessor (to a crime), conspirator [listed in OED
as consisting adj consenting with one citation from 1385 (and one from
1866)];
consentement n. (a) Consent, approval; ?divine grace; (b) an acquiescence,
toleration (of sinful desire); (c) abetting; (d) an abettor;
consentinge ger. 1. (a) Agreement; (b) accord, consistency; 2. (a) Approval,
consent; (b) acquiescence, lack of restraint; yielding, surrender (to lust, etc.);
~ of sinne; (c) connivance, abetting; prive ~ [listed in OED as vbl. sb the
action of the verb consent, the giving of consent from 1380 and as a ppl adj
meaning agreeing or giving consent (to a proposal or course of action
formerly also to an opinion) 13501382+1578];
consentingli adv. Sympathetically, in a considerate manner [OED records the
first usage in 1552];
consentour n. (a) One who consents or approves; (b) one who complies or
acquiesces; (c) one who abets or supports (an offender); an abettor or accom-
plice [OED One who consents, or is a party to anything, 13031380 (two
quotations in the Middle English period)].
It is noteworthy that neither explicitly nor covertly does the OED make
reference to the idea of sin as that which is consented to in its definition
language, whereas the MED makes reference to both crime and sin, and seems
to gesture towards these terms as defining characteristics of that to which
consent may be given in the usages which are defined by the idea of yielding.
Consense
Further consideration will be necessary to uncover why it was that English in

the Middle English period acquired a new term expressing the linked ideas of
permission and agreement. We should look first, perhaps, at the noun consense,
adopted from OF cunsense, consence, ~sense, which seems to have come into
English at the same time as consent v. as a competing term of similar meaning.
Consense is not recorded in English in this sense after c.1380.4 The first citation
refers to consent as one of three steps towards carnal desire. Indeed, the sense
given in the OED for consense is consent. The MED divides consence into: (a)
Yielding (to a sinful desire) and (b) acquiescence (in) or tacit encouragement (of
sinful conduct); and therefore suggests that the term covers only part of the
range of senses of consent. Both dictionaries cite the parallel passages from the
same text, the Ancrene Riwle, to illustrate the first usage of the two lexical items,
consense and consent, in English. The citations for consense are as follows:
OED: a1225 Ancr. R. 228 at we ne beon nout allunge ibrouht erin, mid
kunsence of heorte and mid skiles hettunge. Ibid. 288 reo degrez beo
erinne [carnal desire]. . e uorme is cogitaciun: e oer is affectiun: e
ridde is kunsence.
MED: 1230(?a1200) *Ancr. 62a: Wi consens [Nero: kunscence] of heorte,
wi skiles hettunge. Ibid. 78b: reo degrez beo rin [carnal desire]. . e
forme is cogitatium, e oer is affectiun, e ridde is suscence [Nero:
kunsence]. . Cunsense. . et is skiles hettunge, hwen e delit i e lust is igan se
ouerfor et ter nere nan wi seggunge.
The OED editors, then, consider consense to be synonymous with consent,

while the MED editors suggest that consense is a synonym of consent in one of
its senses only, the one which they suggest is its earliest usage in English. Both
dictionaries note that the term was borrowed into English from French at the
same time as consent, and both cite Ancrene Riwle as the text in which the
terms are first recorded in English. The OED and the MED also offer two
further citations, both from Wycliffes sermons, and here consense collocates
with synne. In the case of the MED a particular sin is not specified, indicating
the reason for the MED definition of the term. The second MED sense division
is supported by citations from Monumenta ritualia ecclesiae Anglicanae and
Wycliffes De Officio Pastorale. The OED entries indicate that consense as a
competing term meaning consent had dropped out of the language by the end
of the fourteenth century. Yet the MED offers citations for the second sense-
branch from a1450 and a1500 (although the composition date in this second
case is, interestingly, estimated at c1378). OED notes, however, that the term
consent faced competition immediately after the medieval period from the
noun consension, listed as meaning Agreement in thought, feeling or opin-
ion, in use between 1593 and 1692, with a revival in North American usage in
1878.
The Lexical Field of Consent in English: History
The loss and replacement of native word stock in the Middle English period is
a complex matter: to try to discover something about the reasons for the
adoption of these terms into Middle English we shall now consider how
CONSENT was lexicalized in Old English.
The concept of CONSENT encompasses in its meaning notions of concord
and agreement, and includes agreement to something, with the implied sense
of permission or agreeing to someone doing something. Was this idea
lexicalized in Old English? Within section 9. Language and Communication,
the TOE contains a small lexical group entitled Consent. The taxonomy of
headings at this point of the classification runs:
09.07.04 Assertion, affirmation
09.07.04.01 Confirmation, agreement See 14.04 ff. Agreement
09.07.04.01.01 Acknowledgement
09.07.04.01.02 Consent
09.07.04.01.03 Unanimity
As we have seen, agreement is the prototypical meaning of consent in

French and Latin, and this is reflected in the presentation of the senses in the
OED and MED. The term consent was not first used in English with the
meaning agreement, however, and this is implicit in the placing of the terms
grouped under the heading Consent in a separate category from those
expressing agreement. Within the structure of the TOE the CONSENT vocabu-
lary is sandwiched between the concepts of ACKNOWLEDGEMENT and UNANIM-
ITY (a group equally suggestive of the idea of agreement), and these groups are
the last in this section. The scope of this paper does not allow investigation of
other lexical fields which might be seen to overlap with that of CONSENT. A
brief examination of the vocabulary which appears in the TOE category under
the headword Agreement does not reveal any clear overlap of lexical items,
for the lexis expressing the ideas of agreement and of permission appears
mainly in the TOE categories 12.03.05 Permission and 14.04 Making of
terms, agreement, convention. We may note, however, that the definition
language of the former group in both Bosworth-Toller and the TOE tends to
include the term consent: geunnan is defined in Bosworth-Toller as To give,
grant, allow, concede, but Tollers Supplement adds (1) To grant a request,
consent; (ge)afian is defined in Bosworth-Toller as To favour, support,
permit, allow, admit, assent, consent, agree, approve, obey, submit to, and
(ge)tiian as To grant, allow, with the Supplement adding (1) To consent to

a request. From the TOE under 12.03.05 Permission, we should note:
.Assent, permission: ti
.Consent, will: est
.Permission, allowance: unne
.To allow, permit, grant, consent to: willan
.To grant a request, consent: geunnan
.To concede, allow, grant: (ge)tiian
.To turn to, assent to, approve: (ge)cierran to
.To condescend, deign: gemedemian
from 06.02.02 Will, disposition:

.By my will: mines ances
.At ones pleasure/will: ances
and from 06.02.02.01 Will, wish, pleasure:

.What pleases, accords pleasure: gemede, gewyrd
.Willingness to give, pleasure in doing: unne
The terms willan and (ge)afian (see definition in Bosworth-Toller for the
latter term, which is listed with the Permission group in the TOE but without
reference to the idea of CONSENT in its defining language) do not appear in the
lexical group of terms expressing CONSENT in either the TOE or the materials
of the Historical Thesaurus of English (HT), which is currently in preparation
at Glasgow University. Nevertheless, most of the remaining terms do appear,
sometimes in variant forms, in the Consent group in the HT materials,
offering further evidence of how the vocabulary of a particular period looks
when viewed diachronically rather than at its particular moment of usage only.
The TOE category appears as follows:
09.07.04.01.02 Consent: gifung, (ge)afung, gewrung
.Consent, agreement: geafsumnesog, geeodsumnesog
.To consent: mideahtianog, midwrianog, geafettanog5
It is immediately obvious that, when viewed synchronically, CONSENT, al-

though lexicalized in Old English, had not achieved the complex of meanings
that the definitions grouped under the term consent in the MED suggest in the
Middle English period. The flags o and g also show that, on the evidence
available in the standard Old English dictionaries, the terms expressing the
concept are rare within the corpus of Old English, appearing mainly in
glossaries or glossed texts. That this group appears at such a low level in the
classification of its TOE category (indicated by the number of preceding sub-

groups) suggests that the concept of CONSENT was not central to the meaning
of this (or any other) category of vocabulary in Old English.
The Lexical Field of CONSENT: Scope
With the help of the raw materials of the HT, it is possible to consider the
question of the lexicalization of the idea of CONSENT in Old English within the
diachronic context of the semantic field of CONSENT. The preliminary classifi-
cation of the HT materials was according to numbers allocated to ideas in
Rogets Thesaurus (Roget). [Roget places the vocabulary expressing CON-
SENT in category 758, its third category in section 2 Special social volition in
Division two, Social volition.] 758 is immediately preceded by 756 Permis-
sion (an arrangement in line with the classification of this vocabulary by the
TOEs editors), and by 757 Prohibition. The other categories in the section,
again arranged in what Roget presents as antonymic pairs, are 759 Offer, 760
Refusal, 761 Request, 762 Deprecation, and 763 Petitioner. This arrangement
seems both to underline the fact that Roget is a classification of ideas rather
than of lexis (see Collier and Kay 198081: 8889), and to indicate that his
meanings are legalistic. In law the concept of CONSENT is defined within the
context of the widely held belief that a sexual relationship is initiated and
formed by the making of an offer (usually by a man) with the intention of
obtaining (a womans) agreement. It is assumed that the man will attempt to
fulfil this intention by a variety of means up to, but normally not including,
physical force (Brundage 1993: 6768; Naffine 1994: 20). The Roget arrange-
ment would seem also to corroborate the implications of the legal definition of
consent.
The identification of those lexical items from the HT corpus classified in
Roget category 758 makes it possible to begin to consider the semantic field of
CONSENT diachronically within the medieval period. These as yet unedited
materials of the HT offer a preliminary view of members of the lexical set of
CONSENT nouns:
est OE; afsumnes OE; consense a1225c1380; assent a1300(1814); con-
sent a1300; accord 1393; condescent c14601689.
In due time it should be possible to undertake a detailed examination of the

lexical items listed under the headword Consent in both TOE and HT and to
consider whether the group of lexical items, apparently prototypical in the Old
and Middle English periods, changes according to whether one is looking at a
synchronic snapshot of the lexis, or at part of a diachronic listing. Such an
examination will enable us to consider the vocabulary in use at the interface
between Old and Middle English, with HT classification allowing close
consideration of the naturalizations of borrowed French terms in English in
the thirteenth century. There is space here to consider in detail only a few of
the lexical items that appear to have expressed the idea of CONSENT immedi-
ately before and during the Middle English period.
Est
The absence of est from the OED indicates that the term did not survive
beyond 1150, and est is not an MED headword. Thus, est would appear to
have given way to other possibilities for the expression of the idea of CONSENT
once that term itself was borrowed into English. Bosworth-Toller lists a large
number of definition terms within two categories, the first two of which, under
category I, are will and consent, thus suggesting that the term est had the
central sense consent as part of its core meaning. The other senses listed
under I are grace, favour, liberality, magnificence, and bounty, and
the sense group II consists of the sense delicacies. The Dictionary of Old
English (DOE) lists three main sense groups: 1. kindness, favour, grace;
bounty, munificence; 2. devotion (to God), piety, prayer, vow; 3. delight(s),
pleasure(s), luxuries. Here the senses most closely related to the concept of
CONSENT occur at 1.a.i ofer Godes/metodes est against the consent/will of
God; 1.a.iii in accordance with the will of God; perhaps 1.a.iv specifically
mercy (of God), glossing misericordia; and perhaps 1.c kindly, bounteously,
gladly. The other senses, involving devotion to God and luxuries, riches, seem
to confirm the impression gained from the Bosworth-Toller definitions that
the word was moving in the direction of the kinds of virtues and pleasures
associated with the magnanimity first of God and then of royalty and the
aristocracy, rather than widening to encompass such areas of meaning as
permission and agreement. The term est appears in six lexical groups in the
TOE, most of which relate to the senses luxurious, delicacy, grace and
favour. The term does not appear at 09.07.04.01.02 Consent, but is found
instead within 12.03.04 Permission.
afsumnes
The term does not appear without prefix in the TOE, which lists only
geafsumnes. Similarly, it appears in Bosworth-Toller as geafsumniss, with
the sense Agreement, consent. The definition indicates that the meaning of
the term is closer to consent in its sense harmony and agreement. The verb
(ge)afian appears in the TOE in the list of words meaning To let, allow,
permit. Bosworth-Toller offers three sense groups for this term, all seemingly
pertinent to the meaning consent: I To consent to, agree with, approve of,
assent to, allow, permit; II to submit to, bear, suffer, endure; III to bear with,
tolerate.6 Cf. geafung; IV to consent, assent. The OE noun form (ge)afung,
at the head of this section of the TOE is given two sense branches in Bosworth-
Toller: I permission; II submission to action, toleration; and a cross-reference
(Cf. ge-afian; III consent). The first quotation cited for this last sense very
strongly resembles that offered for consent in its earliest usage in English in
the OED and the MED: Us is gecynde t we lc yfel on rio wisan
urhtion: urh gespan and urh lustfulnesse and urh geafunga (consensu).
According to OED the etymology is unknown, and the word is not found in
any of the cognate languages (Holthausen 1941; Bammesberger 1948; Hoad
1986). The definition offered for the verb is To consent to; to allow, permit; to
submit to, suffer, endure; to tolerate: a rather wide range of meaning. Cer-
tainly the Middle English reflex thave overlaps with consent in its modern
sense, although perhaps less so in its earliest uses in ME, where the contexts
cited do not make explicit mention of the idea of sin (e.g. a1023 Wulfstan
Hom. iii (Napier) 23 Eal t he for us and for ure lufan afode and olode).
The noun form thaving is given the interpretation permission, consent.
The MED offers one definition of thaven v., subdivided into four
branches:7
(a) To permit an action, consent to something; put up with (a condition or
situation); with that clause: permit (sth. to happen); also, tolerate (sb.); (b) to
grant permission to (sb.) for (an action); also, concede (sth.) to (sb.); (c) to
permit (sb. to do sth.); (d) to bring (sth.) about, manage; arrange; (e) to endure
(a blow), withstand; also, ?suffer illness.
Parts of both the OED and the MED definitions, in particular MED (e),
indicate an interface with the idea of suffering as well as with ideas of
permission and consent. So, similarly, do some of the citations in which the
term thaven collocates with the term tholen, whose meanings include MED:
1 (a) To be made to undergo (a penalty for misdeeds); (b) to submit to

(judgement), 2 (a) To suffer; (b) to suffer the distress of (woe, misfortune,
etc.); (c) to suffer or endure (hunger, pain, etc.); also to suffer (physical
injury); 4 (a) To be patient; forbear; (b) to put up with (injustice, a
complaint, an act, etc.). bear; bear with (someones words); etc., for example,
under (a) St.Marg. 36/3.
The element of suffering among these definitions, not paralleled so directly in

the other permission words, leads me to conjecture that thaven occupies at
least part of the semantic space which came to be filled by consent, but that the
borrowed term offered the refinement of the specific association with consent-
ing to sin.
Assent
This lexical item is particularly interesting within the terms of this paper
because it was borrowed into English at the same time as consent, and it seems
to bear much the same semantic load. As with consent, assent is recorded in
English first in the verb form. Again, the OED notes a number of senses in use
in the Middle English period for both the noun and the verb forms:
1. To give the concurrence of ones will, to agree to (a proposal), to comply
with (a desire) 12971878
2. To come to an agreement as to a proposal; to agree together, determine,
decide c13001591
3. To conform in practice, submit, yield (to) 13401656
4. To give or express ones agreement with a statement or matter of opinion;
to agree to an abstract proposition, or a proposal that does not concern
oneself, or involve ones own action c1380c1450+16121874
The OED editors add a qualifiying note to sense 4: The ordinary modern use
as distinguished from consent. This seems to be in part an attempt to distin-
guish the modern usages of consent and assent, which are deemed to be
different from one another, and to offer an at least partial synchronic as well as
a diachronic view of the sense development of the term assent. The first
definition concludes: Arch. in general sense, and commonly replaced by
consent, exc[ept] as said of the sovereign assenting to a measure, or as in 4.
The diachronic approach of the OED is thus able to offer only a broad outline
of the ways in which sense histories of the two lexical items have gradually
diverged, and more specialist help is required to tease out the differences (if
any) in the Middle English period. Clearly, the meaning of assent relates to
that of est in sense 4 and to that of consent in the other senses.
As with consent in MED the noun form precedes the verb form assenten.
The relevant senses for the noun in MED are:
1. (a) Consent, approval; formal endorsement, c1330a1475
(b) acquiescence, connivance, c1390a1500(a1415)
2. (a) Mutual agreement (of two or more parties), a13331474
(b) at, bi, with on ~, in complete agreement, of one mind, unanimously, etc.,
c1350c1475
3, (a) Sentiment, attitude, opinion; ben at assent, agree in sentiment, c1385
a1500(?a1400)
(b) will, intent, intention; bi fe~, par ~, voluntarily, willingly, c1387/95
a1500
(c) of (at). .assent, associated (with sb.) in sentiment, purpose, or activity
(such as an offence or crime); ben of ~, be an associate or accomplice, c1390
a1500(a1460)
By contrast, the senses for the verb assent parallel those given for the verb
consent rather closely. Both verbs are very well attested. The entry for the verb
form assent occupies well over two columns for the first two senses alone:
1. (a) To assent or consent (to a proposal, plan, etc.); approve of (a suggestion,

idea, etc.); express agreement, agree, c1300a1500(?c1400)
(b) to consent to (an offense, a wrong); condone, abet, c1300a1475(?a1430)
(c) ben assent(ed, be in agreement, consent, approve; condone, a1375
a1500(?1400)
2. Of two or more parties: (a) to come to an agreement, agree, c1330
c1450(?a1400)
(b) to agree upon (a proposal, a choice), c1390a1500(a1415)
(c) ben assent(ed, come to, or be in agreement; be agreed, a1393a1500
(a1415)
3. To yield or submit (to sb.); ben assented, be compliant or accommodating,
c1300a1500(a1415)
The full array of the definitions offered by the MED, perhaps more than any
one of the senses individually, gives the impression of a term whose meaning
in the Middle English period was identical to that of the term consent. Unlike
the OED, the MED is a period dictionary and cannot therefore offer a narrative
to account for or hypothesize about the development of the meaning of a
particular term within the period (except where it is clearly indicated by the
dates of usage), much less about the ways in which terms competed and found
accommodation in relation to one another within the semantic space available.
It is possible in this case that it was the idea of agreement which became focal
in the meaning of consent, enforced by the presence of the Latin etymon con.
Accord
The OED offers ten senses for the verb accord, nine of which are recorded in
use in the medieval period. All centre on the concepts of bringing others,
objects or oneself into agreement, harmony or reconciliation (senses 14 listed
under the heading I To cause to agree, reconcile); or agreeing (senses 510
listed under the heading II To agree). The definitions are almost exactly the
same as those in the first category of the verb consent in the OED. Only sense
10 seems to carry any of the prototypical meaning of the concept of CONSENT:
10. To agree to, consent to, grant (a request), 1393+a16491873
This sense has only one citation for the medieval period, however.
Similarly, the definitions of the noun form, like those of assent, seem to
cover the same ground as those of consent:
1. Reconciliation, agreement, harmony; concurrence of opinion, will, action,
12971800
2. A formal act of reconciliation, or agreement; a treaty of peace; a treaty
generally, 1297+c14401860
4. Agreement or harmonious correspondence of things or their properties, as
of colours or tints, esp. of sounds. Agreement in pitch and tone; harmony,
c13841879
5. Assent to a proposal or request; permission, grant, 13931483+1602
The first term of the definition is the only one which offers a suggestion of
variation in semantic load, adding the terms etymological sense reconcilia-
tion. Again it is the final sense which comes closest to the prototypical
meaning of consent. The OED editors are able to offer only two citations from
the medieval period, however, and these citations seem closer to the ideas of
permission and of will than to consent. The paucity of references seems

to confirm the idea that usage of accord in the sense consent was extremely
limited in the Middle English period.
Examination of the MED corroborates the information given in the OED:
all the definitions for the verb accord in the MED concern coming to agree-
ment, reconciliation, making a treaty, or being harmonious:
1. (a) To come to agreement or understanding; become reconciled, c1300
a1500(?c1400)
(b) to make a formal agreement or settlement; come to terms; make peace,
a1121a1500(c1465)
2. (a) To reconcile (persons); make friends of, a11211483
(b) to bring about reconciliation (between persons), a1398 (one quot.)
3. To agree in sentiment or opinion, be in agreement, c1350?c1450(?a1400)
4. (a) To agree mutually (to do sth.); also, to conspire, c1300a1500(?c1450)
(b) to agree (upon or to a proposal or plan); assent, consent; also, promise,
1384a1500(a1460)
(c) to grant or award (sth.), c1400(c1378)-1421
5. Of things: (a) to be compatible or harmonious, 1340a1475
(b) to correspond; be proportionate; be related or refer (to sth.), 1340a1398
(c) to be comparable, similar or alike (in some respect); be the same, c1400
?c1425
(d) to compare (one word with another), play on words, a1425 (one quot.)
6. (a) To reconcile or harmonize (things); compose (differences), 1340
c1430(a1410)
(b) to make amends; settle (a debt), 1451 (1 quot.)
7. (a) To be fitting or proper (with respect to nature, reason, morals,
customs, etc.); to be becoming, a1393c1475(a1449)
(b) be suitable or good (for a certain purpose), a1398
(c) of medical treatment or diet: be appropriate; hence, beneficial, a1398
a1500(?c1425)
8. Gram. Of forms: to be in agreement or concord, c1400(?a1387)-a1500
Only the sense given as 4(c) seems to coincide with the prototypical meaning
of consent, and here again, the emphasis is on the idea of agreement, with the
element of permission almost wholly lacking.
The senses given for the noun equally reflect the concepts of agreement,
harmony and reconciliation, with the effects of these happy states being
perhaps even more strongly emphasized, particularly in the first definition:
1. (a) Friendly sentiment, attitude, or disposition; good-will, c1330(?c1300)-
c1443
(b) sentiment or opinion; standen of~ agree in sentiment (with sb.); fallen of
~, to consent; bi ~, voluntarily, a1393a1450(c1410)
(c) ben at or of on ~, by mutual agreement; standen in on ~, have the same
attitude, a1250a1500
The second definition in the MED refers to the concept of CONSENT, and links
it explicitly with the notion of permission:
2. Consent, permission, a1393+1405
Here, however, the paucity of material is even more striking: there are
only two illustrations for this sense of accord. The remaining definitions are
exclusively concerned with reconciliation and harmony, in some cases in such
technical senses as the formation of covenants between parties in dispute or at
war, the harmony of musical sounds, and grammatical agreement.
The dictionaries seem to indicate that accord is used with the sense
consent for only a short time (at least as a prototype of that definition), and
the paucity of citations even in the MED indicates that the term was not in
frequent use in this sense in the Middle English period. These definitions
suggest one reason for the survival of the term consent in the prototypical
sense in which it was first used in English. The term accord seems clearly to
have denoted agreement, harmony, concord etc. in the Middle English period:
it is these ideas which form the bases of the definitions in both the OED and
the MED even more firmly than the etymological sense of reconciliation in
which it was first used in English. The term accord seems, therefore, to have
prevented the senses of agreement etc. from becoming the prototypical senses
of the term consent.
Conclusion
The primary focus of this paper has been the definition language used in the
major dictionaries for the term consent. It is clear that the OED offers a more
generalized interpretation of the term consent and that it favours the etymo-
logical meaning in the ordering of the senses, perhaps indicating a suspicion
on the part of the editors that the term was not fully naturalized in English at
the time of its first recorded usage. Agreement, the prototypical meaning in
French and Latin, is thus reflected in the OEDs ordering of the senses. The
MEDs definition for the first usage of the term in English is tied much more
closely to the citation, but its hierarchy of senses again privileges the etymo-
logical meaning.
In terms of the consideration of the legal definition of the concept of
consent which formed the springboard for this discussion, the separation of
the concepts of AGREEMENT and PERMISSION that underlie the concept of
CONSENT is not necessary in the context of the use of the concept of CONSENT as
a defence to a rape charge. The metalanguage of the OED appears to promote
the idea that consent implies a contract normally proposed by one party (in the
male role), who seeks to gain the (perhaps unwilling) agreement of the other;
for example sense 6 Voluntarily to accede to or acquiesce in what another
proposes or desires; to agree, comply, yield. The citations themselves do not
support the OED definition and they should not be used to illustrate the notion
that the term consent is centrally useful to express the opposite of the formula
against her will normally used to denote rape.
A secondary focus of this paper was on the lexicalization of the concept
of CONSENT in Old English, and involved lexical-field development, which led
to the borrowing of the term in Middle English, together with an examination
of some of the competing possibilities for the prototypical term in the field.
The concept of CONSENT encompasses the concepts of concord and agreement,
including the sense of agreement to something, and with the implied sense of
permission or of agreeing to someones doing something. Bosworth-Toller
suggests that the Old English terms for permission include the concept of
CONSENT, but more recently the TOE indicates no evidence for overlap be-
tween the words listed under the headword Agreement and those listed within
the very small category Consent. Therefore, although there is some evidence
that the concept of CONSENT was lexicalized in Old English, the lexical items
are for the most part both rare and hardly central to the Old English corpus.
The problems of dating Old English vocabulary securely mean that it will be
very difficult to work at the interface between Old and Middle English, but
once the work has been done to produce a conceptual classification of Middle
English comparable with that of the TOE, we shall be able to examine the
incoming term consent within its semantic field.8 This preliminary analysis of
the vocabulary offers some understanding of why the term was first used in
English in what remained its prototypical sense and counters the prominence
in the major dictionaries of the senses most strongly influenced by the terms
etymological meaning.
Notes
1. Godefroy (1883) lists two verb forms: consentir with three senses: Act, accorder;
Approuver; and Neutr. tre daccord, and consentre v. a. with the sense accorder.
Godefroy lists consence s. f. et m. with the sense consentement, intelligence, complicit,
complot, with subdivisions Volont, Faire consence quelquun, consentir a son dsir,
and Egard, considration; and consens s. m. with the sense accord, consentenant. We find
consent s. m. meaning consentenent, accord, concert and consente s. f. meaning consen-
tement, complicit. Other forms include consencion s. f., meaning consentenent, permis-
sion, with subdivisions Accord, pact, and Volent, dsir; consen adj. meaning qui a
consenti , qui a connaissance de, qui reconnait; consentable, adj., qui est de concert
avec, complice de; consentance, s. f. consentement; consenteison, s. f. consentement,
permission; consenteor, s. m. celui qui donne son consentenent quelque chose, com-
plice; consentissement, s. m. consentent.
2. The definitions are occasionally shortened, with syntactical information omitted, for the
readers convenience.
3. I set out the dates in the way devised by the editors of the Historical Thesaurus, with +
indicating a gap in the evidence. (The composition dates suggested in the MED some-
times make date strings appear less than meaningful.)
4. The word-shape appears as a hyphenated nonce-word in 1678 with the meaning Joint-
sense (equivalent to consciousness).
5. The TOE editors use four symbols as annotations to the OE entries, the flags relating only
to word forms, not to meanings. In this extract we find the flag o, which the editors
suggest should be viewed as a warning that a particular word form is very infrequent,
and the flag g, which points to word forms that appear generally in glossed texts or
glossaries (TOE: xxi).
6. Toller (1921) offers four sense groups for this term: I to permit; II to permit what is
displeasing to oneself, suffer; III to permit something (troublesome) to be done to oneself,
to submit to, to suffer.
7. It seems worth noting again that the practice of the MED is not to offer one or more main
definitions with sub-groupings, but rather to divide up the definition(s) without implica-
tions of a hierarchy of prototypicality of meaning.
8. Work towards a Middle English thesaurus has begun at Kings College London, sup-
ported by a grant given under the Institutional Fellowship scheme of the British Acad-
emy/HEFCE, to whom I here record my gratitude.
References
Bammesberger, Alfred. 1948. English Etymology. Heidelberg: Carl Winter.

