566 PhonologyPhonetics Interface

that phonetic factors shape phonological systems.

Building on some of the insights of the natural phonology program (Stampe, 1972) and earlier phonetic
research, many phonologists began to explicitly
model phonetic motivations in the formal theory.
This integrated version of phonetics and phonology
took a complementary position to that of traditional
generative grammar by building phonetic explanations directly into the formalism. The advent of Optimality Theory (Prince and Smolensky, 1993), with its
hierarchically ranked well-formedness conditions,
provided a suitable phonological framework for
developing a phonetically informed model of phonology, termed phonetically driven OT (see Hayes et al.,
2004 for a cross-section of works within this framework). Constraints could be grounded in principles
of phonetic naturalness previously postulated by
phoneticians.
The literature in phonetically driven OT has become extensive in recent years and includes both
perceptual and articulatory-based accounts of various
phenomena. A key feature of this body of research is
that it demonstrates the nonarbitrary nature of many
phonological patterns, showing that these patterns
are predictable from independently known facts
about articulation and perception. I discuss some of
this research now.
Archangeli and Pulleyblank (1994) introduce the
notion of phonetic grounding in phonological cooccurrence restrictions, arguing that commonly observed interactions between the feature [ATR] and
vowel height and backness features find substantive
support in phonetic facts. For example, the fact that
low vowels are cross-linguistically [-ATR] is attributed to the retracted position of the tongue root during
the production of low vowels. Hayes (1999) explored
the articulatory basis for obstruent voicing patterns,
finding that a number of voicing asymmetries are
natural from an aerodynamic standpoint: Phonological voicing is most likely for places of articulation,
e.g., bilabial, and contexts, e.g., postnasal, that are
phonetically best suited to sustaining the transglottal
airflow necessary for voicing.
Much of the work in phonetically driven OT
appeals to perceptual factors. Flemming (1995)
invokes a perceptual account of a number of assimilatory phenomena. Building on the theory of acoustic
features advanced in Jakobson et al. (1952),
Flemming proposes a feature set based on vowel
formants to account for the assimilation data. For
example, the triggering of rounding in the high front
vowel /i/ preceding retroflexes in Wembawemba, an
extinct Pama Nyungan language of Victoria, Australia (Hercus, 1986), does not find a straightforward
explanation in acoustic terms but can be interpreted

as assimilation of the third formant, which is lowered

by both retroflexion and lip rounding.
Jun (1996) and Steriade (2001) examine the typology of consonant assimilation, finding asymmetries in
the direction of assimilation dependent on both manner and place of articulation. Jun finds a close correspondence between assimilation patterns and the
results of earlier experiments designed to assess the
relative salience of different perceptual cues to
consonants. Perceptual cues to consonants may be
classified as internal to the consonant (e.g., closure
duration, voicing, energy during the closure) or external (e.g., formant transitions for adjacent vowels,
fundamental frequency values in adjacent sounds,
voice-onset time, burst amplitude). For example, in
consonant clusters, the consonant on the left typically
assimilates to the consonant on the right rather than
vice versa, owing to the greater perceptual salience of
formant transitions coming out of a consonant into a
following vowel relative to transitions from vowel
into a following consonant (see also Ohala, 1990).
Furthemore, sounds with more robust internal cues
(e.g., fricatives) are less likely to assimilate than those
with weaker cues (e.g., nasals and stops).
Steriade (1999) explores the perceptual basis for
laryngeal neutralization, finding that the likelihood
of neutralization increases as the robustness of perceptual cues to laryngeal features decreases. She
showed that her cue-based approach to neutralization
offers a better fit to the observed patterns than a
syllable-based account in which neutralization is
claimed to occur in coda position (Lombardi, 1995).
For example, voicing neutralization in some languages only affects a subset of codas, those occurring
in contexts that are particularly ill-suited to recovering external consonant cues, such as word-finally and
before an obstruent, as in Lithuanian, or only before
another consonant, as in Hungarian. Conversely, neutralization can also affect syllable onsets if the
neutralized consonant occurs before an obstruent, as
in Russian.

Interlanguage Phonetic Variation and

Phonology
Another area of research dealing with the phonologyphonetics interface searches for correlations between
language-specific phonetic properties and crosslinguistic parameterization in phonological patterns.
The phenomenon of weight-sensitive stress has been
a fruitful area of language-specific links between
phonetic and phonological properties. The primary
source of cross-linguistic variation in weight is the
treatment of closed syllables. In some stress systems,
such as the one found in Latin, CVC is treated as

PhonologyPhonetics Interface 567

heavy, whereas in other languages, such as Khalkha

(Halh) Mongolian, CVC is light. Broselow et al.
(1997) examined the phonetic duration of CVC syllables in languages with different weight criteria, including one (Hindi) with the CVC heavy criterion
and one (Malayalam) that treats as light, in order to
determine whether the duration of CVC differs
according to its weight. They found that vowels
are substantially shortened in CVC syllables relative to CV syllables in Malayalam but not in Hindi.
Broselow et al. (1997) attribute the duration differences between Hindi CVC and Malayalam CVC to
differences in moraic structure between the two languages: In Malayalam, a coda consonant shares a
mora with a preceding nucleus, whereas in Hindi, a
coda carries its own mora.
Gordon (2002) also found a close link between
cross-linguistic variation in weight of CVC and
language-specific phonetic properties, although his
data suggests a closer correlation between weight
and a measure of perceptual energy factoring in
both duration and intensity than a measure of duration alone. Gordons work also explores a slightly
different relationship between phonetics and phonology than that assumed in Broselows work. Extending
the hypothesis advanced by other researchers working within the phonetically driven OT framework, he
suggests that language-specific phonetic patterns are
the cause of phonological variation in weight. In
order to demonstrate this directionality of the
phonologyphonetics relationship, he shows that
interlanguage differences in phonetics are rooted
in differences in another phonological property
more basic than weight, syllable structure. A crosslinguistic survey of coda inventories indicates that
languages with heavy CVC overwhelmingly tend to
have more high-energy codas such as sonorants and
voiced consonants than languages with light CVC.
Gordon claims that differences in the net energy profile of CVC assessed over all CVC syllables lead to
variation between languages in the weight of CVC.

phonetic determinism at the synchronic level. For

example, voicing alternations in Tswana and certain
other Bantu (Bantoid) languages suggest a productive
process of postnasal devoicing (Hyman, 2001) that
runs counter to the cross-linguistically dominant and
phonetically natural pattern of postnasal voicing
(Hayes, 1999). Hyman suggests that a phonetically
unmotivated constraint banning voiced obstruents
after nasals accounts for this alternation, which
results from a confluence of sound changes, none of
which by themselves are particularly unnatural phonetically. The incorporation of this unnatural constraint suggests that a synchronic model of phonology
only allowing phonetically natural constraints is overly restrictive in its predictive power. On the other
hand, the fact that children appear to construct systematic and phonetically informed phonologies during the language acquisition process suggests that
speakers do make active use of phonetic constraints
synchronically (see Hayes and Steriade, 2004 for
discussion of the diachronic vs. synchronic role of
phonetics in language).

Summary
There is strong evidence that phonetic factors shape
many aspects of phonological systems, ranging from
phoneme inventories to phonotactic constraints to
prosodic phenomena such as syllable weight. These
phonetic motivations have been incorporated into
the formal phonological framework of phonetically
driven Optimality Theory by many researchers, while
other linguists have assumed that the role of phonetics in phonology is primarily diachronic rather than
synchronic.
See also: Assimilation; Distinctive Features; Generative
Phonology; Natural Phonology; Phonetics, Articulatory;
Phonology: Overview; Quantity; Speech Errors as Evidence in Phonology; Speech Perception; Speech Production; Syllable: Typology.

Phonetics in Phonology: Synchrony or

Diachrony

Bibliography

Despite the growing recognition among phonologists

that phonetics plays an important role in phonological systems, there is disagreement about whether
phonetic grounding should be modeled as a feature
of synchronic phonologies or whether it is merely a
historical force that shapes phonological systems over
time without entering into the linguistic awareness of
current speakers. Evidence for this latter position
potentially comes from languages displaying phonetically unnatural properties that suggest limits to

Archangeli D & Pulleyblank D (1994). Grounded phonology. Cambridge, MA: MIT Press.
Broselow E, Chen S & Huffman M (1997). Syllable weight:
convergence of phonology and phonetics. Phonology 14,
4782.
Chomsky N & Halle M (1968). The sound pattern of
English. New York: Harper & Row.
Cohn A (1993). Nasalization in English: phonology or
phonetics. Phonology 10, 4381.
Flemming E (1995). Auditory representations in phonology.
Ph.D. diss., UCLA.

568 PhonologyPhonetics Interface

Gordon M (2002). A phonetically-driven account of syllable weight. Language 78, 5180.
Hayes B (1999). Phonetically-driven phonology: the role of
optimality theory and inductive grounding. In Darnell
M, Moravscik E, Noonan M, Newmeyer F & Wheatley
K (eds.) Proceedings of the 1996 Milwaukee Conference
on Formalism and Functionalism in Linguistics, vol. 1.
Amsterdam: Benjamins. 243285.
Hayes B & Steriade D (2004). Introduction: the phonetic
bases of phonological markedness. In Hayes B, Kirchner
R & Steriade D (eds.) Phonetically based phonology.
New York: Cambridge University Press. 133.
Hayes B, Kirchner R & Steriade D (eds.) (2004). Phonetically based phonology. New York: Cambridge University
Press.
Hercus L A (1986). Victorian languages, a late survey.
Canberra: Australian National University.
Huffman M (1993). Phonetic patterns of nasalization and
implications for feature specification. In Huffman M &
Krakow R (eds.) Phonetics and phonology 5: nasals,
nasalization, and the velum. San Diego: Academic Press.
303327.
Hyman L (2001). The limits of phonetic determinism in
phonology: *NC revisited. In Hume E & Johnson K
(eds.) The role of speech perception in phonology. San
Diego: Academic Press. 141185.
Jakobson R, Fant G & Halle M (1952). Preliminaries to
speech analysis. Cambridge, MA: MIT Press.
Jun J (1996). Place assimilation as the result of conflicting
perceptual and articulatory constraints. In Camacho J,
Choueiri L & Watanabe M (eds.) Proceedings of the 14th
West Coast Conference on Formal Linguistics. Palo Alto:
CSLI. 221238.
Keating P (1988). Underspecification in phonetics.
Phonology 5, 275292.
Keating P (1990). The window model of coarticulation:
articulatory evidence. In Kingston J & Beckman M (eds.)
Papers in laboratory phonology I: between the grammar and the physics of speech. Cambridge: Cambridge
University Press. 451470.
Liljencrants J & Lindblom B (1972). Numerical simulation
of vowel quality systems: the role of perceptual contrast.
Language 48, 839862.
Lindblom B & Maddieson I (1988). Phonetic universals in
consonant systems. In Hyman L & Routledge C (eds.)
Language, speech and mind: studies in honor of Victoria
A. Fromkin. New York: Routledge. 6280.
Lombardi L (1995). Laryngeal neutralization and syllable
well-formedness. Natural Language and Linguistic
Theory 13, 3974.

Ohala J (1983). The origin of sound patterns in vocal tract

constraints. In MacNeilage P (ed.) The production of
speech. New York: Springer Verlag. 189216.
Ohala J (1990). The phonetics and phonology of aspects of
assimilation. In Kingston J & Beckman M (eds.) Papers
in laboratory phonology I: between the grammar and the
physics of speech. Cambridge: Cambridge University
Press. 258275.
Ohala J (1994). Towards a universal, phonetically-based,
theory of vowel harmony. 1994 Proceedings of the International Congress on Spoken Language Processing.
491494.
Ohala J (1997). The relation between phonetics and phonology. In Hardcastle W & Laver J (eds.) The handbook
of phonetic sciences. Oxford: Blackwell Publishers.
674694.
Ohala J & Ohala M (1993). The phonetics of nasal phonology: theorems and data. In Huffman M & Krakow R
(eds.) Nasals, nasalization, and the velum [phonetics and
phonology 5]. San Diego: Academic Press. 225249.
Parker S (2002). Quantifying the sonority hierarchy. Ph.D.
diss., University of Massachusetts, Amherst.
Pierrehumbert J (1980). The phonology and phonetics of
English intonation. Ph.D. diss., MIT.
Pierrehumbert J & Beckman M (1988). Japanese tone structure. Cambridge, MA: MIT Press.
Prince A & Smolensky P (1993). Optimality theory: constraint interaction in generative grammar. Unpublished
ms. Rutugers University and University of Colorado at
Boulder.
Stampe D (1972). How I spent my summer vacation [a
dissertation on natural phonology]. Ph.D. diss., Ohio
State.
Steriade D (1999). Licensing laryngeal features. UCLA
Working Papers in Phonology 3, 25146.
Steriade D (2001). Directional asymmetries in place assimilation: A perceptual account. In Hume E & Johnson K
(eds.) The role of speech perception in phonology. San
Diego: Academic Press. 219250.
Stevens K & Keyser S J (1989). Primary features and their
enhancements in consonants. Language 65, 81106.
Stevens K (1989). On the quantal nature of speech.
Journal of Phonetics 17, 345.
Westbury J & Keating P (1985). On the naturalness of stop
consonant voicing. UCLA Working Papers in Phonetics
60, 119.
Wright J T (1986). The behavior of nasalized vowels in the
perceptual vowel space. In Ohala J & Jaeger J (eds.)
Experimental phonology. Orlando, FL: Academic Press.
4567.

Photography: Semiotics 569

Photography: Semiotics
S Mazzali-Lurati and L Cantoni, University of
Lugano, Lugano, Switzerland
2006 Elsevier Ltd. All rights reserved.

Semiotic Characters of Photographic

Signs
It is commonly acknowledged that their technological
origin confers to photographs particular semiotic features (cf. definitions by Gubern, 1999: 56; Schaeffer,
1987: 16; Vanlier, 1983: 1315). Unlike other kinds
of visual signs, photographs are more the product of
the mechanical and chemical process of impression
on photosensitive material of light rays than the result
of a human act of creation.
Depending on the importance accorded to the
human intervention in their production, both the relationship of photographs to their referent (degree of
conventionality) and the relationship of the photographic message to the code (degree of codification)
are differently understood.
Degree of Conventionality

A propos of the degree of conventionality of photography, three main theoretical positions have emerged: (a)
the iconic, (b) the indexical, and (c) the conventionalist
position (Dubois, 1983: 2627).
(a) The iconic position characterized the perception
of photography by its first practitioners
(Schaeffer, 1987: 114), for whom, in opposition
to painting, the essence of photography consisted
in mimesis of the real (Dubois, 1983: 26, 33).
Recent approaches to pictorial semiotics restated
the mainly iconic nature of photography
(Gubern, 1999: 57; Sonesson, 1999). Above all,
photographs are images faithfully representing
objects, persons, places, or events through an
objective analogy to reality.
(b) The indexical position (cf. particularly Peirce
2.281; Dubois; Schaeffer; and Vanlier) focused
on the process of production of photographs:
there is a physical contiguity between light rays
and the photographic film, producing a mechanical imprint. Different nuances appeared among
the various proposers. Schaeffer (1987: 59 ss.)
defined photographs as both indexical icons and
iconic indexes, thus underlining that their semiotic nature resides in the compresence of an analogic and an indexical function; however, it is the
latter that constitutes the very nature, the arche of
photography (1987: 27). According to Vanlier
(1983), photographs are for sure indices (that is,

unintentional imprints), but they are not always

indexes (that is, intentional signs).
Between position (a) and position (b) is the
work by Barthes, who defined photography as
analogon of reality (1961), thus stressing the
iconic aspect. In Barthes (1964, 1980) reference
and not analogy became the most distinctive feature of photographs, as well as their power of
evidence (authentication of the existence and
presence of the referent in front of the camera).
The having-been-there constitutes the noema of
photography.
(c) Exponents of the conventionalist position
among them, besides Floch, Lindekens, and
Goodman, can be counted Eco (1984), Metz
(1970), and Groupe m (1979), who held such a
position in their studies on the whole category of
visual signs pointed out that the different modes
and techniques that are necessarily involved in
the production of photographs generate a nontransparency in respect to the real (Goodman,
1968: 15).
Both Floch (1986) and Lindekens (1971) contrasted
the concept of iconicity with the concept of iconization, which defines the process of building analogy to
the real accomplished by the conventional aspects
present in photographs. As such, iconization introduces a gap between reality and the object represented
in the photograph. The photographic analogy is only
relative (Lindekens, 1971: 94).
The advent and socialization of digital photography
raises again the same questions. In fact, in digital photography the decisive moment is not that of release
but that of image manipulation (Berger, 1989: 76;
Sonesson, 1999: 3133); therefore, conventionality
seems to become predominant.
Degree of Codification

Depending on the degree of conventionality conferred

on photography, a different degree of codification is
attributed to the photographic message. Barthes and
Vanlier emphasized the presence of aspects of noncodification. Other authors Floch (1986), who proposed some analysis of photographs from the point
of view of structural semiotics; Lindekens (1971) and
Espe (1983), who both adopted an empirical approach, and on a more general level, Metz (1970)
underlined the codified character of photography.
Barthes (1961) identified in photographs a denotative and a connotative level of signification. At the
denotative level, photographs being a mechanical
analogon of reality, no code mediates the relationship of the image to its referent. The connotative

570 Photography: Semiotics

level corresponds to the coding of the photographic

analogon, generated by different connotation procedures (described in Barthes, 1961, 1964). Paradoxically, in photographs, on the basis of a message without
a code, a connoted (or coded) message develops. A very
similar position was held by Vanlier (1983).
Lindekenss study of the photographic code stressed
the need to distinguish between the intended meaning
(in French, signifie ) coming from the code (the iconic
sense) and the intended meaning coming from the
analogon (the intended meaning of identification)
(1971: 234). In front of a photograph, the most common attitude is looking at the real, thus focusing
on the signified of identification (Lindekens, 1971:
231). However, a perceptual iconic substance exists
in photographs, in which it is possible to identify
minimal distinctive traits (the iconic morphemes,
Lindekens, 1971: 194), the combination of which
constitutes the photographic code. The iconic sense
generated by the code (Lindekens, 1971: 141142)
affects the informative meaning deriving from the
photographic analogon.

Photography as Practice
Pragmatic approaches underlined the connection
existing between the semiotic character of photography and its communicative use and realization. Photographs are considered texts (Eugeni, 1999: 198), the
meaning of which strictly depends on their materiality and on the producers choices (Kress and van
Leeuwen, 1998), but also on their relationship to the
context of use and on the dynamics of their reception.
The relationship to culture, ideology, and institutional context has been particularly explored, mainly
starting from a Marxist perspective (Burgin, 1982;
Sekulla, 1982; Tagg, 1982, 1988). Other approaches
(Dubois, 1983; Floch, 1986; Schaeffer, 1987) focused
more directly on the receivers and their process of
interpretation.
Schaeffers (1987) description of the pragmatic
flexibility of photographs is the most systematic
one. In different communicative contexts the analogical or indexical function of photographs prevails,
thus bringing the images closer to the nature of
index or closer to the nature of icon (Schaeffer,
1987: 101102). Therefore, the semiotic nature of
photography strictly depends on the communicative
norms ruling different communicative contexts
(Schaeffer, 1987: 10).
Following this perspective, the problem of the semiotic nature of photographs can be reconsidered.
For different kinds of use, it is possible to identify
an aspect that plays a major role in the photographic
expression.

Without entering into the details of the classifications proposed by different theoreticians (Berger,
1989: 6264; Schaeffer, 1987: 6874; Sonesson,
1999: 14), five main domains of use of photographs
can be distinguished that imply different kinds of
realization: scientific domains (medicine, criminology,
anthropology, restoration, archaeology, mineralogy,
paleontology, astronomy, meteorology, geography,
and architecture); domains of advertising, economy,
and marketing; information domains; use in private
life; and art photography.
Scientific Domains

The iconic and the indexical aspects play the central

role in realization in the scientific domains. On
the one hand, the aim of such a use of photography
is to document how a given object appears in reality
(e.g., architecture or archaeology, but also topophotography, allowing the visualization of our daily
existential space from unusual perspectives; Zannier,
1982: 78). On the other hand, in these fields photographs aim at letting the receiver see something that
human eyes cannot see without technical means (e.g.,
microphotography, astrophotography, and microscopic analysis in medicine).
Advertising, Economy, and Marketing

The realization of advertising posters, commercial

products catalogues, fashion magazines, and some
kinds of pornographic photography (photographs
used in racy tabloids) is mainly iconic and (as to the
interaction of images with verbal texts) symbolic.
Photographs aim at showing the object itself or some
of its interesting and attractive features. The indexical
aspect plays no role. It is taken for granted at the point
that the receiver identifies the presentation of the
product and the product itself (Schaeffer, 1987: 149).
Information Domains

In these fields (comprising war, propaganda, sociological, documentary, and scandalmongering photography), photographs are consumed in illustrated
publications, newspapers, magazines, and television
news. The widely studied practice of photojournalism
(cf. Barthes, 1961; Lambert, 1986; Schaeffer, 1987)
constitutes a major example. In it, the photographic
realization involves complex dynamics due to the relationship of the image with other codes (especially
the verbal one) and to the problem of objectivity. The
indexical and the symbolic aspects play the most important role. In fact, here photographs mainly aim at
testifying to the real existence of the photographic
referent. From this status of visual testimony
(Schaeffer, 1987: 80) derives the problem of their

Photography: Semiotics 571

objectivity. The knowledge the receiver has of the

presence in photography of an indexical aspect can
produce misunderstandings as to the truth of the
image (Schaeffer, 1987: 139147; Kress and van Leeuwen, 1998). To such misunderstandings can contribute the interaction of the image with the verbal code
(inducing a symbolic realization of the image): in relationship to different texts, the same photograph can
be used as testimony of different (even opposed) facts
or situations (Freund, 1974; Schaeffer, 1987: 81).
Photography in Private Life

The use of photography in private life arises from the

common practice of creating a souvenir of moments
and important steps in ones own life. Its development depended on the wide spread of literacy in the
production of photographs, enabled by the efforts
the photographic industry made in offering easyto-use technology (Zannier, 1982: 207236). The realization of this kind of photograph has a mainly
iconicindexical value. What counts are the iconic representations of persons, places, and events and the
relationship with the represented object, place, or
person (stressed by Dubois, 1983: 8183; Lindekens,
1971: 232). Wedding photographs constitute an interesting case. Although they aim at faithfully (iconically) reproducing the beauty of the wedding day,
they are often retouched by use of various devices
to assure iconicity and the faithfulness that a natural photograph could distort (for such a procedure,
cf. Lindekens, 1971: 229; Barthes, 1961).
Art Photography

The realization of art photography is intrinsically symbolic (cf. Schaeffer, 1987: 150), and in it the theme
of the code is prominent. In fact, as happens in all
forms of art, the elements of the form and the style
(produced through photographic modes, techniques,
and devices) play a major role in shaping the meaning.
See also: Barthes, Roland: Theory of the Sign; Communication: Semiotic Approaches; Context, Communicative;
Eco, Umberto: Theory of the Sign; Iconicity: Theory; Indexicality: Theory; Meaning, Sense, and Reference;
Media: Pragmatics; Media: Semiotics; Multimodality and
the Language of Politics; Politics, Ideology and Discourse;
Reference: Semiotic Theory; Truth Conditional Semantics
and Meaning; Visual Semiotics.

Bibliography
Barthes R (1961). Le message photographique. Communications 1, 127138.
Barthes R (1964). Rhe torique de limage. Communications
4, 4051.

Barthes R (1980). La chambre claire: note sur la photographie.

Paris: Gallimard.
Berger A A (1989). Seeing is believing. An introduction
to visual communication. Mountain View, California:
Mayfield Publishing Company.
Burgin V (1982). Looking at photographs. In Burgin (ed.).
142153.
Burgin V (ed.) (1982). Thinking photography. London/
Basingstoke: Macmillan Press Ltd.
Dubois P (1983). Lacte photografique. Paris: Fernand Nathan.
Eco U (1984). Semiotics and the philosophy of language.
Bloomington: Indiana University Press.
Espe H (1983). Realism and some semiotic functions
of photographs. In Borbe T (ed.) Semiotics unfolding.
Berlin: Mouton. 14351442.
Eugeni R (1999). Analisi semiotica dellimmagine: pittura,
illustrazione, fotografia. Milano: ISU Universita` Cattolica.
Floch J-M (1986). Les formes de lempreinte. Pe rigueux:
Pierre Fanlac.
Freund G (1974). Photographie et socie te . Paris: PointSeuil.
Goodman N (1968). Language of art. Indianapolis: BobbsMerrill.
Groupe m (1979). Iconique et plastique: un fondement de la
rhe torique visuelle. Revue desthe tique 12, 173192.
Gubern R (1974). Mensajes ico nicos en la cultura de masas.
Barcelona: Lumen.
Gubern R (1999). From optical to digital: when the same is
different. Visio 4(1), 5560.
Krauss R (1990). Le photographique: pour une the orie des
e carts. Paris: Ed. Macula.
Kress G & van Leeuwen T (1998). Reading images: the
grammar of visual design. London : Routledge.
Lambert F (1986). Mythographies: la photo de presse et ses
le gendes. Paris: E dition Edilig.
Lindekens R (1971). E le ments pour une se miotique de la
photographie. Paris: Didier.
Metz C (1970). Au-dela` de lanalogie, limage. Communications 15, 110.
Peirce C S (19351966). Collected papers. Hartshorne C,
Weiss P & Burks A W (eds.). Cambridge, MA: Harvard
University Press.
Schaeffer J-M (1987). Limage pre caire: du dispositif photographique. Paris: E ditions du Seuil.
Sekulla A (1982). On the invention of photographic
meaning. In Burgin (ed.). 84109.
Sonesson G (1999). Post-photographie and beyond: from
mechanical production to digital production. Visio 4(1),
1136.
Tagg J (1982). The currency of the photograph. In Burgin
(ed.). 110141.
Tagg J (1988). The burden of representation: essays on
photographies and histories. Amherst: University of
Massachusetts Press.
Vanlier H (1983). Philosophie de la photographie. Paris: Les
Cahiers de la photographie.
Zannier I (1982). Storia e tecnica della fotografia. Bari:
Editori Laterza.

572 Phrasal Stress

Phrasal Stress
H Truckenbrodt, Universitat Tubingen, Tubingen,
Germany
2006 Elsevier Ltd. All rights reserved.

Preliminaries
The Representation of Stress

Word stress (Hayes, 1995) is the strongest stress in a

prosodic word. Phrasal stress is stress assigned beyond word stress in syntactic collocations of words,
such as phrases, clauses, or sentences. Some examples
that are discussed in this article are shown in (1) with
word stress indicated by boldface.
(1a) [Who did you meet?] The brother of Mary
(1b) [Guess
The mayor of Chicago won their
support.
what.]
(1c) [Who came to the The brother of Mary came
to the party.
party?]
(1d) German: [What
did John do?]
Er hat Linguistik
Er hat in Ghana
unterrichtet
unterrichtet
he has linguistics
he has in Ghana
taught
taught
He has taught
He has taught in
linguistics.
Ghana.

In examples such as these, it is felt that the (singly or

doubly) underlined syllables are stressed more than
the other boldface syllables in the same utterance.
Among the underlined syllables, it is felt that the
strongest stress of the utterances is on the element
that is doubly underlined. This (singly or doubly
underlined) addition to (boldface) word stress is
phrasal stress (henceforth also p-stress).
The examples show an important property of the
assignment of p-stress: In the words of Liberman and
Prince (1977), stress is normally preserved under
embedding. Syllables that show p-stress in the examples in (1) are always syllables that are independently
determined to receive word stress by the rules assigning word stress in English or German. P-stress is
therefore assigned in the same representation as
word stress, extending the representation upward,
as shown in Figure 1. Here an element with stronger

Figure 1 P-stress is assigned in the same representation as

word stress, extending the representation upward.

stress than another one is dominated by a relatively

higher grid-column.
The dual representation in Figure 1 in terms of
constituents and grid-marks marking prominence
has grown out of the dual theory of Liberman and
Prince (1977), which employed metrical trees and
grid-columns. Later arguments (e.g., Selkirk, 1980;
Halle and Vergnaud, 1987) have shown that stress
(as represented by the grid-columns) should be seen
as tightly tied to the constituents so that a one-toone relation between a constituent and the element
with the strongest stress in it is assumed (see, e.g., the
Faithfulness Condition in Halle and Vergnaud, 1987;
Hayes 1995). A standard representation for this dual
nature of the theory is the bracketed grid (Halle and
Vergnaud, 1987). I here use the bracketed grid representation of Hayes (1995), in which the gridmark that marks the strongest element (head) of a
constituent is drawn on the same line. (Minimally
different dual representations are found in, e.g.,
Halle and Vergnaud, 1987; Nespor and Vogel,
1986, 1989; Cinque, 1993.) The constituents are
called metrical constituents in accounts that make
claims only about stress (Halle and Vergnaud, 1987;
Cinque, 1993; Zubizarreta, 1998); in these accounts,
there are in principle an arbitrary number of levels of
such constituents, and they are typically thought to be
directly identified with syntactic constituents.
However, the constituents are instead identified
with prosodic constituents in some of the literature
on phrasal phonology (Nespor and Vogel, 1986,
1989). In this literature, there are a small number of
such levels, each with phonological consequences in
individual languages beyond their relevance in stress
assignment. The phenomena investigated in these
terms have led to the conclusion that prosodic constituents at the level of the prosodic word and above
(phonological phrase and intonation phrase) are systematically related to syntactic constituents but not
identical to them (see also Selkirk, 1986, 1995a).
Phrasal Stress and Intonation

Stress and the prosodic/metrical constituents provide

the underpinning for the assignment of intonation
contours in English. In the classical and still important theory of English intonation by Pierrehumbert
(1980) and Beckman and Pierrehumbert (1986),
H and L tones that define the sentence melody are
assigned in one of two forms:
1. As pitch accents (H*, L*, L*H, LH*, H*L,
HL*). These are associated with stressed

Phrasal Stress 573

syllables; the star marks the tone that falls on the

stressed syllable.
2. As edge tones. These anchor with the edge of
intermediate phrases (H-, L-) or intonation
phrases (H%, L%).
These accounts of empirically observable intonation
contours converge with the dual representation previously introduced: Intermediate phrase and intonation phrase are not only anchors for edge tones, but are
also prosodic domains in which an element of strongest stress is defined. This element carries an obligatory pitch accent. The picture is extended to lower levels,
where each accent is taken to be assigned to the
strongest element of an accentual phrase. English
does not have edge tones of the accentual phrase
in this analysis; however, other languages such as
Japanese arguably do (Beckman and Pierrehumbert,
1986; Pierrehumbert and Beckman, 1988).

How Is Phrasal Stress Assigned?

The accounts of the assignment of p-stress agree insofar
they acknowledge that focus as well as the syntactic
structure have a role to play in this assignment.
Phrasal Stress and Focus

Let us first consider focus on single-word constituents.

Core cases of focus involve stress on the asked-for
constituent in an answer sentence (Who likes Mary?
Bill likes Mary; vs. Who does Mary like? Mary likes
Bill), as well as stress by contrast with another constituent (John likes Mary. No, Bill likes Mary; vs.
Mary likes John. No, Mary likes Bill.) As the examples
show, the context in which an utterance is used may
direct the main stress of the sentence to different positions. Since Jackendoff (1972), the influence of focus
is modeled in the theory of grammar with the help of
an abstract feature F that is assigned to a syntactic
constituent. In the case of questionanswer pairs, F is
assigned to the asked-for constituent, thus Who likes
Mary? [Bill]F likes Mary. In the case of explicit contrast, F is assigned to the contrasted constituent, thus
John likes Mary; no, [Bill]F likes Mary. The focus
F then has a semantic/pragmatic interpretation that
relates the assignment of F to the context in which the
sentence is used. At the same time, consequences of
the presence of F for stress assignment are defined.
Following Jackendoff (1972), the strongest stress in
the sentence has to be within the constituent marked
F. Thus, in Who likes Mary? [Bill]F likes Mary, the
answer is correctly predicted to be stressed on the
focused subject: [Bill]F likes Mary.

The Nuclear Stress Rule of Chomsky and Halle

Focus may be assigned to a larger constituent such as

the subject of the sentence in (2a). The limiting case of
this is what is sometimes called an all-new sentence
and is here called a sentence with no narrow focus.
An example is shown in (2b). (For concreteness, F is
here assigned at the sentence level, but it depends on
the theory of focus whether these sentences carry no
focus or one all-embracing focus.) In these cases,
F requires that the strongest stress is within F, but
other rules or principles must come into play to determine where inside F (or inside the clause in (2b)) the
strongest stress is assigned.
(2a) [Who came to
be party?]
(2b) [What happened?/
Guess what:]

[the brother of Mary]F

came to the party
[I met the brother of
Mary]F

An early proposal for English was the Nuclear Stress

Rule (NSR) of Chomsky and Halle (1968). The NSR
assigns stress on the right. (Its original formulation
assigns stress cyclically. I return to the cyclic application later.) In (2a), then, we can say that the strongest
stress must be within F, and that, within F, the NSR
assigns stress in rightmost position, thus on the rightmost word Mary. Similarly in (2b), where F does not
limit the possibilities of stress assignment in the sentence; the NSR here assigns rightmost stress, again on
the word Mary. The two rules of stress assignment
within F, and in rightmost position, make correct
predictions about the location of the strongest
p-stress in an English sentence for a considerable
variety of cases.
Problems with the NSR

Head-Final Syntactic Structures The NSR does not

generalize to languages with head-final structures
in the syntax. Although it correctly predicts strongest stress on the object that follows the verb in
English (3a), it cannot predict strongest stress on an
(unscrambled) object that precedes the verb in Dutch
and German (3b).
(3a) [What did John
do?]
(3b) [What did John
do?]

He was [teaching
linguistics]F
Er hat [Linguistik
unterrichtet]F
he has linguistics taught

However, Dutch and German do not generally assign

p-stress leftmost by a mirror-image rule of the NSR.
In the structures with adjuncts in (4), the rightmost
element is the strongest in both English and German,
although this is the adjunct in English and the

574 Phrasal Stress

verb in German. (See Krifka, 1984; Jacobs, 1993; on

the distinction between arguments and adjuncts in
German; single underlining in (4) is discussed below.)
(4a) [What did John
do?]
(4b) [What did John
do?]

He was [teaching in
Ghana]F
Er hat [in Ghana
unterrichtet]F
he has in Ghana taught

More than One Phrasal Stress in the Sentence Sentences with no narrow focus, such as [the mayor of
Chicago] [won their support] show more than one pstress, as indicated. For a case like this, a cyclic application of the NSR correctly assigns p-stress (Selkirk,
1984): In the subject [the mayor of Chicago], the
NSR assigns rightmost stress; likewise, in the constituent [won their support]. When the two parts are put
together, the NSR applies on the sentence level and
correctly strengthens the rightmost p-stress: [the
mayor of Chicago] [won their support].
Gussenhoven (1983b) observed that the NSR does
not always correctly assign p-stress that is not the
strongest stress of the sentence in English. In (4a), in
which the verb precedes an adjunct, the verb also
receives p-stress. However, in (3a), in which the verb
precedes an argument, the verb does not receive
p-stress. The contrast is subtle, but Gussenhoven
(1983b) conducted an experiment that showed that
the contrast is real.
The Sentence Accent Assignment Rule of
Gussenhoven

Gussenhoven (1983a, 1992) suggested the Sentence

Accent Assignment Rule (SAAR) as a solution to the
problems just reviewed. The SAAR was presented as a
rule assigning accent directly. For the purpose of the
discussion here, the perspective of Pierrehumbert
(1980), Beckman and Pierrehumbert (1986), Hayes
and Lahiri (1991), and Hayes (1995) is adopted by
which (phrasal) stress is assigned first and then serves
as the anchor for pitch accents. The SAAR is accordingly decomposed into two steps here: (1) assignment
of p-stress relative to the syntax, and (2) assignment
of a pitch accent to each element carrying p-stress.
Here step 1 is of primary interest.
The core of the SAAR is this: Within a focus, every
predicate, argument, and modifier must be accented
(here: receive p-stress), with the exception of a predicate that stands next to an accented argument. Thus,
in (4), each adjunct (in Ghana) receives p-stress, and
the verbs (predicate) also receive p-stress. In (3), the
arguments (linguistics/Linguistik) receive p-stress,
but in both languages the verbal heads next to the
stressed arguments do not receive p-stress (teaching/

Figure 2 The rightmost p-stress in the intonation phrase is

strengthened to the strongest stress.

unterrichtet). (Note that, in reducing the SAAR to its

core, I am leaving out a part of the SAAR that, in my
assessment of the empirical situation, concerns the
interaction of stress assignment and movement; see
Bresnan, 1971; Gussenhoven, 1992.)
The Strongest Phrasal Stress

The SAAR not only assigns p-stress correctly in a

wide range of cases, it also provides the correct
input for defining the position of the strongest stress
(as made formally explicit in similar terms by
Uhmann, 1991, for German; by Hayes and Lahiri,
1991, for Bengali; and by Selkirk, 1995b, for
English): The rightmost p-stress in the intonation
phrase is strengthened to the strongest stress. For the
examples in (3b) and (4b), this is shown in Figure 2; it
works similarly in he was teaching in Ghana and
in [the mayor of Chicago] [won their support]. This
suggestion is thus similar to the NSR, but it applies in
the prosodic domain of the intonation phrase rather
than in a syntactic domain.
A syntactically defined alternative is the C-NSR of
Zubizarreta (1998). The accented elements stand in a
syntactic relation of asymmetric c-command relative
to one another. For example, the subject of a clause
asymmetrically c-commands the object. Zubizarreta
assigned stronger stress to the element lower in the
chain of asymmetric c-command, thus, to the object
as opposed to the subject. This suggestion makes
stronger typological predictions, excluding languages
in which stress is assigned leftmost in the intonation
phrase. On the other hand, Szendro i (2001) argued
that Hungarian is a language in which the leftmost
stress in the intonation phrase is strengthened.
The Sentence Accent Assignment Rule and
Depth of Embedding

Cinque (1993) took the implementation of the NSR

in Halle and Vergnaud (1987) as a starting point and
sought to develop a revised cyclic theory that would
account for main stress on the preverbal object in
Dutch and German in examples such as Linguistik
unterrichten. Cinques idea was that the greater
amount of syntactic embedding of the complement
([DP [NP [N Linguistik]]] vs. [V unterrichten]) would
give rise to more cycles and thus, he postulated, to

Phrasal Stress 575

Figure 3 Making the first (indirect) object (IO) arbitrarily heavy

does not prevent assignment of strongest stress on the second
(direct) object (DO).

higher stress than on the following head. Cinques

theory does not seek to account for secondary stress
and thus misses the generalizations captured by the
two-level analysis employing the SAAR that we have
seen previously. It also runs into empirical problems
in cases in which we can rigorously test the claim that
greater depth of embedding leads to greater stress.
Thus, in German examples such as that shown in
Figure 3, we can make the first (indirect) object arbitrarily heavy and the strongest stress will always be on
the second (direct) object. Cinques theory wrongly
predicts that there will be a point at which the indirect
object is more deeply embedded than the following
constituent and then attracts main stress.
This case exemplifies a more general problem for
Cinques account: Among two XPs not contained in
one another (such as subject and VP-adjunct, subject
and object, VP-adjunct and object, object and object),
the depth of embedding never determines which stress
is the strongest. They each receive stress, but the
strongest one is the rightmost p-stress, as predicted
by the accounts in terms of the SAAR plus strengthening of the final one. Thus, depth of embedding does
not generalize from complement-head relations to
other cases as a predictor of main stress. (Cinque
noticed some of the problems arising from this and
proposed an addition to the theory that removes
the prediction on the nonrecursive side of a syntactic
head. This provision does not help in Figure 3 because
the left side must be the recursive side of the German
verb; otherwise, the normal object-verb configuration cannot be derived. But then a further object,
as in Figure 3 is also on the recursive side of the
German verb.)
The Sentence Accent Assignment Rule and
Syntactic XPs

The notion XP plays an important role in a central

line of research in phrasal phonology. Chen (1987)
argued that in Xiamen Chinese (Min Nan Chinese)
the right edge of a syntactic XP coincides with a tone
group boundary, with some systematic exceptions.
Building on a variety of languages, Selkirk convincingly generalized that account and argued that in
some languages the right edge of XP is aligned with
phonological phrase edges (for Xiamen Chinese, Chi

Figure 4 The application of Stress-XP to two core cases of the

SAAR: (A) standard syntactic representation of a syntactic adjunct (B) standard syntactic representation of an argument next to
its selecting head.

Mwi:ni (Swahili), Selkirk, 1986; for Tohono Oodham (Oodham), Hale and Selkirk, 1987) and that in
other languages the left edge of XP is aligned with
phonological phrase edges (for Japanese, Selkirk and
Tateishi, 1991; for Shanghai Chinese (Wu Chinese),
Selkirk and Shen, 1990). The two constraints, cast in
terms of Optimality Theory in Selkirk (1995a), are
Align-XP,R and Align-XP,L. Edge alignment was
extended to Maori by de Lacy (2003), who argued
that Align-XP,L and Align-XP,R are simultaneously
at play in this language, and to the Bantu languages
Chichew
a (Nyanja) and Kimatumbi (Matumbi) by
Truckenbrodt (1995, 1999), who argued that AlignXP,R interacts with a constraint Wrap-XP that punishes dividing XPs into more than one phonological
phrase.
The notion XP, which seems to independently play
a central role in the syntaxphonology interface,
also provides a generalization over predicates, arguments, and modifiers, with the exception of predicates next to an accented argument, in the SAAR.
Thus, a more principled formulation of the SAAR
that is defended in the following is: Each XP is
assigned a beat of phrasal stress. This is the content
of the constraint Stress-XP, proposed in Truckenbrodt
(1995) to account for patterns of phonological
phrasing in languages other than English, Dutch,
and German. (It was not related to the SAAR there.)
P-stress was there construed as stress on the level
of the phonological phrase (Nespor and Vogel,
1986, 1989).
For the purpose at hand, the application of StressXP to two core cases of the SAAR is shown in
Figure 4. Figure 4A is a standard syntactic representation of a syntactic adjunct: The adjunct is a phrase,
and the element it adjoins to (here VP) is likewise
a phrase. In this configuration, Stress-XP demands
p-stress in the adjunct XP as well as p-stress in the
element next to it, which is itself a phrase. (It is
dominated by one of the VP nodes; this VP node, by
Stress-XP, requires p-stress. See Truckenbrodt, 1999,

576 Phrasal Stress

for more detailed discussion of adjunction structures

in the syntaxphonology mapping; the matter is
formally more complex, but, for the case at hand,
the more detailed suggestions there have the same
consequences as the simplified discussion here.)
Figure 4B is a standard syntactic representation of
an argument next to its selecting head. The argument
is itself a maximal projection XP and a sister to a
head, X. Here Stress-XP demands stress in the argument XP, but it does not demand stress inside of its
sister, which is only a head X (here: V). Further, as
long as the argument is stressed, Stress-XP is also
satisfied for the VP in Figure 4B, which contains
stress in the position of the stressed argument.
Thus, the special status of predicates next to an
accented argument in the SAAR finds a simple explanation: Whereas the other categories listed in the
SAAR (in particular arguments and adjuncts) are
XPs, predicates next to an accented argument are
heads (X) that are not affected by Stress-XP. When
the predicate does not have an argument, it stands
alone in its XP like the lower VP in Figure 4A and,
thus, correctly requires stress by Stress-XP. When the
argument of a verb is not accented (as in to [VP read
something], in German [VP etwas lesen]), Stress-XP
also correctly requires p-stress in the VP. In this case,
the object rejects stress, so the p-stress within VP falls
on the verb.
I offer a new application of a proposal of my own
in this review of the literature because I believe that it
is a step forward relative to the other proposals. In
other accounts of p-stress, including, in particular,
Gussenhovens SAAR, Zubizarreta (1998), Bu ring
(2001), and Selkirk (1995b), the special status of the
head-argument relation is written into the rules or
constraints that relate syntactic and prosodic structure. Stress-XP, on the other hand, requires no statement about the head-argument relation or about any
other syntactic relations. Rather, the special assignment of stress in the head-argument configuration
is derived directly from the special syntax of the
head-argument configuration. This is the only case
in which an XP (such as VP) contains another XP
(such as the argument) and a non-XP (such as the
verb). If each XP must contain p-stress, it follows
that the argument XP must contain p-stress but that
the head (non-XP) need not.
Stress-XP shares with Cinques proposal the search
for an account that does not mention heads or arguments in assigning p-stress. It also picks up Cinques
intuition that the presence of more structure on the
side of the argument attracts stress. However, the
factor that generalizes to other cases is the presence
of an XP on the side of the argument, not the depth

of embedding in terms of numbers of nodes. Example

(5) shows how the examples discussed up to this
point and some related ones are derived using
Stress-XP. Brackets indicate XPs (omitting the entire
sentence and omitting pronouns). Boldface brackets
are XPs that contain only a single word. In these
cases, that single word must contain stress by StressXP. With this, the higher XPs, which necessarily
contain such a single-word XP, also satisfy Stress-XP
and contain p-stress. (Recall that, when the lowest
such XP rejects stress, as in to [eat [something]],
stress then defaults to the head of the next higher
XP, here VP. This shows that the application of
Stress-XP to branching XP nodes in (5) is not
vacuous.)
(5a) [the [brother [of
[Mary]]]]

[Mary]s [brother]

(5b) He was [teaching

[linguistics]]
Er hat [[Linguistik]
unterrichtet]

He was [[teaching] [in

[Ghana]]]
Er hat [[in [Ghana]]
[unterrichtet]]

(5c) [the [mayor [of [Chicago]]]] [won [their

[support]]]
(5d) [the [mayor]] [won [their [support]]]
[der [Burgermeister]] hat [[ihre [Unterstu tzung]]
gewonnen]
(5e) [Der [Peter]] hat [[der
[Schwester [des [Freundes
[von [Maria]]]]]]

[eine [Rose]]
geschenkt]

In all cases, the last p-stress is strengthened to the

strongest of the expression.
Stress-XP extends to the observation (Uhmann,
1991; Gussenhoven, 1992) that resultative predicates
need not receive p-stress, whereas secondary predicates must receive p-stress: throw one of the windows
open (Gussenhoven, 1992) vs. eat the porcupine raw.
In a standard small clause structure of resultatives,
their syntax mirrors the causative meaning, such that
[AP [DP [NP windows]] open] is the constituent
that describes the proposition being caused. In
this constituent, windows attracts p-stress by StressXP, whereas open, not an XP of its own, does not
require p-stress. The entire AP contains stress once
windows is stressed. In throw [AP it open], with
the pronoun it rejecting p-stress, we see the effect of
Stress-XP on the AP. Secondary predication does
not involve this thematic embedding of the secondary predicate. Here the object is in its normal
object position, and the secondary predicate is
an adjunct of some sort. In eat [the [porcupine]]
[AP raw], both the object and the AP require p-stress
by Stress-XP, of which the final p-stress is then
strengthened.

Phrasal Stress 577

Deaccenting and the Theory of Focus

The Alternative Semantics of Focus in Rooth
and Deaccenting

Because focus has a strong effect on p-stress, some

predictions about p-stress turn on the theory of the
semantic/pragmatic interpretation of focus. Rooth
(1992) saw a central part of the meaning of F in
the selection of the focused element from a set of
contextually relevant alternatives. In Who likes
Mary? [F Bill] likes Mary, the question defines a relevant range of possible answers; schematically John/
Bill/ . . . /Sue likes Mary. The actual answer asserts one
of these possibilities, and F highlights the part of the
answer to which there are relevant alternatives, here
the subject of the clause. This works similarly in John
likes Mary. No, [F Bill] likes Mary. F marking targets
that part of the second sentence to which there is a
relevant alternative (John) in the context.
The examples of deaccenting by Ladd (1980, 1983)
in (6) show that no contrast of this kind is necessary
for the stress-retracting effect of focus to occur.
(6a) A: What about Fred?

B: I dont like Fred.

(6b) Why dont you have some French toast?

I dont know how to make French toast.

There is no sense in which like in the first answer or

make in the second answer is contrasted or otherwise
juxtaposed with an alternative in some relevant way.
In these cases, as Ladd noted, contextual givenness of
the object seems to be enough for the stress/accent to
retract to an earlier element in the clause.
Two Levels of Focus Interpretation

Is it enough to destress given elements, then? After all,

in the original cases (Who likes Mary? Bill likes Mary
and John likes Mary. No, Bill likes Mary) destressing
of the given material likes Mary also correctly forces
stress to retract to the subject. Schwarzschild (1999)
argued that the theory must be more complex than
that. Example (7a) shows deaccenting of the object
because of its referential givenness in the context
question. In (7b), the object is given by an antecedent
in the same structural configuration. Moreover, in
(7b) both the object and the first part of the answer
without the object (She praised
) are independently given by the context question. With all
elements given in (7b), deaccenting of given elements
cannot predict the correct accent position. The NSR
or the SAAR might correctly predict stress in this
tied situation in (7b). I add example (7c) to
Schwarzschilds examples. It shows a similar case in
which all elements of the answer are independently

given. It establishes that the NSR and the SAAR are

not directly relevant in such givenness ties.
(7a) [What did Johnis motherk do?]
Shek praised himi.
(Or: Shek praised John.)
(7b) [Who did Johnis motherk praise?]
Shek praised himi. (Or: Shek praised John.)
(7c) [Who did Johnis motherk say praised heri?]
(Shei said that) hek (Or: (Shei said that) John
praised heri.)
praised heri.

One conclusion we can draw is that a theory with

two levels of focus is required. Such a theory was
developed by Selkirk (1995b). In this theory, the absence of the feature F is interpreted as contextual
givenness, and presence of the feature FOC (an
F feature not dominated by another F feature) is
interpreted in terms of Rooths alternative semantics
of focus. Both F and FOC affect stress. F helps account for the deaccenting cases in (6). FOC in
this theory correctly allows to predicts that stress in
(7b) and (7c) falls on the element asked for by the
question.
In Selkirks theory, the account in terms of the two
kinds of focus, F and FOC, is integrated with a mechanism of percolation of F that accounts for the generalizations discussed in terms of the SAAR and
Stress-XP. Schwarzschild (1999) showed a number
of empirical problems with the mechanism of focus
feature percolation. They include an argument
(Schwarzschild, 1999: 171) that the default stress
relation between heads and arguments is independent
of the distribution of F and FOC. This is as predicted
by the account in terms of Stress-XP.
Schwarzschild (1999) also explored the notion that
F is interpreted as an exemption from givenness and
that FOC need not be semantically interpreted, that
is, that the semantics of focus need not include a
requirement that the context provides alternatives to
FOC. Schwarzschilds suggestions leave open many
questions in regard to how his suggestions are to be
combined with mechanisms of default assignment of
accent/p-stress.
To connect the discussion of focus with the earlier
discussion in a way that is intuitively accessible in the
framework of this article, I put things together as in
the following list.
1. FOC, semantically interpreted as in Rooths
(1992) theory of focus, contains phrasal stress.
2. Avoid phrasal stress on contextually given
constituents.
3. Stress-XP: Each XP contains phrasal stress.
4. Strengthen the last phrasal stress of the intonation
phrase.

578 Phrasal Stress

Item 1 invokes Rooths theory of focus, plus the

requirement from Schwarzschild that FOC contain
p-stress. Item 2 summarizes the prosodic effect of
givenness, as discussed in this section. Both take
precedence over constraint 3, which connects the
syntactic structure to p-stress.
Jackendoffs requirement that the focus contain the
strongest stress here comes out of item 2: A successful
assignment of FOC in line with Rooths theory will be
surrounded by given material, so that item 2 forces
the main stress of the intonation phrase away from
the given material and into FOC. Item 1 is required
for a number of examples in Schwarzschilds discussion, such as: (John cited Mary) but he dissed Sue.
Here the p-stress on the verb dissed is assigned in
connection with the alternative cited in the preceding clause. This is mediated by FOC, which must
therefore have the effect of introducing p-stress
(Schwarzschild, 1992: 170, 171).
Item 2 is formally not in line with the conceptualization, otherwise adhered to since Jackendoff (1972),
that semantics (givenness) and phonology (p-stress)
are not directly connected but, instead, should be
mediated by focus features in the syntax. In Selkirk
(1995b) and Schwarzschild (1999), this mediation is
taken on by the feature F (not FOC), which is avoided
in the summary list given here to keep things simple.
Pronominal Elements

Definite pronouns such as she and his and indefinite

pronouns such as something and someone are XPs
(DPs) but do not show the stress expected of XPs
(unless narrowly focused): Mary likes John; but She
likes him. Mary likes someone. This can be subsumed
under the general requirement of destressing of given
material (see item 2 of the summary list). The definite
pronouns are referentially given by coreference with
an element in the context (van Deemter, 1994;
Schwarzschild, 1999). The indefinite pronouns something and someone can also be construed as given:
Every context entails the presence of something and
of someone. By the avoidance of F marking in
Schwarzschild (1999) (i.e., by interpreting constituents as given as much as the context allows), these
elements can be interpreted as given, and therefore
they must be interpreted as given. Therefore, they
must be destressed.

Summary
Phrasal stress is assigned in a representation with
constituents and prominence, extending the prosodic
or metrical representation of stress within the word
upward. Focus and syntactic structure both have a

role in shaping the pattern of p-stress in a given

utterance. Contextually given elements avoid phrasal
stress, and elements contrasted with alternatives seem
to require phrasal stress. When focus does not interfere, a combination of two simple generalizations
seems to go a long way in accounting for observed
patterns: Each XP contains a beat of phrasal
stress (generalizing over the core of Gussenhovens
SAAR) and the final phrasal stress thus assigned in
the intonation phrase is strengthened.

See also: Focus; Intonation; Metrical Phonology; Phonological Phrase.

Bibliography
Beckman M E & Pierrehumbert J B (1986). Intonational
structure in Japanese and English. Phonology Yearbook
3, 255309.
Bresnan J (1971). Sentence stress and syntactic transformations. Language 47, 257281.
Bu ring D (2001). Lets phrase it! focus, word order, and
prosodic phrasing in German double object constructions. In Mu ller G & Sternefeld W (eds.) Competition
in syntax. Amsterdam: John Benjamins. 101137.
Chen M Y (1987). The syntax of Xiamen tone sandhi.
Phonology Yearbook 4, 109149.
Chomsky N & Halle M (1968). The sound pattern of
English. New York: Harper and Row.
Cinque G (1993). A null theory of phrase and compound
stress. Linguistic Inquiry 24, 239297.
de Lacy P (2003). Constraint universality and prosodic
phrasing in Maori. In Coetzee A, Carpenter A & de
Lacy P (eds.) Papers in Optimality Theory II. Amherst,
MA: GLSA. 5979.
Gussenhoven C (1983a). Focus, mode and the nucleus.
Journal of Linguistics 19, 377417.
Gussenhoven C (1983b). Testing the reality of focus
domains. Language and Speech 26, 6180.
Gussenhoven C (1992). Sentence accents and argument
structure. In Roca I M (ed.) Themantic structure: its
role in grammar. New York: Foris. 79106.
Hale K & Selkirk E (1987). Government and tonal
phrasing in Papago. Phonology Yearbook 4, 151183.
Halle M & Vergnaud J-R (1987). An essay on stress. Cambridge, MA: MIT Press.
Hayes B (1995). Metrical stress theory: principles and case
studies. Chicago, IL: University of Chicago Press.
Hayes B & Lahiri A (1991). Bengali intonational phonology. Natural Language and Linguistic Theory 9, 4796.
Jackendoff R S (1972). Semantic interpretation in generative grammar. Cambridge, MA: MIT Press.
Jacobs J (1993). Integration. In Reis M (ed.) Wortstellung
und Informationsstruktur. Tu bingen: Niemeyer. 63116.
Krifka M (1984). Focus, Topic, syntaktische Struktur und
semantische Interpretation. Ms., Universita t Mu nchen.

Phraseology 579
Ladd D R (1980). The structure of intonational meaning:
evidence from English. Bloomington, IN: Indiana University Press.
Ladd D R (1983). Even, focus, and normal stress. Journal
of Semantics 2, 257270.
Liberman M & Prince A (1977). On stress and linguistic
rhythm. Linguistic Inquiry 8, 249336.
Nespor M & Vogel I (1986). Prosodic phonology.
Dordrecht: Foris.
Nespor M & Vogel I (1989). On clashes and lapses.
Phonology 6, 69116.
Pierrehumbert J B (1980). The phonology and phonetics
of English intonation. Ph.D. diss., Massachusetts
Institute of Technology.
Pierrehumbert J B & Beckman M E (1988). Japanese tone
structure. Cambridge, MA: MIT Press.
Rooth M (1992). A theory of focus interpretation. Natural
Language Semantics 1, 75116.
Schwarzschild R (1999). Givenness, AvoidF and other constraints on the placement of accent. Natural Language
Semantics 7, 141177.
Selkirk E (1980). The role of prosodic categories in English
word stress. Linguistic Inquiry 11, 563605.
Selkirk E (1984). Phonology and syntax: the relation between sound and structure. Cambridge, MA: MIT Press.
Selkirk E (1986). On derived domains in sentence phonology. Phonology Yearbook 3, 371405.
Selkirk E (1995a). The prosodic structure of function
words. In Beckman J, Dickey L W & Urbanczyk S

(eds.) University of Massachusetts occasional papers 18.

Papers in optimality theory. Amherst, MA: GLSA.
Selkirk E (1995b). Sentence prosody: intonation, stress,
and phrasing. In Goldsmith J (ed.) The handbook
of phonological theory. Cambridge, MA: Blackwell.
550569.
Selkirk E & Shen T (1990). Prosodic domains in Shanghai
Chinese. In Inkelas S & Zec D (eds.) The phonology
syntax connection. Chicago, IL: University of Chicago
Press. 313337.
Selkirk E & Tateishi K (1991). Syntax and downstep in
Japanese. In Georgopolous C & Ishihara R (eds.) Interdisciplinary approaches to language: essays in honor of
S.-Y. Kuroda. Dordrecht: Kluwer. 519543.
Szendro i K (2001). Focus and the syntaxphonology interface. Ph.D. diss., University College London.
Truckenbrodt H (1995). Phonological phrases: their relation to syntax, focus, and prominence. Ph.D. diss.,
Massachusetts Institute of Technology.
Truckenbrodt H (1999). On the relation between syntactic
phrases and phonological phrases. Linguistic Inquiry 30,
219255.
Uhmann S (1991). Fokusphonologie. Tu bingen: Niemeyer.
van Deemter & Kees (1994). Whats new? a semantic
perspective on sentence accent. Journal of Semantics
11, 131.
Zubizarreta M L (1998). Prosody, focus, and word order.
Cambridge, MA: MIT Press.

Phraseology
A Cowie, University of Leeds. Leeds, UK
2006 Elsevier Ltd. All rights reserved.

Introduction
The article devoted to phraseology in the first edition
of the Encyclopedia of language and linguistics (ELL)
reported a marked upsurge of interest in the subject
over the previous 10 years among linguists, lexicographers, and specialists in language acquisition and
language teaching (Cowie, 1994). The opening references in that article were chiefly to anglophone scholars, but the growing activity in the field in Britain,
and to a lesser extent in the United States, owed much
to a substantial, earlier body of East European work
in phraseology, stretching back to the 1940s. In fact,
we can see, if we examine the years as indicated by the
extensive Euralex Bibliography of Phraseology, now
available on the Web in which the most productive
specialists began to publish, that scholars from
throughout the USSR and from East Germany and
Czechoslovakia were the dominant group in the 1960s

and 1970s; that a marked growth of activity could be

observed in the late 1970s and early 1980s among
Austrian, Swiss and West German linguists; and that
the 1980s witnessed the quickening of interest which
I began by referring to: in Britain, but also in Belgium,
Spain, Italy, and Scandinavia. By the end of the 20th
century, something approaching a truly international
body had come into being, with its own professional
association, and regular meetings and publications.
There are still, of course, differences between the
various regional groups as regards the fields of specialization to which they are typically attracted. Thus,
while the first two groups to which I referred and
which incidentally have until recently made up the
bulk of the membership of the European Association
for Phraseology (EUROPHRAS) are often drawn
to historical, dialect, media, and translation studies
(Palm-Meister, 2004), the British and Belgian phraseologists who are active in the field have tended to
be attracted to (EFL) lexicography, the corpus-based
analysis of recurrent word combinations, and the role
of phraseology in language teaching and language

Phraseology 579
Ladd D R (1980). The structure of intonational meaning:
evidence from English. Bloomington, IN: Indiana University Press.
Ladd D R (1983). Even, focus, and normal stress. Journal
of Semantics 2, 257270.
Liberman M & Prince A (1977). On stress and linguistic
rhythm. Linguistic Inquiry 8, 249336.
Nespor M & Vogel I (1986). Prosodic phonology.
Dordrecht: Foris.
Nespor M & Vogel I (1989). On clashes and lapses.
Phonology 6, 69116.
Pierrehumbert J B (1980). The phonology and phonetics
of English intonation. Ph.D. diss., Massachusetts
Institute of Technology.
Pierrehumbert J B & Beckman M E (1988). Japanese tone
structure. Cambridge, MA: MIT Press.
Rooth M (1992). A theory of focus interpretation. Natural
Language Semantics 1, 75116.
Schwarzschild R (1999). Givenness, AvoidF and other constraints on the placement of accent. Natural Language
Semantics 7, 141177.
Selkirk E (1980). The role of prosodic categories in English
word stress. Linguistic Inquiry 11, 563605.
Selkirk E (1984). Phonology and syntax: the relation between sound and structure. Cambridge, MA: MIT Press.
Selkirk E (1986). On derived domains in sentence phonology. Phonology Yearbook 3, 371405.
Selkirk E (1995a). The prosodic structure of function
words. In Beckman J, Dickey L W & Urbanczyk S

(eds.) University of Massachusetts occasional papers 18.

Papers in optimality theory. Amherst, MA: GLSA.
Selkirk E (1995b). Sentence prosody: intonation, stress,
and phrasing. In Goldsmith J (ed.) The handbook
of phonological theory. Cambridge, MA: Blackwell.
550569.
Selkirk E & Shen T (1990). Prosodic domains in Shanghai
Chinese. In Inkelas S & Zec D (eds.) The phonology
syntax connection. Chicago, IL: University of Chicago
Press. 313337.
Selkirk E & Tateishi K (1991). Syntax and downstep in
Japanese. In Georgopolous C & Ishihara R (eds.) Interdisciplinary approaches to language: essays in honor of
S.-Y. Kuroda. Dordrecht: Kluwer. 519543.
Szendroi K (2001). Focus and the syntaxphonology interface. Ph.D. diss., University College London.
Truckenbrodt H (1995). Phonological phrases: their relation to syntax, focus, and prominence. Ph.D. diss.,
Massachusetts Institute of Technology.
Truckenbrodt H (1999). On the relation between syntactic
phrases and phonological phrases. Linguistic Inquiry 30,
219255.
Uhmann S (1991). Fokusphonologie. Tubingen: Niemeyer.
van Deemter & Kees (1994). Whats new? a semantic
perspective on sentence accent. Journal of Semantics
11, 131.
Zubizarreta M L (1998). Prosody, focus, and word order.
Cambridge, MA: MIT Press.

Phraseology
A Cowie, University of Leeds. Leeds, UK
2006 Elsevier Ltd. All rights reserved.