Bellamy, John. 1973. Crime and Public Order in England in the Later Middle Ages.
London: Routledge & Kegan Paul; Toronto: University of Toronto Press.
Bosworth, J. & N. Toller. 1898. An Anglo-Saxon Dictionary. London: Humphrey Milford
(cited as Bosworth-Toller).
Brundage, James A. 1993. Sex, Law and Marriage in the Middle Ages. Aldershot, Hamp-
shire and Brookfield, Vermont: Variorum (first published 1978 as Rape and marriage
in the Medieval Canon Law. Revue de droit canonique, 6275).
Campbell, Alistair. 1972. An Anglo Saxon Dictionary: Enlarged Addenda and Corrigenda
to the Supplement by T. Northcote Toller. Oxford: Clarendon Press.
Carter, John Marshall. 1985. Rape in Medieval England: An Historical and Sociological
Study. Lanham, New York, London: University Press of America.
Collier, L. W. and C. J. Kay. 198081. The Historical Thesaurus of English. Dictionaries
2/3. 88112.
Dor, Juliette. 1992. Post-dating Romance loan-words in Middle English: are the French
words of the Katherine Group English? History of Englishes: New Methods and
Interpretations in Historical Linguistics ed. by Matti Rissanen, Ossi Ihalainen, Terttu
Nevalainen & Irma Taavitsainen. Berlin and New York: Mouton de Gruyter.
Dutch, Robert A., ed. 1962. Rogets Thesaurus. London: Longman.
Godefroy, F. 1883. Dictionnaire de lAncienne Langue Franaise et de tous ses Dialects du
Xe au XVe sicle. Paris.
Hanawalt, Barbara A. 1979. Crime and Conflict in English Communities 13001348.
Cambridge, Mass. and London: Harvard University Press.
Healey, Antoinette diPaolo et al., eds. 1986. Dictionary of Old English. Toronto: Pontifi-
cal Institute of Mediaeval Studies.
Hoad, T. F. 1986. The Concise Oxford Dictionary of English Etymology. Oxford:
Clarendon Press.
Holthausen, Ferdinand. 1949. Etymologisches Wrterbuch der Englischen Sprache. Gt-
tingen: Vandenhoeck and Ruprecht.
Kay, Christian, Iren Wotherspoon, Jane Roberts & M. L. Samuels. forthcoming. The
Historical Thesaurus of English. Oxford: Oxford University Press.
Kurath, Hans & Sherman M. Kuhn, eds. 1954. The Middle English Dictionary. Ann
Arbor: University of Michigan Press.
MacKinnon, Catherine A. 1989. Towards a Feminist Theory of the State. Cambridge, Mass.
& London: Harvard University Press.
Naffine, Ngaire. 1994. Possession: erotic love in the law of rape. Modern Law Review 57.
1037.
Post, J. B. 1978. Ravishment of women and the Statutes of Westminster. Legal Records
and the Historian: Papers presented to the Cambridge Legal History Conference, 710
July 1975, and in Lincolns Inn Old Hall on 3 July 1974 ed. by J. H. Baker. London:
Royal Historical Society.
London: Kings College London Medieval Studies XI.
Simpson, J. A. & E. S. C. Weiner, eds. 1989. The Oxford English Dictionary (second
edition). Prepared from The Oxford English Dictionary, first edited by James A. H.
Murray, Henry Bradley, W. A. Craigie & C. T. Onions, combined with A Supplement to
the Oxford English Dictionary, ed. R. W. Burchfield, 1989. Oxford: Oxford University
Press.
Strang, Barbara M. H. 1970. A History of English. London and New York: Methuen.
Toller, T. N. 1921. An Anglo-Saxon Dictionary: Supplement. London.
The Discourse Motivations for Neologising:
Action Nominalization in
the History of English
Claire Cowie
Dictionary Unit for South African English, Rhodes University
Abstract
This study investigates the role of extralinguistic or contextual factors in the

productivity of derivational morphology. It is argued that increases and de-
creases in morphological productivity are strongly determined by the extent of
neologising, which is a social and cultural activity. The productivity of a
single English affix (t)ion is measured over time in a historical corpus of
English (ARCHER), and its performance contrasted for the different registers
of the corpus. Results show that scientific and medical registers consistently
coin the most action nominalizations in (t)ion. The texts of these registers are
then analysed in detail in order to develop a better analysis of what promotes
neologising in (t)ion. I show that motivations for coining action nominaliza-
tions can range from the pragmatic and functional (organising the presentation
of information) to the purely stylistic.
Introduction
In a recent paper on morphological productivity, Harald Baayen and Antoinette

Renouf are critical of the European structuralist approach to morphological
productivity, in which degrees of productivity are claimed to reflect the extent
to which phonological, morphological and semantic restrictions constrain the
input domain of an affix. Within the input domain defined by these restrictions,
180 CLAIRE COWIE
an affix is claimed to be absolutely productive (1996:87). The authors observe

that without additional qualifications of the restrictive weight of these phono-
logical, morphological and semantic restrictions, the claim that the degree of
productivity and the number of restrictions are inversely related is simply
vacuous (1996:87). To give a simple example, a restriction against adverbial
-ly attaching to adjectives in -ly (*sillyly) is unlikely to rule out large numbers
of complex words, given the small number of adjective bases in -ly. Baayen and
Renoufs most important criticism, however, is that such an approach fails to
take into account that word formation is conceptually driven, and that the
restrictions defining a word formation rule only set the boundary conditions for
word formation (1996:90, my emphasis). The function of word-formation,
they suggest, is to convey (particular shades of) meaning, not simply to
produce forms with a particular structure (ibid).
Despite calls for the investigation of the role of extralinguistic factors in
the productivity of derivational morphology (see for example Dalton-Puffer
1996:223), the topic remains underexplored. Understanding these extralin-
guistic factors is key to understanding change in word-formation. Observing
changes in the increases and decreases in productivity is the microscopic
observation of changes in word-formation, and increases and decreases in
productivity are affected by levels of neologising, which is fundamentally a
social and cultural practice. What Brian Joseph says with regard to morpho-
logical change is particularly true for derivational morphology: any discus-
sion of causes must make reference to the fact that, as is the case with all types
of language change, the spread of morphological innovations is subject to
social factors governing the evaluation of an innovation by speakers and its
adoption by them (1998:364). We can begin to explore the extralinguistic
motivations for neologising, I believe, through a context-sensitive, corpus-
based study of word-formation.
The extralinguistic factor targeted in this case study is register. Baayen
and Renouf add to a history of associations between word-formation and
register when they point out that the way in which words are put to use,
however, may vary substantially from genre to genre, from text type to text
type, and even from author to author, and that there is some evidence that
suggests that the productivity of affixes is similarly subject to variation as a
function of text type and style (1996:90).
In order to develop a fine-grained analysis of the role played by register in
the productivity of derivational morphology, I have focused on one English
DISCOURSE MOTIVATIONS FOR NEOLOGISING 181
affix, namely the deverbal nominalizing suffix -(t)ion, which is typically used
to denote an action or process. I examine the historical evidence of action
nominalization in -(t)ion in the ARCHER corpus1 which runs from 1650 to
1990. I also draw on the Early Modern English section of the Helsinki Corpus
(HCE) for comparison with the sixteenth and early seventeenth centuries. I
suggest that a quantitative assessment of the effect of register on word-
formation must be complemented by a qualitative discourse analysis which
examines the ways in which the discourse apparatus of individual texts
promotes the production of action nominalizations.
Motivations for Neologising
Existing commentary on the motivations for neologising in linguistics is not

based on detailed frameworks of the organisation or stratification of society
and how this affects language change. Laurie Bauer proposes that in order for
a neologism to be formed, the referent of that neologism must satisfy the
requirement of existence (and this can include mythological as well as as
real-world existence) (1983:86).2 The requirement of existence corresponds
to an apparently commonsense notion that a new word is invented when it is
necessary to supply a name for a new object or concept that has entered the
speech community, i.e. a lexical gap exists. This is predicated on a passive
correspondence notion of language which assumes an existing pre-conceptual
structure, instead of what Halliday and Martin term a constructivist ap-
proach, in which language construes human experience rather than simply
reflecting it (1993:8). They reject the view that language does not play a part
in changes in the relationship of human beings to their environment, that it
simply tags along behind, coining new words when new things appear on the
scene but otherwise remaining unaffected in its content plane (its semantics
and its grammar) (1993:10).
Sandra Thompson too claims that productive lexical processes in a
language exist for the purpose, as it were, of providing a way to express
something for which no word is present in the lexicon (1974:2). Such a view
can be found even in Suzanne Fleischmans (1977) study which is overtly
concerned with cultural and linguistic factors in word-formation. She states
that the neological mechanism is oftentimes triggered by a need or deficiency
in the resources of vocabulary (1977:1).3 The passive correspondence notion
182 CLAIRE COWIE
assumes that members of a speech community, who are homogeneous in

knowledge, world view and lexicon, achieve consensus on a new word before
it is introduced into the language. Yet new words are introduced in contexts
with differing influence and then filtered through complex channels. Those
items and concepts which are named are named because they have cultural
salience for a certain group of speakers. A neologism may create knowledge
but does so within a specific discourse. The passive correspondence notion is
embraced in Algeo (1991), but with some additions:
The need for new words is both pragmatic and esthetic. Pragmatically, when
there are new things to talk about, we need new words to name them. Or
sometimes we want to talk about old things in a new way. Changes in society,
whether material or intellectual, call for new words; and the more intense the
social change, the more need we have to name new things or rename old ones.
Thus invention, discovery, exploration, war, commerce and revolution all
breed neology. But language is not limited to the practical values of conceptu-
alization, communication, management, and cooperation. Language is also a
field for play and poetry (Algeo 1991:14).
The renaming of existing concepts and objects is a step away from the
passive correspondence notion, as is Algeos acknowledgement of the poetic
function of word-formation. Well-known treatments of the latter include work
on Shakespeares word-formation and neologising (Salmon 1987, Garner
1987). I would like to propose some categories of motivation between the
need to name an (as yet unnamed) object and the poetic/aesthetic function.
These are the pragmatic or functional role that the new derivation plays in
structuring discourse, and the use of word-formation to achieve a certain
stylistic effect. It is the very fact that word-formation rules are optional, says
Suzanne Romaine, that means they can be exploited as stylistic resources
(1983:199).
Douglas Biber distinguishes between register markers, distinctive fea-
tures found only in particular registers and common or core linguistic fea-
tures which occur with differing frequency in most registers. Differences in
the relative distribution of common linguistic features typically have func-
tional underpinnings, while the use of specialized register markers is often
conventional (1994:33). I would like to suggest that the distinction between
these two types of features is not always clearcut, and that a common linguistic
feature (such as nominalization) may lose its functionality and become associ-
ated with and a marker of a particular style or register.
Finally, let me address the implications of a distinction made by Baayen

and Renouf (1996) between motivations for neologising. The central concern
in their study (the productivity of five affixes in an 80 million word newspaper
corpus over 4 years) is the spontaneous, unintentional and ephemeral use of
productive word-formation, not with conscious and deliberate lexical creativ-
ity in which a novel expression is carefully constructed to express a new
concept intended for repeated use within an often specialized domain.4
The ephemeral, spontaneously produced new words are regarded as typical
of highly productive processes, whereas the deliberately coined words tend to
be associated with semi-productive processes. As the productivity of a pattern
decreases, the likelihood increases that speakers are aware of the fact that
they are coining a new word. They may even exploit the salience of semi-
productive neologisms for foregrounding purposes (1996: 81). Furthermore,
these two categories of neologising are characteristic of different registers.
Deliberate creations are associated with specialised or technical domains.
These dichotomies (ephemeral/deliberate, productive/unproductive, special-
ised/non-specialised) will be questioned in the following examination of
word-formation in context.
Deverbal Nominalization in -(t)ion
The connection between word-formation and register has previously been

explored with regard to deadjectival nominalizations in -ness and -ity by
Romaine (1985) and Riddle (1985). Briefly, these studies were concerned
with derived abstract nouns which entered English through religious and
philosophical discourse from Middle English onward. There has not however
been any systematic comparison of registers with regard to nominalizations.
Part of the definition of action nominals is that they should show reason-
able productivity (Comrie 1976; Koptjevskaja-Tamm 1993:5). But the cat-
egorial status of action nominals can vary greatly across languages, and there
is no obvious criterion for reasonable productivity. In some languages
action nominals are treated morphologically as regular verb forms, and in
others they constitute a group of derived nouns with a number of idiosyn-
cratic features (Koptjevskaja-Tamm 1993:6). Chomsky (1970) famously
argues that gerundive -ing nominals in English do not have to be listed in the
lexicon, but derived nominals in -(t)ion do because of their idiosyncratic
184 CLAIRE COWIE
features. Action nominalization in -(t)ion is nevertheless perceived as a highly

productive process in English (Bauer 1983:221). Even a superficial examina-
tion of ARCHER reveals that the productivity of -(t)ion is considerably higher
than that of other deverbal nominalizing affixes, such as -ment, as well as
deadjectival nominalizing suffixes such as -ness or -ity.
Dalton-Puffer chronicles an explosion of types ending in -(a)cioun in the
last section of the Middle English part of the Helsinki Corpus (13501420).
There are 138 types in this section as opposed to 20 in ME2 (12501350) and
4 in ME1 (11501250) (1996:94). -(t)ion anglicizes Latin -atio (after dropping
the Latin inflectional ending as in accusative constrictionem) as well as
(learned) French -ation, but in Modern English it is largely an independent
suffix with impersonal deverbal substantives (Marchand 1969:259, Nevalainen
forthcoming 7.4.3.1).5 It is impossible to pinpoint exactly when items in -(t)ion
were analysed as English derivations on the basis of borrowed pairs such as
justify/justification, and organize/organisation. Marchand identifies a different
rate of progress for different verbal bases (-ify, -ize, -ate or non-derived) but in
general from the sixteenth century all can have a derivational character in
English.
-(t)ion still attaches almost exclusively to Latinate bases. Nevalainen
comments that this lack of native bases makes it impossible to tell whether a
given form in -(t)ion is the result of borrowing or deverbal derivation in Early
Modern English (forthcoming, 7.5.3.1.4). Given that the starting point of this
study is Early Modern English, no attempt is made to disregard items that
might be loanwords rather than English derivations. This approach allows for
the inclusion of Latinate coinings (words which are not derivations in Latin
and do not have independent English bases), for example fecundation, which
appears in ARCHER and according to the OED is a latinate coining from the
Latin verb fecundare.
Measuring Productivity in a Historical Corpus
Synchronic accounts of the productivity of derivational morphology typically

define productivity as an index of the number of new words a word-formation
process is able to form at a particular time.6 Diachronic productivity, as
defined in Aronoff (1980), is the number of new words a process can form
from T1 to T2. This has been measured over the centuries using first citation
dates from the OED (Anshen and Aronoff 1989, Aronoff and Anshen 1998).
The productivity of an affix can be measured in a historical corpus by
comparing the type frequency (number of word types in an affix) across
different subperiods of the corpus, as Dalton-Puffer (1996) does with Middle
English suffixes for three subperiods of the Middle English part of the
Helsinki corpus.
In addition, the present study adapts methodology from Baayen and
Renouf (1996), which involves counting only the number of new types in an
affix as the sample increases. For a historical corpus, new types are obtained
by counting only the types which do not occur in previous subperiods. This
measure can however be problematic for historical corpora in that they tend to
be far smaller in size than contemporary corpora. A relatively small section
therefore of the lexicon may be sampled in the earlier periods, with new types
being recorded at a much later date than they were originally used. Therefore
with this measure I have employed a starting lexicon, namely the types from
15001640 in HCE, in order to obtain a fair sample of words occurring before
1650. My main concern is with ARCHER, but the table given for HCE
provides some background about the sixteenth and early seventeenth centu-
ries. Tables 1 and 2 show the token and type frequencies of -(t)ion per 100,000
words7 in HCE and ARCHER. Table 2 is graphically illustrated in Figure 1.
Table 1. Types and tokens in -(t)ion per 100,000 words in HCE

Period 15001570 15701640 16401710
Types 121.5 144.4 179.5
Tokens 590.6 615.9 817.9
Table 2. Types and tokens in -(t)ion per 100,000 words in ARCHER

Period 16501700 17001750 17501800 18001850
Types 189.4 202.3 153.2 220.9
Tokens 703.3 1002.5 998.2 1326.3
Period 18501900 19001950 19501990

Types 162.6 244.5 170.3
Tokens 1018.4 979.5 824.8
186 CLAIRE COWIE
Type frequency per 100,000 words for -(t)ion in ARCHER
types
250
200
150
types
100
50
0
1600- 1700- 1750- 1800- 1850- 1900- 1950-
1650 1750 1800 1850 1900 1950
time
Figure 1. Types in -(t)ion per 100,000 words in ARCHER
For the type frequency measure, we find the consistent aggregation

(increase resulting from existing types plus new types for each subperiod) that
might be expected from a productive process from 1500 until 1750 (viewing
the two corpora continuously). The type frequency per 100,000 words for
ARCHER 16501700 in fact corresponds fairly closely to the type frequency
per 100,000 words for HCE 16401710. After 1700, however, the subperiods
of ARCHER increase only alternately.8 Turning to the new types data, Table
3 shows the percentage of new types out of the total number of types for the
ARCHER subperiods, using the HCE starting lexicon. The percentages are
graphically represented in Figure 2, and the proportions yielding these per-
centages are graphically represented in Figure 3.
Table 3. Percentage of new types out of the total number of types in ARCHER with HCE
starting lexicon
Period 16501700 17001750 17501800 18001850
-(t)ion 42.4 26.7 31.3 19.4
Period 18501900 19001950 19501990

-(t)ion 19.4 15.8 23.1
Percentage of new types in -(t)ion for ARCHER
percentage
100
90
80
% new types
70
60
50
40
30
20
10
0
1650- 1700- 1750- 1800- 1850- 1900- 1950-
1700 1750 1800 1850 1900 1950
time
Figure 2. Percentage of new types out of the total number of types in ARCHER subperiods
with HCE starting lexicon
Proportion of new types for -(t)ion in ARCHER
old types new types
700
600
500
400
types
300
200
100
0
1650- 1700- 1750- 1800- 1850- 1900- 1950-
1700 1750 1800 1850 1900 1950
time
Figure 3. Proportion of new types out of total types for ARCHER subperiods with HCE
starting lexicon*
*A chi square test shows that these proportions are not evenly distributed across the periods,
for a significance level of p < 0.05.
188 CLAIRE COWIE
The shape of the graph in Figure 2 shows that the most prominent
difference (15.7%) is between the first two periods of ARCHER. This may be
due to heightened productivity of -(t)ion in the seventeenth century, but there
is a possibility that the measure is still vulnerable to the decreasing effect for
small corpora, despite the starting lexicon. In Figure 2 the percentage of new
types for 17501800 is higher than that for 17001750, disrupting what might
otherwise be a fairly consistent decrease.9 The percentage differences become
smaller and the two halves of the nineteenth century are identical.
Register
The numbers of new types for each register in a subperiod are obtained
through a breakdown of the total new types for that subperiod.10 If these
figures are then normalized (the sizes for each cell of a register are not the
same11 ), then registers can be ranked in order of the most contributions to the
new types for a subperiod, as in Table 4, below:
Table 4. Ranking of registers in ARCHER according to number of new types in -(t)ion

Period 1 2 3 4 5 6 7 8 9
1650 Science Medical News Sermon Letters Journal Drama Fiction -
1700 Medical Science Sermon Letters Fiction Journal Drama News -
1750 Medical Science Letters Legal Fiction News Sermon Drama Journal
1800 Medical Science News Letters Legal Fiction Sermon Journal Drama
1850 Medical Science Letters Sermon News Journal Fiction Legal Drama
1900 Medical Science News Legal Sermon Fiction Letters Drama Journal
1950 Medical Science News Legal Sermon Letters Fiction Drama Journal
From the ranking in Table 4, it is evident that medicine and science consis-
tently produce more new types in -(t)ion than the other registers. At the other
end of the scale, drama and journals tend to produce the least. The remaining
registers variably occupy ranks 3 to 7, making generalisation difficult. Letters,
legal prose and news seem to be higher on the whole than sermons and fiction.
No systematic changes in the ranking of registers over time are observable in
Table 4. It is clear that in terms of action nominalization, medical and
scientific writing were from the seventeenth century already considerably
more prone to nominalizing than other registers.
This grouping loosely corresponds to multidimensional analyses of AR-

CHER, which have shown that from the seventeenth to the twentieth century
the expository, professional registers (science and medical writing, legal
prose and news reportage) have become more informational, whereas more
popular written and speech-based registers, such as drama, letters and jour-
nals, have become more involved (Biber, Finegan and and Atkinson 1994;
Biber and Finegan 1997).12 The expository registers, which have become
increasingly specialized, have also consistently followed a tendency towards a
more literate style. According to Biber and Finegan they have come to
exploit the resources of the written mode in innovative ways, resulting in a
style of discourse not previously attested (1997:273).
There is a danger of assuming that the texts of a register for a subperiod,
especially in the earlier periods, will be homogeneous in terms of style. There
are some facts that can be obtained at the outset about the homogeneity of texts
in the ARCHER registers by looking at the composition of the corpus. For
example, we might speculate that scientific writing is likely to be a more
homogeneous register because the texts, for all periods, are taken from the
same journal. Yet this journal, the Philosophical Transactions of the Royal
Society, has become more homogeneous over time, and the early texts are
diverse. The seventeenth century volumes feature a very wide range of topics,
from records of the weather to descriptions of naturalists walks (see Atkinson
1996). I will therefore be examining these nominalizing registers and their
internal variation closely, through a discourse analysis of nominalizations in
their contexts of occurrence. Corpus linguistics has tended to omit such
textual analysis. The most powerful interpretation emerges, argues Michael
Stubbs, if comparisons of texts across corpora are combined with the analysis
of the organization of individual texts (1996:5). In the next section I explore
models for the analysis of nominalizations in discourse.
Models of Derivation in Discourse
Michael Halliday has used the term grammatical metaphor in relation to

word-formation: Instead of a lexical transformation (of one word to another)
the transformation is in the grammar from one class to another, with the word
remaining the same (Halliday and Martin 1993:13). For example, the deriva-
tion of happiness from happy involves a transition from a quality (adjective)
190 CLAIRE COWIE
to an abstract thing (nominalization) (Stubbs 1996:8687). Nominalization is

probably the most prominent exemplar of grammatical metaphor in Halliday
and Martins textual analyses of scientific writing. They use the following
passage from Newtons Opticks to illustrate the practice of nominalization in
early scientific texts:
If the Humours of the Eye by old Age decay, so as by shrinking to make the
Cornea and Coat of the Crystalline Humour grow flatter than before, the light
will not be refracted enough, and for want of a sufficient Refraction will not
converge to the bottom, of the Eye but to some place beyond it, and by
consequence paint in the bottom of the Eye a confused Picture, and according
to the Indistinctness of this Picture the Object will appear confused. This is
the reason of the decay in the sight of old Men. And shews why their Sight is
mended by Spectacles. For those Convex glasses supply the defect of Plump-
ness in the Eye, and by increasing the Refraction make the Rays converge
sooner, so as to convene distinctly at the bottom of the Eye if the glass have a
due degree of convexity. And the contrary happens in short-sighted Men
whose Eyes are too plump. (Newton, Treatise on Opticks 1704)
In the following pairs, some verb or adjective in the first expression

(underlined above) has been reworded in the second as a noun (in bold):
will not be refracted enough for want of a sufficient Refraction
paint a confused picture according to the Indistinctness of this Picture
make the Cornea grow flatter supply the defect of Plumpness in the Eye
those Convex glasses if the glass have a due degree of convexity
(Halliday and Martin 1993:7).
These nominalizations enable a chunk of discourse that was previously

presented as new information to be re-used as a given in the course of the
succeeding argument.13 The authors remark: Creating a technical term is in
itself a grammatical process; and when the argument is constructed by gram-
mar in this way, the words that are turned into nouns tend thereby to become
technicalized (1993:78). Nominalization is used to achieve an important
discourse effect, namely the packaging of a complex phenomenon into a
single semiotic entity, by making it one element of a clause structure, so that
its rhetorical function its place in the unfolding argument is rendered
fully explicit (Halliday 1993:60). This rhetorical function includes the pre-
sentation of previously given information, in which the nominalization occu-
pies the first position in the clause, and the presentation of new information, or
foregrounding, in which the nominalization appears in the unmarked position
at the end of the clause.
Thus the device of nominalizing, Halliday claims, far from being an

arbitrary or ritualistic feature, is an essential resource for constructing scien-
tific discourse. We see it emerging in the language of this period, when the
foundations of an effective register for codifying, transmitting and extending
the new learning are rapidly being laid down (1993:61). If it is the case that
action nominalization in -(t)ion is at its height in the seventeenth century, this
may go some way toward an explanation. I will use an excerpt from a
scientific text (Huygens and Papin) from the 16501700 period of ARCHER,
to further illustrate the processes described by Halliday.
This Experiment drew another after it, to know, whether the 1
water purged of Air were less fit than common water to make plants
vegetate. For this end I took two Vials full, the one of water purged,
the other of common water, and having put a twigg of
Baulme in each, I left them both in the Air. I found, that the twigg in 5
the common watrer [sic] shot at the end of six daies, and in water
purged shot this time neither but ten daies after it had been put
in.
I repeated this Experiment once more, and I was much surprized
to see, that the twigg in the water freed of Air begun this time 10
to shoot the third day, and the other in the common water, still
the sixth day
Although this Experiment appeared at first contrary to the
precedent, yet it still confirmed the first thought, to wit, that
the Air which is mixed in common water serves for vegetation, 15
considering the little root which the twigg shot in the water
cleansed of Air. Meantime I do not believe, it will be easie to
know the particular reason, which made the first root shoot so
soon.
1675 Hugyens, Cristiaan and M. Papin. Some Experiments made in
the Air Pump upon Plants, Philosophical Transactions 10: 120.
The items I have highlighted are the verb vegetate in line 3 and its
nominalization vegetation in line 15. Vegetation here is a transparent deriva-
tion and appears with the original meaning of growth, as an action or
process, and not with the modern lexicalized meaning of plant matter.14 The
verb vegetate is set out in the beginning in the experimental question. The
experimental question is followed by the narrative of the experiment, and
then the nominalization occurs in the summing up and discussion of the
experiments results (notably without its argument plants) referring to a
192 CLAIRE COWIE
previously described process.