Introduction
The article devoted to phraseology in the first edition
of the Encyclopedia of language and linguistics (ELL)
reported a marked upsurge of interest in the subject
over the previous 10 years among linguists, lexicographers, and specialists in language acquisition and
language teaching (Cowie, 1994). The opening references in that article were chiefly to anglophone scholars, but the growing activity in the field in Britain,
and to a lesser extent in the United States, owed much
to a substantial, earlier body of East European work
in phraseology, stretching back to the 1940s. In fact,
we can see, if we examine the years as indicated by the
extensive Euralex Bibliography of Phraseology, now
available on the Web in which the most productive
specialists began to publish, that scholars from
throughout the USSR and from East Germany and
Czechoslovakia were the dominant group in the 1960s

and 1970s; that a marked growth of activity could be

observed in the late 1970s and early 1980s among
Austrian, Swiss and West German linguists; and that
the 1980s witnessed the quickening of interest which
I began by referring to: in Britain, but also in Belgium,
Spain, Italy, and Scandinavia. By the end of the 20th
century, something approaching a truly international
body had come into being, with its own professional
association, and regular meetings and publications.
There are still, of course, differences between the
various regional groups as regards the fields of specialization to which they are typically attracted. Thus,
while the first two groups to which I referred and
which incidentally have until recently made up the
bulk of the membership of the European Association
for Phraseology (EUROPHRAS) are often drawn
to historical, dialect, media, and translation studies
(Palm-Meister, 2004), the British and Belgian phraseologists who are active in the field have tended to
be attracted to (EFL) lexicography, the corpus-based
analysis of recurrent word combinations, and the role
of phraseology in language teaching and language

580 Phraseology

acquisition (Cowie, 1998; Granger, 1998; Moon,

1998a, 1998b). Recent volumes of conference proceedings, however, suggest that the edges are becoming blurred (e.g., Burger et al., 2003), and it should be
noted especially that the treatment of phraseology in
dictionaries of English has, since the 1980s, been a
beneficiary of achievements in the categorization of
word combinations pioneered in the former Soviet
Union (Cowie, 1998b). It is the connection in a
wider sense between phraseological research and dictionary making that will constitute the major theme
of this article.
There has been a partial resolution, too, of differences of another kind. A decade ago, it was possible
to refer to the continuing influence in collocational
analysis of neo-Firthian lexical theory . . . with its
emphasis on observed frequency of co-occurrence
within stated distances . . . in a large computerized
corpus (Cowie, 1994: 3168), in contrast to the approach based on a framework of descriptive categories that I have just referred to. These positions
are not as firmly entrenched as they were then. As
we shall see later, present-day phraseologists from
quite diverse backgrounds acknowledge the benefits
that can accrue from an approach which combines
the advantages of access to large-scale corpus data
and the value of recognizing, as part of the analytical
process, the grammatical and pragmatic functions
that are served by multiword units (Moon, 1998a,
1998b; Cowie, 1999; Kilgarriff and Tugwell, 2002;
Cermak, 2003).

There were also a number of opportunities and

challenges. Consider, first, collocations and the
forms which English collocational dictionaries have
taken in recent years. The most recent to appear is the
result of a coming together of theoretical insights,
corpus analysis, and user-sensitive design features
(Crowther et al., 2002). Another question on which
light has been shed in recent research is the possible
variability of word combinations. As Moon has
helped us to realize, even the most opaque of idioms
are more often subject to structural and/or lexical
change by no means always the result of creative
manipulation than we have hitherto been led to
believe (Moon, 1998a, 1998b). This phenomenon
has an important bearing on the description of idioms
in phraseological dictionaries, and will also be
explored in more detail below.
Another area in which progress has been made, in
the past decade, has to do with word combinations
that function at the pragmatic level, as distinct from
the grammatical level. Though much work had been
done earlier in Eastern Europe, and elsewhere, on
proverbs as one subcategory at the pragmatic level
as much for their cultural as for their linguistic interest there remained the challenge of such speech
formulae (also referred to as gambits) as you
know what I mean, or to put it another way, which
posed considerable descriptive difficulties. We shall
return to these, also. (For accounts of recent work in
this general area, see Arnaud and Moon, 1993;
Cowie, 2001; Cermak, 2003.)

Progress in Description

Collocations and Collocational Dictionaries

When, in the late 1970s and early 1980s, British

linguists became aware of the considerable progress
already made by their East European colleagues, they
were chiefly impressed by their success in identifying
relevant categories and introducing appropriate criteria for their definition (Cowie et al., 1983). They
had recognized a major division between combinations with a pragmatic function (sometimes called
propositions) and those with a referential function
(nominations) and were aware that the latter were
ranged along a continuum, from transparent free
combinations at one end to relatively fixed and
opaque idioms at the other. This is now part of
received wisdom, though as we shall see, emphasis
needs to be laid on the relatively of relatively fixed
(Howarth, 1996). But there was some unfinished
business, including the need to harmonize the sometimes conflicting analytical schemes and categories
used by various individuals and schools. This was
one of the themes of a conference organized in Leeds
in 1994 (Cowie, 1998d).

I referred, in the first edition of ELL, to collocations

as associations of two or more lexemes (or roots)
recognized in and defined by their occurrence in a
specific range of grammatical constructions. The
phrase recognized in is an acknowledgement of the
fact that collocations are perceived, in speech or
writing, as surface manifestations, or realizations,
of abstract composites (Cowie, 1997: 43). It is this
assumption that enables us to claim that heavy rain
and rain heavily are, as it were, collocation forms
of the same lexemic collocation and to account for
their evident relatedness in meaning (cf. Howarth,
1998: 168). Acknowledging the difference between
the two levels is vitally important, as there is still
a tendency, in the name of empirical soundness,
to regard collocation forms as the proper focus of
description.
A related point has to do with the importance
of analyzing such combinations as hard cash, fresh
air, break someones fall, and corner the market
all restricted collocations with reference to the

Phraseology 581

syntactic constructions in which they function. In

those cases the relevant constructions are attributive
adjective head noun and transitive verb noun
object. The wisdom of adopting this view has been
accepted for some time by the compilers of collocational dictionaries. Indeed, one can go further and
point to the precedent set by Harold Palmer and
A. S. Hornby, who in 1933 published the Second
interim report on English collocations, the first linguistically based analysis of English phraseology ever
to appear, and one in which the various combinations
were organized, to some degree of delicacy, according
to the syntactic patterns in which they functioned
(Cowie, 1998a; Palmer, 1933).
It is interesting, to return to the present, that a
number of computational projects of recent years
concerned with collocation extraction have made
the same assumptions regarding the grammatical
functions of collocations. Specialists working on the
DECIDE project, for example, were concerned with
the development of tools for the extraction of collocations from tagged and raw text (Grefenstette et al.,
1996). The DECIDE team developed tools such as
morphological analyzers, part-of-speech taggers and
low-level parsers each of the tools, note, corresponding to a particular level of abstraction. The
working of the parser was especially interesting.
Given a sentence such as A radiator has been fitted
in the bathroom, the parser was able to identify the
semantic object of the verb (here radiator) even
though this precedes the verb as the grammatical
subject of a passive construction. It therefore overcame the obstacle to recovering the members of a collocation (e.g., fit radiator) created by their syntactic
displacement within the same sentence (Howarth,
1996).
But to return to collocational dictionaries, one need
only refer to a few entries in Selected English collocations (SEC), in its first edition (Kozowska and
Dzierzanowska, 1982), to realize that grammatical
function in all entries provides the basis for an overall
organization whereby the user accesses each collocation first by its constituent noun, which is also a
capitalized headword, and second by its verb or adjective, which appears as part of a labeled list, as in
this short excerpt:
DICTIONARY
V. compile, consult, look up, produce, refer to,
revise, use 
Adj. compact, comprehensive, concise,
excellent, extensive . . . 

As the layout of that entry shows, the dictionary was

designed to serve the needs of the user as writer,
the assumption having been made that, when writing,

he or she moves from the autonomous noun

the base to a collocate verb or adjective
whose selection is problematic (Cowie, 1998c).
There was, though, one problem that SEC did not
address. Ideally, having arranged collocates according
to their word class, the lexicographers should have
introduced a feature designed to help the user locate a
specific semantic choice (forge, say, as opposed to
sever as collocates of the noun link). Such a device is
not present in SEC, where the collocates are arranged
alphabetically, as shown above (cf. Hausmann, 1985).
The BBI dictionary of word combinations (BBI), in
both its editions, went some way towards supplying
this particular need. From the entry for help, noun,
below, it can be seen that there is structuring on two
levels, syntactic and semantic. The first three numbered subsections deal with transitive verb noun
collocations, while numbers 4 and 5 are devoted to
adjective noun collocations. The allocation of the
verb noun collocations to subsections 13, however, is based on meaning differences. The verb collocates at 1, for example, have to do with the giving of
help (give, offer, provide), while those at 2 have
to do with seeking it (call for, seek). This organizational pattern, kept up throughout BBI, enables the
user to home in on highly specific areas of meaning,
and then to make individual selections.
help I n. 1. to give, offer, provide  2. to call for, seek
 3. to cry for, plead for  4. a big  (to) . . . 5. (a)
great, invaluable, tremendous; little 

The more recent publication of the Oxford collocations dictionary for students of English (OCDSE)
(Crowther et al., 2002) represents a further major
advance in the design of English collocational dictionaries. It owes, of course, certain of its features to its
two predecessors. Like SEC, OCDSE is based on
authentic, largely written material, except that in
the latter case the corpus is the immensely bigger
British National Corpus (BNC). Moreover, the orientation of the dictionary, as earlier in both SEC and
BBI, is from an independent base to a dependent
collocate. However, the most significant improvements made by the compilers of OCDSE have to do
with the semantic organization of collocates within
an overall arrangement that is syntactic. In the entry
for enthusiasm, part of which is reproduced below,
there are two verbal patterns, represented as VERB
ENTHUSIASM and ENTHUSIASM VERB, forms which convey the transitiveintransitive distinction without
making too many demands on the grammatical
knowledge of the user. Within the transitive subsection the collocates are broken up into semantic groupings, a feature as we recall of BBI, except that here
the collocates are in bold, there are vertical strokes to

582 Phraseology

mark off the different semantic sets, and examples are

introduced to illustrate some of the combinations:
enthusiasm noun
 VERB ENTHUSIASM be full of, feel, have Her voice
was full of enthusiasm. | convey, express, show She
managed to convey an enthusiasm she did not
feel. . . .
 ENTHUSIASM VERB bubble over/up trying to hide the
boyish enthusiasm bubbling up inside him | grow |
fade, wane, wear off . . .
Idioms and Variability

Phraseologists commonly refer to the formal fixedness of word combinations, and especially of idioms
in the strict sense. Until recently, though, it has been
less widely recognized that, in some corpora, a high
percentage of given multiword units including such
opaque examples as take the biscuit or up the anti do
not have frozen or fixed canonical forms. Working
with a corpus of 18 million words, Moon (1998a,
1998b) has shown that this proportion can be as
high as 40 per cent, a finding which has serious
implications for teaching, lexicography, and such matters as the automatic and computational detection of
such items in corpora (1998b: 92).
Variability differs in type and in degree. One type
to which attention has been drawn is systematic, or
in some cases partly so. It manifests itself as a set
of formally and semantically related variants:
get your eye in
keep your eye in
have your eye in
have (no, an) axe to grind
with(out) an axe to grind

More usual, perhaps, are multiword expressions with

synonymous variants. Here, it is the variants as
wholes that are equivalent: the substituting lexemes
are not necessarily so:
burn ones boats vs. burn ones bridges
hit the roof vs. hit the ceiling

U.S. and British speakers may prefer different variants, as in the case of sweep something under the
carpet (U.K.) vs. brush something under the rug
(U.S.), and indeed contrasting variants may represent
different levels of formality, as witness make a fortune
vs. make a killing (Moon, 1998b: 93).
Particular difficulties are caused by the treatment in
phraseological dictionaries of series of variants such
as be in power, come into power, and put somebody
into power. Because of the different syntactic-semantic patterns involved (they are respectively copular,
inchoative, and causative) they are treated in the

Oxford dictionary of phrasal verbs as separate

expressions, with their relatedness being shown by
means of cross-references (Cowie and Mackin, 1993):
be in power [Vpr] have reached or gained (political)
control. . . . ! come/put into power.

The examples we have been examining have all

involved lexical alternation at one structural point.
However, multiword expressions are by no means
unusual in which complex choices come into play at
two or more places. Large-scale text corpora such as
the BNC are already making an important contribution to the analysis of such items, though in some
cases corpus data may raise as many questions as
they answer. Consider the word combination that
has as one of its possible forms the underlined portion
of this example:
To lock up young car thieves is another example of
bolting the stable door after the horse has fled.

Of the 38 instances of the phrase the stable door

found in the BNC, 13 form part of a longer fixed
stem the stable door after the horse, immediately
preceded by one of a set of five verbs, shut, bolt,
close, lock, and slam, and immediately followed by
one from a choice of three, bolt, flee, and steal. It is
worth noting that the first verb slot is obligatorily
filled, in over half the cases by a present participle
form, and that the last position is usually occupied,
too, almost always by a perfective verb form, though
truncated patterns such as its a bit like shutting the
stable door are occasionally found.
There are two questions that this evidence causes
one to reflect on, apart from its helpful specificity at
several points. The first is that the frequencies in
initial position of shut, bolt, and close, at 4, 4, and 3
respectively, though tiny in a corpus of 100 million
words, are intuitively satisfying and, incidentally,
match the choices indicated in a phraseological dictionary based on a much smaller (written) corpus
gathered by traditional methods (Cowie et al., 1983).
The BNC data also indicate that, whatever the choice
made initially, bolt is by far the preferred selection
in final position. (No support, incidentally, can be
found in the BNC data for the alternatives suggested
for end position in that same phraseological dictionary, which include [have] run away, gone, and
disappeared.)
The second point has to do with the meanings of
the verbs in both positions. Clearly, shut, bolt, and
close are related in sense, and this may have prompted
use of the slightly more unusual but also semantically related lock and slam recorded in the BNC.
Here there is a difficulty, even within a corpus of
100 million words, both for the analyst and for the

Phraseology 583

theoretician tempted to build too strong a case on

frequencies of occurrence. The specific problem is
that faced with frequencies of no more than 4 or 3
for the verbs shut, bolt, close, and the evidence, as
already suggested, of their semantic relatedness, it
seems that their selection is as likely to be accounted
for by generative processes as by repetitive use and
memorization. It is arguable that, in such cases, reference to corpora should be backed up by tests designed
to elicit the implicit knowledge of informants (Cowie,
2003).
Speech Formulae

As we have seen, it is now customary to recognize a

distinction between nominations (also called composites), which function at or below the level of the
simple sentence, and are divisible into idioms and
collocations, and propositions (pragmatic combinations), which are often sentences and which function as proverbs, catchphrases, and slogans (Cowie,
1998a).
Examples of the second major type are Too many
cooks spoil the broth (proverb), The buck stops here
(catchphrase), and All we do is driven by you (slogan)
(Gla ser, 1990, 1998). Alongside those familiar, traditional categories, of which Gla ser provides a detailed
description, we need to consider a class which I shall
refer to as routine formulae and distinguish carefully
from speech formulae (sometimes referred to as
gambits) (Cowie, 2001). Then we need to ask
whether those important types can be further divided.
All this detailed work is of great importance for
dictionary making.
Three of the expressions quoted by Gla ser are mind
the step, many happy returns, and hold your horses
(Gla ser, 1998: 127). These belong to the same category they are all routine formulae. They commonly
occur without any expansion or verbal reaction from
another speaker though note please mind the step,
many happy returns of the day, and hold your horses
a minute and they function as warnings, greetings,
prohibitions, and so on that is, as various kinds of
socially recurrent speech act (Moon, 1998a).
It is tempting to regard I beg your pardon and you
know what I mean, which are speech formulae, as
belonging to the same category as mind the step or
good morning. After all, the first pair, like the second,
can occur as independent sentences, and all four have
a speech act function. The two speech formulae are
used, respectively, as an apology or request for clarification and as a check on understanding, as can be
seen from the following examples:
Were in trouble if it wont fit in. I beg your pardon,
if it wont fit in what?

I had to have what Ive wanted by hook or by crook,

and I dont mean crook in a bad sense I mean one
way or another, you know what I mean.

But those examples suggest crucial differences, both

between the routine formulae and the speech formulae, and between the latter two. The speech formulae,
as here, occur almost as often as not in a wider
sentential context (which mind the step typically
does not). Moreover, I beg your pardon functions
chiefly as a response to something another speaker
has said, while you know what I mean, though not a
reaction to another person, serves to check that he
or she has grasped the meaning of by hook or by
crook (which is made explicit by I mean one way or
another). The second formula thus has reference both
to the listener and to the language of the speakers
own utterance (Cowie, 2001).
Although many happy returns and mind the step
are addressed to a listener or reader, they are not
essentially interactive: they do not require a verbal
response, nor do they serve as one. Speech formulae
are, by contrast, typically interactive, a distinction
that calls for the setting up of a category parallel to
but separate from routine formulae.
The two kinds of formulae, then, constitute separate classes; but there are internal differences, too, as
we have seen. For example, the sentence that
contained I beg your pardon served to express the
speakers response to an earlier utterance by somebody else. Yet it is by no means always the case that a
speech formula is a part of a response to something
another person has said. In fact, very many formulae
have to do with some other kind of interaction. One
example must suffice. It is a case in which the function of the formula and its position are both significant. The formula is are you with me? and its function
is to check whether the listener is following and understanding what is said. But if the check can be
applied at any moment, then the formula can appear
at any point, too, even to the extent of disrupting
main clause structure:
It could be a car, like weve just said, but then again,
Ive never had a new car. It could be are you with
me?

Phraseology as an Academic Subject

There is no doubt that as a result of the developments,
institutional as well as intellectual, of the past
decade, phraseology has now acquired the status of
an academic subject (cf. Pawley, 2001). EUROPHRAS, the professional association which began
with a European base and membership, is now becoming worldwide in its membership and its recognition of shared aims and interests. It may soon have its

584 Phraseology

own yearbook. It is noteworthy, too, that the subject

is being taught to doctorate level in a number of
universities in several countries, of which Spain is an
outstanding example. As Corpas Pastor has shown in
a detailed survey of research undertaken in Spain
since the late 1980s, the subject has impressively
established its credentials in Spanish universities
(Corpas Pastor, 2003). Then, in addition to the conferences held regularly by EUROPHRAS, there have
been a number of workshops organized by interest
groups, e.g., Grossmann and Pantin on collocation
(2003). Finally, and not least, there is wide agreement
among colleagues as to what constitutes a program of
academic studies in phraseology. No student textbook as yet exists that captures all the major subfields
in which phraseologists are active, but we do have a
number of introductory readers: Corpas Pastor
(1997) for Spanish, Burger (1998) for German, and
Cowie (1998d) for English.
See also: Collocations; Computers in Lexicography; Cor-

pus Approaches to Idiom; Lexicography: Overview; Lexicon Grammars.

Bibliography
Arnaud P & Moon R M (1993). Fre quence et emploi des
proverbes anglais et franc ais. In Plantin C (ed.) Lieux
communs: topo, stereotypes, cliches. Paris: Kime .
323341.
Benson M, Benson E & Ilson R F (1997). The BBI
dictionary of English word combinations (2nd edn.).
Amsterdam: John Benjamins.
Burger H (1998). Phraseologie: eine Einfuhrung am Beispiel
des Deutschen. Berlin: Erich Schmidt.
Burger H, Ha cki-Buhofer A & Gre ciano G (eds.) (2003).
Flut von Texten Vielfalt der Kulturen. Baltmannsweiler,
Germany: Schneider.
C erma k F (2003). Paremiological minimum of Czech: the
corpus evidence. In Burger et al. (eds.). 1531.
Corpas Pastor G (1997). Manual de fraseologa espanola.
Madrid: Gredos.
Corpas Pastor G (2003). Diez anos de investigacion en
fraseologa: analisis sintactico-semanticos, contrastivos
y traductologicos. Madrid: Iberoamericana.
Cowie A P (1994). Phraseology. In Asher R E (ed.) The
encyclopedia of language and linguistics. Oxford/New
York: Pergamon Press. 31683171.
Cowie A P (1997). Phraseology in formal academic prose.
In Aarts J, de Mo nnink I & Wekker H (eds.) Studies in
English language and teaching. Amsterdam/Atlanta:
Rodopi. 4356.
Cowie A P (1998a). A. S. Hornby, 18981998: a centenary
tribute. International Journal of Lexicography 11(4),
251268.
Cowie A P (1998b). Introduction. In Cowie A P (ed.). 120.
Cowie A P (1998c). Phraseological dictionaries: some east
west comparisons. In Cowie (ed.). 209228.

Cowie A P (ed.) (1998d). Phraseology: theory, analysis, and

applications. Oxford: Clarendon Press.
Cowie A P (1999). Phraseology and corpora: some implications for dictionary-making. International Journal of
Lexicography 12(4), 307323.
Cowie A P (2001). Speech formulae in English: problems of
analysis and dictionary treatment. In van der Meer G &
ter Meulen A (eds.) Making sense: from lexeme to discourse. Groningen: Center for Language and Cognition
Groningen. 112.
Cowie A P (2003). Exploring native-speaker knowledge of
phraseology: informant testing or corpus research? In
Burger H et al. (eds.). 7381.
Cowie A P & Mackin R (1975). Oxford dictionary of
current idiomatic English I: Verbs with prepositions and
particles. Oxford: Oxford University Press.
Cowie A P & Mackin R (1993). Oxford dictionary of
phrasal verbs. Oxford: Oxford University Press.
Cowie A P, Mackin R & McCaig I R (1983). Oxford
dictionary of current idiomatic English 2: Phrase,
clause and sentence idioms. Oxford: Oxford University
Press.
Cowie A P, Mackin R & McCaig I R (1993). Oxford
dictionary of English idioms. Oxford: Oxford University
Press.
Crowther J, Dignen S & Lea D (eds.) (2002). Oxford
collocations dictionary for students of English. Oxford:
Oxford University Press.
Euralex Bibliography of Phraseology. http://www.ims.unistuttgart.de/euralex/bibweb
Gla ser R (1990). Phraseologie der englischen Sprache (2nd
edn.). Leipzig: VEB Verlag Enzyklopa die.
Gla ser R (1998). The stylistic potential of phraseological
units in the light of genre analysis. In Cowie A P (ed.).
125143.
Granger S (1998). Prefabricated patterns in advanced
EFL writing: collocations and formulae. In Cowie A P
(ed.). 145160.
Grefenstette G, Heid U, Schulze B M, Fontenelle T &
Gerardy C (1996). The DECIDE project: multilingual
collocation extraction. In Gellerstam M, Ja rborg J,
Malmgren S-G, Nore n K, Rogstro m L & Papmehl C R
(eds.) Euralex 96: proceedings, part 1. Gothenburg:
Go teborgs Universitet. 93107.
Grossmann F & Plantin A (2003). Quelques pistes pour le
traitement des collocations. In Grossmann F & Plantin A
(eds.) Les Collocations: analyse et traitement. Amsterdam:
De Wereld. 521.
Hausmann F J (1985). Kollokationen im deutschen
Wo rterbuch: ein Beitrag zur Theorie des lexikographischen Beispiels. In Bergenholtz H & Mugdan J
(eds.) Lexikographie und Grammatik. Tu bingen: Max
Niemeyer. 118129.
Howarth P A (1996). Phraseology in English academic
writing: some implications for language teaching and
dictionary making. Tu bingen: Max Niemeyer.
Howarth P A (1998). The phraseology of learners academic writing. In Cowie A P (ed.). 161186.
Kilgarriff A & Tugwell D (2002). Sketching words. In
Corre ard M-H (ed.) Lexicography and natural language

Phrastic, Neustic, Tropic: Hares Trichotomy 585

processing: a festschrift in honour of B. T. S. Atkins.
Grenoble: Euralex. 125137.
Kozowska C D & Dzierz anowska H (1982). Selected
English collocations. Warsaw: Pan stwowe Wydawnictwo Naukowe.
Kozowska C D & Dzierz anowska H (1988). Selected
English collocations (2nd edn.). Warsaw: Pan stwowe
Wydawnictwo Naukowe.
Moon R M (1998a). Fixed expressions and idioms in
English: a corpus-based approach. Oxford: Oxford
University Press.

Moon R M (1998b). Frequencies and forms of phrasal

lexemes in English. In Cowie A P (ed.). 79100.
Palm-Meister C (ed.) (2004). Europhras 2000. Tu bingen:
Stauffenburg.
Palmer H E (1933). The second interim report on English
collocations. Tokyo: Kaitakusha.
Pawley A (2001). Phraseology, linguistics and the dictionary. International Journal of Lexicography 14(2),
122134.
Sinclair J, Fox G & Moon R (1995). Collins COBUILD
dictionary of idioms. London: Harper Collins.

Phrastic, Neustic, Tropic: Hares Trichotomy

K Allan, Monash University, Clayton, VIC, Australia
2006 Elsevier Ltd. All rights reserved.

The terms phrastic, neustic, and tropic were introduced to the theory of speech acts by philosopher
Richard M. Hare. Hare (1949) had compared pairs
like the imperative in (1) with the declarative in (2):
(1) Keep to the path.
(2) You will keep to the path.

He concluded that (1) and (2) have the same phrastic, but a different neustic, which he characterized
as follows:
(3) [Keeping to the path by you]phrastic[please]neustic
(4) [Keeping to the path by you]phrastic[yes]neustic

It is tempting to symbolize neustic please by !

and yes by , but Hare (1970) gives us reason to
compare (2) with (5) and (6).
(5) You will keep to the path!
(6) Will you keep to the path!(?)

Although (5) has the declarative form of (2), it has

what Hare calls the same subscription (illocutionary
point) as (1). Example (6) also has the same illocutionary point, but expressed in the interrogative form.
Hare (1970) introduced a third operator, tropic, to
capture the mood of the utterance, here identified
with the clause type (see Mood, Clause Types, and
Illocutionary Force). We can now translate (1)(2)
and (5)(6) as follows:
(7) [Keeping to the path by you]phrastic[!tropic]
[please]neustic
(8) [Keeping to the path by you]phrastic[tropic]
[yes]neustic

(9) [Keeping to the path by you]phrastic[tropic]

[please]neustic
(10) [Keeping to the path by you]phrastic[?tropic]
[please]neustic

Hare writes, I shall retain the term phrastic for

the part of sentences which is governed by the tropic
and is common to sentences with different tropics
(1970: 21); this is what Searle (1969) calls the
propositional content of the speech act. A neustic
has to be present or understood before a sentence can
be used to make an assertion or perform any other
speech act (Hare, 1970: 22). Obviously, the inventory of neustics needs to be vastly increased beyond
please and yes to include the extensive number of
illocutionary points to which a speaker may subscribe. As we can see from (7)(10), to share a
phrastic and a tropic will not guarantee a common
neustic; nor will the sharing of a phrastic and a neustic guarantee a common tropic. These three parts of
a speech act are independent.
Hare (1970) uses the distinction between neustic
and tropic to explain the fact that (11) makes an
assertion about what time it is, but no such assertion
is made by the identical sentence as it occurs in the
protasis (if-clause) in (12), nor in the complement
clause in (13).
(11) It is ten oclock.
(12) If it is ten oclock, then Jane is in bed.
(13) Max says that it is ten oclock.

In each of (11)(13) the clause it is ten oclock has

the same phrastic and the same tropic, but the neustic
is a property of the whole speech act.
See also: Mood, Clause Types, and Illocutionary Force;

Speech Acts.

Phrastic, Neustic, Tropic: Hares Trichotomy 585

processing: a festschrift in honour of B. T. S. Atkins.
Grenoble: Euralex. 125137.
Kozowska C D & Dzierzanowska H (1982). Selected
English collocations. Warsaw: Panstwowe Wydawnictwo Naukowe.
Kozowska C D & Dzierzanowska H (1988). Selected
English collocations (2nd edn.). Warsaw: Panstwowe
Wydawnictwo Naukowe.
Moon R M (1998a). Fixed expressions and idioms in
English: a corpus-based approach. Oxford: Oxford
University Press.

Moon R M (1998b). Frequencies and forms of phrasal

lexemes in English. In Cowie A P (ed.). 79100.
Palm-Meister C (ed.) (2004). Europhras 2000. Tubingen:
Stauffenburg.
Palmer H E (1933). The second interim report on English
collocations. Tokyo: Kaitakusha.
Pawley A (2001). Phraseology, linguistics and the dictionary. International Journal of Lexicography 14(2),
122134.
Sinclair J, Fox G & Moon R (1995). Collins COBUILD
dictionary of idioms. London: Harper Collins.

Phrastic, Neustic, Tropic: Hares Trichotomy

K Allan, Monash University, Clayton, VIC, Australia
2006 Elsevier Ltd. All rights reserved.

The terms phrastic, neustic, and tropic were introduced to the theory of speech acts by philosopher
Richard M. Hare. Hare (1949) had compared pairs
like the imperative in (1) with the declarative in (2):
(1) Keep to the path.
(2) You will keep to the path.

He concluded that (1) and (2) have the same phrastic, but a different neustic, which he characterized
as follows:
(3) [Keeping to the path by you]phrastic[please]neustic
(4) [Keeping to the path by you]phrastic[yes]neustic

It is tempting to symbolize neustic please by !

and yes by , but Hare (1970) gives us reason to
compare (2) with (5) and (6).
(5) You will keep to the path!
(6) Will you keep to the path!(?)

Although (5) has the declarative form of (2), it has

what Hare calls the same subscription (illocutionary
point) as (1). Example (6) also has the same illocutionary point, but expressed in the interrogative form.
Hare (1970) introduced a third operator, tropic, to
capture the mood of the utterance, here identified
with the clause type (see Mood, Clause Types, and
Illocutionary Force). We can now translate (1)(2)
and (5)(6) as follows:
(7) [Keeping to the path by you]phrastic[!tropic]
[please]neustic
(8) [Keeping to the path by you]phrastic[tropic]
[yes]neustic

(9) [Keeping to the path by you]phrastic[tropic]

[please]neustic
(10) [Keeping to the path by you]phrastic[?tropic]
[please]neustic

Hare writes, I shall retain the term phrastic for

the part of sentences which is governed by the tropic
and is common to sentences with different tropics
(1970: 21); this is what Searle (1969) calls the
propositional content of the speech act. A neustic
has to be present or understood before a sentence can
be used to make an assertion or perform any other
speech act (Hare, 1970: 22). Obviously, the inventory of neustics needs to be vastly increased beyond
please and yes to include the extensive number of
illocutionary points to which a speaker may subscribe. As we can see from (7)(10), to share a
phrastic and a tropic will not guarantee a common
neustic; nor will the sharing of a phrastic and a neustic guarantee a common tropic. These three parts of
a speech act are independent.
Hare (1970) uses the distinction between neustic
and tropic to explain the fact that (11) makes an
assertion about what time it is, but no such assertion
is made by the identical sentence as it occurs in the
protasis (if-clause) in (12), nor in the complement
clause in (13).
(11) It is ten oclock.
(12) If it is ten oclock, then Jane is in bed.
(13) Max says that it is ten oclock.

In each of (11)(13) the clause it is ten oclock has

the same phrastic and the same tropic, but the neustic
is a property of the whole speech act.
See also: Mood, Clause Types, and Illocutionary Force;

Speech Acts.

586 Phrastic, Neustic, Tropic: Hares Trichotomy

Bibliography
Hare R M (1949). Imperative sentences. Mind 58, Reprinted in Hare (1971), 121.
Hare R M (1970). Meaning and speech acts. Philosophical
Review 79, 324. Reprinted in Hare (1971), 7493.

Hare R M (1971). Practical inferences. London:

Macmillan.
Searle J R (1969). Speech acts. Cambridge: Cambridge
University Press.

Phytosemiotics
J Deely, University of St. Thomas, Houston, TX, USA
2006 Elsevier Ltd. All rights reserved.

The term phytosemiotics was coined by Martin

Krampen in 1981. It appeared as the title of his article, which was the original attempt to define the
action of signs or semiosis in the world of plants, as
contradistinguished from the animal and human
realms. This seminal article undertook to show how
the whole of the biosphere is subtended and unified in
its interdependencies by the action of signs. As Sebeok
later summarized (2001: 14), where there is a feature
to be found in the botanical world there is bound to
be a complementary counter-feature in the zoo semiotic world. Krampen (1992: 217) put the matter
comprehensively: human sign production is deeply
rooted in its symbiosis with plants the meaning of
human (and animal) life thus being indexically contingent on plant life. Krampens convincing argument
swiftly became part of the 20th-century mainstream
development of semiotics, opening the way to a postmodern and global intellectual culture (see Deely,
1982, 2001; Deely et al., 1986; Krampen, 1994,
1997, 2001; Kull, 2000; Sharov, 1998).
Since time immemorial, human beings have divided
the universe between living things and inorganic substances, and have subdivided the living into humans,
animals, and plants. Study of the action of signs in the
specifically human world had come to be called
anthroposemiotics, a term whose provenance is clear
(study of the species-specifically human action of
signs, from anthropos for human being semiosis),
and whose coinage traces to Sebeok (1968: 8). Earlier
(Sebeok, 1963), Sebeok had coined the more
generic term zoo semiotics for the action of signs in
animals as such, whether human or not. Krampens
coinage, thus, was key to completing the notion of
biosemiotics, the architectonic idea proposed by
Thomas A. Sebeok (1990) that semiosis is criterial
of life (see Hoffmeyer and Emmeche, 1999; Kull,
2001). The further idea that semiosis is criterial
of the development of nature as a whole requires a

further term, physiosemiotics (Deely, 1990), to

name knowledge developed from studying the virtual
semiosis evidenced in the development of the inorganic world prior to and preparatory for life.
This implies that semiotics may ultimately provide a
clearer way of accounting for what has heretofore
been called simply evolution (Deely, 1999). Thus,
we have completed the proposal of a terminology to
cover C. S. Peirces grand vision (Deely, 1996) of
semiosis as an action within and between all the
individuals and levels of nature, organic and inorganic alike, with semiotics as the umbrella term generically covering all knowledge derived from study of
that action.
These terms anthroposemiotics, zoo semiotics,
phytosemiotics, physiosemiotics serve philosophically, of course, as but place-markers, so to speak,
for the detailed scientific study needed to make of
semiotics the full intellectual reality demanded to
meet the requirements of a postmodern, global civilization on planet Earth. The contribution of Krampen
in filling in the vegetative sphere of phytosemiotics
must be regarded as one of the major contributions to
global semiotics.
See also: Indexicality: Theory; Peirce, Charles Sanders

(18391914); Sebeok, Thomas Albert (19202001); Semiosis; Semiosphere versus Biosphere.

Bibliography
Deely J (1982). On the notion of phytosemiotics. In
Deely J & Evans J (eds.) Semiotics 1982. Lanham,
MD: University Press of America. 541554.
Deely J (1990). Basics of semiotics. Bloomington, IN:
Indiana University Press.
Deely J (1989). Physiosemiosis and semiotics. In Spinks
C W & Deely J N (eds.) Semiotics 1998. New York:
Peter Lang Publishing, Inc. 191197.
Deely J (1996). The grand vision. In Colapietro V &
Olshewsky T (eds.) Peirces doctrine of sign. Berlin:
Mouton de Gruyter. 4567.
Deely J (2001). Four ages of understanding. Toronto:
University of Toronto Press.

586 Phrastic, Neustic, Tropic: Hares Trichotomy

Bibliography
Hare R M (1949). Imperative sentences. Mind 58, Reprinted in Hare (1971), 121.
Hare R M (1970). Meaning and speech acts. Philosophical
Review 79, 324. Reprinted in Hare (1971), 7493.

Hare R M (1971). Practical inferences. London:

Macmillan.
Searle J R (1969). Speech acts. Cambridge: Cambridge
University Press.

Phytosemiotics
J Deely, University of St. Thomas, Houston, TX, USA
2006 Elsevier Ltd. All rights reserved.

The term phytosemiotics was coined by Martin

Krampen in 1981. It appeared as the title of his article, which was the original attempt to define the
action of signs or semiosis in the world of plants, as
contradistinguished from the animal and human
realms. This seminal article undertook to show how
the whole of the biosphere is subtended and unified in
its interdependencies by the action of signs. As Sebeok
later summarized (2001: 14), where there is a feature
to be found in the botanical world there is bound to
be a complementary counter-feature in the zoosemiotic world. Krampen (1992: 217) put the matter
comprehensively: human sign production is deeply
rooted in its symbiosis with plants the meaning of
human (and animal) life thus being indexically contingent on plant life. Krampens convincing argument
swiftly became part of the 20th-century mainstream
development of semiotics, opening the way to a postmodern and global intellectual culture (see Deely,
1982, 2001; Deely et al., 1986; Krampen, 1994,
1997, 2001; Kull, 2000; Sharov, 1998).
Since time immemorial, human beings have divided
the universe between living things and inorganic substances, and have subdivided the living into humans,
animals, and plants. Study of the action of signs in the
specifically human world had come to be called
anthroposemiotics, a term whose provenance is clear
(study of the species-specifically human action of
signs, from anthropos for human being semiosis),
and whose coinage traces to Sebeok (1968: 8). Earlier
(Sebeok, 1963), Sebeok had coined the more
generic term zoosemiotics for the action of signs in
animals as such, whether human or not. Krampens
coinage, thus, was key to completing the notion of
biosemiotics, the architectonic idea proposed by
Thomas A. Sebeok (1990) that semiosis is criterial
of life (see Hoffmeyer and Emmeche, 1999; Kull,
2001). The further idea that semiosis is criterial
of the development of nature as a whole requires a

further term, physiosemiotics (Deely, 1990), to

name knowledge developed from studying the virtual
semiosis evidenced in the development of the inorganic world prior to and preparatory for life.
This implies that semiotics may ultimately provide a
clearer way of accounting for what has heretofore
been called simply evolution (Deely, 1999). Thus,
we have completed the proposal of a terminology to
cover C. S. Peirces grand vision (Deely, 1996) of
semiosis as an action within and between all the
individuals and levels of nature, organic and inorganic alike, with semiotics as the umbrella term generically covering all knowledge derived from study of
that action.
These terms anthroposemiotics, zoosemiotics,
phytosemiotics, physiosemiotics serve philosophically, of course, as but place-markers, so to speak,
for the detailed scientific study needed to make of
semiotics the full intellectual reality demanded to
meet the requirements of a postmodern, global civilization on planet Earth. The contribution of Krampen
in filling in the vegetative sphere of phytosemiotics
must be regarded as one of the major contributions to
global semiotics.
See also: Indexicality: Theory; Peirce, Charles Sanders

(18391914); Sebeok, Thomas Albert (19202001); Semiosis; Semiosphere versus Biosphere.

Bibliography
Deely J (1982). On the notion of phytosemiotics. In
Deely J & Evans J (eds.) Semiotics 1982. Lanham,
MD: University Press of America. 541554.
Deely J (1990). Basics of semiotics. Bloomington, IN:
Indiana University Press.
Deely J (1989). Physiosemiosis and semiotics. In Spinks
C W & Deely J N (eds.) Semiotics 1998. New York:
Peter Lang Publishing, Inc. 191197.
Deely J (1996). The grand vision. In Colapietro V &
Olshewsky T (eds.) Peirces doctrine of sign. Berlin:
Mouton de Gruyter. 4567.
Deely J (2001). Four ages of understanding. Toronto:
University of Toronto Press.

Piaget, Jean (18961980) 587

Deely J N, Williams B & Kruse F E (eds.) (1986). Frontiers
in semiotics. Bloomington: Indiana University Press.
Hoffmeyer J & Emmeche C (eds.) (1999). Biosemiotics,
Special Issue of Semiotica 127.
Krampen M (1981). Phytosemiotics. Semiotica 36(3/4),
187209.
Krampen M (1992). Phytosemiotics revisited. In Biosemiotics. The semiotic web 1991. Berlin: Mouton de
Gruyter. 213219.
Krampen M (1994). Phytosemiotics. In Sebeok T A et al.
(eds.) Encyclopedic dictionary of semiotics (2nd edn.).
Berlin: Mouton de Gruyter. 726730.
Krampen M (1997). Phytosemiosis. In Posner R, Robering K
& Sebeok T A (eds.) Semiotics: a handbook on the signtheoretic foundations of nature and culture. Berlin:
Walter de Gruyter. 507522.
Krampen M (2001). No plant no breath. In Kull K (ed.).
415421.
Kull K (2000). An Introduction to phytosemiotics: semiotic
botany and vegetative sign systems. Sign System Studies
28, 326350.

Kull K (ed.) (2001). Jakob von Uexku ll: a paradigm for

biology and semiotics, a special issue of Semiotica
134(1/4).
No th W (2001). German-Italian colloquium. The semiotic
threshold from nature to culture. The Semiotics of Nature, a special issue of Sign System Studies 29(1), Kull K
& No th W (eds.).
Sebeok T A (1963). Book review of M. Lindauer, Communication among social bees; W. N. Kellog, Porpoises and
sonar; and J. C. Lilly, Man and dolphin. Language 39,
448466.
Sebeok T A (1968). Goals and limitations of the study of
animal communication. In Sebeok T A (ed.) Animal
communication: techniques of study and results of
research. Bloomington: Indiana University Press. 314.
Sebeok T A (1990). The sign science and the life science. In
Bernard J, Deely J, Voigt V & Withalm G (eds.)
Symbolicity. Lanham, MD: University Press of America.
243252.
Sharov A (1998). From cybernetics to semiotics in biology.
In Semiotica 120(3/4), 403419.

Piaget, Jean (18961980)

W Bublitz, Universitaet Augsburg, Augsburg,
Germany
2006 Elsevier Ltd. All rights reserved.

Jean Piaget (Figure 1), who was born in Neucha tel

(Switzerland) in 1896, received his Ph.D. in biology
from the University of Neucha tel in 1918 and spent
the following semester studying experimental psychology, psychiatry, and psychoanalysis at the University of Zu rich and two more semesters in Paris, where
his research oscillated between developmental psychology, psychopathology, logic, and philosophy. In
1921, he returned to Switzerland, where he worked
and lived until his death in Geneva in 1980. Over the
years, he held a number of chairs (some of them simultaneously) in psychology, sociology, history and philosophy of science, and genetic and experimental
psychology at several universities, among them Neucha tel, Lausanne, Paris, and Geneva. In 1955 he
founded the interdisciplinary Centre Internationale
dEpiste mologie Ge ne tique in Geneva. Piaget was a
member of the Executive Council of the UNESCO and
of 20 Academic Societies, held honorary doctorates
from more than 30 universities, and was awarded
numerous international prizes.
According to the Jean Piaget Archives (Switzerland),
Piaget published more than 50 books and 500
papers as well as 37 volumes in the series Etudes
dEpiste mologie Ge ne tique. This remarkable record

is a mirror of the multidisciplinary nature and the

extraordinary multifacetedness of his work. His exceptional ability to transcend the boundaries of developmental psychology and genetic epistemology
and integrate stimuli from other disciplines inspired
related work in such seemingly remote areas as

Figure 1 Jean Piaget (18961980). Courtesy of Universitaet

Augsburg.

Piaget, Jean (18961980) 587

Deely J N, Williams B & Kruse F E (eds.) (1986). Frontiers
in semiotics. Bloomington: Indiana University Press.
Hoffmeyer J & Emmeche C (eds.) (1999). Biosemiotics,
Special Issue of Semiotica 127.
Krampen M (1981). Phytosemiotics. Semiotica 36(3/4),
187209.
Krampen M (1992). Phytosemiotics revisited. In Biosemiotics. The semiotic web 1991. Berlin: Mouton de
Gruyter. 213219.
Krampen M (1994). Phytosemiotics. In Sebeok T A et al.
(eds.) Encyclopedic dictionary of semiotics (2nd edn.).
Berlin: Mouton de Gruyter. 726730.
Krampen M (1997). Phytosemiosis. In Posner R, Robering K
& Sebeok T A (eds.) Semiotics: a handbook on the signtheoretic foundations of nature and culture. Berlin:
Walter de Gruyter. 507522.
Krampen M (2001). No plant no breath. In Kull K (ed.).
415421.
Kull K (2000). An Introduction to phytosemiotics: semiotic
botany and vegetative sign systems. Sign System Studies
28, 326350.

Kull K (ed.) (2001). Jakob von Uexkull: a paradigm for

biology and semiotics, a special issue of Semiotica
134(1/4).
Noth W (2001). German-Italian colloquium. The semiotic
threshold from nature to culture. The Semiotics of Nature, a special issue of Sign System Studies 29(1), Kull K
& Noth W (eds.).
Sebeok T A (1963). Book review of M. Lindauer, Communication among social bees; W. N. Kellog, Porpoises and
sonar; and J. C. Lilly, Man and dolphin. Language 39,
448466.
Sebeok T A (1968). Goals and limitations of the study of
animal communication. In Sebeok T A (ed.) Animal
communication: techniques of study and results of
research. Bloomington: Indiana University Press. 314.
Sebeok T A (1990). The sign science and the life science. In
Bernard J, Deely J, Voigt V & Withalm G (eds.)
Symbolicity. Lanham, MD: University Press of America.
243252.
Sharov A (1998). From cybernetics to semiotics in biology.
In Semiotica 120(3/4), 403419.

Piaget, Jean (18961980)

W Bublitz, Universitaet Augsburg, Augsburg,
Germany
2006 Elsevier Ltd. All rights reserved.

Jean Piaget (Figure 1), who was born in Neuchatel

(Switzerland) in 1896, received his Ph.D. in biology
from the University of Neuchatel in 1918 and spent
the following semester studying experimental psychology, psychiatry, and psychoanalysis at the University of Zurich and two more semesters in Paris, where
his research oscillated between developmental psychology, psychopathology, logic, and philosophy. In
1921, he returned to Switzerland, where he worked
and lived until his death in Geneva in 1980. Over the
years, he held a number of chairs (some of them simultaneously) in psychology, sociology, history and philosophy of science, and genetic and experimental
psychology at several universities, among them Neuchatel, Lausanne, Paris, and Geneva. In 1955 he
founded the interdisciplinary Centre Internationale
dEpistemologie Genetique in Geneva. Piaget was a
member of the Executive Council of the UNESCO and
of 20 Academic Societies, held honorary doctorates
from more than 30 universities, and was awarded
numerous international prizes.
According to the Jean Piaget Archives (Switzerland),
Piaget published more than 50 books and 500
papers as well as 37 volumes in the series Etudes
dEpistemologie Genetique. This remarkable record

is a mirror of the multidisciplinary nature and the

extraordinary multifacetedness of his work. His exceptional ability to transcend the boundaries of developmental psychology and genetic epistemology
and integrate stimuli from other disciplines inspired
related work in such seemingly remote areas as

Figure 1 Jean Piaget (18961980). Courtesy of Universitaet

Augsburg.

588 Piaget, Jean (18961980)

sociology, philosophy, education, and, though to a

much lesser extent, (psycho-)linguistics.
Piagets central concern was to extend our knowledge about knowledge, its nature and the conditions
of its growth. To answer the key epistemological
questions how do we know?, how do we acquire
knowledge?, how does knowledge grow?, he began
to study the cognitive development of the child. But
child psychology, as he pointed out in his old age, was
merely a vehicle for a much more general goal:
I was extremely happy not to appear only as a child
psychologist. . . . my efforts directed toward the psychogenesis of knowledge were for me only a link between
two dominant preoccupations: the search for the
mechanisms of biological adaptation and the analysis
of that higher form of adaptation which is scientific thought, the epistemological interpretation of
which has always been my central aim. . . . To make
of epistemology an experimental discipline as well as a
theoretical one has always been my goal (Gruber and
Vone`che, 1977: xi f.).

The interdisciplinary breadth of his oeuvre defies a

comprehensive overview of all or only most of its
facets. From a linguistic point of view, however, his
studies of the cognitive development of the child are
most relevant. Starting from the premise that all living organisms interact with the world around them,
he sees the childs construction of knowledge as an
active process. Newly acquired knowledge is related
and adapted to earlier acquired knowledge, which is
thus newly organized. Ideas (e.g., about categories as
space, time, number, causality, with which Piaget was
primarily concerned) have no a priori existence; they
are neither given in the childs brain nor preexisting
and only to be discovered or found in the world.
Rather, they have to be constructed or invented,
though only through the childs (in early years, physical) interaction with his or her environment. This
view has come to be known as constructivism and
separates Piaget both from strict empiricism and radical nativism: he opposes the Kantian claim that our
knowledge is essentially constituted by innate or a
priori ideas as well as the empiricist claim that we
are not born with innate ideas but acquire knowledge
under the impact of our sensory or perceptual experiences with the world, which is thus directly copied.
Constructivism emphasizes the childs autonomy in
the active construction of knowledge (which happens
in relation to the world) during the course of his or her
cognitive development. Construction is a perpetual
process:
The individual and the social group are constantly in
the process of constructing and reconstructing
their views of the world. At a given moment, the most

advanced . . . achievements in the pursuit of knowledge

may play a regulatory role with regard to other achievements still to be made. But this does not give either the
most advanced, or . . . the most primitive, kinds of
knowledge the status of an ultimate reality. . . . the
only way in which we get knowledge is through continual construction, and . . . we can have no enduring
knowledge without actively maintaining this process
(Gruber and Vone`che, 1977: xxii f.).

The claim that the stages through which the

childs cognitive development passes and their fixed
order are universal is central to Piagets theory
(though often criticized because of its neglect of the
specific social and cultural milieu, into which the
child is set).
For the childs linguistic development, the period of
representational thought (which follows the sensorimotor stage and precedes the stages of concrete and
formal operations) is of paramount importance.
However, in Piagets view, language itself is only an
auxiliary means of encoding thought, and linguistic
knowledge is not an autonomous or even overly essential component of the childs overall knowledge.
But even though he was not interested in child
language (or its acquisition) per se, some aspects of
his work fuelled a few (mostly) controversial debates
in (psycho-)linguistics, notably one with Chomsky
(in the late 1970s): while Piaget argues that the development of language is secondary to and dependent on
the childs general cognitive development, Chomsky
regards language as a specific faculty with its own
developmental laws and propounds the conception of
an innate universal grammar.
The Jean Piaget Archives keeps record of Piagets
substantial and still growing influence on psychology,
sociology, education, epistemology, economics, and
related fields.

See also: Bee Dance; Chomsky, Noam (b. 1928); Kant,

Immanuel (17241804).

Bibliography
Bringuier J C (1980). Conversations with Jean Piaget.
Chicago: UP.
Gruber H E & Vone`che J J (eds.) (1977). The essential
Piaget. London: Routledge.
Jean Piaget Archives Foundation (1989). The Jean Piaget
bibliography. Geneva: Jean Piaget Archives Foundation.
Kitchener R (1986). Piagets theory of knowledge. New
Haven: Yale UP.
Mogdil S & Mogdil C (eds.) (1976). Piagetian research.
Compilation and commentary (8 vols). New Jersey:
NFER Publ.

Pictish 589
Piaget J (1924). Le jugement et le raisonnement chez lenfant. Neucha tel: Delachaux et Niestle . (Judgment and
reasoning in the child. London: Routledge, 1928.)
Piaget J (1936). La naissance de lintelligence chez lenfant.
Neucha tel: Delachaux et Niestle . (Origins of intelligence
in the child. London: Routledge, 1953.)
Piaget J (1950). Introduction a` le piste mologie ge ne tique
(3 vols). Paris: Presses Universitaires de France.
Piaget J (1954). Les relations entre laffectivite et lintelligence dans le de veloppement mental de lenfant. Paris:
Centre de Documentation Universitaire.
Piaget J (1975). LE quilibration des structures cognitives.
Paris: Presses Universitaires de France. (Equilibration of
cognitive structures. Chicago: UP, 1985.)
Piaget J (1976). Autobiographie. Revuee Europe enne des
Sciences Sociales 14, 143.
Piaget J (1981 and 1983). Le possible et le ne cessaire I:
Le volution des possibles chez lenfant. Le possible et le
ne cessaire II: Le volution du ne cessaire chez lenfant.
Paris: Presses Univeritaires de France. (Possibility and
necessity, 2 vols., Minneapolis: UP, 1987.)
Piaget J & Garcia R (1983). Psychoge ne`se et histoire
des sciences. Paris: Flammarion. (Psychogenesis and the
history of science. New York: Columbia UP, 1989.)

Piaget J & Garcia R (1987). Vers une logique de la significatio. Ginebra: Murionde. (Towards a logic of meanings.
Hillsdale, NJ: Erlbaum, 1991.)
Piattelli-Palmarini M (ed.) (1979). The ories du langage,
the ories de lapprentissage. Le de bat entre Jean Piaget et
Noam Chomsky. Paris: Seuil. (Language and Learning.
The debate between Jean Piaget and Noam Chomsky.
Cambridge/MA: Harvard UP, 1980.)
Smith L (1992). Jean Piaget: critical assessments (4 vols).
London: Routledge.
Smith L (1996). Critical readings on Piaget. London:
Routledge.
Vuyk R (1982). Overview and critique of Piagets genetic
epistemology 19651980 (2 vols). London: Academic
Press.

Relevant Websites
www.unige.ch/piaget/.
www.piaget.org/biography/biog.html.

Pictish
W Nicolaisen, University of Aberdeen, Aberdeen, UK
2006 Elsevier Ltd. All rights reserved.

Pictish was the language spoken by the Picts, inhabitants of the northeast of Scotland, roughly from the
Forth-Clyde line to the Cromarty Firth, but possibly
also further afield, including the Northern and Western Isles, from the early centuries A.D. until the middle
of the 9th century when, as the result of the merger of
the kingdoms of the Scotti and the Picti under Kenneth
MacAlpine, it was replaced by Gaelic, which had
reached Scotland from Ireland from approximately
500 A.D. onward. The Picts were known as Picti (or
Pecti) to the Roman military, who interpreted their
name in Latin terms as cognate with pictus painted.
They were referred to by the neighboring Anglo-Saxons as Pehtas, Pihtas, Pyhtas, Peohtas, or Piohtas; by
the Norsemen as Pe ttar or Pe ttir (as in Pe tlandsfjorr
the Pentland Firth); and in Middle Welsh as Peithwyr, but it is not known what they called themselves.
No sentence in their language has been recorded,
and our main sources for Pictish are king lists,
inscriptions, and, particularly, place names.
The nature and linguistic affiliation of Pictish has
attracted attention for a long time, and this scholarly,
and sometimes not so scholarly, pre-occupation with

the language(s) of the Picts has led to a comparative

neglect of other aspects of Scottish linguistic history
and prehistory, especially when it comes to the analysis and interpretation of early place names. As far as
Pictish is concerned, however, the fascination for its
linguistic status has resulted in a large variety of
theories that have been offered, and often seriously
defended, right to our own time. As recently as 1998,
for example, Paul Dunbavin regarded the Picts as
Finno-Ugric immigrants from the Baltic coast, basing
his revolutionary conclusion, among other arguments, on the apparent derivation of certain Scottish
river names, mentioned by Ptolemy about 150 A.D.,
from certain Finnish topographic terms. Apart from
the absence of any documentary reference to FinnoUgric people in Pictland, the proposed etymologies
suffer from the frequently encountered flaw in such
studies, the superficial equation of spellings reported
almost 2000 years ago with modern forms of words
in an otherwise unconnected language.
Whereas Dunbavin used toponymic, especially hydronymic, materials to support his proposal, Harald
V. Sverdrup (1995), in the course of classifying and
translating Pictish inscriptions, claimed that it can
be shown . . . that [Pictish] was neither a Celtic nor an
Indo-European language but was distantly related to
Caucasian languages, dating the arrival of the initial

Pictish 589
Piaget J (1924). Le jugement et le raisonnement chez lenfant. Neuchatel: Delachaux et Niestle. (Judgment and
reasoning in the child. London: Routledge, 1928.)
Piaget J (1936). La naissance de lintelligence chez lenfant.
Neuchatel: Delachaux et Niestle. (Origins of intelligence
in the child. London: Routledge, 1953.)
Piaget J (1950). Introduction a` lepistemologie genetique
(3 vols). Paris: Presses Universitaires de France.
Piaget J (1954). Les relations entre laffectivite et lintelligence dans le developpement mental de lenfant. Paris:
Centre de Documentation Universitaire.
Piaget J (1975). LEquilibration des structures cognitives.
Paris: Presses Universitaires de France. (Equilibration of
cognitive structures. Chicago: UP, 1985.)
Piaget J (1976). Autobiographie. Revuee Europeenne des
Sciences Sociales 14, 143.
Piaget J (1981 and 1983). Le possible et le necessaire I:
Levolution des possibles chez lenfant. Le possible et le
necessaire II: Levolution du necessaire chez lenfant.
Paris: Presses Univeritaires de France. (Possibility and
necessity, 2 vols., Minneapolis: UP, 1987.)
Piaget J & Garcia R (1983). Psychogene`se et histoire
des sciences. Paris: Flammarion. (Psychogenesis and the
history of science. New York: Columbia UP, 1989.)

Piaget J & Garcia R (1987). Vers une logique de la significatio. Ginebra: Murionde. (Towards a logic of meanings.
Hillsdale, NJ: Erlbaum, 1991.)
Piattelli-Palmarini M (ed.) (1979). Theories du langage,
theories de lapprentissage. Le debat entre Jean Piaget et
Noam Chomsky. Paris: Seuil. (Language and Learning.
The debate between Jean Piaget and Noam Chomsky.
Cambridge/MA: Harvard UP, 1980.)
Smith L (1992). Jean Piaget: critical assessments (4 vols).
London: Routledge.
Smith L (1996). Critical readings on Piaget. London:
Routledge.
Vuyk R (1982). Overview and critique of Piagets genetic
epistemology 19651980 (2 vols). London: Academic
Press.

Relevant Websites
www.unige.ch/piaget/.
www.piaget.org/biography/biog.html.

Pictish
W Nicolaisen, University of Aberdeen, Aberdeen, UK
2006 Elsevier Ltd. All rights reserved.

Pictish was the language spoken by the Picts, inhabitants of the northeast of Scotland, roughly from the
Forth-Clyde line to the Cromarty Firth, but possibly
also further afield, including the Northern and Western Isles, from the early centuries A.D. until the middle
of the 9th century when, as the result of the merger of
the kingdoms of the Scotti and the Picti under Kenneth
MacAlpine, it was replaced by Gaelic, which had
reached Scotland from Ireland from approximately
500 A.D. onward. The Picts were known as Picti (or
Pecti) to the Roman military, who interpreted their
name in Latin terms as cognate with pictus painted.
They were referred to by the neighboring Anglo-Saxons as Pehtas, Pihtas, Pyhtas, Peohtas, or Piohtas; by
the Norsemen as Pettar or Pettir (as in Petlandsfjorr
the Pentland Firth); and in Middle Welsh as Peithwyr, but it is not known what they called themselves.
No sentence in their language has been recorded,
and our main sources for Pictish are king lists,
inscriptions, and, particularly, place names.
The nature and linguistic affiliation of Pictish has
attracted attention for a long time, and this scholarly,
and sometimes not so scholarly, pre-occupation with

the language(s) of the Picts has led to a comparative

neglect of other aspects of Scottish linguistic history
and prehistory, especially when it comes to the analysis and interpretation of early place names. As far as
Pictish is concerned, however, the fascination for its
linguistic status has resulted in a large variety of
theories that have been offered, and often seriously
defended, right to our own time. As recently as 1998,
for example, Paul Dunbavin regarded the Picts as
Finno-Ugric immigrants from the Baltic coast, basing
his revolutionary conclusion, among other arguments, on the apparent derivation of certain Scottish
river names, mentioned by Ptolemy about 150 A.D.,
from certain Finnish topographic terms. Apart from
the absence of any documentary reference to FinnoUgric people in Pictland, the proposed etymologies
suffer from the frequently encountered flaw in such
studies, the superficial equation of spellings reported
almost 2000 years ago with modern forms of words
in an otherwise unconnected language.
Whereas Dunbavin used toponymic, especially hydronymic, materials to support his proposal, Harald
V. Sverdrup (1995), in the course of classifying and
translating Pictish inscriptions, claimed that it can
be shown . . . that [Pictish] was neither a Celtic nor an
Indo-European language but was distantly related to
Caucasian languages, dating the arrival of the initial

590 Pictish

settlers to the paleoneolithic transition before

7000 B.C. This would predate by several thousand
years any other known nonmaterial evidence, including place names. It is the presumed enigmatic
nature of Pictish that has led Sverdrup to the underlying readings, classification, and translation of the
inscriptions.
A considerably earlier perception of Pictish as a
non-Indo-European language comes from John Rhys,
who in 1892, after discounting the Ugro-Finnish people (Lapps, Finns, and Estonians) and the Ligurians,
felt logically bound to inquire what Basque can do to
help us to an understanding of the Pictish inscriptions. However, 6 years later he revised his own
theory by making it known that he no longer thought
Pictish was related to Basque but rather to be preIndo-European (although not as old as the neolithic
or mesolithic periods) that first came under p-Celtic
influence from the Cumbrians south of the ForthClyde line. This change of direction did not stop
J. B. Johnston, maintaining in 1934 his view (first
expressed in 1892) that the river Urr in southwest
Scotland derives from the Basque ur water, from
falling into the same trap as Dunbavin.
One of the most outspoken opponents of a Celtic
interpretation of the Pictish inscriptions was the
Irish archeologist R. A. S. Macalister, who in 1922
expressed the view: The most reasonable theory
about the Picts was that they were survivals of the
aboriginal pre-Celtic Bronze Age people. Certainly no
attempt at explaining the Pictish inscriptions by
means of any Celtic language could be called successful. John Fraser, too, held in 1927 that, having arrived before the Scots and the Britons, they must at
one time have spoken a non-Indo-European language,
although he took into account the later influence of
(Scots) Gaelic and Brittonic.
Advocates for a non-Indo-European and often
specifically of an anti-Celtic designation of Pictish
represented a wide, fragmented variety of linguistic
affiliations. In contrast, the pro-Celtic camp was divided into two opposing groups: those who regarded
it as a q-Celtic language and those who regarded it as
a p-Celtic language. Following such illustrious predecessors as George Buchanan (1582) and James
Macpherson of Ossian fame (1763), Francis J. Diack
(1944) was one of the strongest proponents of an
uninterrupted Gaelic history of Scotland from the
1st century until today; the Gaelic nature of Pictish
was asserted as recently as 1994 by Sheila McGregor.
The p-Celtic school also has a respectable pedigree
in William Camden (1586) and Father Innes (1729).
W. F. Skene (1868) declared Pictish to be neither
Welsh nor Gaelic but a Gaelic dialect partaking
largely of Welsh forms. One of the first scholars to

put the p-Celtic nature of Pictish on a sound footing

was Alexander Macbain (18911892) in his survey of
Ptolemys geography of Scotland; his stance was
strongly supported by W. J. Watson (1904, 1921,
1926), mainly on the basis of place-name evidence.
This is also at the heart of Kenneth H. Jacksons
(1955) overview of The Pictish language, in the
course of which he presents the first maps of linguistic
Pictland based on the distribution of such place-name
elements as pett (Pit-), aber, carden, lanerc, pert, and
pevr. He also suggested, however, that there may have
been two Pictish languages, one the language of
the pre-Indo-European inhabitants, the other the
Gallo-Brittonic tongue of Iron Age invaders. W. F. H.
Nicolaisen acknowledged the presence of non- or preIndo-European place names in Pictish territory, but
did not regard them as Pictish. In her repudiation of
Jacksons two Pictishes, in her thorough investigation
of Language in Pictland, Katherine Forsyth (1997), a
specialist in Ogham inscriptions, mustered some very
persuasive arguments against Jacksons construct
and, although it is always risky to call anything definitive, her conclusion that the Picts were as fully
Celtic as their Irish and British neighbors is difficult
to dispute, and it is good to see Pictish placed where it
belongs beside other p-Celtic languages such as
Cumbric, Welsh, Cornish, Breton, and Gaulish; by
implication, the firm ascription of Pictish in this linguistic grouping adjudges the Scotti to have brought
the Gaelic language with them from Ireland, a consequence of fundamental importance in a long-running
debate.
See also: Cornish; Finnish; IndoEuropean Languages;
United Kingdom: Language Situation; Scots Gaelic;
Welsh.