Accounts of the role of derivation in discourse are rare, given the way that
derivation is traditionally treated acontextually in morphological theory.15
Kastovsky and Kryk-Kastovsky (1997) explore word-formation as a subtype
of lexical cohesion in discourse, as it involves the partial recurrence of lexical
material.16 They find that nominal forms are more likely to function as
referring items, and verbs and adjectives are more likely to function as target
items (1997:468). Below is an example of a deverbal nominalization acting
cohesively:
Once they hear Settlers are involved, that will only make it worse. Our
involvement! Tonya Walton exclaimed. We had nothing to do with the
attack (Isaac Asimov and Roger MacBride Allan 1993, Caliban, in Kastovsky
and Kryk-Kastovsky 1997:464).
Although word-formation is a relatively frequently used cohesive device,

Kastovsky and Kryk-Kastovsky note that no large-scale systematic investi-
gation of this function of word-formation has been attempted either for a
single language or, what would be even more interesting, contrastively
(1997: 463). They also call for research on the quantitative distribution of
these processes, the relationship between the quantitative distribution and text
type, and the nature of the relationship of the referring item and target
item in terms of referentiality, directionality and syntactic position (1997:
466).
Koptjevskaja-Tamm suggests that action nominals tend to be chosen
from a range of competing expressions for their text-compressing function
(1993:266). Because this function is dependent on communication needs,
nominalizations are unevenly distributed among different styles of speech.
More specifically, the complexity of scientific discourse, reflecting the com-
plexity of interrelated factors in scientific thought, favours nominalizations
and other means of maximizing the amount of information in texts of compa-
rable length (Koptjevskaja-Tamm 1993:266). Another important observation
made by Koptjevskaja-Tamm regarding the function of action nominals is the
possibility of argument reduction that they represent. Action nominals allow
subjects and objects to be deleted in situations where these are generic or
indefinite (1993:270).
Discourse Analysis
In this section I compare two ARCHER scientific texts from the seventeenth
century. I also examine a nineteenth century medical text. The first text,
Anonymous, has a high type frequency compared to other scientific texts in
this subperiod: 18 nominalizations in (t)ion per 1000 words (22 types in 992
words). Hugyens and Papin, from the same subperiod, has only 2 types (one of
which was vegetation) in 1866 words, a type frequency of 1.1 per 1000 words.
The type frequency of an individual text alone however is not a reliable
indicator of productivity.
Vitriol is by the Spagyrical Tribe reputed one of the chief Pillars 1
of Medicine and Alchimy; and is indeed endowed with many excellent
and truly admirable properties; being employed by nature in her
most Curious Mineral operations. Tis it self one of the most noble
and useful productions, and therefore deserves our especial 5
Consideration. I do not pretend to render a Mechanical account of
its generation, or a history of all that may be performed by its
mediation in Medicine or Chymistry; my design is only to furnish
Inquisitive persons with some Observations and Experiments, which
will probably enable them more easily to investigate the Nature of 10
this Protean substance It is often associated with Earth
and Stone, wherein Mettals are contained; and with many natural
recrements of metals, such as Misy, Sory, Chalcitis; from which
tis usually separable by the common method with Water, sometimes
not to be extricated until the Mineral be first calcined or burnt 15
Vitriol consists of Insipid phlegme, Earth or Oker, some
Mettal, Mineral Sulphur, an acid Salt or Spirit, together with some
small portion of the Volatil Aerial Salt.
That it contains Water, needs no great proof, since no Saline
substance can crystallize without it; and distillation will 20
convince any person, that it exceeds in quantity any of the other
Principles.
The Earth or Oker may be thus separated: Dissolve Vitriol in
fair water, immediately a yellow powder will separate, and in a
short time subside: The greater the quantity of water imployed, the 25
more Oker precipitates: The weaker the lixivium, the less able to
support Bodies more ponderous than common water: And the lighter
the Water (as if distilled rain-water, or phlegme of vinous
Spirits,) the more Earthy parts subside, upon the same
Hydrostatical principle I just now mentioned. I have above twenty 30
times repeated this dissolution, seconded by filtration and
194 CLAIRE COWIE
coagulation, and each time separated some quantity of this Earth;

and am Perswaded, had I long continued the operation, the success
would have been the same; only I observed the quantity separated
each time sensibly to diminish: And Basilius Valentinus assures, 35
that at length the Vitriol will let fall no more sediment; and that
then it is the subject of most noble operations by him
particularized; which they who have leisure, and confidence in his
specious promises, may do well to try. I have found a more easie
and expedite way of effecting this separation, which may be of 40
great use to them who work on Vitriol, much abbreviate their
labour, and considerably lessen their expence.
Take a good quantity of the common, Dantzick, or Hungarian,
Vitriol; having powdred it, put it into a slender Cucurbite, place
it in Water, keep under it an equal constant fire three or four 45
days This repeated once or twice, the Vitriol attains unto
a high degree of purity, and is easily capable of many alterations,
whereunto it was not subject before this purification. This
operation will not succeed in a dry digestion: I mean, Ashes,
Sand, Filings of Iron, Steel, open Fire, or even flame of Lamps, 50
whether fed with Oil or Spirit of Wine. This Earth may also
be obtained in a great proportion, though in another form,
if after a long and intense Calcination the Vitriol is freed
from its remaining Salt by frequent ablutions with warm
Water Besides, whereas Salt, Nitre, &c. require in distillation 55
a larg quantity of Earthy substance to disjoyn the Saline parts, and
prevent fusion; Vitriol and Allom need it not; and unquestionable
proof, that Earthy parts abound therein.
1675 Anonymous. Some Observations and Experiments about Vitriol.
Philosophical Transactions 9: 103.
A predilection for action nominalizations emerges from the first para-

graph, with its sequence of operations, productions, consideration, mediation
and generation (lines 48). None of these have a base form in the text, and all
occur in the last position in the clause, yet they do not introduce new informa-
tion. Their role in the introductory paragraph is to sum up that which is already
known about the substance vitriol, and why is it important.
In lines 1415 two methods of obtaining vitriol from a mineral are
identified: separating the vitriol using water, or burning (calcining) the min-
eral. The separation process is described from line 23. The stages of the
separation (dissolution, filtration and coagulation) are listed in lines 3132.
All arguments of the base verbs in those nominalizations are suppressed. The
formulation allows for the economic restatement of a lengthy process with

several stages. The listing of action nominalizations in this way serves (on
most occasions) a text-compressing function. Of these three, only dissolution
has a base in the passage (dissolve).
The separation process is referred to as the operation in line 33 and the
nominalization separation is actually used for the first time only in line 40,
after several participial uses of the verb. The alternative process, in which the
mineral is burnt or calcined, is presented as calcination in line 53. Poten-
tially the only nominalization in the passage which introduces new informa-
tion is digestion in line 49, occurring in the last position in the clause. It is
followed by a colon, and a list of items that might qualify as a dry digestion.
However, the meaning of digestion here is concrete, and once an item has lost
its transparency its cohesive powers are lost. Finally let me point out the
interesting relationship between fusion in line 57, and the verb disjoyn (line
56). It is necessary to disjoin the saline parts, an action which is prolonged by
preventing fusion. The pair show the same antonymic relation between the
bases as Newtons nominalization of grow flatter as defect of plumpness.
The next text, A.I., is more restrained in its use of action nominaliza-
tion. There are 8 types in 759 words, a type frequency of 10.5, which, although
not low compared to, for instance, Huygens and Papin, is low compared to
Anonymous 18.
Reflecting on that Question, Whether Liquids gravitate upon Bodies 1
immersed or not? I came to a Resolution in my own thoughts, that
they do gravitate; and one of the greatest instances that did
occur to me was, that a bubble of Air, rising from the bottom,
does dilate it self all the way to the top; which is caused by the 5
lessening of the weight or pressure of the incumbent water, the
nearer it is to the top. Upon consideration of that instance, the
following conjecture presented it self to my thoughts; That fishes
by reason of the bladder of Air that is within them can sustein or
keep themselves in any depth of water. For the Air in that bladder 10
is like the bubble, more or less compressed, according to the depth
the fish swims at, and takes up more or less space; and
consequently the body of the fish, part of whose bulk this bladder
is, is greater or less according to the several depths, and yet
retains the same weight. The Rule <de infidentibus humido>, is, 15
that a Body that is heavier than so much water as is equal in
quantity to the bulk of it, will sink; a Body that is light, will
swim; a Body of equal weight, will rest in any part of the water.
196 CLAIRE COWIE
Now by this Rule, if the fish in the middle Region of the

water be of equal weight to the water that is commensurate to the 20
bulk of it, the fish will rest here without any tendency upwards
or downwards: And if the fish be deeper in the water, the bulk of
the fish becoming less by the compression of the bladder, and yet
retaining the same weight, it will sink and rest at the bottom: And
on the other side, if the fish be higher than that middle Region, 25
the Air dilating its self, and the bulk of the fish consequently
increasing, but not the weight, the fish will rise upwards, and
rest at the top of the water ?
So far this Conjecture: In reference to which, when it was
propounded to the Honourable Robert Boyle, he, reflecting upon the 30
manner how a Fish comes to rise or sink in water, soon bethought
himself of an Experiment probably to determine, Whether a Fish
makes those motions by constricting or expanding himself? The
Experiment by him suggested was; To take a Bolthead with a wide
neck, and having filld it almost full with water, to put into it 35
some live fish of a convenient size, that is, the biggest that can
be got in, as a Roch, Perch, or the like; and then to draw out the
neck of the Bolthead as slender as you can; and to fill that also
almost with water: Whereupon the fish lying at a certain depth in
the water of the Glass, if upon his sinking you perceive the water 40
at the slender top does subside, you may inferr, he contracts
himself, and if, upon his rising, the water be also raised, you may
conclude, he dilates himself.
1675 A.I. A Conjecture concerning the Bladders of Air that are found in Fishes.
Philosophical Transactions 10: 114.
Compression is the only nominalization for which there is a base in the

text (line 12). The nominalization (line 23) appears in the last position in the
clause, allowing for an economic restatement of the process as the agent of
another process. Note that the nominalization retains one of the arguments
(the patient) of its base verb: the bladder. There are other verbs in this passage
that are likely candidates for nominalization, yet they are not nominalized,
despite appearing in the passage several times in various guises, for example
dilate, which appears as a non-finite verb (line 5), a present participle (line 26)
and finite verb (line 42). The same is true of gravitate, constrict, and expand.
The last two verbs appear as gerunds in line 33. This choice of the gerundive
over action nominalization allows for the complementation of the reflexive
pronoun himself.
There are few passive verbs in this passage, and what further differenti-
ates it from the previous passage, among other things, is the causal co-
ordination and personal pronouns (a feature of involved production). This
comparison of Anonymous and A.I. gives us some sense of how texts which
exploit nominalization as a cohesive device can differ, and also some idea of
the nature of the texts which are likely to exploit it more. My next text,
Sinclair is from the medical register of the 18501900 period of ARCHER.
Case IV. <Accidental Haemorrhage; Death> On May 17th, 1
1866, Mrs. , aet. 40, eight months advanced in her ninth
pregnancy, while about her domestic duties, was seized with a
moderate flow of blood from the vagina, preceded by a sensation
as if something had snapped within her. The haemorrhage soon 5
increased, and she sent for her physician, who found her much
exhausted, with a cold skin, clammy perspiration, and feeble,
rapid pulse. He administered restoratives, and after a while she
rallied and became tolerably comfortable. He ruptured the
membranes and gave ergot, but failed to check the flow entirely. 10
The case was fast becoming so critical that there remained but
one means to give the slightest chance of life to the mother, and
that was to deliver by turning, the cervix being too undeveloped
and rigid to use forceps. At this stage, I was asked to assist.
I found her with a pulse of 120, small and feeble, great physical 15
depression, and blood oozing from the vagina. The foetal head
presented. I concurred with the attending physician that death
threatened, and that it might be advisable not to allow her to
die undelivered. After etherization, the hand was passed into
the uterus with considerable resistance. Version was easily 20
accomplished, but the extraction of the foetus was effected with
unusual difficulty. An enormous quantity of clots was removed
with the placenta from the cavity of the uterus. Uterine
contractions did not come on for some time after delivery,
although the uterus was stimulated by the presence of one hand 25
internally and the other manipulating the abdomen. Ergot was
also repeated. She came out of the ether quietly, and took
stimulants freely; but she was extremely exhausted, and died in
about thirty minutes after delivery.
1868. Sinclair, A.D. Unusual obstetric cases.

Boston Medical and Surgical Journal, 78: 3338.
Etherization in line 19 occurs in the first position of the clause and

appears without any of the arguments of the base verb. It is impossible to
198 CLAIRE COWIE
determine whether these involve the physician as active subject, or the patient
as passive subject.17 We have seen several nominalizations of which the
arguments have been suppressed, but up to this point I have suggested that this
is primarily for reasons of economy. In scientific and medical discourse
however there may be further motives of impersonality in the suppression of
the scientist as agent.18
The nominalizations in this passage occur at an interesting moment. Up to
line 14, the passage is a narrative of events that take place before the speaker
arrives on the scene. Lines 1419 are also narrated events, which take place
after the arrival of the speaker. Verbs in this section are mostly active, past
tense, until line 19. Then, from line 19, there is a sequence of action nominal-
izations, beginning with etherization, followed by version and extraction, in
two subsequent clauses (both are in the first position of the main clause). The
only argument is the foetus (extraction). These nominalizations militate
against the use of agents or personal pronouns. Whereas the passive voice has
been in limited use up to this point, we now get the uterus was stimulated
and ergot was also repeated. In line 27 the personal pronouns return as well
as the active voice. The nominalizations have a distancing effect, along with
other devices, such as the articles one and the which qualify hand. It may be
that the discourse shifts in this way in order for the physician to distance
himself from what others experience as a human trauma, but as Don Chapman
points out,19 this may simply be an attempt to make the discourse generic.
This passage demonstrates the ways in which a medical text can switch
between typical narrative discourse and more specialised patterns.
Finally, let me point out the rephrasing of turning (line 13) as version,
which, according to the OED, has an obstetric meaning from 1853. The
variation with turning in this passage suggests that the term version may not
have been universally adopted. Possibly the other nominalizations in -(t)ion
provoked the use of the Latinate term.
-(t)ion as a Marker of Style
One level at which deverbal nominalization in -(t)ion can be said to have

stylistic and social associations is its membership of the non-native stratum of
English word-formation. In the Early Modern English period, a rapid increase
in Latinate vocabulary resulted in the social stratification of the native and the
non-native levels of the vocabulary and word-formation resources (Grlach

1991:136, Nevalainen forthcoming, 7.1.1., 7.4.3.1.). One consequence of the
flood of new Latin words was that a language barrier was erected within
English in the sixteenth century and that the proper use of the Latinate
portion of English came to replace knowledge of the classical languages alone
as the marker of social class and education (Grlach 1991:162).
While Latin became unfashionable in general use in the Restoration
period, new Latin loans and neo-classical formations became increasingly
associated with technical registers in the eighteenth century (Nevalainen
forthcoming 4.3.1.). Thus the stratification of the vocabulary increasingly has
less to do with social status and more to do with a learned/non-learned
distinction. On the one hand we appear to have the creation of knowledge and
the structuring of scientific discourse through functional word-formation, and
on the other, the use of a word-formation process in scientific discourse to
effect a learned style. It is not entirely possible, however, to tell when the
former has slipped into the latter. In the seventeenth century, the learned
associations developed by -(t)ion are variously alluded to, for instance by
satirists such as Shadwell, whose 1676 play, The Virtuoso, contains the
following lines, uttered by Sir Nicholas Gimcrack, representing the scientist
Robert Hooke:
It comes first to fluidity, then to orbiculation, then fixation, so to angulization,
then crystallization, from thence to germination or ebullition, then vegetation,
then plantamination, perfect animation, sensation, local motion, and the like
(Shadwell, The Virtuoso, 1676).
This passage is assumed to be a parody of a description in Hookes

Micrographia (1665) of how a plum turns blue. But the original version of the
Observation of Blue Mold hardly differs, and we find the same hyperbolic
listing of action nominalizations:
Nor do I imagine that the skips from one to another will be found very great,
if beginning from fluidity, or body without any form, we descend gradually
till we arrive at the highest form of a bruite Animals Soul, making the steps
or foundations of our Enquiry Fluidity, Orbiculation, Fixation, Angulization
or Crystallization, Germination or Ebullition, Vegetation, Plantamination,
Animation, Sensation, Imagination (Hooke, Micrographia, Observation of
Blue Mold, 1665).
Hooke seems to be describing various stages in a natural order, which

suggests that these processes, really states here, have more than just a practical
200 CLAIRE COWIE
descriptive significance. This is a good illustration of the use of word-forma-

tion for the kind of semiotic reification described by Halliday. Marchand
offers further evidence of ridiculing criticism of the growing use of the
learned suffix in the form of this seventeenth century quote which appears
in the OED entry for -(t)ion: But what languages do they speak, servant?
Several languages, as Cawation, Chirpation, Hootation, Whistleation,
Crowation, Cackleation, Shriekation, Hissation.20 This kind of evidence
suggests that -(t)ion was a conspicuous marker of learned register by the
seventeenth century. Today we may recognise the use of action nominaliza-
tion as a marker of style perhaps in bureaucratic or official rather than
scientific discourse:
A newly evolving register is always functional in its context the language
may become ritualized, but it cannot start that way, because to become
ritualized, a feature must first acquire value, and it can acquire value only by
being functional. Thus despite the extent to which scientific English comes to
be ritualized, and carried over as a language of prestige and power into other
contexts where its special features make no sense except as ritual (for example
in bureaucratic discourse), all the characteristics we observed . . . are in origin
functional in the effective construction of reality (Halliday 1993:68).
It is the same development, probably, that leads Nigel Spivey to proffer the
following advice:
New words are bandied about by all in sundry, as soon as they catch on, so it
is only necessary to keep your ears open for this elementary stuff. Expert
consultants do more than this: they invent their own jargon. They use words
in a way which nobody has heard before, thus underlining once again
separateness from their clients The technique for inventing jargon is
simple enough. The trick is to take a noun in common use, and turn it into a
verb; or, to reverse the procedure. Loathsome, but impressive-sounding
words emerge in this barbarous manner. Profit can spawn profitalise; mer-
chandising can develop merchandisation (Spivey 1986:34).
In a corpus survey of register some forms will appear that are plainly
playing on these learned associations but will occur in different registers from
science and medicine. For instance, Shadwells satirical nominalizations
would have come up in drama if the text was in the corpus. In ARCHER we
find botheration in a letter from William Blake to his brother in 1803: I write
in great haste & with a head full of botheration about various projected works
& particularly a work now Proposed to the Public at the End of Cowpers Life,
which will very likely be of great consequence. Botheration is described by
Marchand as a mock-learned formation (1969:261). The word appears in

the OED from 1797, defined as an act of bothering or petty vexation or
annoyance.
Conclusion
The analysis of nominalizations in their surrounding discourse is the practical

realisation of the theoretical claims made earlier: neologisms are created in
context and are not agreed upon in an abstract speech community. We should
also consider what implications the above analyses have for the kind of
distinctions we have seen made between the motivations for neologising, such
as Baayen and Renoufs distinction between the deliberate coining of a new
word (typically unproductive processes and typically in specialist registers),
and a less intentional coining of new words akin to syntactic productivity.
Kastovksy and Kryk-Kastovsky place word-formation processes on a func-
tional scale, one pole of which is purely lexical. The output, lexical items,
provide a designation, a name for a nameworthy segment of extralinguistic
reality (1997:465). The quasi-pronominal function shown by word-forma-
tion processes in cohesion can be placed on the other, syntactic, pole.
Action nominalization in -(t)ion is a highly productive process. Yet it
tends to be used more in specialist and technical registers. Certainly it is used
intentionally to provide designations. But it is questionable in the light of the
above analyses whether the cohesive use can really be regarded as uninten-
tional or ephemeral. Furthermore, should the use of the suffix as a marker of
style be considered deliberate? It is clear that a more elaborate framework of
motivations for neologising is called for.
Notes
1. ARCHER (A Representative Corpus of Historical English Registers) was compiled by