Bibliography
Cox R A V (1997). Modern Gaelic reflexes of two Pictish
words. Nomina 20, 4758.
Cummins W A (1998). The age of the Picts. Stroud, UK:
Sutton.
Diack F C (19201921/1922). Place-names of Pictland.
Revue Celtique 38, 109132; 39, 125174.
Diack F C (1944). The inscriptions of Pictland. Aberdeen,
UK: Third Spalding Club.
Dunbavin R (1998). Picts and ancient Britons: an exploration of Pictish origins. Long Eaton, UK: Third
Millennium Publication.
Forsyth K C (1996). The Ogham inscriptions of Scotland:
an edited corpus. Ph.D. diss., Harvard University.
Forsyth K (1997). Language in Pictland: the case against
non-Indo-European Pictish. Utrecht: de Keltische
Draak.
Fraser J (1927). The question of the Picts. Scottish Gaelic
Studies 2, 172201.

Pictography: Semiotic Approaches 591

Fraser J (1942). Pet(t) in place-names. Scottish Gaelic
Studies 5, 6771.
Henderson I (1967). The Picts. London: Thames and
Hudson.
Henry O (ed.) (1992). The worm, the germ and the thorn.
Balgavies, UK: The Pinkfoot Press.
Jackson K H (1955). The Pictish language. In Wainwright
F T (ed.) The problem of the Picts. Edinburgh, UK:
Nelson. 129166.
Johnston J B (1934). Place-names of Scotland (3rd edn.).
London: John Murray.
Macalister R A S (1940). The inscriptions and language of
the Picts. In Ryan J (ed.) Fe il-scribhian Eo in Mhic
Ne ill: essays and studies presented to Professor Eo in
MacNeill. Dublin, Ireland: At the Sign of the Three
Candles. 184226.
Macbain A (18911892). Ptolemys geography of Scotland. Transactions of the Gaelic Society of Inverness
128, 267288.
MacGregor S (1994). The abers of Perthshire. Pictish Arts
Society Journal (spring), 1219.
Nicolaisen W F H (1996). The Picts and their place-names.
Rosemarkie, UK: Groam House.
Nicolaisen W F H (2001). Scottish place-names: their
study and significance (rev. edn.). Edinburgh, UK:
John Donald.
Nicolaisen W F H (2004). A gallimaufry of languages. In
Van Nahl A, Elmevik L & Brink S (eds.) Namenwelten:
orts and personennamen in historischer sicht. Berlin and
New York: de Gruyter. 233240.

Nicoll E H (ed.) (1995). A Pictish panorama. Balgavies, UK:

The Pinkfoot Press.
Rhys J (18911892). The inscriptions and language of the
northern Picts. Proceedings of the Society of Antiquaries
of Scotland 126, 263351.
Rhys J (18971898). A revised account of the inscriptions
of the northern Picts. Proceedings of the Society of
Antiquaries of Scotland 132, 324398.
Ritchie A (1989). Picts. Edinburgh, UK: H. M. Stationery
Office.
Skene W F (1868). Four ancient books of Wales (Vol. 1).
Edinburgh, UK: Edmonston and Douglas.
Sutherland E (1994). In search of the Picts. London:
Constable.
Sverdrup H (1995). Reports in ecology and environmental
engineering 4: Classifying and translating inscriptions in
the Pictish language. Lund.
Taylor S (1994). Generic-element variation with special
reference to eastern Scotland. Nomina 20, 522.
Wainwright F T (ed.) (1955). The problem of the Picts.
Edinburgh, UK: Nelson.
Watson W J (1904). Place-names of Ross and Cromarty.
Inverness, UK: The Northern Counties Printing and
Publishing Company.
Watson W J (1921). The Picts: their original position in
Scotland. Transactions of the Gaelic Society of Inverness
80, 240261.
Watson W J (1926). The history of the Celtic place-names
of Scotland. Edinburgh, UK: Blackwood.

Pictography: Semiotic Approaches

F Nuessel, University of Louisville, Louisville, KY, USA
2006 Elsevier Ltd. All rights reserved.

Introduction
The etymology of the word pictography is simple:
picto comes from the Latin pictus, which is the past
participle of the verb pingere to paint, and graph
derives from the Greek graphein to write. In simple
terms, a pictograph is an image that denotes a word
or idea. A pictograph is thus a painting or an illustration that bears a pictorial resemblance to its referent.
Thus, the picture or drawing of a cat stands for or
means cat.
In this article, pictography is discussed in relation
to the following subjects:
1. The notion of pictography as a possible stage in the
development of a systematic written representation of language.
2. The semiotics of images.

3. Modern manifestations of pictography (computer

icons, word games, comics, and so forth).

Writing Systems
Pictography is often discussed in relation to writing
systems. John DeFrancis (b. 1911), whose critical
discussion of writing systems has clarified many commonly held misconceptions about them, notes that
there are at least six levels of representation of writing
systems (DeFrancis, 1989: 218): pictographic, ideographic, logographic/morphemic, syllabic, phonemic,
and featural. Although the purpose here is not to
provide a detailed history of writing systems, it is
appropriate to comment briefly on these various
forms of written representations of speech. The first
type of representation, pictographic, is the relevant
one for this article. In his book on writing, DeFrancis
(1989: 219) pointed out that pictographs are not
writing, nor are they forerunners of writing as
asserted by I. J. Gelb (19071985; Gelb, 1963: 190).

Pictography: Semiotic Approaches 591

Fraser J (1942). Pet(t) in place-names. Scottish Gaelic
Studies 5, 6771.
Henderson I (1967). The Picts. London: Thames and
Hudson.
Henry O (ed.) (1992). The worm, the germ and the thorn.
Balgavies, UK: The Pinkfoot Press.
Jackson K H (1955). The Pictish language. In Wainwright
F T (ed.) The problem of the Picts. Edinburgh, UK:
Nelson. 129166.
Johnston J B (1934). Place-names of Scotland (3rd edn.).
London: John Murray.
Macalister R A S (1940). The inscriptions and language of
the Picts. In Ryan J (ed.) Feil-scribhian Eoin Mhic
Neill: essays and studies presented to Professor Eoin
MacNeill. Dublin, Ireland: At the Sign of the Three
Candles. 184226.
Macbain A (18911892). Ptolemys geography of Scotland. Transactions of the Gaelic Society of Inverness
128, 267288.
MacGregor S (1994). The abers of Perthshire. Pictish Arts
Society Journal (spring), 1219.
Nicolaisen W F H (1996). The Picts and their place-names.
Rosemarkie, UK: Groam House.
Nicolaisen W F H (2001). Scottish place-names: their
study and significance (rev. edn.). Edinburgh, UK:
John Donald.
Nicolaisen W F H (2004). A gallimaufry of languages. In
Van Nahl A, Elmevik L & Brink S (eds.) Namenwelten:
orts and personennamen in historischer sicht. Berlin and
New York: de Gruyter. 233240.

Nicoll E H (ed.) (1995). A Pictish panorama. Balgavies, UK:

The Pinkfoot Press.
Rhys J (18911892). The inscriptions and language of the
northern Picts. Proceedings of the Society of Antiquaries
of Scotland 126, 263351.
Rhys J (18971898). A revised account of the inscriptions
of the northern Picts. Proceedings of the Society of
Antiquaries of Scotland 132, 324398.
Ritchie A (1989). Picts. Edinburgh, UK: H. M. Stationery
Office.
Skene W F (1868). Four ancient books of Wales (Vol. 1).
Edinburgh, UK: Edmonston and Douglas.
Sutherland E (1994). In search of the Picts. London:
Constable.
Sverdrup H (1995). Reports in ecology and environmental
engineering 4: Classifying and translating inscriptions in
the Pictish language. Lund.
Taylor S (1994). Generic-element variation with special
reference to eastern Scotland. Nomina 20, 522.
Wainwright F T (ed.) (1955). The problem of the Picts.
Edinburgh, UK: Nelson.
Watson W J (1904). Place-names of Ross and Cromarty.
Inverness, UK: The Northern Counties Printing and
Publishing Company.
Watson W J (1921). The Picts: their original position in
Scotland. Transactions of the Gaelic Society of Inverness
80, 240261.
Watson W J (1926). The history of the Celtic place-names
of Scotland. Edinburgh, UK: Blackwood.

Pictography: Semiotic Approaches

F Nuessel, University of Louisville, Louisville, KY, USA
2006 Elsevier Ltd. All rights reserved.

Introduction
The etymology of the word pictography is simple:
picto comes from the Latin pictus, which is the past
participle of the verb pingere to paint, and graph
derives from the Greek graphein to write. In simple
terms, a pictograph is an image that denotes a word
or idea. A pictograph is thus a painting or an illustration that bears a pictorial resemblance to its referent.
Thus, the picture or drawing of a cat stands for or
means cat.
In this article, pictography is discussed in relation
to the following subjects:
1. The notion of pictography as a possible stage in the
development of a systematic written representation of language.
2. The semiotics of images.

3. Modern manifestations of pictography (computer

icons, word games, comics, and so forth).

Writing Systems
Pictography is often discussed in relation to writing
systems. John DeFrancis (b. 1911), whose critical
discussion of writing systems has clarified many commonly held misconceptions about them, notes that
there are at least six levels of representation of writing
systems (DeFrancis, 1989: 218): pictographic, ideographic, logographic/morphemic, syllabic, phonemic,
and featural. Although the purpose here is not to
provide a detailed history of writing systems, it is
appropriate to comment briefly on these various
forms of written representations of speech. The first
type of representation, pictographic, is the relevant
one for this article. In his book on writing, DeFrancis
(1989: 219) pointed out that pictographs are not
writing, nor are they forerunners of writing as
asserted by I. J. Gelb (19071985; Gelb, 1963: 190).

592 Pictography: Semiotic Approaches

Rather, as DeFrancis (1989: 219) stated, [a]t the

pictographic level, inadequate attention is paid to
sorting out the two possible aspects of pictographic
symbols, namely, that they represent either concepts
divorced from sound, or sounds at one of the various
levels of speech. Hans Jensen (18841973) called
pictography idea writing (Jensen, 1969: 40). Roy
Harris (b. 1931) said that [t]he term pictogram
appears to be used to include almost any type of
non-alphabetic symbol (Harris, 1986: 34). In this
regard, pictography is not truly writing nor is it a
precursor of writing. It is simply a graphic depiction
of certain objects, people, or events that have significance within a particular culture. Such events are
contemporaneously interpretable by persons who
have witnessed them. Subsequently, the pictograph
will have to be interpreted by someone who has acquired knowledge of the events from an eyewitness or
from a person who functions as the oral historian of
that culture. In general, pictographs are simple drawings that encapsulate the essence of a memorable
historical event that has had a significant impact on
a particular community, e.g., American Indians.
DeFranciss remaining five forms of writing may be
described briefly as follows. Ideographic writing consists of symbols that represent meaning, without indicating their pronunciation (Chinese is considered
ideographic). An ideographic system is partly pictographic and partially abstract. Logographic/morphemic (terms that are also associated with the Chinese
writing system) writing consists of graphic symbols
that represent words, whereas morphemic writing
refers to various of forms of writing that break
words down into their component parts. For example, professor consists of the components profess
-or. Syllabic writing represents whole syllables
(Japanese with its two writing varieties, hiragana
and katakana, is considered syllabic). Phonemic
writing is writing that reflects the individual sounds
of a language that is composed of consonants and
vowels (Spanish exemplifies a phonemic writing
system). Featural writing is writing that indicates if
a sound is voiced, aspirate, vocalic, and so forth
(Korean Hangul writing is sometimes called featural).
This overview of the major writing systems is by no
means complete and the reader should consult
DeFrancis (1989) or Jensen (1969) for an in-depth
discussion of the various world writing systems. We
now consider selectively a few of the earliest manifestations of writing (Jensen, 1969: 82122), i.e.,
cuneiforms, hieroglyphs, and pictographs.
Sumerian writing is called cuneiform (cuneus
wedge), a term coined by the Oxford professor
Thomas Hyde (16361703) in 1700 because of the
use of wedge-shaped signs in the writing system.

Knowledge of this form of writing first came to the

attention of Europeans in the 17th century when an
Italian traveler returned from the Middle East with
some cuneiform etchings on bricks. Decipherment of
the script took place over time. The Sumerian writing
system is the earliest complete writing system and it
served as a medium of communication for speakers
of various languages. Its script functioned for many
cultures over several millennia. The cuneiform system is not uniform, i.e., different languages utilized
the symbols in various ways. Among the cuneiform scripts are the SumerianBabylonianAssyrian
form (Mesopotamia, Chaldea, Asia Minor) and the
Elamitic, Old Persian, and Ugaritic forms.
Hieroglyphic writing refers to picture symbols
employed in Egyptian writing. The etymology of
this word is from the Greek hieros holy and glyphein
to carve. Until more was learned about this form of
writing, it was assumed that hieroglyphs were merely
pictographs. Discovered by Napoleons invading
forces in Egypt in 1798, the Rosetta stone had inscriptions in hieroglyphic, demotic, and Greek. This trilingual text facilitated translation of the hieroglyphic
writing system. Jean-Francois Champollion (1790
1832) was responsible for deciphering Egyptian
hieroglyphic writing. An examination of the progression of the Egyptian writing system from a semantic
one to a phonetic one shows an evolution of pictographic-like symbols (fish, bird, stone jug, and so
forth) to what became alphabetic symbols. A hieroglyph can consist of two elements: (1) an ideogram
that represents objects or ideas and (2) a phonogram that represents one to three consonantal sounds.
Hieroglyphs usually consisted of both types of
elements.
A precise chronology of the meaningful representation of visual materials is elusive and ultimately may
never be known with certainty. Archaeology offers
clues to the customs of previous cultures; dating of
visual records (pictographs) of common and momentous historical events provides important evidence of
previous customs. What compelled the cave painters
of northern (Altamira) Spain and Lascaux, France
to preserve their primitive milieu with paint may
never be known. Those pictures, however, now offer
us some notion of that prehistoric period (30 000
to 40 000 B.C.E). Other archaeological excavations
from other locations worldwide provide further evidence, including paintings, sculptures, carvings, and
etchings.
Jensen (1969: 24) noted two requisites for the first
steps to writing: 1. Its production by an act of drawing, painting or scratching on a durable writing material, and 2. the purpose of communicating (to others
or, as an aid to the memory, to the writer himself).

Pictography: Semiotic Approaches 593

Object writing is a primitive first step toward writing.

It may include, for example, notched sticks, such as
Australian messenger sticks with various notches, to
symbolize certain information; the knotted cords,
or quippu, of the Inca Indians; American Indian
wampum belts; and the aroko of the Jebos of Lagos,
Nigeria, which is a reed cord with various symbolic
elements. Jensen used the expression picture-writing
to refer to paintings found in the caves of northern
Spain. These are akin to contemporary icons, such as
those used in signage to indicate traffic directions,
prohibitions, and hazards.
DeFrancis (1989: 281) defined a pictograph as
a symbol which depicts things or actions. Pictographic symbols have two dimensions. First, they
either represent concepts unrelated to speech, or
they represent sounds beyond the phonemic, or single-sound, level, i.e., they represent the discourse level
of speech, beyond words, phrases, and sentences. In
some sense, primitive pictographs represent a narrative structure. If, however, a pictograph represents
a specific level of speech beyond the phonetic level,
this must be acknowledged. DeFrancis (1989: 219)
noted that pictographic representation is generally
not considered to be a forerunner of writing. He
argued in favor of this notion because these visual
symbols fail to capture the featural, phonemic, syllabic, logographic/morphemic, or ideographic level of
speech. Thus, Amerindian and Yukaghir (fishers and
hunters in the lower areas of the Kolyma River, near
the Arctic Ocean, in the former Yakut Autonomous
Soviet Socialist Republic) pictographic systems
were not precursors of writing. On the other hand,
Sumerian, Chinese, and Mayan pictographs are
forerunners of writing because they utilized sound
representation in their scheme. De Francis (1989:
220221) corrected the commonly held misconception that Chinese is strictly pictographic. Chinese is
not now pictographic because its current system of
characters does not contain identifiable pictures.
Nevertheless, in its earliest form, the characters represented actual objects, i.e., there is a progression from
a pictographic representation to an abstract one
(Jensen, 1969: 166167). Thus, what is considered
to be the most primary group of Chinese characters
comprises picture signs. These signs show a close
resemblance between the object represented and the
object. The perspective of these signs may be from the
front, the side, or the top.
An exemplar par excellence of a pictograph comes
from the Dakota Indian winter-count buffalo robe
in Garrick Mallerys (18311894) volume (Mallery,
1886: plate VI) on North American Indians; the pictographs, drawn on a buffalo hide by Lone-Dog, of
the Yanktonais Dakota tribe, are a record of the

period from winter 1800/1801 to winter 1870/1871.

The pictographic images spiral out from the center of
the robe. Among the events that are recorded, to
name but a few of the incidents detailed on this
garment, are the outbreak of smallpox (18011802),
depicted by a human-like figure with pock marks; a
Canadian log cabin built of wood (18171818),
depicted by a primitive shelter with a chimney and
an adjacent tree; flood-related drawings of many
Indians (18251826), depicted by round objects
(heads) atop flat lines (local rivers); and an eclipse of
the sun (18691870), depicted by stars and a black
circle. All of the episodes are rendered by pictographs
that include visual icons, some of which are easily
understood and some of which require interpretation
by eyewitnesses or the tribal historian (see Figure 1).

Images and Semiotics

In this section, we discuss briefly the semiotic aspects
of visual signs, especially as they relate to pictography. Charles Sanders Peirces (18391914) theoretical

Figure 1 Lone Dogs Winter Count Before 1886. From Bureau of

American Ethnology Annual Report 4 (1886), Plate VI. National
Anthropological Archives. Smithsonian Institution Museum
Support Center.

594 Pictography: Semiotic Approaches

perspective of the sign has three elementary parts: (1)

the representamen (that part of the sign that represents), (2) the object (the concept encoded by the
sign), and (3) the intepretant (the meaning of the
sign assigned by a person) (for a related discussion
of Peirces notion of the sign, see Danesi and Perron
(1999: 7275) and Danesi (2004: 2333); see also
Farias and Queiroz (2003)). In Peirces system, the
representamen has three components: (1) the qualisign, which calls attention to certain qualities of its
referent (color, shape, size, and so forth), (2) the sinsign, which draws attention to an object in time or
space (deictic pronouns such as this and that), and (3)
the legisign, which is a conventional sign (symbols;
words for concepts).
Peirce specified three ways to represent objects:
(1) icons, which include representation by resemblance, (2) indication, which is gestural indication of
location, and (3) symbols, such as green as a symbol
of hope or fertility (in some cultures). In semiotic
terms, Peirce employed a classification system for
signs. His basic sign was the icon, which is a sensebased form of representation. These signs belong to
the category of firstness because they are physical
substitutes for their referents. Icons arise from the
cultures in which they develop, with the result that
their cross-cultural manifestations may differ to a
greater or lesser extent.
In recognition of the distinctive property of icons,
Peirce employed the term hypo-icon to mean those
icons that are culturally determined. The hypo-icon
refers to three basic types of icons. First, there is the
image, which has a similarity of expression and content. Second, there are diagrams, which involve analogous relations. Third, there is metaphor, which
involves the comparison of qualities between two
objects that may not be otherwise comparable. In
general, a hypo-icon may be understood by cultural
outsiders with some assistance from the members of
the culture in question.
Our interest here is the visual icon. The word icon
derives from the Greek eikon image, representation.
Marcel Danesi (b. 1946) defined an icon as a sign
that is made to resemble its referent through some
form of replication, resemblance, or simulation
(Danesi, 2000: 115). Icon in this article means that
there is representation through resemblance. In one
of their most common manifestations, icons represent people and events associated with Christianity,
e.g., visual depictions of sacred persons, such as
saints. In computer technology, icons represent everything visual all that can be displayed visually on
a computer screen, including icons that represent
programs or commands. Technically, iconic signs
may occur in any sensual modality. Thus, music and

onomatopoeia are auditory icons. Perfume is used in

certain societies generally as an olfactory icon to
enhance the human mating process. Certain chemicals that may be used to intensify the flavor of food
are gustatory signs. Finally, jigsaw puzzles are tactile
icons intended to produce an ultimate visual icon.
Again, our interest here is the visual icon. We now
examine various contemporary examples of visual
icons that serve as visual signs representative of the
real-world counterparts.
A visual sign may resemble quite closely the original that it represents, as in a photograph or in a
painting from the photo-realistic school of art, or it
may be much less real, as in the drawings of school
children. As noted previously, some visual signs may
be visual but not iconic, and others may be iconic but
not visual (audition, taste, smell, touch). Most visual
signs are iconic, though some are aniconic; aniconicity is a term that was introduced by Thomas
A. Sebeok (19202001) in an article on iconicity
(Sebeok, 1979: 111; cf. Liungman, 1991: 7). Examples of aniconic signs include the symbols of a typewriter or computer keyboard, Morse code, Braille,
certain traffic signs, and so forth. Aniconic signs are
often used conventionally, e.g., as decorative patterns, on certain ritual clothing, and on flags. There
is a tendency for iconic signs to become aniconic
through a process known as deiconization or conventionalization. For example, the Chinese system of
writing and hieroglyphic writing originated as visual
imitations of the objects to which they refer, but they
ultimately became conventionalized into symbols
that stood for words or syllables (Jensen, 1969:
166167).

Modern Manifestations of Pictography

The pictograph is a form of representation in which
the sign bears a pictorial resemblance to its referent.
Numerous contemporary icons have this property,
and in some ways, they may be labeled as modern
pictographs. In a multilingual world in which communication is immediate and global transportation is swift, the need for visual icons to convey
important (if not urgent) messages becomes a necessity. Humans now try to encode this vital information with signs, colors, shapes, and other formats to
convey what is hoped are universal and not culturespecific meanings, although this is quite difficult.
Some examples include computer icons, emoticons,
traffic signs, word games, and comics, to name
but a few important examples. These modern pictographs are called ideograms, i.e., they are visual
signs that stand for ideas or concepts (Shepherd,
1971).

Pictography: Semiotic Approaches 595

Table 1 Emoticons to indicate emotional responses

Table 2 Emoticons to represent people, animals, or objects

Meaning

Emoticon

Meaning

:-) :- :-]

Smile
Frown
Big smile
Anger
Pain
Wink

O-O

oo

Glasses
Normal face
Face with glasses
Bat
Cat
Robot

Emoticon

-)

>_<
.,

Computer Icons

Computers utilize icons to specify commands. Certain computer icons are, in fact, aniconic, because
they are conventionalized, and usually culture specific. A quick look at the desktop of an iMac computer
(Apple Computer, Inc.) illustrates this. The iMac, for
example, contains a trash barrel icon ( delete),
a magnifying glass icon with the English word
Sherlock ( a command to locate certain types of
information in the system), a printer icon ( print
command), and so forth. Other symbols that may
appear at certain times while using this brand of
computer include a bomb symbol (to indicate malfunction) and a clock face (to indicate that a command will take a certain amount of time to
accomplish). These icons thus indicate certain conventionalized computer functions through visual
means only.
Emoticons

Though a relatively recent phenomenon, computermediated communication, especially electronic mail

(e-mail), is now ubiquitous. One of its aspects is the
use of abbreviation to expedite the messages conveyed, in accordance with G. K. Zipfs (19021950)
law that hypothesizes that linguistic forms tend to be
abbreviated or reduced to minimize the labor required to produce them (Zipf, 1949). This occurs
with acronyms, e.g., FAQ ( frequently asked questions) and ?4U ( question for you). Because computer communication is a visual medium, it also uses
conventional symbols to create visual messages. It is
possible to use the conventional symbols found on
typewriter keys and computer keyboards to generate
certain icons or pictographs. These are called emoticons (emotion icon). Scott Fahlman (b. 1948)
is credited with popularizing the emoticon on the
Carnegie Mellon University electronic bulletin board
system between 1981 and mid-1982. This creative
combination of keyboard symbols can be used to
indicate the users affective reactions to a situation.
The emoticon is a conventionalized rendition of the
third type of nonverbal communication that Paul
Ekman (b. 1934) and W. V. Friesen (b. 1933) labeled

[_]

affect displays. These displays are primarily facial

indicators of emotional responses to certain stimuli
(Ekman and Friesen, 1969: 6981). Such responses
are a pan-cultural programmatic response to certain
affect antecedents (environment, expectations, memory) and they represent happiness, anger, surprise,
fear, disgust, sadness, and so forth. Because of their
quasi-universal nature, they may be encoded visually
through depictions of the face. Table 1 illustrates this
usage. In their volume on basic and applied studies of
the Facial Action Coding System (FACS), Ekman and
Friesen (1978) showed that there are certain facial
gestures that correspond to basic emotions. With
this system, Ekman and Friesen illustrated that it is
possible to measure and code facial reflexes of certain
basic emotions. These affect displays appear to be
universal and the primary affects include happiness,
surprise, fear, sadness, anger, disgust, and interest
(Ekman and Friesen, 1969: 71).
Another combination of keyboard symbols may
also be used to represent a sort of primitive drawing
or pictograph, to imitate the appearance of people,
animals, or objects, as shown in Table 2. These configurations are recognizable drawings of people, animals, and objects and are a modern retroflex of the
primitive cave paintings found in various archaeological sites. These emoticons resemble the shapes of the
original subjects and they are intended to represent
them, using only the repertoire of available symbols
on a typical keyboard.
Wong-Baker FACES Pain Rating Scale

The use of faces to depict degrees of pain provides

health care personnel and their patients with a means
of specifying the amount of suffering experienced. It
also provides health care providers with an objective
way of treating pain effectively. The Wong-Baker
FACES Pain Rating Scale provides a scale from 0 to
5 or 0 to 10 that shows line art faces with differing
degrees of pain, as illustrated in Figure 2. Research
indicates that simplicity is key to understanding the
amount of pain experienced by children, and the faces
used in the Wong-Baker FACES Pain Rating Scale are
the most efficient and accurate way to obtain this

596 Pictography: Semiotic Approaches

information from children, who may not have the

verbal sophistication to indicate their level of pain.
Traffic Signs

Traffic signs are iconic forms that contain coded visual information about traffic regulations and related
information. Unique visual signs enhance comprehension of the rules of the road, and color is key to
conveying certain types of information, as described
in Table 3. Traffic sign shapes also provide specific
information for the driver (see Figure 3). Efforts have
been made to internationalize traffic signs and signals
in order to provide drivers worldwide with a uniform
set of regulatory signs.

the word for the European nation, Finland. The

key to solving visual word games is to pronounce in
sequence each individual word that a picture or symbol represents (see Figure 5).
Visual icons are used to create other types of word
games. The well-known concrete poems became
quite popular in the middle of the 20th century, and
many of them are to be found in Mary Ellen Solts (b.
1920) anthology (Solt, 1968). The visual typographical format in concrete poems imitates the shape actual shape of the semantic content of the poem. The
Portuguese poet Salette Tavares (19221991) created
a good example of this type of iconography. His

Word Games

Table 3 Color meaning in traffic signs

Certain word games involve pictographs or icons.

The classic example of this is the rebus (Latin for
by means of things), which is a word puzzle in
which an icon stands for a phonetic value. Thus, the
picture of a bee (iconic sign) and the Arabic numeral 9
(aniconic sign) represent the English word benign
(see Figure 4). Another common example is the picture of a fish fin (iconic sign) plus the letter L and the
symbol & (aniconic signs), which visually represent

Color

Meaning

Red

Informs the driver of requirements that the driver must

follow STOP, YIELD, DO NOT ENTER
Warns the driver of road conditions and dangers
ahead SOFT SHOULDERS, ICE
Informs the driver about traffic regulations, such as
speed limits
Informs the driver about directions, such as exits and
distances

Yellow
White
Green

Figure 2 Wong-Baker FACES Pain Rating Scale. Reproduced with permission from Wong D L, Hockenberry-Eaton M, Winkelstein M L,
and Schwartz P (2001). Wongs essentials of pediatric nursing (6th edn.), p. 1301. Copyrighted by Mosby, Inc., St. Louis.

Figure 3 Shape meaning in U.S. traffic signs. (A) Octagon, STOP; (B) triangle, YIELD; (C) square, NO LEFT TURN (square
regulatory signs indicate rules that a driver must follow; superimposition of a red circle with a diagonal bar means NO); (D) circle
(RAILROAD CROSSING); (E) pentagon, CHILDREN CROSSING (pentagon, or house, shape indicates a school zone).

Pictography: Semiotic Approaches 597

poem, Arranhisso (Solt, 1968: 190), utilizes the

Portuguese word for spider (aranha) and its verbal
derivatives (not all of which are real Portuguese
words) to form the shape of a spider.
The forerunners of 20th-century concrete poetry
were the emblematic poems of earlier times. One
example is George Withers (15881677) Rhomboidal dirge (Bombaugh, 1961: 94), which has the shape
of a rhomboid and is a sort of life review. Its shape
reflects quite succinct recollections that represent the
narrators life. It begins and ends with the interjection
Farewell! There is also the anonymous shape poem
entitled A wine glass, the verses of which create the
profile of a wine goblet (Bombaugh, 1961: 93).

Figure 4 Rebus (English word benign). Source: http://

www.enchantedlearning.com.rhymes/topics.animals/shtml

Figure 5 Rebus (English word Finland).

Another example, the emblematic poem by Joseph

Miller (b. 1939) entitled If the Earth is illustrated
in Figure 6. Lewis Carrolls (pseudonym of Charles
Lutwidge Dodgson, 18321898) Rats tale narrative
(see Figure 7) in Alice in wonderland (1865) is an
example of a mini-story that assumes the shape of
its theme; this tale (a pun on the word tail) has the
shape of a rats tail.
Comics

Comics are sequences of related artwork. The technical designation for this narrative art form is lexipictogram. The related sequences may be four
contiguous panels in daily newspapers, or a dozen,
in Sunday papers. Comic books, which are 24 to 32
pages in length, may be a single story or a collection
of four different stories. Graphic novels, a relatively
new phenomenon, may be several hundred pages in
length, and tell full-length novelesque stories in pictures. The best qualitative example of this genre is Art
Spiegelmans (b. 1948) 1992 Pulitzer Prize-winning
Maus, a riveting tale about the Nazi concentration
camps, with mice as the protagonists. Most often,
comics are a combination of text and image, though
some comics are completely visual, such as Jim
Woodrings (b. 1952) The Frank book, in which
the reader understands the meaning through visual

Figure 6 Shape poem If the Earth. Copyright by Joe Miller, 1975, Moab, Utah. Reproduced with permission.

598 Pictography: Semiotic Approaches

Figure 7 The narrative Rats tale, from Lewis Carrolls Alice in

wonderland. The shape is emblematic of the semantic meaning.

material only. In fact, however, it is possible to describe this material verbally and to create a verbal
narrative account of the readers interpretation of
the visual content. Will Eisner (b. 1911) calls comics
sequential art (Eisner, 1985). Since comics frequently contain text in the form of dialogue balloons emanating from the mouths of the characters (humans or
animals), they are also narrative art.
Comics have been the subject of semiotic research
since the 1960s. In this regard, Umberto Eco (b.
1932) analyzed visual metaphors (Eco, 1994). Likewise, Roma n Gubern (b. 1934; Gubern, 1972) and
Fresnault-Deruelle (b. 1943; Fresnault-Deruelle,
1972) examined this form and its governing codes.
Comic content and form obey certain codes, including shape and size of the panels, typeface, and pen
and ink strokes, all of which are intended to reflect the
mood (somber, humorous, frightful) of the narrative
and the emotion (anxiety, joy, sadness) of the
protagonists of the story line. Comics utilize verbal
and nonverbal means to convey their messages.
Certain conventions are employed in the creation of a comic. The visual component of the comic
follows a left-to-right chronological sequence with a
set of visual iconic and linguistic materials. Comics
consist of frames or panels that may be of equal or
differing size and shape according to the action
depicted. The pictorial elements include the scenario
and the protagonists, whose states of mind are represented in their facial gestures and body language.
Comics thus display a whole range of kinesic and
proxemic behaviors through the use of distinct artistic techniques, such as mobilgrams, or motion lines,
to specify direction and rapidity. The verbal component of the comic follows fairly strict codes, thus
allowing the reader to understand the formalities
of the medium once the meanings of the artistic techniques have been established. First, there are locugrams (speech balloons) with a connecting line to
the mouth or face of the protagonist. Next is the
line that envelops the locugram, the perigram,
and its format depicts metaphorically the state of
mind of the protagonist. Third, the psychopictogram
(dream balloon) expresses the thoughts or dreams of
the protagonist. Conventionalized ideograms, sometimes called sensograms, are iconic symbols that convey the emotions of the protagonist, which are
culturally determined conventions; examples include
the light bulb to indicate a sudden insight, or a heart
with broken lines to indicate the sad outcome of a
romantic relationship, and so forth. Phonosymbolism, or onomatopoetic protocol, is common in
comics. The phonosymbols Bam! and Wack!, for example, represent the impact of fists in a fistfight
(Chapman, 1984).

Pictography: Semiotic Approaches 599

Among the techniques employed in the comic are

visual perspective, parallel actions, flashbacks, and
so forth. The ability to decipher the meaning of
comics is culturally determined, though many of the
conventions used in comics are shared in Western and
Japanese cultures. Several books about comics provide excellent introductions to this medium, especially those by Davies (1995), Will Eisner (1985), Robert
C. Harvey (1996), and Scott McCloud (1985).
Cartoons, found in many popular magazines, are
distinct from comics because they are usually one
panel in length and their content is usually political
satire or commentary (Press, 1981). In addition
to their function as political parody, single-panel
cartoons may also simply be humorous.
See also: Comics: Semiotic Approaches; ComputerMediated Communication: Cognitive Science Approach;
Face; Iconicity: Literary Texts; Iconicity: Sign Language;
Media and Language: Overview; Word and Image; Word
Recognition, Written; Writing and Cognition.

Bibliography
Bombaugh C C (1961). Oddities and curiosities of words
and literature (gleanings for the curious). New York:
Dover Publications.
Chapman R (1984). The treatment of sounds in language
and literature. Oxford: Blackwell.
Danesi M (2000). Encyclopedic dictionary of semiotics,
media, and communications. Toronto: University of
Toronto Press.
Danesi M (2002). The puzzle instinct: the meaning of puzzles in human life. Bloomington, IN: Indiana University
Press.
Danesi M (2004). Messages, signs, and meanings: a basic
textbook in semiotics and communication theory. Toronto: Canadian Scholars Press, Inc.
Danesi M & Perron P (1999). Analyzing cultures: an
introduction and handbook. Bloomington, IN: Indiana
University Press. [This is an excellent introduction to
semiotic theory and concepts.]
Davies L J (1995). The multidimensional language of
the cartoon: a study of aesthetics, popular culture, and
symbolic interaction. Semiotica 104, 165211.
DeFrancis J (1989). Visible speech: the diverse oneness of
writing systems. Honolulu: University of Hawaii Press.
[This is an excellent introduction to history and development of writing systems.]
Eco U (1994). Apocalypse postponed: essays by Umberto
Eco. Lumley R (ed.) Bloomington, IN: Indiana University
Press.

Eisner W (1985). Comics and sequential art. Tamarac, FL:

Poorhouse Press.
Ekman P & Friesen W V (1969). The repertoire of nonverbal behavior: categories, origins, usage and coding.
Semiotica 1, 4998.
Ekman P & Friesen P (1978). The facial action
coding system. Palo Alto, CA: Consulting Psychologists
Press.
Farias P & Queiroz J (2003). On diagrams for Peirces 10,
28, and 66 classes of signs. Semiotica 147, 165184.
Fresnault-Deruelle P (1972). La bande desine e, essai danalyse se miologique. Paris: Hachette.
Gelb I J (1963). A study of writing. Chicago: University of
Chicago Press.
Gubern R (1972). El lenguaje de los comics. Barcelona:
Pennsula.
Harris R (1986). The origin of writing. La Salle, IL: Open
Court Press.
Harvey R C (1996). The art of the comic book: an
aesthetic history. Jackson, MS: University Press of
Mississippi.
Jensen H (1969). Sign, symbol and script: an account
of mans efforts to write (3rd rev. edn.). Unwin G
(trans.). New York: G. P. Putnams Sons. [This is an
excellent and comprehensive account of world writing
systems.]
Liungman C G (1971). Dictionary of symbols. Santa Barbara, CA: ABC-Clio.
Mallery G (1886). The Dakota winter counts and Pictographs of the North American Indians. In Fourth annual report of the Bureau of American Ethnology.
Washington, D.C.: Smithsonian Institution. 89146,
plates VILI. [Reprinted by J & L Reprint Company,
Lincoln, NE, 1987.]
McCloud S (1993). Understanding comics: the invisible art.
Northampton, MA: Kitchen Sink Press.
Peirce C S (19311958). Collected papers of Charles Sanders Peirce, vols 18. In Hartshorne C & Weiss P (eds.).
Cambridge, MA: Harvard University Press.
Press C (1981). The political cartoon. Rutherford, NJ:
Farleigh Dickinson University Press.
Sebeok T A (1979). Iconicity. In Sebeok T A (ed.) The sign
and its masters. Austin, TX: University of Texas Press.
107127.
Sebeok T A (1994). Signs: an introduction to semiotics.
Toronto: University of Toronto Press.
Shepherd W (1971). Shepherds glossary of graphic signs
and symbols. London: J. M. Dent and Sons.
Solt M E (1968). Concrete poetry: a world view. Bloomington, IN: Indiana University Press.
Wallis M (1975). Arts and signs. Bloomington, IN: Indiana
University Press.
Zipf G K (1949). Human behavior and the principle of the
least effort. Cambridge, MA: Addison-Wesley.

600 Pidgins and Creoles: Overview

Pidgins and Creoles: Overview

S Romaine, Oxford University, Oxford, UK
2006 Elsevier Ltd. All rights reserved.

Definitions
European colonization during the 17th to 19th centuries created a classic scenario for the emergence of
new language varieties called pidgins and creoles out
of trade between the native inhabitants and Europeans. The term pidgin is probably a distortion of
English business and the term creole was used in
reference to a nonindigenous person born in the
American colonies, and later used to refer to customs,
flora, and fauna of these colonies. Many pidgins and
creoles grew up around trade routes in the Atlantic
or Pacific, and subsequently in settlement colonies on
plantations, where a multilingual work force comprised of slaves or indentured immigrant laborers
needed a common language. Although European
colonial encounters have produced the most well
known and studied languages, there are examples of
indigenous pidgins and creoles predating European
contact such as Mobilian Jargon (Mobilian), a now
extinct pidgin based on Muskogean (Muskogee), and
widely used along the lower Mississippi River valley
for communication among native Americans speaking
Choctaw, Chickasaw, and other languages (see
Mobilian Jargon).
The study of pidgins and creoles raises fundamental
questions about the evolution of complex systems,
since pidgins, in particular, have been traditionally
regarded as simple systems par excellence. The usual
European explanation given for the simplicity, and
lack of highly developed inflectional morphology
in particular, was that it reflected primitiveness, native mental inferiority, and the cognitive inability of
the natives to acquire more complex European languages. Thus, for example, Churchill (1911: 23) on
Bislama, the pidgin English spoken in Vanuatu: the
savage of our study, like many other primitive thinker,
has no conception of being in the absolute; his speech
has no true verb to be (see Bislama).
Hampered by negative attitudes for many years,
scholars ignored pidgins and creoles in the belief
that they were not real languages, but were instead
bastardized, corrupted, or inferior versions of the
European languages to which they appeared most
closely related. Although scholars still do not agree
on how to define pidgins and creoles, or the nature
of their relationship to one another, most linguists
recognize such a group of languages, whether defined
in terms of shared structural properties and/or sociohistorical circumstances of their genesis. Striking

similarities across pidgin and creole tense-moodaspect (TMA) systems (see Tense, Mood, Aspect:
Overview) were noted by some of the earliest scholars
in the field such as Hugo Schuchardt, generally
regarded as the founding father of creole studies (see
Schuchardt, Hugo (18421927)). TMA marking
became a focal point of debate among creolists as a
result of the bioprogram hypothesis (Bickerton, 1981,
1984), according to which creoles held the key to
understanding how human languages originally
evolved many centuries ago (see Evolutionary Theories of Language: Previous Theories and Evolutionary Theories of Language: Current Theories). This
theory led not only to an increase in research on
these languages, but also a great deal of attention
from scholars in other fields of linguistics, such as
language acquisition and related disciplines such as
cognitive science.

Classifying Pidgins and Creoles

The standard view that pidgins and creoles are mixed
languages with the vocabulary of the superstrate (also
called the lexifier or base language) and the grammar
of the substrate (the native languages of the groups in
contact) has been the traditional basis for classifying
these languages according to their lexical affiliation.
English-lexicon pidgins and creoles such as Solomon
Islands Pijin spoken in the Solomon Islands or Jamaican Creole English (Southwestern Carribean Creole
English) in Jamaica comprise a group of languages
with lexicons predominantly derived from English.
Haitian Creole French and Tayo, a French creole
of New Caledonia, are French-lexicon creoles drawing most of their vocabulary from French. Such
groupings are, however, distinctly different from the
genetically-based language families established by
the comparative historical method (see ). Pidgins or
creoles as a group are not genetically related among
themselves, although those with the same lexifier
usually are.
There is a great deal of variation in terms of the
extent to which a particular pidgin or creole draws
on its lexifier for vocabulary, and a variety of problems in determining the sources of words, due to
phonological restructuring. Compare the lexical
composition of Sranan and Saramaccan, two of six
English-lexicon creoles spoken in Surinam, in what
was formerly the Dutch-controlled part of Guyana.
About 50% of the words in Saramaccan are from
English (e.g., wa ka walk), 10% from Dutch (e.g.,
strei fight <strijd), 35% from Portuguese (e.g., disa
quit < deixar), and 5% from the African substrate

Pidgins and Creoles: Overview 601

languages (e.g., toto mbot woodpecker. By contrast,

only 18% of Sranan words are English in origin, with
4.3% of African origin, 3.2% of Portuguese, 21.5%
of Dutch; 4.3% could be derived from either English
or Dutch. Innovations comprise another 36%, and
12.7% have other origins. African words are concentrated in the semantic domains of religion, traditional
food, music, diseases, flora, and fauna. Words from
the other languages do not concentrate in particular
semantic domains. Numbers, for instance, draw on
both English and Dutch. Sranan and Saramaccan are
not mutually intelligible, and neither is mutually intelligible with any of the input languages. Other languages show a more equal distribution between two
main languages, such as Russenorsk, a pidgin once
spoken along the Arctic coast of northern Norway
from the 18th until the early 20th century. Its vocabulary is 47% Norwegian, 39% Russian, 14% other
languages including Dutch (or possibly German),
English, Saami, French, Finnish, and Swedish (see
Russenorsk).
Many creoles, like Lesser Antillean (Lesser Antillean Creole French), a French-based creole spoken in
the French Antilles, started out with a far more mixed
lexicon than they possess today. Where contact with
the main European lexifier was permanently terminated, as in Surinam, the lexicon retains a high degree
of mixture to the present day; where such contact
continued, as in the Lesser Antilles, items from the
main lexifier tended gradually to replace items from
other sources. Depending on the circumstances, a
creole may adopt more items from the superstrate
language due to intense contact. In Tok Pisin spoken
in Papua New Guinea, some of the 200 German elements as well as words from indigenous languages,
are now being replaced by English words. Thus, beten
(German pray) is giving way to English pre, and
Tolai (Kuanua) kiau to English egg (see Tok Pisin).

Relationships between Pidgins and

Creoles
The question of the genetic and typological relationship between pidgins and creoles and the languages
spoken by their creators continues to generate controversy. Pidgins and creoles challenge conventional
models of language change and genetic relationships
because they appear to be descendants of neither the
European languages from which they took most of
their vocabulary, nor of the languages spoken by their
creators. The conventional view of the languages and
their relationship to one another found in a variety of
introductory texts (Hall, 1966; Romaine, 1988) has
been to assume that a pidgin is a contact variety
restricted in form and function, and native to no

one, which is formed by members of at least two

(and usually more) groups of different linguistic backgrounds, e.g., Krio in Sierra Leone (see Krio). A creole
is a nativized pidgin, expanded in form and function
to meet the communicative needs of a community of
native speakers, e.g., Haitian Creole French.
This perspective regards pidginization and creolization as mirror image processes and assumes a prior
pidgin history for creoles. This view implies a twostage development. The first involves rapid and drastic restructuring to produce a reduced and simplified
language variety. The second consists of elaboration
of this variety as its functions expand, and it becomes
nativized or serves as the primary language of most
of its speakers. The reduction in form characteristic
of a pidgin follows from its restricted communicative
functions. Pidgin speakers, who have another language, can get by with a minimum of grammatical
apparatus, but the linguistic resources of a creole
must be adequate to fulfill the communicative needs
of human language users.
The degree of structural stability varies, depending
on the extent of internal development and functional
expansion the pidgin has undergone at any particular
point in its life cycle. Creolization can occur at any
stage in the development continuum from rudimentary jargon to expanded pidgin. If creolization occurs
at the jargon stage, the amount of expansion will
be more considerable than that required to make an
expanded pidgin structurally adequate. In some cases,
however, pidgins may expand without nativization.
Where this happens, pidgins and creoles may overlap
in terms of the structural complexity, and there will
be few, if any, structural differences between an expanded pidgin and a creole that develops from it.
Varieties of Melanesian Pidgin English (a cover term
for three English-lexicon pidgins/creoles in the southwest Pacific comprising Tok Pisin, Solomon Islands
Pijin and Vanuatu Bislama) are far richer lexically and
more complex grammatically than many early creoles
elsewhere. Their linguistic elaboration was carried
out primarily by adult second language speakers
who used them as lingua francas in urban areas.
Creolization is thus not a unique trigger for complexity, and the same language may exist as both pidgin
and creole.
Debate continues about the role of children vs.
adults in nativization and creolization. Other scholars
have emphasized the discontinuity between creoles
and pidgins on the basis of features present in certain
creoles not found in their antecedent pidgins. They
argue that ordinary evolutionary processes leading to
gradual divergence over time may not be applicable
to creoles. Instead, creoles are born again nongenetic languages that emerge abruptly ab novo via a break

602 Pidgins and Creoles: Overview

in transmission and radical restructuring (Thomason

and Kaufman, 1988).

Origins
Because pidgins and creoles are the outcome of
diverse processes and influences in situations of language contact where speakers of different languages
have to work out a common means of communication, competing theories have emphasized the importance of different sources of influence. Few creolists
believe that one theory can explain everything satisfactorily, and there are at least four theories accounting for the genesis of creoles: substrate, superstrate,
diffusion, and universals.
Substrate

The substrate hypothesis emphasizes the influence

of the speakers ancestral languages. Structural affinities have been established between the languages of
West Africa and many of the Atlantic creoles. Scholars have also documented substantial congruence
between Austronesian substratum languages (see
Austronesian Languages: Overview) and Pacific pidgins as compelling evidence of the historically primary role of Pacific Islanders in shaping a developing
pidgin in the Pacific. Substrate influence can be seen
in the pronominal systems of Melanesian Pidgin English such as the personal pronouns in Tok Pisin. The
forms are rather transparently modeled after English,
yet incorporate grammatical distinctions not found in
English, but widely present in the indigenous languages forming the substrate.

first person
second person
third person

Personal pronouns in Tok Pisin

singular
plural
mi
I
mipela we (exclusive)
yumi we (inclusive)
yu
you
yupela you
em he/she/it ol they

Almost all Oceanic languages distinguish between

inclusive (referring to the speaker and addressee(s),
I you) and exclusive first-person pronouns (referring to the speaker and some other person(s), I
he/she/it/they). Thus, yumi consists of the features
[speaker, hearer, other] and mipela, [speaker,
hearer, other]. There are also dual and trial forms,
e.g., yumitupela we two (inclusive), i.e. [speaker,
hearer, other], mitripela we three (exclusive),
etc., although these distinctions are not always made
consistently. As English provides no lexical forms for
the inclusive/exclusive and dual distinctions or you
plural, these are created by forming a compound from
you me to give yumi and yumitupela, and by using
the suffix-pela (fellow) to mark plurality in yupela.
The third-person singular form em is derived from the

unstressed third person singular him and the third

person plural form ol from all.
A more controversial variant of the substrate
hypothesis is incorporated into the notion of relexification, a process that applies to the words/structures
of substrate language and matches them with phonological representations from the lexifier language.
Haitian Creole French gade shares some meanings
with the French verb garder to watch over/take care
of/to keep, from which it derives its phonetic form,
but it has an additional meaning to take care of/
defend oneself. The semantics of gade is very similar
to that of the substrate Fongbe (Fon-Gbe) verb kpo n
to watch over/take care of/to keep/to look. Haitian
Creole French gade also means to look, while in
French that meaning is expressed by regarder. These
similarities have led some scholars to regard Haitian
Creole French as a French relexification of African
languages of the Ewe-Fon (or Fongbe) group
(Lefebvre, 1998).
Superstrate

The superstrate hypothesis traces the primary source

of structural features to nonstandard varieties of the
lexifiers, and to evolutionary tendencies already observable in them (Chaudenson, 1992). According to
this scenario, early plantation slaves acquired a normally transmitted variety of the lexifier directly from
Europeans, but this imperfectly acquired variety was
subsequently diluted over time as successive generations of slaves learned from other slaves rather than
from Europeans. Creoles thus represent gradual continuous developments with no abrupt break in transmission from their lexifiers. This evidence eliminates
the assumption of a prior pidgin history and accepts
creoles as varieties of their lexifiers rather than as
special or unique new languages. That is, there are
no particular linguistic evolutionary processes likely
to yield (prototypical) creoles; they are produced by
the same restructuring processes that bring about
change in any language. Creoles are neither typologically nor genetically unique, but advanced varieties
of the lexifiers.
Linguistic evidence supporting this hypothesis can
be found in morphemes or constructions chosen for
specific grammatical functions that start from models
available in the lexifiers. Haitian Creole French m pu
alle I will go may not be a totally new and radical
departure from French but could instead be derived
from regional French je suis pour aller.
Diffusion

Another explanation for some of the similarities among

pidgins and creoles is diffusion of a pre-existing pidgin.

Pidgins and Creoles: Overview 603

According to this hypothesis, a pre-existing English

or French pidgin was transplanted from Africa rather
than created anew independently in each territory.
Support for this hypothesis can be found in historical
evidence that sailors diffused not only words with
nautical origins from one part of the world to another,
but also items that were more generally part of regional and nonstandard usage. Thus, capsize was
probably originally a nautical term meaning to overturn a boat. Today, kapsaitim in Melanesian Pidgin
English means to spill or overturn anything. Traders,
missionaries, and early settlers were also responsible
for diffusing certain elements. Words from Portuguese
such as savvy (<sabir to know/understand, first
attested in 1686) are found widely around the world.
Scholars have traced the paths of diffusion of so-called
worldwide features found in Anglophone pidgins
and creoles from the Atlantic to Pacific (Baker and
Huber, 2001). Words from indigenous languages are
also widespread, e.g., African nyam eat/food and
Hawaiian kanaka person/man, a term that came
to be used, often derogatorily, to refer to Pacific
Islanders.
Universals

This theory actually comprises a variety of sometimes

opposing viewpoints because universals have been
conceived of in a variety of ways within different
theoretical perspectives. Its central assumption is
that creoles are more similar to one another than the
languages to which they are otherwise most closely
related due to the operation of universals. Although it
has become fashionable to refer to a common creole
syntax or creole prototype, not all creolists agree on
the nature or extent of the similarities or the reasons
for them. If creoles form a synchronically definable
class, then there should be more similarities between
Haitian Creole French and Guyanese Creole English
than between Haitian Creole French and French, or
between Guyanese Creole English and English. One
kind of universalist claim is that creoles reflect more
closely universal grammar and the innate component
of the human language capacity (see Linguistic Universals, Chomskyan). Another, however, is grounded
within a different notion of universals derived from
crosslinguistic typology and theories of markedness
(see Linguistic Universals, Greenbergian). The observation that creoles tend to be isolating languages even
when the contributing languages show a different
typology has a long history predating modern typological theories (see Morphology in Pidgins and
Creoles). Kituba, for example, emerged almost exclusively from contact among Bantu languages that are
agglutinative.

The notion of creoles as the simplest instantiation of

universal grammar is at the heart of Bickertons
(1981) bioprogram hypothesis, which applies to
radical creoles, i.e., those that have undergone a sudden creolization without further major superstrate
influence. It is based to a large extent on similarities
between Hawaii Creole English, Guyanese Creole
English, Haitian Creole French, and Sranan. Evidence
from Hawaii Creole English has been the cornerstone
of the bioprogram because creolization has been more
recent there than in many other cases, and because
the language lacked an African substrate, yet was
strikingly similar to other creoles (see Hawaiian Creole English). This similarity is explained by assuming
that creoles represent a retrograde evolutionary movement to a maximally unmarked state.
Bickerton (1981) proposed a list of 13 features
shared by creoles that were not inherited from the
antecedent pidgins, and therefore must have been
created by children as a result of the bioprogram.
1. Focused constituents are moved to sentence initial
position, e.g., Haitian Creole French se mache
Jan mache al lekol John walked to school.
2. Creoles use a definite article for presupposed specific noun phrases, indefinite articles for asserted
specific noun phrases, and zero for nonspecific
noun phrases. Hawaii Creole English uses definite article da for presupposed specific noun
phrases, e.g., she wen go with da teacher she
went with the teacher, indefinite article one typically for first mention, e.g., he get one white truck
he has a white truck, and no article or maker of
plurality for other noun phrases, e.g., young guys
they no get job Young people dont have jobs.
3. Three preverbal morphemes express tense (anterior), mood (irrealis), and aspect (durative) in
that order, e.g., Haitian Creole French li te
mache he walked, lav(a) mache he will walk,
lap mache he is walking.
4. Realized complements are either unmarked or
marked with a different form than the one used
for unrealized complements, e.g., Mauritian Creole French (Morisyen) il desid al met posoh ladah
she decided to put a fish in it vs. li ti pe ale aswar
pu al bril lakaz sa garsohla me lor sime ban
dayin lin atake li He would have gone that evening to burn the boys house, but on the way he
was attacked by witches.
5. Creoles mark relative clauses when the head
noun is the subject of the relative clause, e.g.,
Hawaii Creole English some they drink make
trouble Some who drink make trouble.
6. Nondefinite subjects, nondefinite verb phrase constituents, and the verb must all be negated in

604 Pidgins and Creoles: Overview

7.

8.

9.

10.

11.

12.

13.

negative sentences, e.g., Guyanese Creole English

non dag na bait non kyat no dog bit any cat.
Creoles use the same lexical item for both existentials and possessives, e.g., Hawaii Creole English get one wahine she get one daughter There
is a woman who has a daughter.
Creoles have separate forms for each of the semantically distinct functions of the copula (i.e.,
locative and equative), e.g., Sranan a ben de na
ini a kamra (s)he was in the room. vs. mi na
botoman I am a boatman.
Adjectives function as verbs, e.g., Jamaican
Creole English di pikni sik the child is sick.
This function explains the absence of the copula
in this construction.
There are no differences in word order between
declaratives and questions, e.g., Guyanese Creole
English i bai di eg dem means he bought the eggs
or did he buy the eggs?, depending on intonation.
Questions particles are optional and sentence
final, e.g., Tok Pisin yu tok wanem? what did
you say. Question words are often bimorphemic, e.g., Haitian Creole French ki kote where
(French qui cote which side), and Tok Pisin
wanem which/what (English what name).
Formally distinct passives are typically absent,
e.g., Jamaican Creole English dem plaan di tree
they planted the tree vs. di tree plaan the tree
was planted.
Creoles have serial verb constructions in which
chains of two or more verbs have the same subject,
e.g., Nigerian Pidgin English (Pidgin, Nigerian)
dem come take night carry di wife, go give di man
They came in the night and carried the woman to
her husband. (see Serial Verb Constructions).

There are also many similarities in the source morphemes used by creoles to express these distinctions.
The semantics of the grammatical morphemes are
highly constant as are their etymologies; in almost
all cases, they are drawn from the superstrate language. The indefinite article is usually derived from
the numeral one, the irrealis mood marker from a
verb meaning go, the completive marker from a verb
meaning finish, the irrealis complementizer from a
reflex of for, etc.
Support for the uniqueness of these features to
creoles is, however, weakened by the existence of
some of the same traits in pidgins as well as in the
relevant substrates and superstrates. The relexification hypothesis argues that the typological traits of
Haitian Creole French display more in common
with those of the substrate language Fongbe than
with French. If so, then the supposed creole typology
results from the reproduction of substratum

properties rather than from the operation of universals. Bimorphemic question words are also found in
many of the African substrate languages, and English
has what time when, how come why, etc. It is
also well within the norms of colloquial French
and English to use intonation rather than word
order to distinguish questions from declaratives, e.g.,
youre doing what? The absence of passives may also
reflect the lack of models in some of the substrate and
superstrate languages.
Closer study of the particulars of individual TMA
systems in creole languages has engendered increasing dissatisfaction with the bioprogram hypothesis
(Singler, 1990). For one thing, the claims were originally formulated on the basis of data from creoles
whose superstrate languages are Indo-European. Secondly, it is also unclear how much creole TMA systems might have changed over time after creolization.
The bioprogram assumes that the creoles in question have not departed from their original TMA prototype and that the present day systems provide
evidence of relevance for its operation. Thirdly, even
the defining languages do not conform entirely to
predictions on closer examination. The TMA system
of Hawaii Creole English is not crosslinguistically
unique or even unusual; the overwhelming majority
of its TMA categories are common in languages of
world (Velupillai, 2003). More detailed investigations of historical evidence indicate that Bickertons
scenario of nativization bears little resemblance to
what actually happened in Hawaii (Roberts, 2000).
The typology of creoles might also be largely a
result of parameter settings typical of languages
with low inflectional morphology (see Principles and
Parameters Framework of Generative Grammar).
Thus, features such as preverbal TMA markers, serial
verbs, and SVO word order fall out more generally
from lack of inflections and unmarked parametric
settings. McWhorter (1998) attempts to vindicate
creoles as a unique typological class by proposing a
diagnostic test for creolity based not on specific
shared structural features such as TMA markers, serial verbs, etc., but on a combination of three traits
resulting from a break in transmission: little or no use
of inflectional affixation, little or no use of lexical
tone, and semantically regular derivational affixation. McWhorters explanation for why these traits
cluster essentially reiterates the conventional assumption that pidgins are languages that have been
stripped of all but the bare communicative necessities
in order to speed acquisition. Because creoles are new
languages that emerge from pidgins, they have not
had the time to develop many of the complexities
found in other languages that have developed gradually over a much longer time period. Thus, he predicts

Pidgins and Creoles: Overview 605

that features such as ergativity (see Ergativity), a

distinction between alienable and inalienable possession, switch reference marking (see Switch Reference), noun class or grammatical gender marking
(see Gender, Grammatical), etc. will never be found
in creoles. This theory means that not only are creoles
typologically unique, but also that they are the simplest languages. Those who stress the role of substrate
influence and relexification, however, have argued
that the reason why these features do not surface in
creoles even where they are present in the substrate is
because there are no appropriate phonetic strings in
the superstrate to match them with.
The question of how to measure simplicity and
complexity is theory-dependent and therefore controversial. McWhorters (2001) complexity metric is
based on degree of overt signalling of various phonetic, morphological, syntactic, and semantic distinctions. From this perspective, a phonemic inventory
can be considered more complex if it contains more
marked members than some other (see Phonological
Universals). Markedness is interpreted in terms of
frequency of representation among the worlds languages. Ejectives and clicks are more marked than
ordinary consonants because they occur less frequently. The presence of rarer sounds in an inventory also
presupposes the existence of more common or less
marked ones. However, there may be other dimensions of simplicity/complexity to consider, such as
syllable/word structure. Much less is known about
the phonology of pidgins and creoles than about
their syntax and lexicon. Syntax is rendered more
complex by the additional of rules that make it
more difficult to process, e.g., different word orders
for main and subordinate clauses. Inflectional marking is assumed to be more difficult than the use of free
morphemes. However, there is no universally accepted account of syntactic rules nor an agreed theory of
processing. Semantically, creoles are more transparent and adhere more closely to the principle of one
formone meaning.
There are problems with this view too, because
creoles do not share their features universally or exclusively. There are examples of noncreole languages
with the assumed typical creole-like features, and
some examples of languages with no known creole
history that are less complex than some creoles.
Given that language change may also lead to simplification, some languages that are older than creoles
may also be less complex than creoles. Similarities
among creoles may be the result of chance similarities
among unrelated substrates. Although the absence
of inflection is perhaps the most often cited typological feature of creoles, it may be the accidental result

of limited typological spread of the contributing

languages.
Yet another interpretation of the universalist approach involves the assumption that common processes of restructuring apply in situations of language
contact to produce common structural outcomes. The
effects of contact may operate to differing degrees
depending on the social context, e.g., number and
nature of languages involved, extent of multilingualism, etc. The fact that pidgins and creoles share some
structural features with each other and with other
language varieties that are reduced in function such
as koines, learner varieties, etc., indicates that the
same solutions tend to recur to some degree wherever
acquisition and change occurs, regardless of contact,
but especially in cases of contact. The entities called
pidgins and creoles are salient instances of the processes of pidginization and creolization respectively,
although they are not in any sense to be regarded as
unique or completed outcomes of them. From this
point of view, pidgins represent a special or limiting
case of reduction in form resulting from restriction
in use.
This statement brings us back to the position that
the only thing special about creoles is the sociohistorical situation of language contact in which they
emerge. Even that may not be so special when we
consider the history of so-called normal languages,
most of which are hydrid varieties that have undergone restructuring to various degrees depending on
the circumstances. Even normal languages such as
English have been shaped by heavy contact with nonGermanic languages and thus can be thought of as
having more than one parent. If universal grammar is
a mental construct, or an innate predisposition to
develop grammar, then in so far as there is no psychological continuity between the mental representations
of one generation of speakers of a language and the
next, all grammars are created anew each generation.
There will always be a certain amount of discontinuity between the grammars of parents and children,
and acquisition is always imperfect. Thus, the supposed dichotomy between normal and abrupt transmission is spurious because normal transmission is in
fact abrupt.

Directions for Future Research

Resolution of some of the debates about pidgins and
creoles, their origins, and their relationships to one
another as well as to the languages spoken by their
creators is hampered by lack of knowledge of the
relevant substrate languages as well as insufficient
knowledge of the history of the nonstandard varieties

606 Pidgins and Creoles: Overview

of European languages that formed the lexifiers.

There are few detailed grammatical descriptions of
pidgins and creoles available for sophisticated typological analysis. More sociohistorical research is also
needed. Earlier scholarship often overstated the similarities among creoles and ignored key properties
unique to individual ones.

See also: Austronesian Languages: Overview; Bislama;

Congo, Democratic Republic of: Language Situation; Ergativity; Evolutionary Theories of Language: Current
Theories; Evolutionary Theories of Language: Previous
Theories; Gender, Grammatical; Guyana: Language
Situation; Haiti: Language Situation; Hawaiian Creole
English; Jamaica: Language Situation; Krio; Linguistic
Universals, Chomskyan; Linguistic Universals, Greenbergian; Mauritius: Language Situation; Mobilian Jargon;
Morphology in Pidgins and Creoles; New Caledonia: Language Situation; Nigeria: Language Situation; Norway:
Language Situation; Papua New Guinea: Language Situation; Phonological Universals; Principles and Parameters
Framework of Generative Grammar; Russenorsk; Russian
Federation: Language Situation; Schuchardt, Hugo (1842
1927); Serial Verb Constructions; Sierra Leone: Language
Situation; Solomon Islands: Language Situation; St Lucia:
Language Situation; Suriname: Language Situation;
Switch Reference; Tense, Mood, Aspect: Overview; Tok
Pisin; United States of America: Language Situation;
Vanuatu: Language Situation.
Language Maps (Appendix 1): Maps 47, 48.

Bibliography
Baker P & Huber M (2001). Atlantic, Pacific, and worldwide features in English-lexicon contact languages.
English World Wide 22(2), 157208.
Bickerton D (1981). Roots of language. Ann Arbor:
Karoma.
Bickerton D (1984). The language bioprogram hypothesis.
Behavioral and Brain Sciences 7, 173221.
Bickerton D (1988). Creoles languages and the bioprogram. In Newmeyer F J (ed.) Linguistics: the Cambridge
survey 2: Linguistic theory: extensions and implications.
Cambridge: Cambridge University Press. 268284.