Douglas Biber and Edward Finegan and I am grateful for their permission to use it at
Northern Arizona University.
2. In his discussion of the role of pragmatics in word-formation Bauer (1983) also posits a
nameability requirement, which holds that the appropriate derivational morphology be
available to express together with the base the given concept or item. This second
condition is much more interesting than the first, and raises questions about what
202 CLAIRE COWIE
derivational morphology can be found across languages. In this paper, however, I deal
only with the first requirement.
3. Fleischman qualifies her remarks, however, by observing that this situation prevails
whenever the significatory means which a language has at its disposal at a given moment
fall short of the sum total of influences and impulses which require the lexicon to be
augmented or certain of its constituent elements to be redeployed (1977:1), which,
although vague, allows for a fairly broad range of motivations for neologising.
4. Baayen and Renouf therefore rely primarily on hapax legomena in their measurements of
productivity: The hapax legomena in our corpus tend to be the prototypical instantia-
tions of ephemeral word-formation that is most similar to the productivity of syntactic
constructions Although a neologism can be created to fill a lexical gap (as when a new
technical term is introduced to describe a novel concept, in which case it is likely to be
used more than once), this appears to be a rare phenomenon in our corpus (1996: 7879).
5. Kastovsky argues that this word-formation rule must be represented in synchronic
grammars as -ation, as it is necessary to distinguish between nouns which presuppose a
Latin or French original such as construction, action, revolution and conversion, and
nouns produced by the English word-formation rule such as permutation, qualification
and specification (1986:589, 1992:291). Two examples from ARCHER suggest that -
ification and -ization may have formed independent units. Rustification, which is not in
the OED, occurs instead of rustication in the Manchester Guardian (1959). Privatization,
which appears in The Times (1989), is first cited in 1959 in the OED. The verb privatize,
however, is first cited in 1969.
6. This is to drastically oversimplify. Synchronic productivity is either a psycholinguistic
measure of potential i.e. the number of words a speaker can form from a list of given
bases (see Bauer 1983; Anshen and Aronoff 1989) or it is a measure based on existing
words, whether in a dictionary (Aronoff 1976, Anshen and Aronoff 1989), or a corpus
(Baayen 1992, 1993 and Baayen and Lieber 1991). Productivity measured using a corpus
has recently been referred to as quantitative productivity (Aronoff and Anshen 1998).
7. Normalization of the results is necessary as the subperiods of both HCE and ARCHER
are not the same size.
8. A potential cause of this halfcentury effect may be that the periods which consist of
British and American texts (the last half-centuries of ARCHER: 17501800, 18501900
and 19501990) could be less productive per 100,000 words than the alternate periods
which have British texts only. I have checked this by obtaining separate measurements of
the British and American texts in these subperiods. Biber records suggestions of a greater
tendency to use nominalizations in American English (1987:100). For action nominaliza-
tions in ARCHER this is true of the twentieth century, but for the last half of the
eighteenth century, British texts have the higher type frequency.
9. The result for 17501800 could be attributed either to regional differences (the combined
texts being more productive this time) or to the probability of a greater sample size
picking up more rare types. The same explanation might be offered for the 19501990
result, which differs from 19001950 by 7.3%. Yet 18501900 does not show this effect
at all, and it happens to be the largest sample by approximately 50,000 words. There is
also no statistical correlation between percentage of new types and sample size.
10. To treat each register as an individual corpus and obtain new types for that register only
would result in a more exaggerated case of the problems associated with small corpus
size.
11. Unfortunately the normalisation means that it is not possible to compare the performance
of an individual register over time, but only the relationship between registers over time.
See Appendix for raw figures.
12. The description of prose as informational is related to the earlier nominal style,
which, for Rulon Wells, supposedly set off writing as esoteric, specialized, technical.
Wells claims that nominal style in English can be used to play the role (although much
less conspicuously and effectively) that Latin played several hundred years ago (1970:
303).
13. This applies to deadjectival nominalizations and -ing gerunds as well as deverbal nomi-
nalizations in -(t)ion. I have found in my doctoral thesis on derived nominalization in the
history of English (Cowie 1999) that deverbal and deadjectival nominalizations share
some discourse functions, but that there are also subtle differences.
14. The first citation of vegetation in the OED is 1564. The first concrete sense is dated as
1683.
15. Although see Hopper and Thompson (1984) on how nominalizations, by their nature, are
discourse dependent and backgrounded; they also suggest that nominalization enables
events to be treated as concrete rather than abstract, as human cognition is better able to
deal with concrete entities. Merlan (1976) and Mithun (1984) discuss noun incorporation
in discourse. Fincke (1997) examines the discourse motivations for verbalization in
Bikol. Productive verbalization can present new and contrastive information because the
stem of the verb is morphosyntactically rich (i.e. a non-verb) and not predictable from the
other sentence consituents, for example: dai mo pig-ki-Kimbies You dont (put her in)
Kimbies (nappies/diapers)? Lit. You dont Kimbies (her)?
16. This function is also mentioned in Kastovsky 1982, 1986.
17. The OED identifies etherization as an English derivation (from 1851) meaning the
administration of ether as an anaesthetic or narcotic or the process of becoming, or
condition of being, etherized.
18. The suppression of agency through nominalization in scientific discourse can be moti-
vated by the same factors as the suppression of agency through passive constructions (see
Atkinson 1996).
19. Personal communication.
20. Randolph, Amyntas (1638, OED).
References
Algeo, J. ed. 1991. Fifty years among the new words: a dictionary of neologisms 1942
1991. Cambridge: Cambridge University Press.
204 CLAIRE COWIE
Anshen, F. and M. Aronoff, 1989. Morphological Productivity, Word Frequency and the
OED. Language Change and Variation. ed. by R. Fasold & D. Schriffin. Amsterdam/
Philadelphia: John Benjamins.
Aronoff, M. 1976. Word Formation and Generative Grammar. Massachusetts: MIT Press.
Aronoff, M. 1980. The relevance of productivity in a synchronic theory of word forma-
tion. In J. Fisiak. ed. 7183.
Aronoff, M. and F. Anshen. 1998. Morphology and the Lexicon: Lexicalization and
Productivity. A. Spencer and A. M. Zwicky. 1998. 237248.
Atkinson, D. 1996. The Philosophical Transactions of the Royal Society of London,
16751975: A sociohistorical discourse analysis. Language in Society 25. 333371.
Baayen, R. H. 1992. Quantitative aspects of morphological productivity. Yearbook of
Morphology 1991 ed. by G. Booij and J. Van Marle. Dordrecht: Kluwer. 109149.
Baayen, R. H. 1993. On frequency, transparency and productivity. Yearbook of morphol-
ogy 1992 ed. by G. Booij and J. Van Marle. Dordrecht: Kluwer. 181208.
Baayen, R. H. and R. Lieber. 1991. Productivity and English derivation: a corpus-based
study. Linguistics 29. 801843.
Baayen, R. H. and A. Renouf. 1996. Chronicling The Times: Productive Lexical Innova-
tions in an English Newspaper. Language 72: 1. 6996.
Bauer, L. 1983. English Word-Formation. Cambridge: Cambridge University Press.
Biber, D. 1987. A textual comparison of British and American Writing. American Speech
62: 2. 99119.
Biber, D. 1994. An Analytical Framework for Register Studies. Sociolinguistic Perspec-
tives on Register. ed. by D. Biber & E. Finegan. Oxford: Oxford University Press.
Biber, D. and E. Finegan. 1997. Diachronic relations among Speech-based and Written
Registers in English. In T. Nevalainen and L. Kahlas-Tarkka eds. 253276.
Biber, D., E. Finegan, and D. Atkinson. 1994. ARCHER and its challenges: Compiling
and exploring a representative corpus of historical English registers. Creating and
Using English Language Corpora. ed. by U. Fries, G. Tottie & P. Schneider.
Amsterdam/Atlanta: Rodopi.
Chomsky, N. 1970. Remarks on Nominalization. Readings in English Transformational
Grammar. ed. by R. A. Jacobs & P. S. Rosenbaum. Waltham, MA: Ginn & Company.
Comrie, B. 1976. The Syntax of Action Nominals: a Cross-Linguistic Study. Lingua 40.
177201.
Cowie, C. 1999. Diachronic word-formation: a corpus-based study of nominalization in
the history of English. Ph.D dissertation, University of Cambridge.
Dalton-Puffer, C. 1996. The French Influence on Middle English Morphology: A Corpus-
based study of Derivation. Berlin: Mouton de Gruyter.
Fincke, S. 1997. Discourse Motivations for Productive Verbalizations in Bikol. Paper
presented at the 23rd Berkeley Linguistics Society Meeting, Berkeley, February 1997.
Fisiak, J. ed. 1980. Historical Morphology. The Hague: Mouton.
Fisiak, J. ed. 1985. Historical Semantics, Historical Word Formation. Berlin: Mouton.
Fleischman, S. 1977. Cultural and Linguistic Factors in Word Formation: An Integrated
Approach to the Development of the suffix age. Berkeley: University of California
Press.
Garner, B. A. 1987. Shakespeares Latinate Neologisms. In V. Salmon & E. Burness eds.
207228.
Grlach, M. 1991. Introduction to Early Modern English. Cambridge: Cambridge Univer-

sity Press.
Halliday, M. A. K. 1993. On the Language of Physical Science. M. A. K. Halliday & J. R.
Martin 1993. 5468.
Halliday, M. A. K. 1994. An introduction to functional grammar. 2nd edition. London:
Edward Arnold.
Halliday, M. A. K. and J. R. Martin 1993. General Orientation. M. A. K. Halliday & J. R.
Martin 1993. 221
Halliday, M. A. K. and J. R. Martin, eds. 1993. Writing Science: Literacy and Discursive
Power. London and Washington, D. C.: The Falmer Press.
Hooke, R. [1665] Micrographia. B. Vickers, 1987. 99132.
Hopper, P. and S. A. Thompson. 1984. The Discourse Basis for Lexical Categories in
Universal Grammar. Language 60:4. 703772.
Joseph, B. 1998. Diachronic Morphology. A. Spencer and A. M. Zwicky. 1998. 351
373.
Kastovsky, D. 1982. Word-formation: A functional view. Folia Linguistica 16. 181198.
Kastovsky, D. 1986. The problem of productivity in word-formation. Linguistics 24.
585600.
Kastovsky, D. 1992. The formats change the problems remain: Word-formation theory
between 1960 and 1990. Ptz. ed. 285310.
Kastovsky, D. and B. Kryk-Kastovsky. 1997. Morpholexical and Pragmatic Factors in
Text Cohesion. Language in Time and Space: Studies in Honour of Wolfgang Viereck
on the Occasion of his 60th Birthday. ed. by H. Ramisch & K. Wynne. Stuttgart: Franz
Steiner. 462475.
Koptjevskaja-Tamm, M. 1993. Nominalizations. London: Routledge.
Marchand, H. 1969. The Categories and Types of Modern English Word Formation. 2nd
edition. Wiesbaden: Otto Harrassowitz.
Merlan, F. 1976. Noun incorporation and discourse reference in Modern Nahuatl. Inter-
national Journal of American Linguistics 42. 17791.
Mithun, M. 1984 . The Evolution of Noun Incorporation. Language 60:4. 874894.
Nevalainen, T., forthcoming. Early Modern English Lexis and Semantics. The Cam-
bridge History of the English Language, Vol. III: Early Modern English 14761776. ed.
by R. Lass. Cambridge: Cambridge University Press.
Nevalainen, T. & L. Kahlas-Tarkka, eds. 1997. To Explain the Present: Studies in the
changing English language in honour of Matti Rissanen. Helsinki: Socit Neo-
philologique.
Putz, Martin. 1992, ed. Thirty years of linguistic evolution: studies in honour of Ren
Dirven. Amsterdam & Philadelphia: Benjamins.
Oxford English Dictionary. 1989. 2nd edition. ed. by J.A. Simpson & E. S. C. Weiner.
Oxford: Clarendon Press.
Riddle, E. 1985. A historical perspective on the productivity of the suffixes -ness and -ity.
J. Fisiak. 1985. 435461.
Romaine, S. 1983. On the productivity of word formation rules and limits of variability in
the lexicon. Australian Journal of Linguistics 3. 177200.
Romaine, S. 1985. Variability in Word Formation Patterns and Productivity in the History
206 CLAIRE COWIE
of English. Papers from the Sixth International Conference of Historical Linguistics.

ed. by J. Fisiak. John Benjamins/Adam Mickiewicz University Press. 451465.
Salmon, V. [1970] 1987. Some functions of Shakespearian word-formation. Reader in
the Language of Shakespearean Drama ed. by V. Salmon and E. Burness. Amsterdam/
Philadelphia: John Benjamins. 193206.
Shadwell, T. [1676] 1966. The Virtuoso ed. by M. H. Nicolson and D.S. Rodes. London:
Edward Arnold.
Spencer, A. and A. M. Zwicky, eds. 1998. The Handbook of Morphology. Oxford:
Blackwell.
Spivey, N. 1986. Bluff your way in Consultancy. Horsham: Ravette.
Stubbs, M. 1996. Text and Corpus Analysis. Oxford: Blackwell.
Thompson, S. A. 1974. On the issue of productivity. Approaches to the Lexicon. ed. by
S.A. Thompson & C. Lord. 125.
Vickers, B. ed. 1987. English Science, Bacon to Newton. Cambridge: Cambridge Univer-
sity Press.
Wells, R. 1970. Nominal and Verbal Style. Linguistics and Literary Style. ed. by D. C.
Freeman. New York: Holt, Rinehart and Winston. 297306.
Appendix
Table 5. Word counts for subperiods and registers of ARCHER*

Register 1650 1700 1750 1800 1850 1900 1950
1700 1750 1800 1850 1900 1950
Drama 32391 24421 43161 33265 71161 26101 67352
Fiction 31175 44165 90367 58413 89543 49177 107037
Journal 21912 21847 44708 22826 41954 13437 9253
Legal 43558 34073 29093 21951 26 111
Letters 13506 14111 24751 14581 23544 13 388 27 367
Medical 8156 17103 6979 26467 49272 20 855 9620
News 24660 21894 48864 23301 45900 22 419 50 996
Science 18928 21571 21097 21528 22631 22 675 23 226
Sermons 11372 8854 24537 4561 29022 4273 27 864
Totals 162100 173966 348022 225892 402120 194275 348826
* Note that the second halves of the eighteenth, nineteenth and twentieth centuries are
designed to have both American and British texts for purposes of comparison, and this
explains some of the irregularities in size. Given that the regional aspect is not explored in
this paper, I have simply normalized the results to 100,000 words.
Table 6. Distribution of new types in -(t)ion across registers in ARCHER (raw figures)**
New types Period
across
Register 1650 1700 1750 1800 1850 1900 1950
Drama 5 3 9 2 1 3 7
Fiction 12 11 29 12 10 9 15
Journal 5 4 5 3 6 0 0
Legal - - 17 7 1 5 11
Letters 4 5 11 6 9 2 6
Medical 9 28 7 24 42 19 13
News 11 0 15 13 10 10 27
Science 28 18 13 15 16 19 22
Sermons 4 8 6 2 5 2 7
Totals 130 94 167 97 127 75 137
** The figures given for individual registers in each period do not add up to the total number
of new types for each period, as I have excluded those items which appear in more than one
register.
The Vernacularization of the Negative
Prefix dis- in Early Modern English
R. W. McConchie
Abstract
This article deals with the processes of vernacularization which affected the
negative prefix dis- in Early Modern English. It attempts to identify the period
at which the extension of its use as a productive English prefix took place, and
to find some possible sources of the spread of its use, firstly among medical
writers of the period, and secondarily among some of the figures associated
either closely or loosely with humanism and the Sidney circle. Although
various sources of data have been used, including electronic corpora, diction-
aries, and concordances, some attempt has been made to reconcile their
differences in order to determine the use to which they can be put.
Introduction
Prefixation in English has been discussed previously in general terms, but

little has been done to explore particular prefixes and the way they have been
used. Discussion has also tended to be concentrated on the morphosemantic
aspects of prefixation. Questions such as the extent of their use, the degree of
their restriction, or their individual ability to create new words have been little
considered, especially within a diachronic perspective. The careers of prefixes
over the centuries may however ultimately reveal some deeper insights into
the word-creation process generally what Ladislav Zgusta called the
immense role of productivity (Zgusta 1973: 18) and into the relation
210 R.W. McCONCHIE
between sociolinguistic and statistical significance in lexical studies in par-

ticular.
In these areas, however, we are still exploring the territory, rather than
seeking answers. This paper sets out to explore the assimilation into English
and rise of the prefix dis- in the early modern period as a fully productive affix
creating a considerable number of new words in the sixteenth century. A
forthcoming paper deals with this within very restricted terms (McConchie
2000), but I now propose to account for all occurrences of dis- prefix words in
the works under consideration, not just first citations in the OED. This paper
examines works by William Clever, William Turner, John Jones, John Caius
and excerpts from William Bullein and Gabriel Harvey, in addition to the
material already available in the OED, the Helsinki Corpus, and concordance
material for Sidney, Spenser, and Marlowe. The problem of the period of
reanalysis of this prefix is intriguing, and the present data will perhaps allow
some tentative steps towards clarifying it.
Could historical sociolinguistics entertain and perhaps quantify the theory
that there is a connection between the rarity of a word and the sociolinguistic
significance of its occurrence? Very frequent terms are used as necessities, but
rarer words are chosen far more consciously for various reasons. Consider
examples from Spensers works such as disentrail, dispaint, dismayfully, or
dispurveyance. The case of dispaint will serve to illustrate the point. This rare
Spenserian word, perhaps his own neologism, clearly parallels the Chaucerian
depaint, as in Chaucer Boethius ix, 1, 111 [Th]e cercle of the sterres in alle the
places there as the shining nyg[h]t is depeynted. Spenser writes, His chamber
was dispainted all within / With sondry colours (Faerie Queene 2, ix, 50),
while at the same period the original term is certainly still in use, Bartholomew
Yong using it in his translation of Montemayors Diana (1598: 468) Let now
each meade with flowers be depainted. Spenser presumably wanted to be
understood as using a rare and, one supposes, prestigious literary form sanc-
tioned by the authority of Chaucer.
Likewise, Gabriel Harveys language in a work such as Pierces Super-
erogation might be understood as a conscious attempt to compete with, outdo
or parody Thomas Nashes, so that if Nashe uses such forms as a stylistic
device, Harvey might be expected to do something similar. We understand
such usages as these reasonably satisfactorily in literary terms explicable from
the text itself. Much the same considerations might apply to specialized
scientific or technical text types.
VERNACULARIZATION OF DIS- 211
Is it then likely that the infrequency of a word, other circumstances

permitting, suggests that its occurrence in a second author is more likely to
indicate direct influence or adoption? After all, this is in a loose sense an
assumption upon which editors of Shakespeare and others have based much of
their comparative apparatus for many years, without particularly questioning
its validity.
It is far less obvious, however, why William Clever, an obscure writer
with no obvious literary aspirations, and writing a medical text of no particular
technical sophistication, should use terms such as disapetite, disfashion,
disseason, or disornate. This is especially so when parallels for such terms are
difficult, if not impossible, to find in the record of the lexicon.
This paper will report on the significance of the statistical rise in the
appearance of such words in the sixteenth century and on the sociolinguistic
implications of this phenomenon. The particular hypotheses I will discuss are
(i) that Clevers usage of dis- words is characteristic of medical writings in the
third quarter of the sixteenth century; (ii) that the rise in the prefix dis- takes
place in the sixteenth century and can be described statistically as an s-curve;
and (iii) that the rise in the use of this prefix is related to the later stages of
literary humanism in England.1 The paper also raises by implication some of
the difficulties concerning the nature of the evidence presently available.
This evidence suggests that the prefix dis- changed from one restricted to
use simply in borrowings, finding its way into English directly from Latin and
French and perhaps to a small extent from Spanish and Italian, to fully
productive status by the end of the sixteenth century. Present evidence,
however, here means primarily the Oxford English Dictionary, whose six-
teenth-century evidence is unchanged since its first edition.2 This research has
attempted in a small way to get beyond this source in order to sidestep the
OEDs inevitable shortcomings by restricting the inquiry to a limited number
of source texts which have been used in their entirety rather than excerpted, as
far as possible, and by extracting all instances of dis- words.
My previous research concentrated on those occurrences of dis- words
which were first citations for the OED, but made no attempt to classify them
more narrowly by their origin. The present paper uses all occurrences in the
texts concerned, limiting the search as before to those to which OED ascribes
a first date no earlier than 1520 and no later than 1620, or where the present
instance is an antedating of a later date. Part of this study has also been
restricted to words falling into the OEDs category II, as a living prefix, with
212 R.W. McCONCHIE
privative force, encompassing and evidencing the productive use of the

prefix fully naturalized in English as against borrowings, and listed in the
OED as senses 610 of dis-.
William Clever
The impetus for this research ultimately arose from the apparent occurrence of
a rather large number of such words in a medical text by William Clever, The
flower of phisicke, 1590.3 This work is a relatively long but not very special-
ised account of the influence of the inherent physical characteristics of indi-
viduals, their temperament or temperature as he calls it, and of climate and
other external factors on human health. Clever uses 31 dis- words which fall
into the period 15201620, more than half of which are either antedatings for
the OED (9 instances) or unrecorded (9 instances). The sheer rarity of many of
these words calls attention to their presence.
The first hypothesis tested was whether their use was in any way charac-
teristic of medical writing more generally, and whether comparable numbers
could be found in other such works. A second consideration was the emer-
gence of this trend in the lexicon, so that, in order to narrow the possible time
of introduction of these terms, the texts chosen predate Clever by about a
generation and a half, rather than being contemporaneous. The works exam-
ined are by John Caius, William Turner, William Bullein and John Jones, the
latest of them dating from the mid-1560s.
Dis- Words in Medical Texts
The results of excerpting all dis- words first recorded in OED between 1520
and 1620 from these medical texts, or antedatings which place the word in that
period, are set out below. All instances of such words have been listed and
tabulated, as well as whether the word is an absolute first citation in OED or
not (column 5), and the number of occurrences (column 7). A yes in
brackets in column 5 indicates that the word is unrecorded and thus would
have been an absolute first citation if registered in the OED.4 The gross
number of words in each text is also noted (column 2).5
Table 1. Dis- words in some mid- to late sixteenth-century medical authors

1 2 3 4 5 6 7
no. of OED 1st un- no. of
Author words dis- words category cit. recorded occurrences
Caius 13010 distemperature n yes 1
Turner Bathes 20180 nil nil no nil
Bullein 12260 disgrace v no 1
displace v no 3
Jones6 7230 nil nil no nil
Turner Wines 14900 nil nil no nil
Clever 440007 disapetite v (yes) yes 1
discourse v no no 1
discoverture n yes no 1
discrepant a no no 1
disequal v yes no 1
disfashion n (yes) yes 1
disfashionment n (yes) yes 1
disfavour v no no 1
disflourish v yes no 1
disframe v yes no 6
disframed ppl a (yes) yes 1
disgrace n no no 1
disgrace v no no 2
disliking ppl a yes no 1
dismatch v yes no 1
disoner v (yes) yes 1
disordered ppl a no no 2
disornate v (yes) yes 1
displace v no no 1
dispropriate v yes no 1
disquiet v no no 2
disseason v yes no 2
disseasonable v (yes) yes 2
disseperate v (yes) yes 2
dissimilitude n no no 1
dissociate v yes no 1
dissunder v no no 2
distemper n no no 1
distemperature n no no 7
disure v (yes) yes 1
The actual occurrences are listed with contextual extracts, in the order of
mention, in the appendix.
214 R.W. McCONCHIE
The evidence accumulated so far thus suggests that either Clever was
exceptional in his use of these words compared to other medical writers or, if
he was to some to degree typical, that the increase in their use occurred after
the mid-1560s. While the occurrence of both antedatings and unrecorded
words suggests that Clever was not entirely typical, and that this was an
idiosyncrasy, the three words recorded in Bullein and Caius are all used by
Clever as well.
Borrowed or Productive Prefix?
I now return to the information contained in the OED and new data added to it
(McConchie 1997, appendix 1). The next point to be clarified is how many
and which of these terms belonged to the OEDs category II. Accepting for the
purposes of the exercise that the OEDs ascription of the etymologies of these
words is correct,8 we can remove all words stated to be borrowed from Latin,
Italian, French and Spanish from the list. This leaves only those words
representing the naturalized and fully creative function of this prefix in the
sixteenth century. Adding the words listed as examples in the OEDs explana-
tory preamble as well as any more recent discoveries, the final list amounts to
459 in all. These will be referred to hereafter as dis- II words.
It is immediately apparent that a considerable number of these are re-
corded in the OED as hapax legomena, 162 or 36.3% in all.9 This presumably
represents a high rate of inventiveness by those authors using them, although
the fact that the OED ascribes only one citation to a word is no guarantee of its
being a hapax or nonce-word. It may simply represent the state of the data then
available to the editors, and perhaps not even that very accurately. However,
while it is difficult to speculate about the rarity of individual words, the
general trend for these words to be neologisms seems clear enough. The
vernacular lexicographers such as Florio and Cotgrave account for a good
many of the neologisms, though this phenomenon (witness Clever) is not
limited to their activities.
Analysis of the classes of word represented by these neologisms reveals
that verbal and deverbal forms predominate, as suggested by Jespersen (1942:
26.5, 481). The position is outlined in Table 2, which uses the OED classifica-
tion as a basis.
Table 2. Word class totals and percentages (OED); total 461

Word class total percentage
Adjective 43 9.33
Adverb 4 0.97
Noun 78 16.70
Past participle 3 0.65
Participial adjective 41 8.89
Verb 275 59.65
Verbal noun 17 3.69
In all, the total for verbs and deverbal forms is 336, or 72.88%.
It seems reasonable to assume that the prefix had moved from relative
inflexibility during Middle English to a state of highly innovative freedom by
the early seventeenth century, during which period it could be affixed to a
seemingly limitless number of native and naturalized words, especially verbs.
The immediate effect of deleting from the data borrowings recorded in
our four medical writers, however, is to remove all three remaining dis- forms.
The evidence thus suggests not merely that medical writings had no part in the
introduction of this class of word to the lexicon in the third quarter of the
sixteenth century, but that the existing current of such neologisms, many of
which were already established, was either ignored or perhaps even resisted
by these writers.
A Sixteenth-century Lexical s-curve?
The second hypothesis was that the rise in the sixteenth-century use of the
prefix dis might show itself statistically as an s-curve beginning some time in
the sixteenth century and perhaps peaking in the late sixteenth or early
seventeenth century. The OED data and the additions from Clever suggest that
this rise peaks in the late sixteenth century. As recorded here, however, this
would show not as an s-curve as in other morphological and grammatical
studies but, since what is noted here is essentially introductions of words, as a
sine-curve, as introductions presumably reach a peak and fall away numeri-
cally.10 The establishment of words in the lexicon and their continued use
does not of course show here, and the OEDs record of usage cannot be
assumed to be either consistent or reliable in any case, except perhaps in the
most general sense. The OED was not so much concerned about well-estab-
lished, relatively invariable usage, but about lexical and semantic change and
216 R.W. McCONCHIE
variation, and it records this at the expense of continuity. Only corpora which
record all occurrences of words would produce an s-curve.
A graph of dis- II words 15201620 consolidated into decades is given below.
Dis- II words in numbers per decade, 15201620
120
100
80
number
60
40
20
0
1 2 3 4 5 6 7 8 9 10
decade
Figure 1. Graph of dis- II words from OED and more recent additions; numbers per decade
This appears to show the kind of distribution we are seeking. However,

the more detailed breakdown of the figures given in Figure 2 shows that the
position is much more uncertain than this graph may make it seem.
Dis- II words by year, 15201620
45
40
35
30
number
25
20
15
10
5
0
1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97
year (1-100 = 1520-1620)
Figure 2. Graph of dis- II words from OED and more recent additions.11
Both these graphs of dis- II words, while exhibiting a rise during the
century, are unfortunately inconclusive in at least one respect. Although a
steep and obvious rise occurs from about 1575 on, there is a longish plateau
from about 1525 which invites investigation of the preceding generation or so.
Rather more confusingly, the rise in the last quarter of the century shown in
Figure 2 is explained by the dominating presence of only five authors: Florio,
Cotgrave, Shakespeare, Clever, and Sylvester. One must wonder whether
more detailed examination of less-researched sixteenth-century writers would
even out the curve and make it look more satisfactory, or would merely throw
up more oddities. In this respect one must keep in mind the research on
Shakespeare and Nashe in the OED by Jrgen Schfer, and the cautions he
offers about both their differential treatment and that of others (Schfer 1980).
It is worthwhile returning now to the larger picture to see which authors
appear in the OEDs first citations of dis- II words in absolute terms over the
whole period under discussion.
Table 3. Authors or works yielding most dis- II words 15201620 (3+)