Bickerton D & Muysken P (1988). A dialog concerning the

linguistic status of creole languages. In Newmeyer F J
(ed.) Linguistics: the Cambridge survey 2: Linguistic theory: extensions and implications. Cambridge: Cambridge
University Press. 302306.
Chaudenson R (1992). Des les, des hommes, des langues:
essais sur la cre olisation linguistique et culturelle. Paris:
LHarmattan.
Churchill W (1911). Beach-La-Mar, the jargon trade speech
of the Western Pacific. Washington, DC: Carnegie Institution Publication No. 164.
Hall R A Jr. (1966). Pidgin and creole languages. Ithaca,
NY: Cornell University Press.
Holm J (1989). Pidgins and creoles (2 vols). Cambridge:
Cambridge University Press.
Keesing R (1988). Melanesian Pidgin and the Oceanic substrate. Stanford: Stanford University Press.
Lefebvre C (1998). Creole genesis and the acquisition
of grammar: the case of Haitian Creole. Cambridge:
Cambridge University Press.
McWhorter J (1998). Identifying the creole prototype:
vindicating a typological class. Language 74, 788818.
McWhorter J (2001). The worlds simplest grammars are
creole grammars. Linguistic Typology 5(2), 125166.
Mufwene S S (1986). Les langues cre oles peuvent-elles e tre
de finie s sans allusion a` leur histoire? Etudes Cre oles 9,
135150.
Muysken P (1988). Are creoles a special type of language?
In Newmeyer F J (ed.) Linguistics: the Cambridge
survey 2: Linguistic theory: extensions and implications.
Cambridge: Cambridge University Press. 285301.
Roberts S J (2000). Nativization and the genesis of Hawaiian Creole. In McWhorter J (ed.) Language change and
language contact in pidgins and creoles. Amsterdam:
John Benjamins. 257300.
Romaine S (1988). Pidgin and creole languages. London:
Longman.
Romaine S (1992). Language, education and development:
urban and rural Tok Pisin in Papua New Guinea.
Oxford: Oxford University Press.
Singler J V (ed.) (1990). Pidgin and creole tense-moodaspect systems. Amsterdam: John Benjamins.
Thomason S G & Kaufman K (1988). Language contact,
creolization and genetic linguistics. Berkeley: University
of California Press.
Velupillai V (2003). Hawaii Creole English: a typological
analysis of the tense-mood-aspect system. Basingstoke:
Palgrave Macmillan.

Pike, Kenneth Lee (19122000) 607

Pike, Kenneth Lee (19122000)

F Robbins, Dallas, TX, USA
2006 Elsevier Ltd. All rights reserved.

Kenneth Lee Pike was born in East Woodstock,

Connecticut, a country doctors next-to-youngest
child (of eight children). He studied at Gordon
College in Boston, Massachusetts (Th.B. in theology
and missions, 1933), and later at the University of
Michigan under Charles Fries (Ph.D. in linguistics,
1942). He served for some 30 years on the Michigan
faculty (19471977). Pike received the Distinguished
Faculty Achievement award at Michigan (1966).
He held the Charles C. Fries Professorship in Linguistics from l974 to 1977. He chaired the Department
of Linguistics for two years (l9751977) and served
as director of the English Language Institute of the
University of Michigan (19761977). After his retirement from Michigan he moved to the SIL center
in Dallas. Pike was made Adjunct Professor at the
University of Texas at Arlington in 1981.
To understand Pike the linguist, one must understand Pike the man. He considered every aspect of
life related to every other aspect, in ways orchestrated
by God. For Pike, there was no conflict in this perspective. His first commitment was to God, and his
scholarly work was service to God his faith and
learning fully integrated.
Pike spent the summer of 1935 studying under
W. Cameron Townsend at Camp Wycliffe, a small
training institute in linguistics in Sulphur Springs,
Arkansas (subsequently called the Summer Institute
of Linguistics). Committing himself to serve the minority language communities of the world, he went
with Townsend to Mexico and began to study and
learn the Mixtec language of San Miguel el Grande,
toward translating the New Testament. Pikes analysis of Mixtec, particularly the tone, was seminal in
his whole approach to language. In 1935, Pike met
Mr. Townsends niece, Evelyn Griset. In 1938, Pike
and Griset were married. They made a formidable
team in both training and research in linguistics,
while rearing their three children.
Pikes interests and research went far beyond linguistics, as reflected in Headlands listing (in Wise
et al., 2003: 13) by decades of emphases in his
publications:
1940s: sounds of language: phonetics and phonemics, tone and intonation
1950s: anthropology and language in relation to
culture, developing his holistic view
1960s: mathematics

1970s: grammatical analysis

1980s: philosophy.
Convinced of the need for scholarly excellence
in the translation endeavor, Townsend sent Pike to
study with Edward Sapir at the Linguistics Institute of
the Linguistic Society of America (University of
Michigan, summer 1937). Later, during Pikes 30
years on Michigans faculty, he divided his time between the University of Michigan and SIL: directing
the Summer Institute of Linguistics at the University
of Oklahoma, starting other SIL schools, helping
SIL colleagues around the world to analyze the languages they were studying, and encouraging them to
publish the results.
Following his holistic view Pike developed tagmemics, in which the tagmeme is a paradigmatic slotclass unit within a syntagmatic unit. Ken and Evelyn
Pike co-authored Grammatical analysis (1972),
which has been used as a tagmemics textbook in
several countries.
Pike urged SIL colleagues to study at universities
where theories opposed to tagmemics were taught,
and to incorporate insights from all theories into whatever one they adopted. He lectured in 42 countries and
studied well over a hundred indigenous languages, including languages in Australia, Bolivia, Cameroon,
Co te dIvoire, Ecuador, Ghana, India, Indonesia,
Mexico, Nepal, New Guinea, Nigeria, Peru, the
Philippines, and Togo.
Pikes honors include:
Presidential Medal of Merit from the Philippines
Deans Medal at Georgetown University
Nobel Peace Prize nominee, 15 successive years
Templeton Prize nominee, three times
Member of the National Academy of Sciences
President of Summer Institute of Linguistics (now SIL
International) for 37 years
President Emeritus of SIL
President of the Linguistic Society of America (1961)
President of the Linguistic Association of Canada and
the United States (19771978).
Pike received honorary doctorates and professorships from universities around the world, including the University of Chicago, Universite Rene
Descartes, the University of Lima, and AlbertLudwigs-Universita t in Freiburg. Not long before his
death in 2000 a banquet was held in a Dallas hotel
to honor Pike.
The most complete bibliography of Pike is Spanne,
J. & Wise, M. R. (2003). The writings of Kenneth

608 Pike, Kenneth Lee (19122000)

Pike, in Wise et al. (2003: 5781). Over the period of

65 years, Pike authored and edited more than 30
books, over 200 scholarly articles, 90 articles for
popular magazines, 8 poetry collections, Scripture
translations, individual poems, instruction workbooks, and video and audio recordings, all reflecting
his integration of faith and scholarly endeavor.
See also: Fries, Charles Carpenter (18971967); Phoneme;

Sapir, Edward (18841939); Summer Institute of Linguistics; Tagmemics.

Bibliography
Headland T N, Pike K L & Harris M (eds.) (1990). Frontiers of Anthropology 7: Emic and etics: The insider/
outsider debate. Newberry Park: Sage.
Heimbach S (compiler) (1997). Seasons of life: a complete
collection of Kenneth L. Pike Poetry (5 vols). Dallas:
Summer Institute of Linguistics.
Pike K L (1943). University of Michigan publications in languages and literature 21: phonetics, a critical analysis of
phonetic theory and a technic for the practical description
of sounds. Ann Arbor: University of Michigan.

Pima

Pike K L (1967a). Janua Linguarum, series maior 24: Language in relation to a unified theory of the structure of
human behavior (2nd rev. edn.). The Hague: Mouton.
Pike K L (1967b). Stir, change, create. Grand Rapids: Wm.
B. Eerdmans.
Pike K L (1971). Mark my words. Grand Rapids: Wm.
B Eerdmans.
Pike K L (1993). Talk, thought, and thing: the emic road
toward conscious knowledge. Dallas: Summer Institute
of Linguistics.
Pike K L & Pike E G (1977). Grammatical Analysis. Summer Institute of Linguistics and the University of Texas at
Arlington.
Pike K L, Stark D S & Me re cias A (1951). El Nuevo
Testamento de nuestro sen or Jesucristo. Cuernavaca:
Tipografa Indigena.
Spanne J & Wise M R (2003). The writings of Kenneth
Pike. In Wise et al. (eds.). 5781.
Wise M R, Headland T N & Brend R M (eds.) (2003).
Publications in linguistics 139: Language and life: essays
in memory of Kenneth L. Pike. Dallas: SIL International
and The University of Texas at Arlington.
Young R E, Pike K L & Becker A L (1970). Rhetoric:
discovery and change. New York: Harcourt, Brace &
World.

See: Tohono Oodham.

Pitcairn Island: Language Situation

Editorial Team
2006 Elsevier Ltd. All rights reserved.

A group of volcanic islands in the South Pacific

Ocean, the Pitcairn are situated about midway between Peru and New Zealand. Although the islands
show signs of a previous settlement by people presumably of Polynesian extraction, they were uninhabited when they were settled in 1790 by a group of
British mutineers from His Majestys armed ship
Bounty. The men took with them twelve Tahitian
women to be their spouses, and six Tahitian men to
be their helpers. Relations among the islanders were
not always of the best, the Tahitians tried to escape on
a few occasions, but by 1800 all the survivors had
resigned themselves to life on Pitcairn and already
24 children had been born there. Because resources

were thought insufficient to sustain life on Pitcairn in

the long term, the community, made up of a modest
number of islanders despite the occasional arrival of
new settlers, relocated twice to other islands. In 1831
all the inhabitants of Pitcairn moved to Tahiti, but the
comparatively laxer customs they found there and the
deficient immunity to new diseases drove the islanders back to Pitcairn. In 1856, all 194 islanders moved
to Norfolk Island, a previous British penal colony by
then unoccupied, but many were unhappy there and a
good number of families returned to Pitcairn in the
following years. Pitcairn Island has been permanently
settled ever since, although the population has been
fluctuating between high peaks such as the one of
1937 with 233 islanders and low peaks such as the
present one with only approximately 50 inhabitants.
Pitcairn Island is now a British overseas territory and
English is the official language, used in all formal

608 Pike, Kenneth Lee (19122000)

Pike, in Wise et al. (2003: 5781). Over the period of

65 years, Pike authored and edited more than 30
books, over 200 scholarly articles, 90 articles for
popular magazines, 8 poetry collections, Scripture
translations, individual poems, instruction workbooks, and video and audio recordings, all reflecting
his integration of faith and scholarly endeavor.
See also: Fries, Charles Carpenter (18971967); Phoneme;

Sapir, Edward (18841939); Summer Institute of Linguistics; Tagmemics.

Bibliography
Headland T N, Pike K L & Harris M (eds.) (1990). Frontiers of Anthropology 7: Emic and etics: The insider/
outsider debate. Newberry Park: Sage.
Heimbach S (compiler) (1997). Seasons of life: a complete
collection of Kenneth L. Pike Poetry (5 vols). Dallas:
Summer Institute of Linguistics.
Pike K L (1943). University of Michigan publications in languages and literature 21: phonetics, a critical analysis of
phonetic theory and a technic for the practical description
of sounds. Ann Arbor: University of Michigan.

Pima

Pike K L (1967a). Janua Linguarum, series maior 24: Language in relation to a unified theory of the structure of
human behavior (2nd rev. edn.). The Hague: Mouton.
Pike K L (1967b). Stir, change, create. Grand Rapids: Wm.
B. Eerdmans.
Pike K L (1971). Mark my words. Grand Rapids: Wm.
B Eerdmans.
Pike K L (1993). Talk, thought, and thing: the emic road
toward conscious knowledge. Dallas: Summer Institute
of Linguistics.
Pike K L & Pike E G (1977). Grammatical Analysis. Summer Institute of Linguistics and the University of Texas at
Arlington.
Pike K L, Stark D S & Merecias A (1951). El Nuevo
Testamento de nuestro senor Jesucristo. Cuernavaca:
Tipografa Indigena.
Spanne J & Wise M R (2003). The writings of Kenneth
Pike. In Wise et al. (eds.). 5781.
Wise M R, Headland T N & Brend R M (eds.) (2003).
Publications in linguistics 139: Language and life: essays
in memory of Kenneth L. Pike. Dallas: SIL International
and The University of Texas at Arlington.
Young R E, Pike K L & Becker A L (1970). Rhetoric:
discovery and change. New York: Harcourt, Brace &
World.

See: Tohono Oodham.

Pitcairn Island: Language Situation

Editorial Team
2006 Elsevier Ltd. All rights reserved.

A group of volcanic islands in the South Pacific

Ocean, the Pitcairn are situated about midway between Peru and New Zealand. Although the islands
show signs of a previous settlement by people presumably of Polynesian extraction, they were uninhabited when they were settled in 1790 by a group of
British mutineers from His Majestys armed ship
Bounty. The men took with them twelve Tahitian
women to be their spouses, and six Tahitian men to
be their helpers. Relations among the islanders were
not always of the best, the Tahitians tried to escape on
a few occasions, but by 1800 all the survivors had
resigned themselves to life on Pitcairn and already
24 children had been born there. Because resources

were thought insufficient to sustain life on Pitcairn in

the long term, the community, made up of a modest
number of islanders despite the occasional arrival of
new settlers, relocated twice to other islands. In 1831
all the inhabitants of Pitcairn moved to Tahiti, but the
comparatively laxer customs they found there and the
deficient immunity to new diseases drove the islanders back to Pitcairn. In 1856, all 194 islanders moved
to Norfolk Island, a previous British penal colony by
then unoccupied, but many were unhappy there and a
good number of families returned to Pitcairn in the
following years. Pitcairn Island has been permanently
settled ever since, although the population has been
fluctuating between high peaks such as the one of
1937 with 233 islanders and low peaks such as the
present one with only approximately 50 inhabitants.
Pitcairn Island is now a British overseas territory and
English is the official language, used in all formal

Pitjantjatjara/Yankunytjatjara 609

settings, from politics to education. However, Pitcairn

(Pitcairn-Norfolk), which has developed out of the
English/Tahitian pidgin first used by the Bounty mutineers and their wives, is still used by the islanders as

a second language, as it serves to strengthen group

identity.
See also: Norfolk Island: Language Situation.

Pitjantjatjara/Yankunytjatjara
C Goddard, University of New England, Armidale,
Australia
2006 Elsevier Ltd. All rights reserved.

Pitjantjatjara, along with its neighboring dialect

Yankunytjatjara, are part of the Western Desert Language (WDL) a vast dialect continuum located in
the arid and sparsely populated central and western
inland of Australia (see Australian Languages; Australia: Language Situation). The two dialects are commonly referred to jointly as P/Y. There are about 2500
P/Y speakers. Since it is still being acquired by children, P/Y counts as one of the less endangered of
Australian languages.
As a typical Pama-Nyungan language, WDL is agglutinative (chiefly suffixing), with well developed
systems of nominal and verbal inflection. Canonical
constituent order is S (O) (PP) V, but can vary rather
freely. Ellipsis of third person arguments, when the
referent can be understood in context, is common.
Nominal inflection is of the split ergative variety. The
verbal system has eight tense-aspect-mood categories,
with complex allomorphy governed by a system of
four conjugational classes. Serial verb constructions
(see Serial Verb Constructions) of several types,
abound. P/Y has a switch-reference system in several
subordinate clause types, and in coordinate constructions. Aspects of P/Y have been studied by a number
of linguists. There are three major grammars and a
substantial dictionary, a range of pedagogical material, and a variety of specialized linguistic
studies. A wide range of vernacular texts of traditional stories, ethnoscience, and oral histories has
been published locally.

Sociocultural and Historical Aspects

The similarity between the two dialects has been
reinforced by shared historical experiences in the
wake of European intrusion early last century, including long periods of co-residence on mission and
government settlements, and, subsequently, in selfmanaged Aboriginal communities. Most speakers

now co-reside on Aboriginal-owned lands in the

northwest of the state of South Australia and adjacent areas in the Northern Territory. For a long time
Pitjantjatjara was the prestige variety, because it had
been adopted by missionaries at Ernabella for Bible
translation and use in Christian worship, and was
subsequently used in bilingual education programs
in local primary schools. Lately the two dialects
have been moving towards parity of esteem. Many
Australian and international tourists encounter P/Y
when they visit the Uluru National Park in central
Australia.
Though the two dialects share about 80% vocabulary, there are a number of prominent dialect-specific
words, and these form the basis of the traditional
WDL system for referring to speech varieties.
Pitjantja-tjara and Yankunytja-tjara are based on
alternative (nominalized) forms of verbs meaning
come/go (suffix -tjara means having). Northern
and southern varieties of Yankunytjatjara can be
termed Mulatjara and Matutjara, respectively, based
on alternative forms of the adverb true. In earlier
times, there was a multiplicity of such terms in use.
The system was highly relativistic, allowing for crosscutting categorization and for different levels of
inclusiveness, which suited the traditional mobile
and dynamic social economy. These days the terms
Pitjantjatjara and Yankunytjatjara have acquired
more stable and name-like sociopolitical functions.
Traditional P/Y culture is replete with symbolism
(totemism) and religious myth. There are hundreds
of Dreaming stories, songs, and ceremonies. There is
a large body of traditional folktales for children.
Many P/Y speech practices have parallels in the
other languages of Australia. These include the existence of hortatory rhetoric (alpiri), elaborate verbal
indirectness practiced with certain categories of kin
(and total avoidance with others), and prescribed joking relationships characterized by mock
insult and abuse. There is a taboo against using
the names of recently deceased persons in the presence of bereaved relatives. An auxiliary register, i.e., a
special vocabulary (termed anitji), is used during
ceremonial times.

Pitjantjatjara/Yankunytjatjara 609

settings, from politics to education. However, Pitcairn

(Pitcairn-Norfolk), which has developed out of the
English/Tahitian pidgin first used by the Bounty mutineers and their wives, is still used by the islanders as

a second language, as it serves to strengthen group

identity.
See also: Norfolk Island: Language Situation.

Pitjantjatjara/Yankunytjatjara
C Goddard, University of New England, Armidale,
Australia
2006 Elsevier Ltd. All rights reserved.

Pitjantjatjara, along with its neighboring dialect

Yankunytjatjara, are part of the Western Desert Language (WDL) a vast dialect continuum located in
the arid and sparsely populated central and western
inland of Australia (see Australian Languages; Australia: Language Situation). The two dialects are commonly referred to jointly as P/Y. There are about 2500
P/Y speakers. Since it is still being acquired by children, P/Y counts as one of the less endangered of
Australian languages.
As a typical Pama-Nyungan language, WDL is agglutinative (chiefly suffixing), with well developed
systems of nominal and verbal inflection. Canonical
constituent order is S (O) (PP) V, but can vary rather
freely. Ellipsis of third person arguments, when the
referent can be understood in context, is common.
Nominal inflection is of the split ergative variety. The
verbal system has eight tense-aspect-mood categories,
with complex allomorphy governed by a system of
four conjugational classes. Serial verb constructions
(see Serial Verb Constructions) of several types,
abound. P/Y has a switch-reference system in several
subordinate clause types, and in coordinate constructions. Aspects of P/Y have been studied by a number
of linguists. There are three major grammars and a
substantial dictionary, a range of pedagogical material, and a variety of specialized linguistic
studies. A wide range of vernacular texts of traditional stories, ethnoscience, and oral histories has
been published locally.

Sociocultural and Historical Aspects

The similarity between the two dialects has been
reinforced by shared historical experiences in the
wake of European intrusion early last century, including long periods of co-residence on mission and
government settlements, and, subsequently, in selfmanaged Aboriginal communities. Most speakers

now co-reside on Aboriginal-owned lands in the

northwest of the state of South Australia and adjacent areas in the Northern Territory. For a long time
Pitjantjatjara was the prestige variety, because it had
been adopted by missionaries at Ernabella for Bible
translation and use in Christian worship, and was
subsequently used in bilingual education programs
in local primary schools. Lately the two dialects
have been moving towards parity of esteem. Many
Australian and international tourists encounter P/Y
when they visit the Uluru National Park in central
Australia.
Though the two dialects share about 80% vocabulary, there are a number of prominent dialect-specific
words, and these form the basis of the traditional
WDL system for referring to speech varieties.
Pitjantja-tjara and Yankunytja-tjara are based on
alternative (nominalized) forms of verbs meaning
come/go (suffix -tjara means having). Northern
and southern varieties of Yankunytjatjara can be
termed Mulatjara and Matutjara, respectively, based
on alternative forms of the adverb true. In earlier
times, there was a multiplicity of such terms in use.
The system was highly relativistic, allowing for crosscutting categorization and for different levels of
inclusiveness, which suited the traditional mobile
and dynamic social economy. These days the terms
Pitjantjatjara and Yankunytjatjara have acquired
more stable and name-like sociopolitical functions.
Traditional P/Y culture is replete with symbolism
(totemism) and religious myth. There are hundreds
of Dreaming stories, songs, and ceremonies. There is
a large body of traditional folktales for children.
Many P/Y speech practices have parallels in the
other languages of Australia. These include the existence of hortatory rhetoric (alpiri), elaborate verbal
indirectness practiced with certain categories of kin
(and total avoidance with others), and prescribed joking relationships characterized by mock
insult and abuse. There is a taboo against using
the names of recently deceased persons in the presence of bereaved relatives. An auxiliary register, i.e., a
special vocabulary (termed anitji), is used during
ceremonial times.

610 Pitjantjatjara/Yankunytjatjara

Structure

Nominal Morphology

Phonology

The case system includes nominative, ergative, accusative, genitive/purposive, locative, allative, ablative,
and perlative cases. Typically a case-marker is applied
only to the final word of an NP. Since modifiers
generally follow their heads, a typical multi-word
NP looks like: wati pulka kutjara-ku [man big twoPURP] for two big men. Like most other Pama-Nyungan languages, there is a split marking system for the
core cases. For both nouns and pronouns, the nominative case is unmarked. With nouns, accusative
case goes unmarked but there is a marked ergative
form (with -ngku/-lu or a variant). With pronouns,
the ergative goes unmarked but there is a marked
accusative (with -nya). Split case-marking is sometimes described in terms of two distinct case systems:
nominative-accusative for pronouns and ergativeabsolutive for nouns. Aside from being less economical, such an analysis has difficulty with various
complex NP constructions involving both nouns and
pronouns. For example, inalienable possession constructions can bring body-parts and pronouns into a
single NP, and inclusive constructions can bring
names and pronouns into a single NP. For example,
to say that someone hit me on the head, one uses the
NP ngayu-nya kata [1SG-ACC head:ACC] me head.
To say that Kunmanara and someone else did something to someone, one uses the NP Kunmanara-lu
pula [name-ERG 3DL:ERG].
Ergative and locative case allomorphy depends on
whether the word to be marked is vowel- or consonant-final, and on whether the NP is an ordinary
noun-phrase, on the one hand, or a pronoun or proper noun, on the other. Ergative is -ngku (common) or
-lu (proper) with vowel-final words, and otherwise
-Tu (where T is a homorganic stop). Locative is -ngka
(common) or -la (proper) with vowel-final words, and
otherwise -Ta. Genitive/purposive case is marked
with -ku (nouns) or -mpa (pronouns, except for 1SG
ngayu-ku). Locative also expresses instrumental and
comitative functions; e.g., punu-ngka [stick-LOC]
with a stick, untal-ta [daughter-LOC] with (my)
daughter.
Pronouns distinguish singular, dual, and plural
numbers (see Table 2). Most WDL dialects also

P/Y has 17 consonant phonemes: see Table 1. There

are five places of articulation, each with a stop and a
nasal. There are two series of apicals, i.e., consonants
pronounced with the tongue tip as active articulator: alveolar and post alveolar (retroflex). There is a
single laminal series, with the tongue blade as active
articulator. There are three vowels (a, i, u), each with
a length distinction, though long vowels are not
common and are confined to initial syllables.
P/Y phonotactics stipulate that a word must have
at least two vowels, with long vowels counting as
two for this purpose. Several morphophonemic rules
refer to whether a stem has an odd or even number
of vowels, making it convenient to work in terms of
morae, with long vowels counting as two morae.
Words usually start with a single consonant and
never with more than one. Inside a word, CC clusters
occur subject to strict limitations. Most common are
homorganic nasal/lateral stop sequences. Only a
very limited set of consonants (n, ny, n, ly, r) is permitted word-finally, and then only in the Yankunytjatjara
variety. In Pitjantjatjara, consonant-final words are
blocked by addition of the syllable-pa.

Morphology and Syntax

A number of these features are illustrated in the text
extract at the end of this section.

Table 1 Pitjantjatjara/Yankunytjatjara consonant phonemes, in

standard orthography (Goddard, 1985: 11)
Apical

Stops
Nasals
Laterals
Tap
Glides

Laminal

Alveolar

Postalveolar

Dental

Bilabial

Dorsal

t
n
l
r

t
n
l

tj
ny
ly

p
m

k
ng

Table 2 Pitjantjatjara/Yankunytjatjara subject free pronouns (Goddard, 1996: xi)

Subject

Singular (sg)

First person
Second person
Third person

ngayu(lu)a
nyuntu
palu(ru)

Dual (du)

I
you
he, she, it

ngali
nyupali
pula

The syllables in parentheses are dropped when case suffixes are added.

Plural (pl)

we two
you two
they two

nganana
nyura
tjana

we
you
they

Pitjantjatjara/Yankunytjatjara 611

have enclitic or bound pronouns that can be used

instead of or in addition to free pronouns. They
appear attached to the first phrase of a sentence,
conjunctions counting as phrases for this purpose.
P/Y has the following defective set nominative/
ergative: -na 1SG, -n 2SG, -li 1DU, -la 1PL, -ya 3PL;
accusative: -ni/-tja 1SG, -nta -2SG, -linya 1DU, -lanya
1PL. Bound pronouns are not obligatory in P/Y,
though they are common. There are four demonstrative stems: nyanga this, pala that, nyara that over
there, and the anaphoric demonstrative panya that
one, you know which.
Verbs

All WDL dialects share a similar system of tenseaspect-mood categories and four conjugational classes, though the details differ from dialect to dialect.
The P/Y categories are: present, past, past imperfective, future, imperative, imperative imperfective, and
characteristic. In addition, there are serial and nominalized verb forms. Each verbal category is manifested
by up to four different allomorphs (e.g., imperative:
-, -la, -wa, -ra), depending on the conjugational
class. The P/Y system is economically analyzed
in terms of three stem types: a simple stem which
functions as a base for perfective categories, an augmented stem for imperfective categories, and an additional augmented stem for the aspect-neutral
forms: see Table 3. The augmented forms were
probably inflected words in an earlier stage of the
language, with the present-day forms resulting from
double-marking.
The -class and l-class are open, with predominantly intransitive and transitive memberships, respectively. The ng-class and n-class are likewise
predominantly intransitive and transitive respectively, but they have only a handful of basic roots each.
These roots, furthermore, are the only monosyllabic
verb roots in the language: n-class: ya- go, tju- put,
ma- get; ng-class: pu- hit, nya- see and yugive (examples from Yankunytjatjara). The overall

membership of the ng-class and n-class is very large,

however, because numerous verbs are formed by
compounding with the basic roots or via derivational
affixation. Derivational processes are sensitive to
mora parity, as well as to the transitivity preference
of the verb class. For example, the main intransitive
verbaliser is suffix -ri/-ari. The derived stem belongs
to the ng-class if it has an even number of morae, and
to -class if it has an odd number of morae.
Complex Sentences

A single clause may contain more than one verb, if the

subsidiary verbs are suffixed with the serial ending. It
is common in narratives for clauses to contain several
serial verbs, as well as the main finite verb. The
grammar of serial verbs and their associated NPs and
modifiers is quite complex. Typically for WDL, subordinate clauses are formed by adding case suffixes to a
nominalized clause. For example, a purposive clause
is formed with suffix -ku (identical with purposive
case), e.g., kungka-ngku mai pau-ntja-ku [womanERG food bake-NOML-PURP] so the woman could
cook food. Inside the subordinate clause, the subject,
object, and any other NPs occur with the same casemarking as they would have in a simple clause. The
circumstantial clause is formed in Yankunytjatjara
with suffix -la (one of the locative suffixes), e.g.,
kungka-ngku mai pau-ntja-la [woman-ERG food
bake-NOML-LOC] while/because the woman cooked
the food. The Pitjantjatjara circumstantial is nyangka, which has likely descended from an earlier
*-nytja-ngka (simplification of the first of two
nasal-stop clusters is common in WDL phonology).
Another subordinate type is the aversive clause,
which identifies an outcome to be avoided or
prevented.
P/Y purposive and circumstantial clauses comply
with a switch reference constraint, i.e., they can
only be used if the subordinate clause subject refers
to a different individual to the main clause subject.
If the subjects are the same, a different subordinate

Table 3 Pitjantjatjara/Yankunytjatjara verbs (Goddard, 1985: 90)

Imperative
Past (perfective)
Imperative (imperfective)
Present (imperfective)
Past (imperfective)
Future
Characteristic
Serial form
Nominalized form

()
talk

(l )
bite

(ng)
hit

(n)
put

wangka
wangkangu
wangkama
wangkanyi
wangkangi
wangkaku
wangkapai
wangkara
wangkanytja

patjala
patjanu
patjanma
patjani
patjaningi
patjalku
patjalpai
patjara
patjantja

puwa
pungu
pungama
punganyi
pungangi
pungkuku
pungkupai
pungkula
pungkunytja

tjura
tjunu
tjunama
tjunanyi
tjunangi
tjunkuku
tjunkupai
tjunkula
tjunkunytja

612 Pitjantjatjara/Yankunytjatjara

structure is used in place of the purposive, with the

intentive suffix -kitja. An interesting feature of
the intentive construction is that the clause as a
whole takes an ergative suffix (-ngku) if the verb of
the main clause is transitive, e.g., mai pau-ntji-kitja
(-ngku) [food bake-NOML-INTENT-(ERG)] (wanting) to
cook food. Actor agreement of this kind is also
found with adverbs of manner and emotion (better
regarded as active adjectives), and with frequency
expressions.
There are three coordinating conjunctions: ka and,
but, munu and, and palu but, even though. Unusually for Australian languages, switch-reference operates for coordination. Normally, ka can only be used
as a conjunction if the subject of the new clause
refers to a different individual to the subject of the
preceding clause; otherwise, munu is used. A range
of free and clitic particles express illocutionary and
discourse-related meanings.
Pitjantjatjara Text Extract. From Wati Tjangarangku Iti Intiritjunanyi Theres an Ogre Pinching
the Baby!, told by Anmanari Alice. Revised edition
published by NW Resource Centre, Ernabella.
Ka-l
CONTR-QUOT

minyma-ngku panya
woman-ERG
THAT.ONE

pata-ra
wait-SERIAL

watja-nu.
tell-PAST
Then the waiting woman told him.
Panya tjangara-na pungku-la wanti-kati-ngu.
that.one ogre-1SG:ERG hit-SERIAL leave-PROCESS-PAST
I killed that ogre and got away.
Nyangatja-na puli-ngka nyina-nyi,
this-1SG:NOM
hill-LOC
sit-PRES,
nyuntu-mpa pata-ra.
2SG:NOM-PURP wait-SERIAL
Ive been sitting here on the hill, -waiting for you
(to get back).
Munu
ADD

Ka

wangka-ngu, Palya
nyangatja-n
say-PAST
good
here-2SG:ERG
pu-ngu. Munu-li-nku a-ra-lta.
hit-PAST ADD-1DU-REFL go-IMP-and.then
He replied, You did well to kill it here. Lets get
out of here.
CONTR

Munu

pula

ADD

3DU:NOM

ma-pitja-ngu
away-go-PAST

ngura
place

kutjupa-kutu.
other-ALL
And so away they went to some other place.