No. Author/work No. Author/work
59 Florio, John 3 Bacon, Francis
24 Cotgrave, Randle 3 Berners, Lord
20 Shakespeare, William 3 Eden, Richard
19 Clever, William 3 Fenton, Sir Geoffrey
15 Sylvester, Josuah 3 Fletcher, Giles
9 Daniel, Samuel 3 Hakluyt, Richard
9 Jonson, Ben 3 Hall, Edward
8 Holland, Philemon 3 Heywood, John
8 Spenser, Edmund 3 Heywood, Thomas
6 Foxe, John 3 Newton, Thomas
6 Nashe, Thomas 3 Puttenham, George
6 Warner, William 3 Sandys, Edwin, Abp.
5 Acts Henry VIII 3 Shelton, Thomas
5 Donne, John 3 Sidney, Mary, Countess of Pembroke
5 Palsgrave, John (Jehan) 3 Sidney, Sir Philip
4 Chapman, George 3 Speed, John
4 Drant, Thomas 3 Stanyhurst, Richard
4 Phaer,Thomas 3 Tindale, William
4 Sandys, Sir Edward 3 Udall, Nicholas
3 Wright, Thomas
218 R.W. McCONCHIE
The dominance of two vernacular lexicographers, Florio and Cotgrave, in

the list in an age which admired copiousness in dictionaries, and in languages
which provided ample scope for calquing dis- forms, must make one hesitate
to declare this a classic s-curve, regarding it perhaps as an artefact of their
activities instead. At the same time, however, since this graph does indirectly
represent the active introduction of terms, and discounts their continued use,
better usage data might still produce some kind of s-curve.
As to medical writings used by the OED as sources, the instances of dis-
II words as first citations in the OED for this period seem only to amount to
ten, and thus do not easily admit of interpretation. These are: discommend-
able, discarve, disproportionated, disbalass, disdiet, disknowledge, disoppi-
late, discommodable, dishabit, distrustiness, and dischest. This list produces
Figure 3, which represents the situation shown in Figure 1 very incompletely
and unsatisfactorily.
OED dis- II first citations (plus antedatings) 15201610 from

medical writings
10
8
number
6
4
2
0
1 2 3 4 5 6 7 8 9
decade
decade
Figure 3. Dis- II first citations from OED plus antedatings from Thomas Hill (1) and
William Clever (9)
While the OED might represent this discipline insufficiently or incom-

pletely, the results from Caius, Turner, Bullein and Jones presented earlier
suggest that this gap in mid-century might be genuine, medical writers for
some reason eschewing the slowly growing body of dis II words for a
generation or so. Whether the isolated instances of discommendable in the
Vertuous boke of distillacion and Coplands discarve are coincidental or
might indicate wider use of such terms, perhaps a little closer to the locus of
reanalysis, is still to be determined. It is possible that this hiatus in mid-

century, if confirmed by more extensive research, might ultimately arise from
a change in the perception of the nature and function of vernacular medical
texts.
Clever and the Sidney Circle
Parallels for Clevers use of dis- II words, assuming they exist, must therefore
be found elsewhere than in medical writers, although some increase certainly
occurs among them. Since the OED data suggest that many of the dis- words
were used by members of the Sidney circle,12 comparison between the practices
of Clever and those of some of these writers seemed an obvious next step. Direct
comparisons between Clever and Sidney, Spenser, Marlowe and Gabriel
Harvey were thus made in order to test my third thesis, that the rise in the use
of the prefix dis- is related to the later stages of literary humanism in England.
It would be most interesting to add Shakespeare (and of course many
others), but to do so here would upset the balance, in limited data, between
those closely or loosely associated with the Sidney circle. Shakespeares
repute alone is insufficient reason for inclusion.13 Others more intimately
concerned with humanism such as Thomas Drant and Thomas Phaer also
suggest themselves as obvious candidates for research.
Words used in common between Clever and Sidney, Spenser, Marlowe
and Harvey are indicated by boldface, and dis II words by italics. Concor-
dances to the works of Sidney, Spenser, and Marlowe were used, while the
Harvey data is derived from the Corpus of Early English Correspondence
(CEEC; Nevalainen and Raumolin-Brunberg).
Table 4. Dis- words in Sir Philip Sidney (total 14) and William Clever
Sidney Clever
1. disaster n 1. disapetite v
2. disburden v 2. discourse v
3. discontent n 1 3. discoverture n
4. discourse v 4. discrepant a
5. disdainfulness n 5. disequal v
6. disgrace n 6. disfashion n
7. disgrace v 7. disfashionment n
8. disgraceful a 8. disfavour v
9. disorder n 9. disflourish v
220 R.W. McCONCHIE
10. displace v 10. disframe v

11. displacing vbl n 11. disframed ppl a
12. disquiet v 12. disgrace n
13. dissembler n 13. disgrace v
14. distinguish v 14. disliking ppl a
15. dismatch v
16. disoner v
17. disordered ppl a
18. disornate v
19. displace v
20. dispropriate v
21. disquiet v
22. disseason v
23. disseasonable a
24. disseperate v
25. dissimilitude n
26. dissociate v
27. dissunder v
28. distemper v
29. distemperature n
30. distinguish v
31. disure v
Correspondences in bold; dis- II words in italics.
Here the rate of correspondence (words used by both writers expressed as

a percentage of the total minus the correspondences) between Sidney and
Clever is six out of twenty-eight (21.4%). Only three of Sidneys words,
however, are dis-II terms, suggesting a relative lack of interest in such neolo-
gisms. They only share one dis- II word. By and large, Sidneys usage in this
respect is rather predictable and conventional.
Clever and Spenser
The author whose works produce by far the most dis- words amongst those
surveyed here, however, is Edmund Spenser.
Table 5. Dis- words in Edmund Spenser (total 56) showing correspondence with William
Clever
disastrous a dismount v
disattire v disorder n
disburden v disordered ppl a
discard v disorderly adv
discontent n1 dispace v
discontentment n dispaint v
discordful a dispart v1
discourteous a dispart v2
discourtesy n dispersed ppl a
discovery displace v
disdainful a dispread v
disdainfully adv disprofess v
disentrail v dispurveyance n
disentrailed ppl a disquiet a
disgrace n disquietness n
disgrace v disquiet n
dishearten v disrobe v
dishonourable a dissemblance n
disleal a dissemble
dislike v dissembling ppl a
dislike n dissent n
disliking n? distent ppl a?
disloign v disthronize v
dismay n distinguish v
dismay v distort ppl a
dismayed ppl a distraught v
dismayful a distressed ppl a
dismayfully adv distrustful a
However, the correspondence rate is the lowest of all, a mere five out of
eighty-three, or 6.0%, and they share no dis- II words at all, which might rule
out any suggestion of influence. To what extent Spensers well-known ar-
chaizing, especially in the Faerie Queene, produced forms such as dismay-
fully, is a question deserving separate treatment elsewhere.
Clever and Gabriel Harvey
Gabriel Harvey was loosely associated with the Sidney circle (Hannay 1989:
1824), and became involved in the Marprelate controversy and the conse-
222 R.W. McCONCHIE
quent pamphlet war with Thomas Nashe. His writings show a great propensity
both to coin his own words, and to pick up on those used, usually against him,
by the inventive Nashe. For the present purposes, however, a group of his
private letters totalling about 5000 words was used as a source. These pro-
duced the following (see appendix for quotations):
Table 6. Dis- words in Gabriel Harvey (total 4) showing correspondence with William
Clever
Word
1. discourse v
2. disdainful a
3. disdaining vbl n
4. disquiet v
Only one dis- II word is shared, in fact the same one. Perhaps the outcome
in this table is a little disappointing, especially from a writer so apparently
willing to neologize in his published works, but this may well indicate the
influence of the private letter text type upon his style.14 In any case this is a
rather small sample as it stands.
Clever and Marlowe
The occurrence of dis- II words in the works of Shakespeare suggests that other
literary figures, especially other dramatists, might yield interesting results. A
concordance to the works of Christopher Marlowe was thus consulted. This
produced a list of 22, roughly comparable in raw numbers to Clever.
Table 7. Dis- words in Christopher Marlowe (total 22) and William Clever
Marlowe
1. discontent n 1 12. dismount v
2. discourse v 13. disordered ppl a
3. discoursing vbl n 14. dispensive a [a 1 cit]
4. discoursive a 15. displace v
5. disdainful a 16. displeased ppl a
6. disdainfully adv 17. dispose n
7. disgrace n 18. dispread v
8. disgrace v 19. disquiet a
9. dislike v 20. dissembling ppl a
10. dismay v 21. distinguish v
11. dismembered ppl a 22. distressed ppl a
Closer examination reveals that while the size of Clevers and Marlowes
lists is comparable, the correspondence rate seems rather limited, since only
five of the forty-seven lemmas occur in both (10.6%). While this disparity
might arise from the nature of the text types involved, it is worth remembering
that many of Clevers terms are not specific to medicine, none of them being
restricted to medical use, with the possible exception of distemperature, and
those that do have an obvious application in medicine, such as disordered,
often have more general currency as well. Note however that my assumption
that five out of forty-four is a low figure arises more from the lack of lexical
frequency studies of this kind than from confidence that it must be so. In
Marlowes case the number of dis II words is four, rather higher than
Sidneys number, and none are shared with Clever. At any rate, Clevers
usage seems a little closer on the whole to Marlowes than to Sidneys, despite
the lower correspondence rate.
Conclusion
The answers to the three hypotheses posed at the outset appear to be: (i) Clever
is unlikely to have been typical of medical writing generally; (ii) that the rise
of dis- takes place in the sixteenth century and forms an s-curve is probable
within the limits of data available; and (iii) the association with literary
humanism is still possible but must be demonstrated in detail. At a more
general level this paper has been an attempt to exploit and to reconcile sources
of data which are essentially different. The result is inevitably unsatisfactory,
but may be sufficient to reveal the need for more complete corpora, and more
extensive studies at the same level of detail. The sociolinguistic implications
seem to be that idiosyncrasy may be so marked in certain registers and levels
of rarity as to distort present data.
It is apparent that more satisfactory conclusions can only be based on
fuller data, including, for example, the Shakespeare corpus. It must also be
apparent that, while we are perhaps a little closer to identifying the critical
chronological locus and lexical vector of reanalysis of dis- II terms, they
cannot be located in the limited data of this study. At this stage it is only
possible to suggest a period, given the patchy nature of the evidence. A fuller
study based on a very much larger database which will permit confidence as to
inclusiveness, both chronologically and in terms of text-type, is obviously a
desideratum.
224 R.W. McCONCHIE
Notes
1. These hypotheses merely reflect the data so far compiled, and do not represent an attempt
to provide comprehensive answers.
2. The latest Shorter Oxford English Dictionary, while not ascribing exact dates, is actually
more up-to-date in this respect.
3. Unfortunately practically nothing is known of Clever himself.
4. The works have been excerpted using Concordancer 1.76.
5. Word counts are rounded down to the nearest 10.
6. Note however Bviir [?]iaforithka in Latin, discutie[n]tia, or opening, termed in greke.
The form discutient does occur in the OED.
7. This estimate is still inaccurate to the extent that there are several unreadable passages in
my copy which have not been deciphered as yet, amounting to perhaps two or three
pages.
8. They may well not be. This whole matter is the subject of extensive revision for OED3.
9. Clearly Clevers terms could not be accurately included in this calculation, which has
therefore been done assuming that a third of his terms might have been so treated had they
appeared in the OED. This guess is an attempt to allow for the discrepancy between what
information the OED slips actually contained and what the editors did, and for the fact
that the OED tagging is not consistent.
10. My intuition from study of the OED, short of complete analysis, is that this does not
happen, or only to a limited degree. Introductions seem to continue at a substantial level
in the seventeenth century, and there seems to be another surge in this category of
introduction in the nineteenth, but these hunches remain to be investigated.
11. Note that for convenience in constructing this graph, Knolles disseveration listed as
16.. has been omitted for want of a more accurate date, Adams dishallow has been
allotted 1620, Barclays disprovide has been ascribed to 1525, Donnes disconduce and
disconducing have been ascribed 1605, and that the words listed under the dis- lemma in
OED for Florio have all been dated 1610, although some are certainly earlier.
12. McConchie, forthcoming, 2000.
13. Shakespeare usages which are OED first citations (though not necessarily absolute first
citations) form an extensive list, including disbench, discandy, disedge, disgraceful,
dishearten, dishorn, disliken, dislimn, dislocate, disorb, displant, dispossess, disprop-
erty, dispurse, disrelish, disseat, and distaste. For a discussion, see McConchie forthcom-
ing, 2000.
14. Research is in progress on his Pierces Superogation, which may present a rather
different picture. A manual search of this work has to date yielded many tokens and a
larger list, but few neologisms, most being familiar from late Middle English, and few
dis- II words.
References
Bullein, William. 1562. Bulwarke of defe[n]ce.

Caius, John. 1552. A boke against the sweate.
Clever, William. 1590. The flower of phisicke.
The Compact Edition of the Oxford English Dictionary. 1979. 2 vols. London: Book Club
Associates.
Donow, Herbert S. 1975. A Concordance to the Poems of Sir Philip Sidney. Ithaca: Cornell
University Press.
Fehrenbach, Robert J., Lea Anne Boone & Mario A. Di Cesare, 1982. A Concordance to the
Plays, Poems, and Translations of Christopher Marlowe. Ithaca: Cornell University
Press.
Hannay, Margaret. 1989. Philips Phoenix: Mary Sidney, Countess of Pembroke. New
York: Oxford University Press.
Harvey, Gabriel. Letters. CEEC Corpus, in preparation.
Jespersen, Otto. 1942. A Modern English Grammar on Historical Principles: Part VI
Morphology. Copenhagen: Einar Munksgaard.
Jones, John. 1566. A dial for all agves.
McConchie, R.W. 1997. Lexicography and Physicke: The Record of Sixteenth-Century
English Medical Terminology. Oxford: Clarendon Press.
McConchie, R.W. 2000. Fashionable idiolects: the use of the negative prefix dis- 1547
1613. The History of English in its Social Context, forthcoming.
Osgood, Charles Grosvenor. 1915. A Concordance to the Poems of Edmund Spenser.
Washington: The Carnegie Institute of Washington.
Schfer, J. 1980. Documentation in the O.E.D.: Shakespeare and Nashe as test cases.
The Shorter Oxford English Dictionary on Historical Principles. 1993. 2 vols. Oxford:
Oxford University Press.
Turner, William. 1562. A booke of the bathes in England.
Turner, William. 1568. A new boke of wines.
Zgusta, Ladislav. 1973. Lexicology: generating words. Lexicography in English: Annals
of the New York Academy of Science, ed. by Raven I. McDavid & Audrey R. Duckert,
vol. 211, 1420.
Appendix
Caius, John A boke against the sweate 1552

Bivv either by simple distemperature, or by infection and
Bullein, William Bulwarke of defe[n]ce 1562 (extract)
f. 20 you be ware, that ye disgrace not your self thorowe rashenes, but be
f. 19 that was loste, and to displace an appostumacion, if it be concurraunt
f. 15 how thesame must be displaced and remoued, certainly my self did
f. 17 will not so easly be displaced, as I haue talked of, but require
226 R.W. McCONCHIE
It is immediately obvious that these texts produced practically nothing, only Bullein and
Caius yielding any words at all. This compares strikingly with Clevers list:
77 (L2r) the head, and disapetiteth the stomach: and the rather if
31 (E3r) same is no consequent discouerture. But it cannot be denied in those
87 (M3r) at large. Auycen discourseth and trauelleth most highlie heerein
122 (Q4v) actions vnequall and discrepant one from another, and their
45 (G2r) and vnperfectly disequalled, Euen as there is manifestly shewed
65 (I4r) of what fashion or disfashion soeuer the body is: but the power of
102 (O2v) alteration, and disfashionment of hayres, which inwardly
79 (L3r) doo thereby vtterlie disfauour complexion, corrupt bloud, and
20 (D1v) of substance: or els disflorisheth the laudable generation of bloud
42 (F4v) as when they were disframed and dissociated from their equall
43 (G1r) although they bee of difframed[sic] conditions and qualities: or els we
47 (G3r) the elementes were disframed with heate, by an extreame and
98 (N4v) they be naturallie disframed, as also for that their passage from
3 (B1r) mination ouerruleth, disframeth and disseperateth those bodies from
(To the reader) in the elements disframing the bodily health of man vppon earth
120 (O3v) could I here vtter in disgrace of the Pandect, for false exposition of
126 (R2v) such medicines which disgrace the disease, are ordinarie and of high
122 (Q4r) in the one, and disgraceth the other. So likewise of Cel[?]lo[
901 (O4vN1r) thinke the same farre disagreeing fro[m] this our purpose: that is
123 (R1r) eyther of liking or disliking propertie. For some are of equall
72 (K3v) bodie with good blood, disonereth the body from raw excrementes
37 (F2r) there was some disordered surfet before health setled in
83 (N1r) and such like disordered persons. Galen affirmeth, verie
86 (M2v) Nature wolde not disornate the beautie of the face with hair[e
3 (A4v) weaken the force, as displace and expell the poyson of the disease
45 (G2r) passiue actions were dispropriated and vnperfectly disequalled, Euen
17 (C4r) disease consisteth, disquieteth both the disease, and the patient
26 (D4v) dangerously crept in, disquieting the naturall disposition of such
108 (P1v) thereby distemper and disseason the inferiour causes of the earth. So
22 (D2v) brest, by reason of a disseasonable winter or vnnaturall spring
23 (D3r) vppon contrarie and disseasonable operations of times: But these
11 (C1r) of moisting showers, disseasoning the earthlie fruites of mans mortall
3 (B1r) disframeth and disseperateth those bodies from due
45 (G2r) if there were such a dissimilitude in the elementes, as that not one
100 (O1v) astonieth hearing, dissmacheth tasting, and stencheth smelling
42 (F4v) were disframed and dissociated from their equall places: As when
26 (D4r) themselues may dissunder in operation, and ingender a seuerall
23 (D3r) other, the disease is dissundered and easily auoyded downward: it
88 (M3v) or corrupt blood, or distempered by the contagion of colde
36 (F1v) of the drinesse and distemperature of the body. Many good writers
50 (G4v) so indangered by the distemperature of Autumne doe drawe dangers
50 (G4v) being subiect to the distemperature of an intermixt meridian poynt
96 (N3v) the temperature or distemperature of the yeare: so may we gather
50 (G4v) the temperatures or distemperatures of the foure seasons holde a

52 (H1v) and as all maling[sic] distemperatures haue conspired herewith, so
77 (L2r) both giddinesse distempereth the head, and disapetiteth the stomach
44 (G1v) qualitie must be disured and fall away therewith, for that the
OED medical sources

1527 discommendable Andrewe Vertuous booke of distillacioun
1541 discarve Copland The questyonary of cyrurgyens
1572 disproportionated Jones, John The bathes of bath
1576 disbalass Newton Lemnies touchstone of complexions
1576 disdiet Newton Lemnies touchstone of complexions
1576 disknowledge Newton Lemnies touchstone of complexions
1577 disoppilate Frampton, John Monardes ioyfull newes out of the newe founde
worlde
1579 discommodable Twyne, Thomas Petrarcas phisicke against fortune
[1579 dishabit Hill, Thomas (anteding of Shakespeare from my own database.)]
1579 distrustiness Twyne, Thomas Petrarcas phisicke against fortune
1579 dischest Jones, John The arte and science of preseruing bodie and soule in
healthe [only marginally a medical text]
Gabriel Harveys letters (CEEC)

ful hardly drawn out of cumpani. And as for disdaining of others, I wuld thai culd have
found (A 1573 T GHARVEY 4:Heading)
most tru, that I have not shoun mi self so surli towards mi inferiors as M. Nevil hath shoun
him self disdainful towards his (A 1573 T GHARVEY 4:Heading)
and a few the like, more fit for schollars declamations to discurs uppon then semli for
masters problems to dispute uppon: (A 1573 T GHARVEY 11:Heading)
and disquiet the Church as I now do the Chappel. Wel, I hope thai shal have no great caus as
long as I busi mi self in no other (A 1573 T GHARVEY 11:Heading)
Brook Symposium on the Revised OED and
English Historical Lexicography: A Report
Edited by Christian Kay
Edmund Weiner, Oxford University Press: Oxford English Dictionary

(OED)
Penny Silva, Dictionary Unit for South African English: Dictionary of SA
English (DSAE)
Frances McSparran, University of Michigan: The Middle English Compen-
dium (MEC)
Jane Roberts, Kings College London: A Thesaurus of Old English (TOE)
Louise Sylvester, Kings College London: Middle English Thesaurus (TME)
Iren Wotherspoon, University of Glasgow: Historical Thesaurus of English
(HTE)
Christian Kay, University of Glasgow, Chair: Roundtable
Introduction
The final session of the G.L. Brook Symposium was a workshop on electronic
resources, with particular emphasis on the work being done at Oxford Univer-
sity Press on the revised OED. Representatives of each project described their
work, illustrating it as appropriate with text samples. This was followed by a
lively roundtable discussion, with members of the audience commenting and
asking questions. Thanks are due to the University of Manchester for setting
up the computing facilities, and to Jean Anderson and Flora Edmonds of
Glasgow for ensuring that everything went well on the day.
230 CHRISTIAN KAY
Edmund Weiner, Oxford University Press: Oxford English Dictionary

(OED)
Penny Silva, Dictionary Unit for South African English:
Dictionary of SA English (DSAE)
The Making of the OED, Revised Edition
The presentation outlined the OEDs approach to the documentation of Eng-

lish in the past, and demonstrated the current revision programme. This is the
first full-scale revision since the dictionary was published between 1884 and
1928, although new words and senses were added in the OED Supplement
(197286; now incorporated in OED2). The editorial policy for the revised
edition encompasses a complete overhaul of the etymologies; expansion of the
coverage of variant forms and pronunciation; the addition of thousands of new
citations and the standardization of thousands of existing ones; and the updat-
ing of definitions, grammatical treatment, and labelling. The existing reading
programme has been expanded to encompass pre-1800 English; new words
and senses (some of which are chronologically old) continue to be added to the
database. Use is being made of the many new resources which are now
available to the editors, but interestingly many antedatings are found through
electronic searching of the OED text itself. The revision is expected to be
complete in 2010. OED2 and the recently published Additions volumes will be
made available online in 2000; successive sections of the revised OED (in-
cluding new unpublished entries) will be published quarterly online, thus
offering a modern version of the original publication by fascicles.
The OEDs Approach to English Varieties
The speakers then described the various regional initiatives for the recording
of the varieties of English around the world, and reported on one case that
of the OED and the Dictionary of South African English on Historical Prin-
ciples as an example of how co-operation assists with the effective cover-
age of English, both in the OED and in the regions.
James Murray admitted in his General Explanations in the front matter
of the OED that the circumference of English is not discernible. His diagram
labelled as the circle of the English language shows the common core of the
language at the centre, shading from the literary to the colloquial. Leading
BROOK SYMPOSIUM 231
away from the common core are five areas of the language technical,
scientific, foreign, dialectal, and slang. Murray believed that the further one
moved outwards from the centre, the less should the English lexicographer be
concerned with documenting words in these categories. Although the OED
included words from the colonies words such as the South African tickey
and trek Murray did not enumerate regional varieties as being a part of the
periphery of his circle of English.
Lexicographers in the various English-speaking regions were, however,
independently compiling regional historical dictionaries many well ahead
of the OED: Jamieson in Scotland (1808), Bartlett (1848) and Thornton
(1912) in America, Yule and Burnell in India (1886), Morris in Australasia
(1898), and Pettman in South Africa (1913). And the OED Supplement
(197283) expanded the coverage of world English considerably.
OED editorial policy today is that English is no longer a hierarchy, with
British (or even English) English as the standard and the other varieties as less
significant derivatives. English is seen rather as a community of varieties in
which British English is one of many forms to be documented. It is recognized
that no single project can cover all varieties effectively that both the broad,
inclusive overview of the OED and the local detail of the regional dictionary
are essential. Regional expertise is important to the OED for several reasons:
regional coverage in the OED is uneven; local judgement is required for
providing the OED with important local words and senses which are not yet
included; the full prior history of a word might be regional, and the historical
record is incomplete without this information; and regional dictionaries serve
as linguistic magnifying glasses, recording the fine detail which the OED
cannot cover.
Co-operation in Documenting World English
The case of South Africa provides an example of how co-operation can assist
in the work of both the OED and the regional dictionary. Since the 1970s,
when the South African historical dictionary was initiated, there has been a
good working relationship between the two projects. From 1990, when the
final editing of the South African historical dictionary commenced, Oxford
and Grahamstown were in daily contact via email exchanging data, and
(more significantly) expertise. Initially the South African project was depen-
dent on the OED for being taught historical methodology, but they now supply
232 CHRISTIAN KAY
the OED with new entries, and with expertise on the local vocabulary. The
outside eye proved to be very useful to both projects for the recognition of
words which were peculiar to each variety. Once the South African text was
completed, it was made available to the OED editors online, providing a
useful corpus for searching.
If all existing historical dictionary projects shared their expertise, the new
circle of the English language could be significantly different from the old:
the full range of English varieties serving as the spokes of a wheel, with the
OED serving as the hub, for the effective collection and dissemination of
information on the English vocabulary world-wide.
See further http://www.oed.com/inside/revision.htm
http://www.rhodes.ac.za/affiliates/dsae/
Frances McSparran, University of Michigan:

The Middle English Compendium (MEC)
The Middle English Compendium is a new set of electronic resources for

medievalists, for use both in research and teaching. It links three different
types of electronic resources in ways that extend the functions, flexibility and
applications of each. Three major electronic resources are being developed: an
electronic version of the Middle English Dictionary (MED), an electronic
HyperBibliography of Middle English texts, manuscripts, editions, and other
material used in the MED, and a Corpus of Middle English Prose and Verse.
The first two components are now almost completed: the letters A-W(1),
corresponding to 107 fascicles of the print MED, are on line, and the remain-
ing fascicles will be added as they are published through 2001; bibliographical
entries for all MED stencils beginning with the letters A, B, and J through Y,
the Chaucer entries, and the incipits, are all on line, and the remaining entries
will be available by December 1999.
The print Middle English Dictionary, now nearing completion (at about
15,000 pages), is the only scholarly dictionary devoted to the language of the
period. Converting it into electronic form makes it infinitely more powerful as
a research tool, since the entire dictionary (which contains over 50,000 head-
words and about 1,000,000 Middle English quotations) becomes a huge
database, searchable in entirely new ways by historians of medieval culture
and society, literary scholars, linguists and dialectologists. The impact of other
languages and cultures on Middle English, the development of specialized
BROOK SYMPOSIUM 233
vocabulary in various fields, semantic innovation, syntactic change, develop-

ment of phrases and collocations, the influence of major authors on the
language all of these can now be investigated within the MED in ways
impossible hitherto. Users can follow up their findings in the electronic MED
by hypertext links from every quotation to the relevant entry in the Hyper-
Bibliography, and move thence to the Corpus and other related resources.
The HyperBibliography both in its range of content and in its functioning
as a mechanism for switching between related resources is something quite
new. It is a rich and searchable store of information about the editions used by
the MED, the manuscripts cited, the scribal dialects of the more than 6000
copies of medieval texts (literary, documentary, medical, historical) quoted in
the MED. Current scholarship on codicology and LALME localizations of
scribal dialects are being used to enrich this material, and electronic distribu-
tion on the web makes it possible to expand and update the content in the light
of new developments. Links from the HyperBibliography to the third major
component, a modest but growing Corpus of Middle English Prose and Verse,
enable users of the MED to move to complete electronic versions of texts cited
in the MED for additional evidence or further research, using either the Corpus
as a whole, or single texts, or user-selected groups of texts.
NEH and University of Michigan funding is enabling us to develop the
electronic version of the MED and the associated HyperBibliography in a two-
year grant period, but we have still to undertake significant development of the
third component, a major, reliable and richly-tagged text collection. Our
Corpus of Middle English Prose and Verse at present contains 54 texts, and
needs a dramatic expansion of material if it is to function as intended in the
Compendium. No dictionary, even one as comprehensive as the MED, can give
a full record of the language. Within a given entry, quotations have been
selected from a huge corpus of material to illustrate the range and development
of senses, forms, and uses over space and through time. Yet scholars or
students who need the full record of occurrences within a given work, author,
dialect area, or period, will want to move beyond the MED to specialized
electronic subsets of texts selected according to their needs. The Compendium
is designed to enable this. We have secured funding to convert about 50 major
texts into electronic form for our own Corpus as a first step towards expanding
the Corpus. Longer term plans and hopes include establishing links with other
major electronic editing projects, and other electronic lexicographical projects.
The Middle English Compendium is available by subscription from the
234 CHRISTIAN KAY
University of Michigan Press. Details may be found at the following URL:

http://www.hti.umich.edu/mec
Jane Roberts, Kings College London:

A Thesaurus of Old English (TOE)
Michael Samuels, at the outset of the Historical Thesaurus of English (HTE)

project, remarked on the need, for research purposes, of a listing of Old
English, to point to the nature of Old English vocabulary omitted from the
OED, which is essentially a record of post-Conquest English. The Thesaurus
of Old English (TOE), published in 1995 as a pilot study for the forthcoming
HTE, draws together Old English word-senses, principally from Bosworth-
Toller and Meritts edition of Clark Hall.
Volume I of the TOE contains the Old English vocabulary organised in a
conceptual classification. Following the overall organisation of its parent
thesaurus, the TOE categories 1 to 5 relate to the External World (SECTION I
of the HTE), categories 6 to 11 to the Mind and to aspects of behaviour on an
abstract level (SECTION II), and categories 12 to 18 to Society and its
adaptation of the physical world (SECTION III).
The overall structure of the classification is hierarchical, proceeding in
each category from the most general terms to the most specific. Thus, the
meaning of a word at any particular point in the hierarchy is defined not only
by its own heading but by the headings above and below it in the structure.
Within each heading, subordination or co-ordination may be indicated, with
unnumbered dots allowing subordination to a further degree of specificity. My
example, illustrating some of the words used of ecclesiastical garb, comes
from Category 16 Religion:
16.02.05.06 Ritual clothing:
. A chrismale (baptismal): crisma, crismalo, crismclao
. A sackcloth/penitential garment: cilicog, hre, scc, witehrglop
. Church vestments: ciricwdo
. Priestly vestments: preostreafo
. . Vestments used at mass: mssereaf
. Episcopal robes: bisceopgegyrelano
. A deacons robe: diacongegyrelao
. Epistolers vestments: pistolclao, pistolrocco
BROOK SYMPOSIUM 235
Cross-referencing alerts the reader to look for more general clothing vocabu-
lary in Category 04 Material Needs. Small superscript letters flag some word
forms as infrequent within the Old English corpus, for example o for words
that occur once only, and p for words found only in the poetic corpus. Volume
II of the TOE contains an alphabetic index, listing the category or categories in
which word-senses are to be found.
The incorporation of the TOE into the HTE should make it easier both to
identify lexical loss in the early Middle English period and to recognise and
follow up continuities. The flags point to noteworthy or puzzling forms that
may present particular difficulties in the editing of the TOE materials into the
HTE. Essentially our purpose was to make the word-senses attested from Old
English available within a classificatory system (Volume I), as is plain from
the Index (Volume II), which serves to list alphabetically, together with a
finding formula, each and every word-sense presented in the thesaurus classi-
fication.
Louise Sylvester, Kings College London:

Middle English Word Studies: A Word and Author Index
With the support of a British Academy/HEFCE grant awarded in 1995 we

began work on a Middle English Word Studies project at Kings College
London. Our plan was first to produce a bibliography of Middle English word
studies on the model of Old English Word Studies: A Preliminary Word and
Author Index (1983). The book will be in two main parts, the first of which is
an annotated bibliography of lexical studies for Middle English. Our reading
programme for the bibliography centres on journal articles from the 1950s
onwards, together with some essays published in collections, and a few books.
We are mainly collecting pre- and post-datings of OED and MED, missing
entries, ghost words, new readings in manuscripts, and the reinterpretation of
textual cruces. For every article or book read, a slip is made stating the author,
title, publication details and the words discussed with page references for
each, as well as a brief summary of the contents and approach of the study.
Each slip is entered into a PAPYRUS database, which currently contains
approximately 3,000 entries. One advantage of the database is that it enables
the kind of interrogation of the data which is not possible with only a paper
archive.
In September last year, Jane Roberts and I decided to conduct some
236 CHRISTIAN KAY
investigations, as a result of which we decided to go for completeness in

reading Notes and Queries and English Language Notes, and to provide only
the best possible sampling that we could for everything else. This is partly, of
course, a time constraint, but bibliographies are open-ended, and we did not
want to continue with our reading programme forever. We also plan to check
our reading against the items listed in Burnley and Tajimas bibliography
which look as if they contain information useful for word studies.
The second part of the volume will consist of a word register, alphabeti-
cally ordered by word, in which an individual word or phrase can be refer-
enced. The reader will be directed to substantial discussions of the word or
phrase in the secondary literature on Middle English. In this context substan-
tial need not mean lengthy, only that the author has supplemented the informa-
tion available in the OED, the MED and/or notes and glossaries in edited texts.
We also plan to include an extensive introductory section offering critical
analyses of the word studies, and a prefatory list of scholarly dictionaries of
Middle English, concordances, and important glossaries for specific texts. Our
intention is to use the information gleaned from the reading programme to
produce a volume of semantic studies which makes use of the HTE taxonomy
to show which semantic fields have been studied and which remain to be
investigated. One model for such a volume is Strites Old English Semantic-
Field Studies (1989). Both of our studies are intended to provide a platform for
the work needed to produce a self-standing Thesaurus of Middle English.
Iren Wotherspoon, University of Glasgow:

Historical Thesaurus of English (HTE)
The Historical Thesaurus of English (HTE) displays the vocabulary of Eng-

lish from Old English to the present in hierarchically ordered conceptual
categories. The words within each category are in chronological order of date
of first recorded use. Style and status labels such as slang, dialectal, historical,
are included. The primary aim of the project is to offer new material for
research on lexical innovation and loss, and on semantic change. Underlying
this aim is the view that such matters can better be understood if words are
examined in the context of their neighbours within a semantic field or sub-
field. In addition to this, because it draws together the vocabulary of specific
areas of human interest or activity, HTE offers a powerful resource for
scholars in other fields, such as literature, ethnography and history. These
BROOK SYMPOSIUM 237
points were demonstrated by live displays of the fields of Anger, Food, and
Funerals, and participants in the Symposium were able to see the results of
queries which they submitted, e.g. to find Scots words for ill-tempered at a
specified period.
Data for the project come from the OED, supplemented by the TOE
materials described in Roberts above. Because of changing word-meanings
and field boundaries, matching one to the other is by no means a simple
process. Hand-written paper slips hold our material and are used in classify-
ing, but we have also passed through several computer revolutions. We have a
legacy database in dBase3, which is still used for data entry. In addition, the
data are now held in Ingres with Hummingbird BI/Query as user interface, and
in Access, which, for technical reasons, was the database used at the Sympo-
sium.
The project was started by Michael Samuels in 1965 and is about 75%
complete overall, presently comprising 409,942 meanings, with 124,236 cat-
egory and sub-category headings. In common with many dictionary editors,
we have a problem in deciding when to stop collecting data, with new words
appearing almost daily in the language. However, we have now decided to add
the OED Additions volumes and then stop. Both electronic and book publica-
tion are planned by Oxford University Press (but we are not yet sure in which
order). Funding over the years has come from the British Academy, the
Leverhulme Trust, the Carnegie Trust for the Universities of Scotland, and the
University of Glasgow.
see further http: //www.arts.gla.ac.uk/EngLang/thesaur/level2.htm
Christian Kay, University of Glasgow, Chair: Roundtable
The Brook Symposium was designed as a showcase for major projects in

historical lexicography, clustering round the central OED presentation. Inevi-
tably in this day and age, it was also a showcase for the effect technology has
had on developing lexical resources. Attendance was gratifyingly high through-
out we had been worried about how much appeal such a session would have,
especially as an optional extra at the end of a busy conference. The level of
interest and awareness was also gratifying, with an exchange of questions and
opinions throughout, as well as at the final round-table.
In the context of several presentations, there was a good deal of discus-
sion of such things as antedatings and the dating of texts, especially the
238 CHRISTIAN KAY
different approaches of OED2 and MED, and the implications of these for
users. The revision of definitions, sometimes into more politically correct
formats, provoked interest, as did the problems associated with defining and
accommodating World Englishes, and the role of the parent language. There
were some fascinating examples of how common words in South African
English have diverged from what we may still, if somewhat apologetically,
call standard English; equally intriguing, and without any real solution, was
the debate on when words may be classified as regional rather than standard.
Following on from this came a discussion of ephemerality; how decisions are
made about the likely staying-power of words, especially those seized on by
the media as words of the moment. An equally difficult question is how much
of the burgeoning technical and scientific vocabulary should be included.
A further theme was the issues raised by electronic publication. There are
huge advantages for the user in the flexibility of the information offered, and
for the editors in being able to update and expand the work much more easily
than in the past. As MEC demonstrated, the internet offers the opportunity to
develop beyond the electronic dictionary into a complete scholarly resource:
the virtual library. However, several speakers pointed out that there are also
issues of control and copyright which are far from being resolved, and that
quality control of the internet remains a serious problem. There is also the
problem that resources which are easily available to us may become increas-
ingly, rather than less, beyond the reach of those in less-privileged parts of the
world.
The availability of the thesaurus-style resources, TOE and HTE, actually
or potentially, also inspired a good deal of interest, as did plans by Jane
Roberts and Louise Sylvester for a Thesaurus of Middle English. Delegates
seemed intrigued by the amount of material available in HTE for certain
categories of meaning. Many had used TOE in their research, and expressed
the wish that it might be available electronically before too long. Interest was
also expressed in the idea of generating a dictionary from TOE.
The new millennium is clearly going to be an exciting one for historical
lexicography. Many projects will reach completion and be presented in new
and challenging formats. Others are only just beginning.
BROOK SYMPOSIUM 239
References
Bosworth, J. & T. N. Toller. 1898. An Anglo-Saxon Dictionary. London.

Burnley, David & Matsuji Tajima. 1994. The Language of Middle English Literature
(=Annotated Bibliographies of Old and Middle English 1). Cambridge: D. S. Brewer.
Cameron, Angus, Ashley Crandall Amos & Allison Kingsmill. 1983. Old English Word
Studies: A Preliminary Word and Author Index. Toronto Old English Series.
Cameron, A. F., A. C. Amos, A. diP. Healey, Sharon Butler, Joan Holland, David
McDougall & Ian McDougall, eds. 1986. Dictionary of Old English (DOE). Toronto:
PIMS. (Published for the Dictionary of Old English Project, Centre for Medieval
Studies, University of Toronto.)
Clark Hall, J. R. & H. D. Meritt. 1960. A Concise Anglo-Saxon Dictionary. Fourth edition.
Cambridge: Cambridge University Press.
Healey, A. diPaolo. 1997. The mapping of meaning and Torontos Dictionary of Old
English. To Explain the Present. Studies in the Changing English Language in Honour
of Matti Rissanen ed. by Tertuu Nevalainen & Leena Kahlas-Tarkka. (=Mmoires de la
Socit Nophilologique de Helsinki 52), 3549. Helsinki: Socit Nophilologique.
Kay, Christian & Iren Wotherspoon. 1997. Historical Thesaurus of English. Dictionar-
ies of Medieval Germanic Languages, ed. by K.H. Van Dalen-Oskam, K.A.C. Depuydt,
W.J.J. Pijnenburg and T.H. Schoonheim. (=International Medieval Research 2), 4954.
Turnhout: Brepols.
Kay, Christian. 1998. Historical Thesaurus of English: Progress Report. Henry Sweet
Society Bulletin. May. 9599.
Oxford English Dictionary Additions Series, vols 1 & 2. 1993, ed. by John Simpson &
Edmund Weiner. Vol 3. 1997, ed. by M. Proffitt. Oxford: Oxford University Press.
(=Kings College London Medieval Studies 11: 2 vols). London: Kings College Centre
for Late Antique and Medieval Studies.
Roberts, Jane & Louise Sylvester. 1997. A Thesaurus of Middle English. Dictionaries of
Medieval Germanic Languages: A Survey of Current Lexicographical Projects, ed. by
K. H. van Dalen-Oskam, K. A. C. Depuydt, W. J. J. Pijnenburg & T. H. Schoonheim,
4145. Turnhout: Brepols.
Silva, Penny, Wendy Dore, Dorothea Mantzel, Colin Muller & Madeleine Wright, eds.
1996. A Dictionary of South African English on Historical Principles. Oxford: Oxford
University Press, in association with the Dictionary Unit for South African English,
Grahamstown.
Simpson, John, & Edmund Weiner, eds. 1989. The Oxford English Dictionary (2nd ed.: 20
vols). Oxford: Oxford University Press.
Strite, Vic. 1989. Old English Semantic-Field Studies. New York: Peter Lang.
Sylvester, Louise & Jane Roberts. 1996. Middle English Word Studies. Medieval English
Studies Newsletter 34 . 811.
Toller, T. N. 1921. An Anglo-Saxon Dictionary: Supplement. London.
Author and Subject Index
A Backhouse, A.E. 64
accord 167, 172-174 bagpipe 83
actress 33 ballum-rancum 83
affixation 88, 179-180, 183 Bang-up Dictionary 74, 75, 84
-aholic 28 Barber, C. 35, 98, 138
Aitchison, J. 2, 134 Barnhart, C.L. 28
Algeo, J. 182 Baron, D. 17
Allerton, D. 17 Bauer, L. 181, 184, 201, 202
alsatian 7 bear 4
ambiguity 96, 128 Berlin, B. 2
amphibious 83 Biber, D. 140, 182, 189, 201
analogy 19-21, 32, 33 Biggam, C. 64
anger 58-60, 61 bikini 28, 33
animal 3-4, 7-8, 11, 12, 17 billy-goat 4
animals, terms for 4-5, 6-8, 16, 17 bird 7
Anshen, F. 185, 202 bitch 4
ant 7 blind 6, 9
anthropology 54, 55-56 boar 4
antonymy 2, 9 bord 109
Archer corpus 179-207 borrowings see loan-words
architecture 103-125 Bosworth-Toller 165, 166, 168, 169,
Ardener, E. 21 175, 176, 234-235
Aronoff, M. 184, 185, 202 Boyle, R. 97-98
artificial intelligence 46, 54 bred 109
asphalt 21-22, 27, 32 breden 109-110
assent 167, 170-172 British National Corpus 137
Athanasiadou, A. 64 Brook, G.L. vii, ix-xi
Atkinson, D. 189 brother 5
aunt 5, 10 Brown Corpus 128-156
authentication 76-79 buck-bean 26, 27
Awdeley, J. 70, 78 Buckingers boot 84
Ayto, J. 17 building 103-125
bull 4-5
B Bullein, W. 210, 214, 218
Baayen, H. 179-180, 183, 185, 201, 202 Burnley, D. 236
Babel, tower of 80 burstow 28
Bacharach 28 butterfly 7
242 Author and Subject Index
C componential analysis 56, 65

Caius, J. 210, 214, 218 components 57, 62; semantic ~ 65, 129
calf 4 compounds 29, 32, 154
calques 32, 94, 97 Comrie, B. 183
cant & cant dictionaries 69-86 conceit 41-45
carcass 3-4 condescent 167
Carew, B.-M. 70 conscience 44-45
cat 4, 7-8 consense 158, 163-164, 167
categories 38, 55, 59-61, 64, 65, 129, consent 157-178
130, 131, 165 Considine, J. 65
categorization 55, 56, 57 context 48, 64, 130, 134, 135, 143
cattle 16 conversion 94, 97
Cawdrey, R. 70 co-ordination 234
chance-medley 22 core: ~ meanings 37-38, 59, 62; ~
change: lexical ~ 1, 2; linguistic ~ 1-18, members 63; ~ senses 61, 62
20, 32; semantic ~ 2, 35-52, 62 corpora 137, 185, 202, 203, 209, 216,
Chapman, D. 198 223; historical ~ 185; Archer corpus
chick 4 179-207; British National Corpus
chicken 4-5 137; Brown Corpus 128-156; Helsinki
child 5 Corpus 36, 38, 128-156, 179-207,
choice, lexical 87-101 210; Lancaster-Oslo/Bergen Corpus
Chomsky, N. 3, 10, 16, 183-184 128-156; London-Lund Corpus 137;
citations 59, 230 Michigan Corpus 36
clam 110 corpse 3-4
Clark-Hall, J.R. 118, 119, 121, 234-235 Cotgrave, R. 214, 217, 218
classical 83 cousin 5
classification 57, 65, 165, 234-235, 236- cow 4-5
237 Cowie, C. 203
Clever, W. 210, 211, 212, 214, 215, crawl 8
217, 219, 220, 221, 222-223 creature 7
clipper 23 Crostwick 28
clusters, semantic 63 Cruse, D. 1-4, 6-8, 10, 11-12, 14, 16
Coates, R. 28, 29, 30 crypel 113-114
cock 4-5 cub 4
cod 7 cunning 42-45
cognition 1, 9-12, 13; nouns of ~ 35-52
cognitive: ~ linguistics vii, 54, 56; ~ D
science 1, 61; ~ semantics 53, 54, 56, afsumnes 167, 169
57, 61, 62, 64 Dalton-Puffer, C. 180, 184, 185
coining 97 databanks 59
Coleman, J. 65 daughter 5
Coles, E. 71, 72 deaf 6, 9
collie 7 Deane, P. 63
colour terms 2, 16 decomposition 62; lexical ~ 56
complementaries 1, 8-9 deer 21
definition 57, 59, 60, 62, 64, 131, 157- resources vii, 229-239
178, 230, 238; sense division 58 elephant 7
Defoe, D. 73 Elzevir type 23
Dekeyser, X. 37, 40 emotion 58-59, 64, 127-156
Dekker, T. 70, 71, 72, 74, 76-77, 78, 79, Empson, W. 35
80, 81-82, 84 encyclopaedic meaning 63
el 109 engine 43, 45, 46
derivation 4, 29, 154, 184, 201-202, 203 entries 83
derivational gaps 16 -ess 33
diachronic: ~ corpora 185; ~ lexicogra- est 167, 168
phy 229-239; ~ lexicology 64; ~ etymology 23, 58-59, 65, 154, 175, 230;
linguistics vii, 1, 2, 55, 59; ~ produc- folk ~ 19-34
tivity 184; ~ semantics vii, 53-68; ~ ewe 4
sociolinguistics 210
dictionaries 54, 56, 57, 65, 69-86, 157- F
178; Dictionary of Old English 114, fall 15
121; Dictionary of South African family resemblance structures 58, 135
English 230-232; dictionary users 59; fancy 43, 45
cant & slang dictionaries 69-86; father 5, 33
Bosworth-Toller 165, 166, 168, 169, feel 6
175, 176, 234-235; Clark-Hall 118, female 26-27, 32
119, 121, 234-235; Middle English fenester 113, 114
Dictionary 157-178, 232-234, 235- fennc 120
236, 238; synonym dictionaries 62 feorstuu 110-111
directionality 35, 37, 40, 46, 50 feverfew 25, 27
dis- 209-227 fieldwork 68, 77, 82, 84
displeasure 58, 60-61 figurative language see metaphor
dissatisfaction 60 Fincke, S. 203
dog 4-5, 6-8 Finegan, E. 140, 189, 201
Drant, T. 219 first 117
dumb 6, 9 Fischer, O. 32
durstod 111-112 fish 7
durstodl 111-112 Fisiak, J. 121
duru 111 Fleischman, S. 181, 202
gedyre 111 Florio, J. 214, 217, 218
fly 8
E foal 4
E., B. 71, 83 folk: ~ etymology 19-34; ~ linguistics
eagduru 113 56; ~ taxonomy 7-8
eagle 7 foreduru 116
eagyrel 113 form abstraction 28
Early Modern English 35-52, 136, 138, formal semantics 63
141, 143-147, 152-153, 179-207, 209- fotstan 107
227 Francis, W. 154
Electronic: ~ publication 55, 238; ~ freebooter 24
fuzziness 36, 38, 51, 61, 130, 131, 133, 207, 210
135, 143, 148-151 hen 4-5
heorr 112
G herring-sue 24
Galileo, G. 87-101 hierarchies 1, 2, 6, 177, 231, 234
gammon 28 hill 6
gander 4 hillock 6
Garner, B.A. 182 historical: ~ corpora 185; ~ lexicography
gate 28 229-239; ~ lexicology 64; ~ linguis-
geat 111 tics vii, 1, 2, 55, 59; ~ productivity
Geeraerts, D. 2, 37, 54, 55, 62, 65, 130, 184; ~ semantics vii, 53-68; ~
134 sociolinguistics 210
genius 45, 48 Historical Thesaurus of English 55, 56,
goat 4 59, 67-68, 166-168, 169, 176, 234-
goose 4 235, 236-237, 238
Grlach, M. 35, 128, 138, 199 hit 21
gosling 4 Hock, H. H. 19
Gotti, M. 83, 98 homonymy 58, 60
grammar: synchronic ~ 202; transforma- hop 8
tional-generative ~ 55, 56; universal ~ Hopper, P. 47, 203
68 horn 117-119
Greene, R. 70, 71, 72, 73-74, 77 horngeap 117-118
Grose, F. 70, 71, 73, 74, 75, 79, 81, 83, hornpic 118
84 hornreced 117-119
growth, lexical 87, 98 hornsele 119
grund 104-106 horngestreon 117-118
grundstan 106-107 horse 4
grundweall 104-106 Hotten, J. C. 75, 77, 83
Grundy, L. 103 hrefan 117
hrier 13
H hrof 117
Halliday, M.A.K. 181, 189-191, 200 hrofstan 120
hamster 57, 64 hrycg 117, 120
Hanks, P. 47 Hughes, G. 83
Hargevik, S. 28, 30 Hllen, W. 65
Harman, T. 70, 73, 77, 78, 80, 81 human 3-4
Harris, M. 83 Hurford, J. 63
harvest 14-15 hyponymy 2, 63
Harvey, G. 210, 219, 221-222 hyrdel 110
he 13-14 hyrnstan 108, 118
Head, R. 70, 77, 83
hear 6 I
Heasley, B. 63 imagination 45
Hell upon Earth 77-78 indeterminacy 53, 61
Helsinki Corpus 36, 38, 128-156, 179- ingenuity 43-45
ingeny 43-45 L
inland sea 6 L., G. 75-76
insect 7 lake 6
intellect 44-45 Lakoff, G. 1
intelligence 43-45 lamb 4
interdisciplinary semantics 64 lame 6, 9
intuition 62 Lancaster-Oslo/Bergen Corpus 128-156
invisible hand theory 31 Lass, R. 14
inwit 45 Latroon, M. 70
ire 58, 60 Lehrer, A. 3, 16, 36, 47
Lent 14-15
J Lewis, C.S. 35, 48, 127, 129, 136, 149
jargon 83 lexical: ~ change 1, 2; ~ choice 87-101;
Jespersen, O. 214 ~ configurations 1-18; ~ decomposi-
Jewess 33 tion 56; ~ feature analysis 16, 53, 55,
Johnson, M. 1 57; ~ gaps 1-18; ~ growth 87, 98; ~
Johnson, S. 130, 131, 132 item 64; ~ semantics 54, 62, 64, 65; ~
Jones, J. 210, 218 studies vii, 210, 235-236
Joseph, B. 180 lexicalization 2
jump 8 lexicography vii, 53-68, 69-86, 229-239;
diachronic ~ 229-239
K lexicology vii, 54, 63; diachronic ~ 64;
Kastovsky, D. 192, 201, 202, 203 synchronic ~ 64
Kay, C. 55, 65, 103 Lieber, R. 202
Kay, P. 2 linguistic change 1-18, 20, 32
Keller, R. 31 linguistics; cognitive ~ vii, 54, 56;
Kellerman, G. 15 diachronic ~ vii, 1, 2, 55, 59; folk ~
key 23 56; observational ~ 68; synchronic ~
kid 4 55, 59; theoretical ~ vii
-kini 28 lion 4, 17
kinship terms 5, 10, 12, 56 lioness 4, 17, 33
kitten 4 Lipka, L. 36
Kjellmer, G. 128 loan-words 15, 32, 36, 44-45, 48, 65, 88,
Kleparski, G. 64 98, 113-114, 115, 158, 159, 175, 184,
knowledge 43, 45 202, 211
knowledge 64; cultural ~ 21-22, 26, 32, loan-translations see calques
65; meta-linguistic ~ 2; terms of ~ 47 locomotion, verbs for 8, 10-11, 12
Koivisto-Alanko, P. 120, 134 London-Lund Corpus 137
Koptjevskaja-Tamm, M. 183, 192 love 127-156
Kvecses, Z. 145, 146, 149 Lovric, M. 83
Kronenfeld, D. B. 31 Lyons, J. 2, 63, 154
Kryk-Kastovsky, B. 192, 201
Kyt, M. 36, 138, 154 M
Kucera, H. 154 male 26-27
mammal 17
man 21, 26-27 mute 6

Marchand, H. 184, 200
mare 4 N
Marlowe, C. 210, 219, 222-223 nanny-goat 4
Martin, J.R. 181, 189-191 Nashe, T. 210, 217, 222
mathematics 54 necessary conditions 61
Matsell, G. 83 negress 33
McConchie, R.W. 210, 214, 224 neologism 94, 95, 97, 99, 100, 179-207,
McMahon, A. 19, 21 210, 230
McSparran, F. xiv nephew 5, 10
meaning: encyclopaedic ~ 63; core ~ 37- Nevalainen, T. 184, 199, 219
38, 59, 62. See also semantic New Canting Dictionary 72
MED see Middle English Dictionary new words see neologism
members, core 63 Newton, I. 97, 190, 193
memorious 17 niece 5, 10
memory 45 -nomics 28
Merlan, F. 203 nominalization 91, 179-207
meronymy 6 norportic 116
metalanguage 59 numb 6
meta-linguistic knowledge 2 Nyman, M. 31
metaphor 1, 46, 49, 51, 54, 95-97, 99
metonymy 42-43, 46, 47, 49-51, 59, 62, O
97 obscenity 74-76, 82, 83, 84
Michigan Corpus 36 observational linguistics 68
Microfiche Concordance to Old English OED 13, 16, 17, 21, 23, 28, 32, 33, 36,
111, 115 40, 41, 42, 44, 55, 57, 58, 59, 60, 62,
Middle English 13-15, 35-52, 60, 157- 65, 66, 83, 92, 93, 95, 96, 97, 99, 100,
178, 184, 215, 224, 232-234, 235- 117, 122, 130, 131-133, 139, 154,
236; Middle English Compendium 157-178, 184, 185, 198, 201, 202,
232-234, 238; Middle English 203, 209-227, 229, 230-232, 235,
Dictionary 157-178, 232-234, 235- 236, 237, 238
236, 238; Middle English Word oferdyre 112
Studies 235-236; Thesaurus of Middle ofergeweorc 117
English 177, 236, 238 oferhrefan 117
mind 44-45 oferslege 112-113
Mithun, M. 203 Old English x, 14-15, 55, 60, 65, 103-
Modern English 14-15, 184. See also 125, 165, 166, 168, 169, 175, 176,
Present Day English 234-235; Bosworth-Toller 165, 166,
monokini 28, 33 168, 169, 175, 176, 234-235; Clark-
Morrissey, M. 15 Hall, 118, 119, 121, 234-235;
mother 5, 33 Dictionary of Old English 114, 121;
mound 6 Microfiche Concordance to Old
mountain 6 English 111, 115; Thesaurus of Old
move 8 English 55, 103, 158, 165, 166-167,
Munsell 55 168, 169, 176, 234-235, 236-237, 238
onomasiology 55, 57, 58, 65, 158 65, 177

opacity 29-32 psycholinguistics 2
origins: ~ of cant 79- 81; ~ of language psychological salience 1, 10-12, 17, 56,
68 60
stman, J.-O. 128 psychology 54, 64
outwit 45 puddle 6
Oxford English Dictionary see OED puppy 4
P Q
Palmer, A. S. 20, 23, 24-25 quagmire 23
Palmer, F.R. 63 quay 23
parent 5, 10, 13, 16 -quel 29-29
Parker, G. 72-73
participant analysis 143, 149-151 R
passion 58, 60 R., S. 76-77, 84
peacock 17 Radden, G. 15
peahen 17 rage 58, 60, 61, 63
perceptual salience 1, 11-12 ram 4
peripheral meanings 37-38, 61, 62, 63, Reibel, D. 80
146-147, 151 Renouf, A. 179-180, 183, 185, 201, 202
personification 47 Riddle, E. 183
Phaer, T. 219 riscen 120
philology vii, 54 Rissanen, M 137
philosophy 54 Roberts, J. 55, 103, 235-236, 236-237
physiology 64 robin 7
pig 4 Rogets Thesaurus 63, 166
piglet 4 Romaine, S. 182, 183
plant 3-4 rooster 4-5
poetess 33 Rosch, E. 11-12
polysemy 14, 35, 36, 37, 58, 59, 63 Roscommon, W. 84
pond 6 rother 13
portic 115-116 ruff-peck 84
poultry 16 run n 64
pragmatics 54, 201 run vb 8, 11
prefixation 209-227 Rundblad, G. 33
Present Day English 136, 138, 140, 145-
147, 150, 152-153. See also Modern S
English Saaed, J. 55
productivity 202, 209; diachronic ~ 184 salience; perceptual ~ 1, 11-12; psycho-
proportional series 1-18 logical ~ 1, 10-12, 17, 56, 60
prototype theory vii, 2, 11-12, 15, 35-38, Salmon, V. 182
44, 46-47, 49-51, 57, 58, 59, 60, 61, Salusbury, T. 87-101
62, 128, 129, 133-135, 136, 138, 140, Saman 80
143, 151-153, 165, 168, 173-174, 175 Samuels, M. 14, 55, 65, 234, 237
prototypicality 1, 2, 11-12, 17, 53, 60, Saussure, F. de 30
Schfer, J. 217 spaniel 6-8

Schmid, H.-J. 2, 15 speak 6
Schwyter, J.R. 128 specialization 88, 93, 97
science 43, 45 Spenser, E. 210, 219, 220-221
science, cognitive 1, 61 Spevack, M. 154
scindel 120 Spivey, N. 200
scingal 120 spring 14-15
Scoundrels Dictionary 71, 78 staca 107
seasons, terms for 13-15 stnen 108
see 6, 9 stallion 4
semantic: ~ change 2, 35-52, 62; ~ stangefog 108
clusters 63; ~ components 16, 129; ~ stanlesung 109
feature analysis 16, 53, 55, 57; ~ stangetimbre 108
fields 16, 35-52, 55, 65, 127-156, 176, stantorr 114
236; ~ fractals 48-51; ~ relationships starling 7
55, 58, 63 stereotype 57
semantics 20, 53-68; cognitive ~ 53, 54, stipel 114
56, 57, 61, 62, 64; diachronic ~ vii, stoccen 110
53-68; formal ~ 63; interdisciplinary ~ Strite, V. 236
64; lexical ~ 54, 62, 64, 65; truth- structuralism 55, 179
conditional ~ 55. See also prototype Stubbs, M. 189, 190
theory subjectification 36-37, 42, 47, 48, 50
semasiology 2, 54-55, 57, 65, 158 subordination 6-8, 59, 234
sense 44-45 suportic 116
sense: ~ division 58; core ~ 61, 62 suffixation 94
sets 1 summer 14-15
Shakespeare, W. x, 128-156, 211, 217, superordination 6-8, 57, 59
219, 222, 224 Svartvik, J. 137
Sharp, J. 70 Sweetser, E. 35, 37, 40, 50
she 13-14 swim 8
sheep 4-5 Sylvester, J. 217
Shirley, J. 81 synaesthetic adjectives 2
sibling 5, 10, 13 synchronic: ~ grammar 202; ~
Sidney, P. 210, 219, 220, 221, 223 lexicology 64; ~ linguistics 55, 59
sighted 17 synonym dictionaries 62
sing 21 synonymy 2, 36, 62-63, 96, 97, 158, 164
sister 5 syntax 91-92, 94, 201
slang 69-86, 231; ~ dictionaries 69-86
smell 6, 9 T
Smith, A. 72, 79 tabby-cat 4, 16
Smith, J. 154 Tajima, M. 236
sociolinguistics, historical 210 taste 6
son 5 taste terms 64
South African English 230-232 taxonomy 6-9, 17, 55, 165, 236; folk ~
sow 4 7-8
Taylor, J. 1, 37, 55, 130, 134 uncle 5, 10

terminology 15, 32, 83, 87-101, 103- underdifferentiation 4-5
125, 179-207, 209-227, 238 understanding 44-45
c 117 Ungerer, F. 2, 15
ctigel 120 unidirectionality 47
theoretical linguistics vii universal grammar 68
thesauruses 57, 58, 61, 238; Historical
Thesaurus of English 55, 56, 59, 67- V
68, 166-168, 169, 176, 234-235, 236- Van Dalen-Oskam, K.H. 55
237, 238; Thesaurus of Old English vernacularization 209-227
55, 103, 158, 165, 166-167, 168, 169, vulgar 83
176, 234-235, 236-237, 238; Thesau-
rus of Middle English 177, 236, 238; W
Rogets Thesaurus 63, 166 wpnmann 26, 33
they 14 walk 6, 8, 11
Thompson, S. 181, 203 Warren, B. 46
thrush 25 watel 110
yrel 113 Watergate 28
tigel 120 Wells, R. 203
tigress 33 wer 33
Tissari, H. 154 wif 26, 33
tom-cat 4 wifmann 26, 33
tommy 25 Wild, J. 70
torr 114 Williams, J. 2
transformational-generative grammar 55, windelstan 114-115
56 windung 110
translation 87-101 Winer, L. 22
Traugott, E. C. 35, 37, 40-41, 47, 50 Winstanley, W. 81
treowen 109 winter 14-15
Trier, J. 47 wit 35-52
trout 7 woman 26-27, 64
truth-conditional semantics 55 Wotherspoon, I. 55
Tufts, H. 81 wrath 58, 60, 61
tur 114
Turner, M. 1 Y
Turner, W. 210, 218 Yokels Preceptor 73
twiddle-diddles 84
typicality 134 Z
Zgusta, L. 209
U Ziegler, J. 3, 5, 16
Ullmann, S. 47, 154
Old English ge- is ignored for the purposes of alphabetization. is alphabetized as ae, as
th. ~ denotes the repetition of the head-word. Italicized entries are cited words, or the titles
of books or electronic resources.
CURRENT ISSUES IN LINGUISTIC THEORY
E. F. K. Koerner, Editor
Department of Linguistics, University of Ottawa
OTTAWA, Canada K1N 6N5
koerner@uottawa.ca
The Current Issues in Linguistic Theory (CILT) series is a theory-oriented series which welcomes
contributions from scholars who have significant proposals to make towards the advancement of
our understanding of language, its structure, functioning and development. CILT has been
established in order to provide a forum for the presentation and discussion of linguistic opinions
of scholars who do not necessarily accept the prevailing mode of thought in linguistic science. It
offers an alternative outlet for meaningful contributions to the current linguistic debate, and
furnishes the diversity of opinion which a healthy discipline must have. In this series the following
volumes have been published thus far or are scheduled for publication:
1. KOERNER, Konrad (ed.): The Transformational-Generative Paradigm and Modern Linguistic

Theory. 1975.
2. WEIDERT, Alfons: Componential Analysis of Lushai Phonology. 1975.
3. MAHER, J. Peter: Papers on Language Theory and History I: Creation and Tradition in
Language. Foreword by Raimo Anttila. 1979.
4. HOPPER, Paul J. (ed.): Studies in Descriptive and Historical Linguistics. Festschrift for Winfred
P. Lehmann. 1977.
5. ITKONEN, Esa: Grammatical Theory and Metascience: A critical investigation into the
methodological and philosophical foundations of autonomous linguistics. 1978.
6. ANTTILA, Raimo: Historical and Comparative Linguistics. 1989.
7. MEISEL, Jrgen M. & Martin D. PAM (eds): Linear Order and Generative Theory. 1979.
8. WILBUR, Terence H.: Prolegomena to a Grammar of Basque. 1979.
9. HOLLIEN, Harry & Patricia (eds): Current Issues in the Phonetic Sciences. Proceedings of the
IPS-77 Congress, Miami Beach, Florida, 17-19 December 1977. 1979.
10. PRIDEAUX, Gary D. (ed.): Perspectives in Experimental Linguistics. Papers from the Univer-
sity of Alberta Conference on Experimental Linguistics, Edmonton, 13-14 Oct. 1978. 1979.
11. BROGYANYI, Bela (ed.): Studies in Diachronic, Synchronic, and Typological Linguistics:
Festschrift for Oswald Szemrenyi on the Occasion of his 65th Birthday. 1979.
12. FISIAK, Jacek (ed.): Theoretical Issues in Contrastive Linguistics. 1981. Out of print
13. MAHER, J. Peter, Allan R. BOMHARD & Konrad KOERNER (eds): Papers from the Third
International Conference on Historical Linguistics, Hamburg, August 22-26 1977. 1982.
14. TRAUGOTT, Elizabeth C., Rebecca LaBRUM & Susan SHEPHERD (eds): Papers from the
Fourth International Conference on Historical Linguistics, Stanford, March 26-30 1979. 1980.
15. ANDERSON, John (ed.): Language Form and Linguistic Variation. Papers dedicated to Angus
McIntosh. 1982.
16. ARBEITMAN, Yol L. & Allan R. BOMHARD (eds): Bono Homini Donum: Essays in
Historical Linguistics, in Memory of J.Alexander Kerns. 1981.
17. LIEB, Hans-Heinrich: Integrational Linguistics. 6 volumes. Vol. II-VI n.y.p. 1984/93.
18. IZZO, Herbert J. (ed.): Italic and Romance. Linguistic Studies in Honor of Ernst Pulgram.
1980.
19. RAMAT, Paolo et al. (eds): Linguistic Reconstruction and Indo-European Syntax. Proceedings
of the Colloquium of the Indogermanischhe Gesellschaft. University of Pavia, 6-7 September
1979. 1980.
20. NORRICK, Neal R.: Semiotic Principles in Semantic Theory. 1981.
21. AHLQVIST, Anders (ed.): Papers from the Fifth International Conference on Historical
Linguistics, Galway, April 6-10 1981. 1982.
22. UNTERMANN, Jrgen & Bela BROGYANYI (eds): Das Germanische und die Rekonstruktion
der Indogermanischen Grundsprache. Akten des Freiburger Kolloquiums der Indogermanischen
Gesellschaft, Freiburg, 26-27 Februar 1981. 1984.
23. DANIELSEN, Niels: Papers in Theoretical Linguistics. Edited by Per Baerentzen. 1992.
24. LEHMANN, Winfred P. & Yakov MALKIEL (eds): Perspectives on Historical Linguistics.
Papers from a conference held at the meeting of the Language Theory Division, Modern
Language Assn., San Francisco, 27-30 December 1979. 1982.
25. ANDERSEN, Paul Kent: Word Order Typology and Comparative Constructions. 1983.
26. BALDI, Philip (ed.): Papers from the XIIth Linguistic Symposium on Romance Languages,
Univ. Park, April 1-3, 1982. 1984.
27. BOMHARD, Alan R.: Toward Proto-Nostratic. A New Approach to the Comparison of Proto-
Indo-European and Proto-Afroasiatic. Foreword by Paul J. Hopper. 1984.
28. BYNON, James (ed.): Current Progress in Afro-Asiatic Linguistics: Papers of the Third
International Hamito-Semitic Congress, London, 1978. 1984.
29. PAPROTT, Wolf & Ren DIRVEN (eds): The Ubiquity of Metaphor: Metaphor in language
and thought. 1985 (publ. 1986).
30. HALL, Robert A. Jr.: Proto-Romance Morphology. = Comparative Romance Grammar, vol.
III. 1984.
31. GUILLAUME, Gustave: Foundations for a Science of Language.
32. COPELAND, James E. (ed.): New Directions in Linguistics and Semiotics. Co-edition with
Rice University Press who hold exclusive rights for US and Canada. 1984.
33. VERSTEEGH, Kees: Pidginization and Creolization. The Case of Arabic. 1984.
34. FISIAK, Jacek (ed.): Papers from the VIth International Conference on Historical Linguistics,
Poznan, 22-26 August. 1983. 1985.
35. COLLINGE, N.E.: The Laws of Indo-European. 1985.
36. KING, Larry D. & Catherine A. MALEY (eds): Selected papers from the XIIIth Linguistic
Symposium on Romance Languages, Chapel Hill, N.C., 24-26 March 1983. 1985.
37. GRIFFEN, T.D.: Aspects of Dynamic Phonology. 1985.
38. BROGYANYI, Bela & Thomas KRMMELBEIN (eds): Germanic Dialects:Linguistic and
Philological Investigations. 1986.
39. BENSON, James D., Michael J. CUMMINGS, & William S. GREAVES (eds): Linguistics in a
Systemic Perspective. 1988.
40. FRIES, Peter Howard (ed.) in collaboration with Nancy M. Fries: Toward an Understanding
of Language: Charles C. Fries in Perspective. 1985.
41. EATON, Roger, et al. (eds): Papers from the 4th International Conference on English Historical
Linguistics, April 10-13, 1985. 1985.
42. MAKKAI, Adam & Alan K. MELBY (eds): Linguistics and Philosophy. Festschrift for Rulon S.
Wells. 1985 (publ. 1986).
43. AKAMATSU, Tsutomu: The Theory of Neutralization and the Archiphoneme in Functional
Phonology. 1988.
44. JUNGRAITHMAYR, Herrmann & Walter W. MUELLER (eds): Proceedings of the Fourth
International Hamito-Semitic Congress. 1987.
45. KOOPMAN, W.F., F.C. Van der LEEK , O. FISCHER & R. EATON (eds): Explanation and
Linguistic Change. 1986
46. PRIDEAUX, Gary D. & William J. BAKER: Strategies and Structures: The processing of relative
clauses. 1987.
47. LEHMANN, Winfred P. (ed.): Language Typology 1985. Papers from the Linguistic Typology
Symposium, Moscow, 9-13 Dec. 1985. 1986.
48. RAMAT, Anna G., Onofrio CARRUBA and Giuliano BERNINI (eds): Papers from the 7th
International Conference on Historical Linguistics. 1987.
49. WAUGH, Linda R. and Stephen RUDY (eds): New Vistas in Grammar: Invariance and
Variation. Proceedings of the Second International Roman Jakobson Conference, New York
University, Nov.5-8, 1985. 1991.
50. RUDZKA-OSTYN, Brygida (ed.): Topics in Cognitive Linguistics. 1988.
51. CHATTERJEE, Ranjit: Aspect and Meaning in Slavic and Indic. With a foreword by Paul
Friedrich. 1989.
52. FASOLD, Ralph W. & Deborah SCHIFFRIN (eds): Language Change and Variation. 1989.
53. SANKOFF, David: Diversity and Diachrony. 1986.
54. WEIDERT, Alfons: Tibeto-Burman Tonology. A comparative analysis. 1987
55. HALL, Robert A. Jr.: Linguistics and Pseudo-Linguistics. 1987.
56. HOCKETT, Charles F.: Refurbishing our Foundations. Elementary linguistics from an ad-
vanced point of view. 1987.
57. BUBENIK, Vt: Hellenistic and Roman Greece as a Sociolinguistic Area. 1989.
58. ARBEITMAN, Yol. L. (ed.): Fucus: A Semitic/Afrasian Gathering in Remembrance of Albert
Ehrman. 1988.
59. VAN VOORST, Jan: Event Structure. 1988.
60. KIRSCHNER, Carl & Janet DECESARIS (eds): Studies in Romance Linguistics. Selected
Proceedings from the XVII Linguistic Symposium on Romance Languages. 1989.
61. CORRIGAN, Roberta L., Fred ECKMAN & Michael NOONAN (eds): Linguistic Categoriza-
tion. Proceedings of an International Symposium in Milwaukee, Wisconsin, April 10-11, 1987.
1989.
62. FRAJZYNGIER, Zygmunt (ed.): Current Progress in Chadic Linguistics. Proceedings of the
International Symposium on Chadic Linguistics, Boulder, Colorado, 1-2 May 1987. 1989.
63. EID, Mushira (ed.): Perspectives on Arabic Linguistics I. Papers from the First Annual
Symposium on Arabic Linguistics. 1990.
64. BROGYANYI, Bela (ed.): Prehistory, History and Historiography of Language, Speech, and
Linguistic Theory. Papers in honor of Oswald Szemrenyi I. 1992.
65. ADAMSON, Sylvia, Vivien A. LAW, Nigel VINCENT and Susan WRIGHT (eds): Papers from
the 5th International Conference on English Historical Linguistics. 1990.
66. ANDERSEN, Henning and Konrad KOERNER (eds): Historical Linguistics 1987.Papers from
the 8th International Conference on Historical Linguistics,Lille, August 30-Sept., 1987. 1990.
67. LEHMANN, Winfred P. (ed.): Language Typology 1987. Systematic Balance in Language.
Papers from the Linguistic Typology Symposium, Berkeley, 1-3 Dec 1987. 1990.
68. BALL, Martin, James FIFE, Erich POPPE &Jenny ROWLAND (eds): Celtic Linguistics/
Ieithyddiaeth Geltaidd. Readings in the Brythonic Languages. Festschrift for T. Arwyn Watkins.
1990.
69. WANNER, Dieter and Douglas A. KIBBEE (eds): New Analyses in Romance Linguistics.
Selected papers from the Linguistic Symposium on Romance Languages XVIIII, Urbana-
Champaign, April 7-9, 1988. 1991.
70. JENSEN, John T.: Morphology. Word structure in generative grammar. 1990.
71. OGRADY, William: Categories and Case. The sentence structure of Korean. 1991.
72. EID, Mushira and John MCCARTHY (eds): Perspectives on Arabic Linguistics II. Papers from
the Second Annual Symposium on Arabic Linguistics. 1990.
73. STAMENOV, Maxim (ed.): Current Advances in Semantic Theory. 1991.
74. LAEUFER, Christiane and Terrell A. MORGAN (eds): Theoretical Analyses in Romance
Linguistics. 1991.
75. DROSTE, Flip G. and John E. JOSEPH (eds): Linguistic Theory and Grammatical Description.
Nine Current Approaches. 1991.
76. WICKENS, Mark A.: Grammatical Number in English Nouns. An empirical and theoretical
account. 1992.
77. BOLTZ, William G. and Michael C. SHAPIRO (eds): Studies in the Historical Phonology of
Asian Languages. 1991.
78. KAC, Michael: Grammars and Grammaticality. 1992.
79. ANTONSEN, Elmer H. and Hans Henrich HOCK (eds): STAEF-CRAEFT: Studies in Ger-
manic Linguistics. Select papers from the First and Second Symposium on Germanic Linguistics,
University of Chicago, 24 April 1985, and Univ. of Illinois at Urbana-Champaign, 3-4 Oct. 1986.
1991.
80. COMRIE, Bernard and Mushira EID (eds): Perspectives on Arabic Linguistics III. Papers from
the Third Annual Symposium on Arabic Linguistics. 1991.
81. LEHMANN, Winfred P. and H.J. HEWITT (eds): Language Typology 1988. Typological
Models in the Service of Reconstruction. 1991.
82. VAN VALIN, Robert D. (ed.): Advances in Role and Reference Grammar. 1992.
83. FIFE, James and Erich POPPE (eds): Studies in Brythonic Word Order. 1991.
84. DAVIS, Garry W. and Gregory K. IVERSON (eds): Explanation in Historical Linguistics. 1992.
85. BROSELOW, Ellen, Mushira EID and John McCARTHY (eds): Perspectives on Arabic
Linguistics IV. Papers from the Annual Symposium on Arabic Linguistics. 1992.
86. KESS, Joseph F.: Psycholinguistics. Psychology, linguistics, and the study of natural language.
1992.
87. BROGYANYI, Bela and Reiner LIPP (eds): Historical Philology: Greek, Latin, and Romance.
Papers in honor of Oswald Szemernyi II. 1992.
88. SHIELDS, Kenneth: A History of Indo-European Verb Morphology. 1992.
89. BURRIDGE, Kate: Syntactic Change in Germanic. A study of some aspects of language change
in Germanic with particular reference to Middle Dutch. 1992.
90. KING, Larry D.: The Semantic Structure of Spanish. Meaning and grammatical form. 1992.
91. HIRSCHBHLER, Paul and Konrad KOERNER (eds): Romance Languages and Modern
Linguistic Theory. Selected papers from the XX Linguistic Symposium on Romance
Languages,University of Ottawa, April 10-14, 1990. 1992.
92. POYATOS, Fernando: Paralanguage: A linguistic and interdisciplinary approach to interactive
speech and sounds. 1992.
93. LIPPI-GREEN, Rosina (ed.): Recent Developments in Germanic Linguistics. 1992.
94. HAGGE, Claude: The Language Builder. An essay on the human signature in linguistic
morphogenesis. 1992.
95. MILLER, D. Gary: Complex Verb Formation. 1992.
96. LIEB, Hans-Heinrich (ed.): Prospects for a New Structuralism. 1992.
97. BROGYANYI, Bela & Reiner LIPP (eds): Comparative-Historical Linguistics: Indo-European
and Finno-Ugric. Papers in honor of Oswald Szemernyi III. 1992.
98. EID, Mushira & Gregory K. IVERSON: Principles and Prediction: The analysis of natural
language. 1993.
99. JENSEN, John T.: English Phonology. 1993.
100. MUFWENE, Salikoko S. and Lioba MOSHI (eds): Topics in African Linguistics. Papers from
the XXI Annual Conference on African Linguistics, University of Georgia, April 1990. 1993.
101. EID, Mushira & Clive HOLES (eds): Perspectives on Arabic Linguistics V. Papers from the Fifth
Annual Symposium on Arabic Linguistics. 1993.
102. DAVIS, Philip W. (ed.): Alternative Linguistics. Descriptive and theoretical Modes. 1995.
103. ASHBY, William J., Marianne MITHUN, Giorgio PERISSINOTTO and Eduardo RAPOSO:
Linguistic Perspectives on Romance Languages. Selected papers from the XXI Linguistic Sympo-
sium on Romance Languages, Santa Barbara, February 21-24, 1991. 1993.
104. KURZOV, Helena: From Indo-European to Latin. The evolution of a morphosyntactic type.
1993.
105. HUALDE, Jos Ignacio and Jon ORTIZ DE URBANA (eds): Generative Studies in Basque
Linguistics. 1993.
106. AERTSEN, Henk and Robert J. JEFFERS (eds): Historical Linguistics 1989. Papers from the 9th
International Conference on Historical Linguistics, New Brunswick, 14-18 August 1989. 1993.
107. MARLE, Jaap van (ed.): Historical Linguistics 1991. Papers from the 10th International
Conference on Historical Linguistics, Amsterdam, August 12-16, 1991. 1993.
108. LIEB, Hans-Heinrich: Linguistic Variables. Towards a unified theory of linguistic variation.
1993.
109. PAGLIUCA, William (ed.): Perspectives on Grammaticalization. 1994.
110. SIMONE, Raffaele (ed.): Iconicity in Language. 1995.
111. TOBIN, Yishai: Invariance, Markedness and Distinctive Feature Analysis. A contrastive study of
sign systems in English and Hebrew. 1994.
112. CULIOLI, Antoine: Cognition and Representation in Linguistic Theory. Translated, edited and
introduced by Michel Liddle. 1995.
113. FERNNDEZ, Francisco, Miguel FUSTER and Juan Jose CALVO (eds): English Historical
Linguistics 1992. Papers from the 7th International Conference on English Historical Linguis-
tics, Valencia, 22-26 September 1992.1994.
114. EGLI, U., P. PAUSE, Chr. SCHWARZE, A. von STECHOW, G. WIENOLD (eds): Lexical
Knowledge in the Organisation of Language. 1995.
115. EID, Mushira, Vincente CANTARINO and Keith WALTERS (eds): Perspectives on Arabic
Linguistics. Vol. VI. Papers from the Sixth Annual Symposium on Arabic Linguistics. 1994.
116. MILLER, D. Gary: Ancient Scripts and Phonological Knowledge. 1994.

117. PHILIPPAKI-WARBURTON, I., K. NICOLAIDIS and M. SIFIANOU (eds): Themes in
Greek Linguistics. Papers from the first International Conference on Greek Linguistics, Reading,
September 1993. 1994.
118. HASAN, Ruqaiya and Peter H. FRIES (eds): On Subject and Theme. A discourse functional
perspective. 1995.
119. LIPPI-GREEN, Rosina: Language Ideology and Language Change in Early Modern German. A
sociolinguistic study of the consonantal system of Nuremberg. 1994.
120. STONHAM, John T. : Combinatorial Morphology. 1994.
121. HASAN, Ruqaiya, Carmel CLORAN and David BUTT (eds): Functional Descriptions. Theorie
in practice. 1996.
122. SMITH, John Charles and Martin MAIDEN (eds): Linguistic Theory and the Romance
Languages. 1995.
123. AMASTAE, Jon, Grant GOODALL, Mario MONTALBETTI and Marianne PHINNEY:
Contemporary Research in Romance Linguistics. Papers from the XXII Linguistic Symposium
on Romance Languages, El Paso//Jurez, February 22-24, 1994. 1995.
124. ANDERSEN, Henning: Historical Linguistics 1993. Selected papers from the 11th International
Conference on Historical Linguistics, Los Angeles, 16-20 August 1993. 1995.
125. SINGH, Rajendra (ed.): Towards a Critical Sociolinguistics. 1996.
126. MATRAS, Yaron (ed.): Romani in Contact. The history, structure and sociology of a language.
1995.
127. GUY, Gregory R., Crawford FEAGIN, Deborah SCHIFFRIN and John BAUGH (eds):
Towards a Social Science of Language. Papers in honor of William Labov. Volume 1: Variation
and change in language and society. 1996.
128. GUY, Gregory R., Crawford FEAGIN, Deborah SCHIFFRIN and John BAUGH (eds):
Towards a Social Science of Language. Papers in honor of William Labov. Volume 2: Social
interaction and discourse structures. 1997.
129. LEVIN, Saul: Semitic and Indo-European: The Principal Etymologies. With observations on
Afro-Asiatic. 1995.
130. EID, Mushira (ed.) Perspectives on Arabic Linguistics. Vol. VII. Papers from the Seventh
131. HUALDE, Jose Ignacio, Joseba A. LAKARRA and R.L. Trask (eds): Towards a History of the
Basque Language. 1995.
132. HERSCHENSOHN, Julia: Case Suspension and Binary Complement Structure in French. 1996.
133. ZAGONA, Karen (ed.): Grammatical Theory and Romance Languages. Selected papers from
the 25th Linguistic Symposium on Romance Languages (LSRL XXV) Seattle, 2-4 March 1995.
1996.
134. EID, Mushira (ed.): Perspectives on Arabic Linguistics Vol. VIII. Papers from the Eighth
135. BRITTON Derek (ed.): Papers from the 8th International Conference on English Historical
Linguistics. 1996.
136. MITKOV, Ruslan and Nicolas NICOLOV (eds): Recent Advances in Natural Language
Processing. 1997.
137. LIPPI-GREEN, Rosina and Joseph C. SALMONS (eds): Germanic Linguistics. Syntactic and
diachronic. 1996.
138. SACKMANN, Robin (ed.): Theoretical Linguistics and Grammatical Description. 1996.
139. BLACK, James R. and Virginia MOTAPANYANE (eds): Microparametric Syntax and Dialect
Variation. 1996.
140. BLACK, James R. and Virginia MOTAPANYANE (eds): Clitics, Pronouns and Movement.
1997.
141. EID, Mushira and Dilworth PARKINSON (eds): Perspectives on Arabic Linguistics Vol. IX.
Papers from the Ninth Annual Symposium on Arabic Linguistics, Georgetown University,
Washington D.C., 1995. 1996.
142. JOSEPH, Brian D. and Joseph C. SALMONS (eds): Nostratic. Sifting the evidence. 1998.
143. ATHANASIADOU, Angeliki and Ren DIRVEN (eds): On Conditionals Again. 1997.
144. SINGH, Rajendra (ed): Trubetzkoy's Orphan. Proceedings of the Montral Roundtable
Morphophonology: contemporary responses (Montral, October 1994). 1996.
145. HEWSON, John and Vit BUBENIK: Tense and Aspect in Indo-European Languages. Theory,
typology, diachrony. 1997.
146. HINSKENS, Frans, Roeland VAN HOUT and W. Leo WETZELS (eds): Variation, Change,
and Phonological Theory. 1997.
147. HEWSON, John: The Cognitive System of the French Verb. 1997.
148. WOLF, George and Nigel LOVE (eds): Linguistics Inside Out. Roy Harris and his critics. 1997.
149. HALL, T. Alan: The Phonology of Coronals. 1997.
150. VERSPOOR, Marjolijn, Kee Dong LEE and Eve SWEETSER (eds): Lexical and Syntactical
Constructions and the Construction of Meaning. Proceedings of the Bi-annual ICLA meeting in
Albuquerque, July 1995. 1997.
151. LIEBERT, Wolf-Andreas, Gisela REDEKER and Linda WAUGH (eds): Discourse and Per-
spectives in Cognitive Linguistics. 1997.
152. HIRAGA, Masako, Chris SINHA and Sherman WILCOX (eds): Cultural, Psychological and
Typological Issues in Cognitive Linguistics. 1999.
153. EID, Mushira and Robert R. RATCLIFFE (eds): Perspectives on Arabic Linguistics Vol. X.
Papers from the Tenth Annual Symposium on Arabic Linguistics, Salt Lake City, 1996. 1997.
154. SIMON-VANDENBERGEN, Anne-Marie, Kristin DAVIDSE and Dirk NOL (eds): Recon-
necting Language. Morphology and Syntax in Functional Perspectives. 1997.
155. FORGET, Danielle, Paul HIRSCHBHLER, France MARTINEAU and Mara-Luisa
RIVERO (eds): Negation and Polarity. Syntax and semantics. Selected papers from the Collo-
quium Negation: Syntax and Semantics. Ottawa, 11-13 May 1995. 1997.
156. MATRAS, Yaron, Peter BAKKER and Hristo KYUCHUKOV (eds): The Typology and
Dialectology of Romani. 1997.
157. LEMA, Jos and Esthela TREVIO (eds): Theoretical Analyses on Romance Languages.
Selected papers from the 26th Linguistic Symposium on Romance Languages (LSRL XXVI),
Mexico City, 28-30 March, 1996. 1998.
158. SNCHEZ MACARRO, Antonia and Ronald CARTER (eds): Linguistic Choice across Genres.
Variation in spoken and written English. 1998.
159. JOSEPH, Brian D., Geoffrey C. HORROCKS and Irene PHILIPPAKI-WARBURTON (eds):
Themes in Greek Linguistics II. 1998.
160. SCHWEGLER, Armin, Bernard TRANEL and Myriam URIBE-ETXEBARRIA (eds): Ro-
mance Linguistics: Theoretical Perspectives. Selected papers from the 27th Linguistic Symposium
on Romance Languages (LSRL XXVII), Irvine, 20-22 February, 1997. 1998.
161. SMITH, John Charles and Delia BENTLEY (eds): Historical Linguistics 1995. Volume 1:
Romance and general linguistics. 2000.
162. HOGG, Richard M. and Linda van BERGEN (eds): Historical Linguistics 1995. Volume 2:
Germanic linguistics.Selected papers from the 12th International Conference on Historical
Linguistics, Manchester, August 1995. 1998.
163. LOCKWOOD, David G., Peter H. FRIES and James E. COPELAND (eds): Functional
Approaches to Language, Culture and Cognition. 2000.
164. SCHMID, Monika, Jennifer R. AUSTIN and Dieter STEIN (eds): Historical Linguistics 1997.
Selected papers from the 13th International Conference on Historical Linguistics, Dsseldorf, 10-
17 August 1997. 1998.
165. BUBENK, Vit: A Historical Syntax of Late Middle Indo-Aryan (Apabhramsa). 1998.
166. LEMMENS, Maarten: Lexical Perspectives on Transitivity and Ergativity. Causative construc-
tions in English. 1998.
167. BENMAMOUN, Elabbas, Mushira EID and Niloofar HAERI (eds): Perspectives on Arabic
Linguistics Vol. XI. Papers from the Eleventh Annual Symposium on Arabic Linguistics,
Atlanta, 1997. 1998.
168. RATCLIFFE, Robert R.: The Broken Plural Problem in Arabic and Comparative Semitic.
Allomorphy and analogy in non-concatenative morphology. 1998.
169. GHADESSY, Mohsen (ed.): Text and Context in Functional Linguistics. 1999.
170. LAMB, Sydney M.: Pathways of the Brain. The neurocognitive basis of language. 1999.
171. WEIGAND, Edda (ed.): Contrastive Lexical Semantics. 1998.
172. DIMITROVA-VULCHANOVA, Mila and Lars HELLAN (eds): Topics in South Slavic Syntax
and Semantics. 1999.
173. TREVIO, Esthela and Jos LEMA (eds): Semantic Issues in Romance Syntax. 1999.
174. HALL, T. Alan and Ursula KLEINHENZ (eds): Studies on the Phonological Word. 1999.
175. GIBBS, Ray W. and Gerard J. STEEN (eds): Metaphor in Cognitive Linguistics. Selected papers
from the 5th International Cognitive Linguistics Conference, Amsterdam, 1997. 1999.
176. VAN HOEK, Karen, Andrej KIBRIK and Leo NOORDMAN (eds): Discourse in Cognitive
Linguistics. Selected papers from the International Cognitive Linguistics Conference, Amster-
dam, July 1997. 1999.
177. CUYCKENS, Hubert and Britta ZAWADA (eds): Polysemy in Cognitive Linguistics. Selected
papers from the International Cognitive Linguistics Conference, Amsterdam, 1997. 2001.
178. FOOLEN, Ad and Frederike van der LEEK (eds): Constructions in Cognitive Linguistics.
Selected papers from the Fifth International Cognitive Linguistic Conference, Amsterdam, 1997.
2000.
179. RINI, Joel: Exploring the Role of Morphology in the Evolution of Spanish. 1999.
180. MEREU, Lunella (ed.): Boundaries of Morphology and Syntax. 1999.
181. MOHAMMAD, Mohammad A.: Word Order, Agreement and Pronominalization in Standard
and Palestinian Arabic. 2000.
182. KENESEI, Istvn (ed.): Theoretical Issues in Eastern European Languages. Selected papers from
the Conference on Linguistic Theory in Eastern European Languages (CLITE), Szeged, April
1998. 1999.
183. CONTINI-MORAVA, Ellen and Yishai TOBIN (eds): Between Grammar and Lexicon. 2000.
184. SAGART, Laurent: The Roots of Old Chinese. 1999.
185. AUTHIER, J.-Marc, Barbara E. BULLOCK, Lisa A. REED (eds): Formal Perspectives on
Romance Linguistics. Selected papers from the 28th Linguistic Symposium on Romance Lan-
guages (LSRL XXVIII), University Park, 16-19 April 1998. 1999.
186. MIESKA TOMIC, Olga and Milorad RADOVANOVIC (eds): History and Perspectives of
Language Study. 2000.
187. FRANCO, Jon, Alazne LANDA and Juan MARTN (eds): Grammatical Analyses in Basque
and Romance Linguistics. 1999.
188. VanNESS SIMMONS, Richard: Chinese Dialect Classification. A comparative approach to
Harngjou, Old Jintarn, and Common Northern Wu. 1999.
189. NICHOLOV, Nicolas and Ruslan MITKOV (eds): Recent Advances in Natural Language
Processing II. Selected papers from RANLP 97. 2000.
190. BENMAMOUN, Elabbas (ed.): Perspectives on Arabic Linguistics Vol. XII. Papers from the
Twelfth Annual Symposium on Arabic Linguistics. 1999.
191. SIHLER, Andrew L.: Language Change. An introduction. 2000.
192. ALEXANDROVA, Galina M. and Olga ARNAUDOVA (eds.): The Minimalist Parameter.
Selected papers from the Open Linguistics Forum, Ottawa, 21-23 March 1997. 2001.
193. KLAUSENBURGER, Jurgen: Grammaticalization. Studies in Latin and Romance morphosyntax.
2000.
194. COLEMAN, Julie and Christian J. KAY (eds): Lexicology, Semantics and Lexicography.
Selected papers from the Fourth G. L. Brook Symposium, Manchester, August 1998. 2000.
195. HERRING, Susan C., Pieter van REENEN and Lene SCHSLER (eds): Textual Parameters in
Older Languages. 2000.
196. HANNAHS, S. J. and Mike DAVENPORT (eds): Issues in Phonological Structure. Papers from
an International Workshop. 1999.
197. COOPMANS, Peter, Martin EVERAERT and Jane GRIMSHAW (eds): Lexical Specification
and Insertion. 2000.
198. NIEMEIER, Susanne and Ren DIRVEN (eds): Evidence for Linguistic Relativity. 2000.
199. VERSPOOR, Marjolijn H. and Martin PTZ (eds): Explorations in Linguistic Relativity. 2000.
200. ANTTILA, Raimo: Greek and Indo-European Etymology in Action. Proto-Indo-European *ag-
. 2000.
201. DRESSLER, Wolfgang U., Oskar E. PFEIFFER, Markus PCHTRAGER and John R.
RENNISON (eds.): Morphological Analysis in Comparison. 2000.
202. LECARME, Jacqueline, Jean LOWENSTAMM and Ur SHLONSKY (eds.): Research in
Afroasiatic Grammar. Papers from the Third conference on Afroasiatic Languages, Sophia
Antipolis, 1996. 2000.
203. NORRICK, Neal R.: Conversational Narrative. Storytelling in everyday talk. 2000.
204. DIRVEN, Ren, Bruce HAWKINS and Esra SANDIKCIOGLU (eds.): Language and Ideology.
Volume 1: cognitive theoretical approaches. 2001.
205. DIRVEN, Ren, Roslyn FRANK and Cornelia ILIE (eds.): Language and Ideology. Volume 2:
cognitive descriptive approaches. 2001.
206. FAWCETT, Robin: A Theory of Syntax for Systemic-Functional Linguistics. 2000.
207. SANZ, Montserrat: Events and Predication. A new approach to syntactic processing in English
and Spanish. 2000.
208. ROBINSON, Orrin W.: Whose German? The ach/ich alternation and related phenomena in
standard and colloquial. 2001.
209. KING, Ruth: The Lexical Basis of Grammatical Borrowing. A Prince Edward Island French case
study. 2000.
210. DWORKIN, Steven N. and Dieter WANNER (eds.): New Approaches to Old Problems. Issues
in Romance historical linguistics. 2000.
211. ELK, Viktor and Yaron MATRAS (eds.): Grammatical Relations in Romani. The Noun
Phrase. 2000.
212. REPETTI, Lori (ed.): Phonological Theory and the Dialects of Italy. 2000.
213. SORNICOLA, Rosanna, Erich POPPE and Ariel SHISHA-HALEVY (eds.): Stability, Variation
and Change of Word-Order Patterns over Time. 2000.
214. WEIGAND, Edda and Marcelo DASCAL (eds.): Negotiation and Power in Dialogic
Interaction. n.y.p.
215. BRINTON, Laurel J.: Historical Linguistics 1999. Selected papers from the 14th International
Conference on Historical Linguistics, Vancouver, 9-13 August 1999. 2001.
216. CAMPS, Joaquim and Caroline R. WILTSHIRE (eds.): Romance Syntax, Semantics and L2
Acquisition. Selected papers from the 30th Linguistic Symposium on Romance Languages,
Gainesville, Florida, February 2000. n.y.p.
217. WILTSHIRE, Caroline R. and Joaquim CAMPS (eds.): Romance Phonology and Variation.
Selected papers from the 30th Linguistic Symposium on Romance Languages, Gainesville,
Florida, February 2000. n.y.p.
218. BENDJABALLAH, S., W.U. DRESSLER, O. PFEIFFER and M. VOEIKOVA (eds.): Morphol-
ogy 2000. Selected papers from the 9th Morphology Meeting, Vienna, 25-27 February 2000. n.y.p.
219. ANDERSEN, Henning (ed.): Actualization. Linguistic Change in Progress. Selected papers from
the Workshop on Patterns of Actualization in Linguistic Change, Vancouver, B.C., 14 August
1999. n.y.p.
220. CRESTI, Diana, Christina TORTORA and Teresa SATTERFIELD (eds.): Current Issues in
Romance Languages. Selected papers from the 29th Linguistic Symposium on Romance Lan-
guages (LSRL), Ann Arbor, 8-11 April 1999. n.y.p.

(Current Issues in Linguistic Theory 194) Julie Coleman (Ed.), Christian J. Kay (Ed.) - Lexicology, Semantics and Lexicography - Selec

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

(Current Issues in Linguistic Theory 194) Julie Coleman (Ed.), Christian J. Kay (Ed.) - Lexicology, Semantics and Lexicography - Selec

Transféré par

Droits d'auteur :

Formats disponibles

LEXICOLOGY, SEMANTICS AND LEXICOGRAPHY

AMSTERDAM STUDIES IN THE THEORY AND

Series IV CURRENT ISSUES IN LINGUISTIC THEORY

Advisory Editorial Board

Raimo Anttila (Los Angeles); Lyle Campbell (Christchurch, N.Z.)

Julie Coleman and Christian J. Kay (eds)

Lexicology, Semantics and Lexicography

JOHN BENJAMINS PUBLISHING COMPANY

National Standard for Information Sciences Permanence of Paper for Printed

Library of Congress Cataloging-in-Publication Data

These papers were originally presented at the 10th International Conference

Dr C.P.Biggam Prof Maurizio Gotti

Dr Gabriella Rundblad Penny Silva

Dr Louise Sylvester Edmund Weiner

The cognitive turn in linguistics has led linguists to re-examine many

Structure, or how to Locate Lexical Gaps

Proportional series, according to Cruse (1986:11819), consist of four or more

(4) having a not having a

animal, an animal is a kind of creature, and so on. Cruse (1986:146) postu-

animal bird fish insect

dog elephant robin eagle cod trout ant butterfly

collie spaniel robin blackbird starling

dog cat bird

collie spaniel robin blackbird starling

According to Cruse (1986:118, cf. 146) (6b) or (6c) is the taxonomy of

run walk crawl hop jump

According to Cruses (1986:151) intuition there is a covert category in this

Cognition, or how to Explain Lexical Gaps

Psychological Salience or Emotional Involvement

logical salience. A human beings natural way of locomotion is moving on land,

Perceptual salience may be less important as a factor, and it may go hand in

Prototypicality is based on the notion of prototypes introduced by the psy-

informants judge some members of taxonomic categories to be better or

Diachrony, or what may Happen to Lexical Gaps

The cognitive explanation suggested in the previous section immediately

opened up and then closed again.

Kungl. Vitterhets Historie och Antikvitets Akademien. 181188.

Gabriella Rundblad and David B. Kronenfeld

Folk-etymology is a very popular phenomenon. Folk-etymologies often begin

Folk-Etymologies and Linguistic Analogy

underlying linguistic or cultural pattern. For example, the mistakes with

the combustion of any substance).1

Folk-etymology, as defined by Winer, is a popular but false hypothesis for a

also be divided up into sub-groups depending on what part of a word triggered

As mentioned earlier, the discussions and conclusions of this paper mainly

Another example is bogbean, a common water plant (Menyanthes

Variations in Cultural Significance

Cultural knowledge and insight can similarly be provided by folk-etymologies

The hypothesis meets no phonological resistance since *womb-man [wu:m

Deeper Folk Processes New Productive Derivational Affixes

An investigation into the tendency for reinterpretation into meaningful

The overall number of Class II folk-etymologies is so small that conclusions

the creation and reinterpretation of words.

Underlying Linguistic and Cultural Forces

Language, as Saussure long ago noted, is a property of social groups of

communication (whether content, attitude toward the content, or relevant

Palmer, A.Smythe. 1890. Folk-Etymology. A Dictionary of Verbal Corruptions or Words

This article concerns the directions and mechanisms of semantic change in

with the models of directionality used in the context of grammaticalization. A

A. CORE AND PERIPHERAL MEMBERS

C. TWO OVERLAPPING CENTRES

Figure 1. Prototypes formation of new prototypical centres

Figure 1 shows a word a category which has one prototypical