See also: Australia: Language Situation; Australian Languages; Ergativity; Serial Verb Constructions; Switch
Reference.

Bibliography
Bowe H (1990). Categories, constituents, and constituent order in Pitjantjatjara, an Aboriginal language of
Australia. London: Routledge.
Douglas W H (1958/1964). An introduction to the Western
Desert Language. Sydney: Oceania Linguistic Monographs No 4 [Revised].
Eckert P & Hudson J (1988). Wangka wiru: a handbook for
the Pitjantjatjara language learner. Underdale: South
Australian College of Advanced Education.
Goddard C (1986). Yankunytjatjara grammar. Alice
Springs: Institute for Aboriginal Development.
Goddard C (1990). Emergent genres of reportage and
advocacy in the Pitjantjatjara print media. Australian
Aboriginal Studies 2, 2747.
Goddard C (1992). Traditional Yankunytjatjara ways of
speaking a semantic perspective. Australian Journal of
Linguistics 12(1), 93122.
Goddard C (1996). Pitjantjatjara/Yankunytjatjara to English dictionary. [Revised 2nd edn.]. Alice Springs:
Institute for Aboriginal Development.
Goddard C & Kalotas A (eds.) (1985). Punu: Yankunytjatjara plant use. Sydney: Angus and Robertson. [Reprinted, IAD Press, Alice Springs, 1995, 2002.].
Klapproth D M (2004). Narrative as social practice:
Anglo-Western and Australian Aboriginal oral traditions.
Berlin: Mouton de Gruyter.
Langlois A (2004). Alive and Kicking: Areyonga Teenage
Pitjantjatjara. Canberra: Pacific Linguistics.
Rose D (2001). The Western Desert code: an Australian
cryptogrammar. Canberra: Pacific Linguistics.

Pitman, Isaac, Sir (18131897)

M K C MacMahon, University of Glasgow,
Glasgow, UK
2006 Elsevier Ltd. All rights reserved.

Despite his eminence as a Victorian businessman, the

driving force in Pitmans life was education. His
shorthand system arose from the need to equip

young people with a valuable practical skill. In addition, he recognized that the irregularities and vagaries
of English orthography could be a handicap to the
young, and this led to his devising various reformed
spelling systems for English.
He was born in Trowbridge, Wiltshire (England)
on January 4, 1813. At the age of 19 he began work as
a school-teacher, and his Stenographic sound-hand, a

612 Pitjantjatjara/Yankunytjatjara

structure is used in place of the purposive, with the

intentive suffix -kitja. An interesting feature of
the intentive construction is that the clause as a
whole takes an ergative suffix (-ngku) if the verb of
the main clause is transitive, e.g., mai pau-ntji-kitja
(-ngku) [food bake-NOML-INTENT-(ERG)] (wanting) to
cook food. Actor agreement of this kind is also
found with adverbs of manner and emotion (better
regarded as active adjectives), and with frequency
expressions.
There are three coordinating conjunctions: ka and,
but, munu and, and palu but, even though. Unusually for Australian languages, switch-reference operates for coordination. Normally, ka can only be used
as a conjunction if the subject of the new clause
refers to a different individual to the subject of the
preceding clause; otherwise, munu is used. A range
of free and clitic particles express illocutionary and
discourse-related meanings.
Pitjantjatjara Text Extract. From Wati Tjangarangku Iti Intiritjunanyi Theres an Ogre Pinching
the Baby!, told by Anmanari Alice. Revised edition
published by NW Resource Centre, Ernabella.
Ka-l
CONTR-QUOT

minyma-ngku panya
woman-ERG
THAT.ONE

pata-ra
wait-SERIAL

watja-nu.
tell-PAST
Then the waiting woman told him.
Panya tjangara-na pungku-la wanti-kati-ngu.
that.one ogre-1SG:ERG hit-SERIAL leave-PROCESS-PAST
I killed that ogre and got away.
Nyangatja-na puli-ngka nyina-nyi,
this-1SG:NOM
hill-LOC
sit-PRES,
nyuntu-mpa pata-ra.
2SG:NOM-PURP wait-SERIAL
Ive been sitting here on the hill, -waiting for you
(to get back).
Munu
ADD

Ka

wangka-ngu, Palya
nyangatja-n
say-PAST
good
here-2SG:ERG
pu-ngu. Munu-li-nku a-ra-lta.
hit-PAST ADD-1DU-REFL go-IMP-and.then
He replied, You did well to kill it here. Lets get
out of here.
CONTR

Munu

pula

ADD

3DU:NOM

ma-pitja-ngu
away-go-PAST

ngura
place

kutjupa-kutu.
other-ALL
And so away they went to some other place.

See also: Australia: Language Situation; Australian Languages; Ergativity; Serial Verb Constructions; Switch
Reference.

Bibliography
Bowe H (1990). Categories, constituents, and constituent order in Pitjantjatjara, an Aboriginal language of
Australia. London: Routledge.
Douglas W H (1958/1964). An introduction to the Western
Desert Language. Sydney: Oceania Linguistic Monographs No 4 [Revised].
Eckert P & Hudson J (1988). Wangka wiru: a handbook for
the Pitjantjatjara language learner. Underdale: South
Australian College of Advanced Education.
Goddard C (1986). Yankunytjatjara grammar. Alice
Springs: Institute for Aboriginal Development.
Goddard C (1990). Emergent genres of reportage and
advocacy in the Pitjantjatjara print media. Australian
Aboriginal Studies 2, 2747.
Goddard C (1992). Traditional Yankunytjatjara ways of
speaking a semantic perspective. Australian Journal of
Linguistics 12(1), 93122.
Goddard C (1996). Pitjantjatjara/Yankunytjatjara to English dictionary. [Revised 2nd edn.]. Alice Springs:
Institute for Aboriginal Development.
Goddard C & Kalotas A (eds.) (1985). Punu: Yankunytjatjara plant use. Sydney: Angus and Robertson. [Reprinted, IAD Press, Alice Springs, 1995, 2002.].
Klapproth D M (2004). Narrative as social practice:
Anglo-Western and Australian Aboriginal oral traditions.
Berlin: Mouton de Gruyter.
Langlois A (2004). Alive and Kicking: Areyonga Teenage
Pitjantjatjara. Canberra: Pacific Linguistics.
Rose D (2001). The Western Desert code: an Australian
cryptogrammar. Canberra: Pacific Linguistics.

Pitman, Isaac, Sir (18131897)

M K C MacMahon, University of Glasgow,
Glasgow, UK
2006 Elsevier Ltd. All rights reserved.

Despite his eminence as a Victorian businessman, the

driving force in Pitmans life was education. His
shorthand system arose from the need to equip

young people with a valuable practical skill. In addition, he recognized that the irregularities and vagaries
of English orthography could be a handicap to the
young, and this led to his devising various reformed
spelling systems for English.
He was born in Trowbridge, Wiltshire (England)
on January 4, 1813. At the age of 19 he began work as
a school-teacher, and his Stenographic sound-hand, a

Place Names 613

phonetically based shorthand, appeared in 1837.

A second edition, entitled Phonography, followed in
1840. Its success, especially in educational circles,
was almost instantaneous. The commercial world
soon turned to it too. In 1843, Pitman was compelled
to give up his teaching post in order to concentrate on
the business side of shorthand and spelling reform,
and in 1847 he set up the Phonetic Institute, the administrative headquarters for his publishing business in
Bath.
His desire that there should be a more logical and
consistent reformed spelling system for English led, in
1844, to Phonotypy, the first of several reformed
alphabets, some of which he devised in conjunction
with Alexander John Ellis (see Ellis, Alexander John
(ne Sharpe) (18141890)).
Pitman was knighted in 1894, and died in Bath on
January 22, 1897.
Although not a phonetician in the modern sense
of the word, Pitman based both his shorthand
and his reformed spellings on a conscious awareness

of the phonemic contrasts of English. His support

for phonetic and applied phonetic endeavors in
Victorian Britain helped to create a climate of opinion
about linguistic matters in which the work of more
intellectually distinguished figures could prosper.

See also: Ellis, Alexander John (ne Sharpe) (18141890);

Phonetic Transcription: History; Phonology: Overview;
Spelling Reform.

Bibliography
Abercrombie D (1937). Isaac Pitman: A pioneer in the
scientific study of language. London: Sir Isaac Pitman
and Sons Ltd. [Repr. in Abercrombie D (ed.) Studies in
phonetics and linguistics. London: Oxford University
Press].
Triggs T D (2004). Pitman, Sir Isaac. In Oxford Dictionary of National Biography. Oxford: Oxford University
Press.

Place Names
C Hough, University of Glasgow, Glasgow, UK
2006 Elsevier Ltd. All rights reserved.

Place names occupy an unusual position within linguistics in that they were often coined in languages or
forms of languages that predate those in contemporary use. Although many originate as literal descriptions of the places concerned, they come to be used as
lexically meaningless labels and as such are transferred easily from one group of speakers to another.
This means that place names frequently survive the
transition between languages when territories are
taken over by new settlers, preserving evidence for
successive stages of population movement and linguistic history. River names have the highest survival rate,
followed by the names of other major topographical
features such as hills and mountains. Settlement
names are generally younger, but may still be well
over 1000 years old, with minor names such as field
and street names being among the most recent.

Population Movement
Hydronyms

Because of the high antiquity of river names, they

represent primary evidence for population movement

in prehistoric times. Parallels between ancient river

names in Britain and continental Europe as for
instance the Aar in Belgium, the Ahr in Germany,
the Ara in Spain, and the Ayr in Scotland, or the
Isar in Germany, the Ise`re in France, the Iserna in
Switzerland, and the Ure in England point to
an origin in a common language, sometimes referred
to as Old European. This in turn suggests that the
earliest recorded immigrants to the British Isles the
Celts, whose arrival is estimated to a few centuries B.C.
were preceded by immigrants from the continent
who brought with them an established system of
hydronymy. The linguistic material has been subjected to detailed analysis in attempts to establish
whether these immigrants were speakers of an IndoEuropean or non-Indo-European language. Proponents of the latter theory have put forward a range
of suggestions, including vigorous arguments in favor
of Basque (Vennemann, 1994). However, conservative scholarship continues to prefer an IndoEuropean origin; and the high concentration of Old
European river names in an area of Europe bounded
north and south by the Baltic and the Alps, and east
and west by the Don and the Rhine, supports a
hypothesis that this may have been the original
homeland of the Indo-European speakers (Kitson,
1996).

Place Names 613

phonetically based shorthand, appeared in 1837.

A second edition, entitled Phonography, followed in
1840. Its success, especially in educational circles,
was almost instantaneous. The commercial world
soon turned to it too. In 1843, Pitman was compelled
to give up his teaching post in order to concentrate on
the business side of shorthand and spelling reform,
and in 1847 he set up the Phonetic Institute, the administrative headquarters for his publishing business in
Bath.
His desire that there should be a more logical and
consistent reformed spelling system for English led, in
1844, to Phonotypy, the first of several reformed
alphabets, some of which he devised in conjunction
with Alexander John Ellis (see Ellis, Alexander John
(ne Sharpe) (18141890)).
Pitman was knighted in 1894, and died in Bath on
January 22, 1897.
Although not a phonetician in the modern sense
of the word, Pitman based both his shorthand
and his reformed spellings on a conscious awareness

of the phonemic contrasts of English. His support

for phonetic and applied phonetic endeavors in
Victorian Britain helped to create a climate of opinion
about linguistic matters in which the work of more
intellectually distinguished figures could prosper.

See also: Ellis, Alexander John (ne Sharpe) (18141890);

Phonetic Transcription: History; Phonology: Overview;
Spelling Reform.

Bibliography
Abercrombie D (1937). Isaac Pitman: A pioneer in the
scientific study of language. London: Sir Isaac Pitman
and Sons Ltd. [Repr. in Abercrombie D (ed.) Studies in
phonetics and linguistics. London: Oxford University
Press].
Triggs T D (2004). Pitman, Sir Isaac. In Oxford Dictionary of National Biography. Oxford: Oxford University
Press.

Place Names
C Hough, University of Glasgow, Glasgow, UK
2006 Elsevier Ltd. All rights reserved.

Place names occupy an unusual position within linguistics in that they were often coined in languages or
forms of languages that predate those in contemporary use. Although many originate as literal descriptions of the places concerned, they come to be used as
lexically meaningless labels and as such are transferred easily from one group of speakers to another.
This means that place names frequently survive the
transition between languages when territories are
taken over by new settlers, preserving evidence for
successive stages of population movement and linguistic history. River names have the highest survival rate,
followed by the names of other major topographical
features such as hills and mountains. Settlement
names are generally younger, but may still be well
over 1000 years old, with minor names such as field
and street names being among the most recent.

Population Movement
Hydronyms

Because of the high antiquity of river names, they

represent primary evidence for population movement

in prehistoric times. Parallels between ancient river

names in Britain and continental Europe as for
instance the Aar in Belgium, the Ahr in Germany,
the Ara in Spain, and the Ayr in Scotland, or the
Isar in Germany, the Ise`re in France, the Iserna in
Switzerland, and the Ure in England point to
an origin in a common language, sometimes referred
to as Old European. This in turn suggests that the
earliest recorded immigrants to the British Isles the
Celts, whose arrival is estimated to a few centuries B.C.
were preceded by immigrants from the continent
who brought with them an established system of
hydronymy. The linguistic material has been subjected to detailed analysis in attempts to establish
whether these immigrants were speakers of an IndoEuropean or non-Indo-European language. Proponents of the latter theory have put forward a range
of suggestions, including vigorous arguments in favor
of Basque (Vennemann, 1994). However, conservative scholarship continues to prefer an IndoEuropean origin; and the high concentration of Old
European river names in an area of Europe bounded
north and south by the Baltic and the Alps, and east
and west by the Don and the Rhine, supports a
hypothesis that this may have been the original
homeland of the Indo-European speakers (Kitson,
1996).

614 Place Names

Settlement Names

Place-Name Structures Whereas ancient river names

are often based on a single root with a derivational
suffix, it is more common for European settlement
names to be compound in structure, with one element (the generic) defining the type of place and
the other (the qualifier) describing an aspect of it.
The generic usually comes second in place names
from the Germanic and early Celtic languages, but
first in Celtic place names from about the 6th century
A.D.; in both types, the main stress falls on the qualifier. Thus, the P-Celtic generic aber river mouth gives
place names such as %Aber"deen mouth of the (River)
Don, while its Old English equivalent mua gives
place names such as "Wey%mouth mouth of the
(River) Wey.
Most generics are either habitational, referring
to a building or group of buildings, or topographical, referring to a landscape feature, although a small
proportion are folk names, transferring the name of
a tribe to the area where they lived. Many Germanic
generics are habitational, as with OE tun farmstead,
village and the equivalent ON by in place names
such as Easton east farmstead, and Kirkby village
with a church. Almost all Celtic place names, however, are topographical, as with Perth from Pictish
*pert wood, copse, and Strathclyde valley of the
(River) Clyde from Gaelic srath valley. The topographical nature of the names used to be taken to
indicate that they were originally the names of natural features transferred to later settlements and that
no genuine settlement names had survived. It is now
realized that the Celtic peoples characteristically defined their settlements in terms of nearby topographical features, so that these are indeed original
settlement names.
Place names introduced by European settlers in
other parts of the world are sometimes commemorative, as with Carolina in America, named by
French colonists in honor of Charles IX, and
(Queen) Adelaide, (Charles) Darwin, and (Viscount)
Melbourne in Australia. Many represent transfers of
names from the mother country, as with Plymouth
in America, recalling the arrival of the Pilgrim Fathers
from Plymouth in England, and Perth in Australia,
named for Perth in Scotland. By this means,
many European place names have been replicated in
America, Canada, South Africa, and Australasia.
The indigenous names that they replaced or overlaid are often based on entirely different principles.
The aboriginal place names of Australia are not
primarily descriptive, but relate to the mythology of
the Dreaming, as with Coonowrin, one of the Glasshouse Mountains, literally meaning crooked neck in

allusion to an injury inflicted on the eponymous character by his father Tibrogargan for failing to take care
of his mother Beerwah both of them also names of
mountains within the group. In New Zealand, Maori
place names form sequences preserving the memory
of important events in cultural history, such as the
journeys of early explorers, including Kupe, Paikea,
Tamatea and Turi; in South Africa, place names from
the indigenous Khoekhoen and African languages
characteristically use locative prefixes or suffixes.
Settlement Patterns The linguistic origins of place
names reflect the languages of the peoples who coined
them and so can be used to identify areas settled by
different groups of speakers. Areas of Finnish and
Swedish settlement in Finland are differentiated by
place names from the two languages, while the
nationalities of early European colonists in different
parts of North America are revealed by place names
from English, French, Russian, Spanish, and Dutch
the last, for instance, mostly in the vicinity of New
York, which, as New Amsterdam, was part of the
Dutch colony of New Netherland in the 17th century.
In the British Isles, much research has focused on the
geographical distribution of generics from different
languages. The area of historical Pictland in what is
now northeast Scotland is defined by over 300 place
names beginning in Pit- from Pictish *pett piece of
land. These are in complementary distribution to
place names from Cumbric cair fort, stockaded
farm, so the two generics are taken to demarcate
the regions inhabited by the Pictish P-Celts and the
Cumbric P-Celts, respectively. Areas of Danish
and Norwegian settlement in mainland Britain are
defined largely through place names ending in -by,
-thorp, and other reflexes of generics characteristic
of Scandinavian toponymy; conversely, the remarkable dearth of such names within the historical county
of Rutland suggests that this territory was excluded
from the areas allocated to the Danes at the division
of the Mercian kingdom in 877. The comparison of
cognate generics, such as Scandinavian heimr territory in Norway, southern Sweden, and Jutland and
OE ham homestead in England and Scotland, has
been utilized to trace the movement of Germanic
tribes during the Migration Period.
Place name qualifiers are also important in this
respect. Those comprising tribal names generally
refer to minority groups whose presence was sufficiently distinctive to contrast with the main population. Examples from England include Cummersdale
(Cumbrian Britons), Denby (Danes), Englefield
(Anglians), Irby (Irishmen), Friston (Frisians),
Normanton (Norwegians), Saxham (Saxons), Scotby

Place Names 615

(Scots), Swaffham (Swabians), and Walton (Britons).

As with other types of source material, the evidence
is often limited and must be handled with care.
Place names such as Irby and Ireby may refer not to
Irishmen but to Norwegians who arrived in Britain via
Ireland, and although formations such as Friston and
Frizington point to a Frisian presence in Anglo-Saxon
England, they cannot be dated closely enough to
establish Frisian involvement in the early settlements.
Settlement Chronology Because some terms were in
use as place-name forming elements earlier than
others or went out of use from an earlier date
they can help establish a chronology of settlement, as
for the Anglo-Saxon settlements of the mid-5th century onward. The earliest habitational terms used by
the Anglo-Saxon settlers in what is now southeast
England have been identified as OE ha m homestead,
OE -ingaha m homestead of the followers of or settlers at, and OE -ingas people of, yielding such
place names as Egham, Gillingham, and Hastings.
Other indicators of early Anglo-Saxon settlement
are place names containing Latin loan words, such
as Wickham from OE *wc-ham homestead associated with a Romano-British settlement (Latin
vicus) and Cheshunt from OE *funta spring characterized by Roman building work (Latin fons, fontis);
and those referring to pagan temples or gods, such as
Harrow from OE hearg temple and Thundersley
from the god Thunor. Somewhat later are place
names indicative of established farming communities,
such as Berwick (barley farm), Cheswick (cheese
farm), and Gatwick (goat farm) from OE wc
specialized farm, as well as those containing qualifiers datable to the later Anglo-Saxon period, such as
Scandinavian loan words and late personal names.
Comparative evidence has been used to trace the
course of the advance northward into what is now
southern Scotland and is usefully summarized by
Nicolaisen (2001: 88108). Again, however, this evidence requires careful treatment. For instance, the
distribution of place names referring to AngloSaxon paganism is virtually confined to southern
England, so although their absence from Scotland
has traditionally been taken to indicate that the
Scottish settlements postdated the conversion to
Christianity (c. 627), comparison with the situation
in northern England suggests that this may not
be axiomatic. As regards the Scandinavian settlements, the rarity of the ON generic setr dwelling in
Icelandic place names may suggest that it had gone
out of use before the colonization of Iceland (c. 870),
pointing to an earlier date for the setr-names of the
Northern Isles of Scotland; alternatively, the different

naming patterns may result from differences in local

conditions.
Analysis of place-name structures may also
throw light on chronology. The P-Celtic generic tref
homestead is found in two main areas of Scotland,
occurring as a first element in the south and southwest, where it forms names such as Traprain (with
Cumbric pren tree), but as a second element in the
east, where it forms names such as Capledrae (with
Gaelic capull horse). These may reflect two distinct
phases of settlement, respectively post- and predating the reversal of Celtic element order during
the 6th century.
Microtoponymy

Whereas the names of major settlements identify the

dominant language groups within an area, minor
names present a more detailed picture of the linguistic
makeup of individual communities. The proportion
of names from different languages reflects, at least to
some extent, the mix of speakers of these languages,
throwing additional light on population studies. The
value of this approach is illustrated by recent work on
the density of Danish settlement in eastern England
(Cameron, 1996b) and on German/Slavic contact in
northern Germany (Debus and Schmitz, 2001). Although it has been argued that the imposition of
Scandinavian settlement names within the English
Danelaw might reflect the influence of an elite minority, Camerons analysis of the range and extent of
Scandinavian vocabulary and personal names in the
field names of northeast Lincolnshire supports a very
heavy Danish presence within the area; while Debus
and Schmitzs investigation of minor names including
hybrids from the Slavic and German languages
demonstrates the importance of this type of material
as evidence for contact between linguistic groups.

Language Contact
Types of Communication

For a place name to survive for centuries or even

millennia means that there has been an unbroken
chain of communication, often involving speakers of
different languages. However, names can be taken
over in different ways, reflecting different types of
language contact and varying degrees of mutual intelligibility. At one extreme, the place-name material
may demonstrate knowledge not only of vocabulary
but also of morphology. Dover in England is an
Anglo-Saxon adaptation of the British name Dubris
waters. The Old English form dofras (c. 700) is
also a plural, suggesting that the Anglo-Saxons were

616 Place Names

sufficiently familiar with the British language to recognize and to translate a plural form. There are many
counter-examples, however, where the development
of folk etymologies testifies to a lack of understanding. An instance is York, from the Romano-British
name Eburaco yew-tree estate, where later spellings
such as eoforwic (c. 1060) reflect confusion with an
Old English compound boar farm.
Phonological Adaptation In some instances, an
existing name is preserved intact, with minor adjustments to conform to the phonetic structures of the
host language. Some names in areas of Scandinavian
settlement in Britain have been Scandinavianized
by the substitution of /k/ for /ts/, as in Keswick <
Cheswick cheese farm, or /sk/ for /s/ as in Skipton
< Shipton sheep farm. Contact with English has
resulted in the anglicization of names from the Celtic
languages of the British Isles and of many names
in other parts of the world. Irish baile townland
is anglicized to bally in Irish place names such as
Ballymena and Ballymoney; Gaelic beinn mountain
is anglicized to ben in Irish and Scottish mountain names such as Ben Gorm and Ben Nevis; and
Dutch hoek corner is anglicized to hook in place
names such as Sandy Hook in New Jersey and Hook
of Holland on the southwest Netherlands coast. New
Zealand is an anglicized form of an earlier Dutch
name Nieuw Zeeland new sea land (replacing the
original name Staaten Landt land of the States given
by the Dutch explorer Abel Tasman in 1642), and
Canberra in Australia is probably an anglicized form
of an aboriginal name Nganbirra meeting place.
Anglicized forms of native American names include
Chicago, Connecticut, Kitty Hawk, Manhattan,
Michigan, and Tennessee. Mutual influence can
also occur. Contact between German and Slavic
has resulted both in the Germanization of some Slavic
names, as with Leipzig (Slavic Lipsk < lipa lime tree)
and Dresden (Slavic Drezdzany < drenzga forest),
and in the Slavicization of some German names, as
with Brno in Moravia (German Bru nn < ?Brunnen
spring).
Hybrid Names Hybrid names, which contain elements from more than one language, are particularly
important as evidence of language contact, and in
some cases of bilingualism. In northeast Scotland,
extant place names from the Pictish language
almost always comprise hybrids in combination
with a Gaelic qualifier, testifying to a high degree of
interaction between speakers of the two languages. It
is uncertain, however, whether these represent place
names coined by Gaelic speakers using Pictish loan
words, or Pictish place names taken over and adapted

by the Gaels. It is not uncommon for an existing name

to be used as a qualifying element with the addition
of a new generic from the host language. Again,
the older name is often anglicized or otherwise
altered beyond recognition, as with native American
names used as qualifiers in formations such as
Cape Hatteras, Potomac River, and Chesapeake Bay.
The Anglo-Saxon names of many Roman towns
comprise the Romano-British name with the addition of OE ceaster walled town, as in Cirencester,
Gloucester, Mancetter, Manchester, and Winchester.
During the advance northward into Cumbric territory in the early 7th century, Northumbrian Angles
captured and renamed the stronghold of Din Eidyn
sloping-ridge fortress, retaining the second element
as a qualifier but replacing the first with OE burh
fortification, and transposing them into Germanic
element order to give the place name Edinburgh.
Here, some degree of understanding is reflected in
the translation of Cumbric din fortress by the equivalent burh. By the 12th century, however, scribal introduction of an unetymological <w> into spellings of
the qualifier (Edwinesburg c. 1128) reflects confusion
with the name of the Anglo-Saxon King Edwin, indicating that the original meaning of the name was no
longer understood.
Epexegetic Compounds Epexegetic compounds
also suggest limited communication. Such instances
as Penhill in England, containing both the British and
Old English words for hill, and Glenborrodale in
Scotland, containing both the Gaelic and Old Norse
words for valley, suggest that the older place name
has been taken over as a qualifying element by a
group of new settlers who understood it simply as a
name, rather than as a meaningful description. Similar examples from Finland and Sweden are discussed
by Embleton (19941995: 127128).
Grimston Hybrids Many place names within the
Danelaw of eastern England comprise a Scandinavian
personal name with OE tu n farmstead. From their
position in relation to English-named settlements and
other evidence, these are believed to represent villages
taken over and partly renamed by the Danes after the
allocation of land to leaders of the Danish armies
during the late 9th century. Although the common
place name Grimston is no longer thought to be an
example of the formation, the term Grimston hybrid
is so well established that it continues to be used to
identify this name type.
Inversion Compounds Contact between Norse and
Gaelic speakers is attested by the occurrence in northwest England and southwest Scotland of inversion

Place Names 617

compounds using Scandinavian vocabulary with

Celtic element order. This corpus of names has been
extensively studied for the light it throws on patterns
of lexical borrowing and substratum transfer, as
well as on settlement history and sociolinguistics.
Many problematic issues remain, however, and it is
still uncertain whether the names were coined by
Scandinavian, Celtic, or bilingual speakers.
Renaming

Naming is closely linked to possession, and it is not

uncommon for places to be renamed to reinforce the
authority of a ruling power. Historically, the political
dominance of England led to the imposition of
English names in other parts of the British Isles and
the former colonies, just as the influence of Spain in
the Basque country from the 16th century onward
resulted in Spanish names for Basque places. In
some instances, this renaming leads to competing
forms of the same toponym, so that many Basque
place names now have both Basque and Spanish spellings. Similarly, some South African place names
have equivalent forms in English and Afrikaans, as
with Cape Town/Kaapstad. In Ireland, renaming took
place wholesale under British rule and is now being
systematically reversed, with the publication of a concise bilingual gazetteer (Brainse Logainmneacha,
1989) representing part of an ongoing research program to provide Irish forms for all major place names.
Alternatively, unrelated names for the same place
may survive in different languages. English speakers
use the Scandinavian name Anglesey (Ongulls island) for the island known to Welsh speakers as
Mo n (mountain), and Gaelic names are preserved
in oral tradition for Irish and Scottish places now
known officially by English toponyms, as with An
Gearasdan for Fort William. Some places even have
three or more names, as with the capital of the Slovak
Republic, known variously by the Slovakian name
Bratislava, the German name Pressburg, and the
Hungarian name Pozsony.
Noncontact Situations It is so unusual for an existing place-name stratum to be wiped out completely
that where this appears to have happened, the implications need to be weighed carefully. In England,
the large-scale replacement of Celtic by AngloSaxon toponymy was at one time thought to indicate
that the pre-Anglo-Saxon inhabitants had been exterminated or forced into the border areas of Scotland
and Wales. This theory is no longer considered tenable as archaeological and other evidence points to the
survival of a Celtic population. Indeed, some AngloSaxon place names, such as Walton and Eccles,

contain such elements as OE walh Briton or PrW

*egle s Romano-British church referring to British
people or institutions. Efforts are now focusing
on the identification of potential Celtic names
previously assumed to be from Old English (Coates
and Breeze, 2000). Even less trace remains of the preScandinavian place names of the Northern Isles and
the district of La Hague in Normandy. Again, a small
number of scholars take this to reflect the extermination of the previous inhabitants, but the majority
view is that the earlier Pictish population of Orkney
and Shetland was enslaved or absorbed by the
Scandinavian incomers, while the absence of GalloRomance names from La Hague may result from a
deliberate policy of changing the toponymy to make
it difficult for former owners to lay claim to their
property. Where even the names of natural features
do not survive, as with the pre-Scandinavian names
on the Isle of Man, in which the only ancient river
name is Douglas, the scale of immigration must have
been overwhelming.

Language History
Languages and Dialects

Because place names originate in, and are transmitted

through, spoken language, they often preserve evidence for forms of language that predate those in
written sources. The earliest records of literary Scots
date from the late 14th century, but place names from
the 12th century onward make it possible to trace the
development of the language during the preliterary
period. Some languages and dialects are attested almost solely through place names. All that is known of
the Pictish language is based on personal names, place
names, and inscriptions. Neither is there any literature in the East Anglian dialect of Old English, but
place-name evidence makes it possible to reconstruct
distinctive features of the language. These include the
use of rather than Anglian e (<WGerm a ), attested
in early spellings with <a> of place names such as
Stratton and Stratford from OE stre t/strt road
(Kristensson, 2001).
Dialectology and Word Geography

Whereas many manuscripts and texts are of uncertain

provenance, place names are precisely locatable and
offer primary evidence for the study of dialectology
and word geography. The complementary distribution of Old English terms such as burna and bro c
stream, cnoll and cnp summit, p and stg
upland path and other pairs in charter boundaries
and place names from Anglo-Saxon England has
been mapped and discussed by Kitson in a major

618 Place Names

contribution to the study of dialect isoglosses (1995).

As regards the Old East Anglian dialect mentioned
above, place-name evidence also reveals a linguistic
boundary testifying to a division between Northern
and Southern East Anglian (Kristensson, 2001). Recent work on eclipsis (the mutation of initial consonants) in place names from Irish and Scottish Gaelic
suggests a division between Northern and Southern
Gaelic, uniting Galloway with Irish and Manx rather
Maolalaigh, 1998).
than with the rest of Scotland (O
Dialectal characteristics may be retained in place
names longer than in other areas of language. Examples include the use of (retroflex l) // < ON /r/ in
place names on the coast of Nordmre in Norway
in contrast to appellatival /r/, and the assimilation /l/
C > C attested in name forms, such as Amli <
Almhl from alm elm in vre Telemark in Norway
(Stemshaug, 2002: 235).
Vocabulary and Semantics

The oral origin of place names also means that they

preserve evidence for demotic forms of language, and
for areas of vocabulary that may be represented poorly in extant literature. These include terms for indigenous flora and fauna, and for landscape features. In
some instances, place names preserve vocabulary that
is otherwise unattested, or attested only from a later
date. In others, they reflect a wider or more subtle
range of meaning than is evident from written
sources. OE pc, the ancestor of present-day English
pike, is recorded only with the meaning point,
pointed tool, but occurrences as a place-name generic describing pointed hills in northern England, and as
a qualifier in combination with terms for water in
place names such as Pickmere and Pickburn, testify
to its early development as a hill term and fish name.
Fieldwork is often crucial here. Most place names are
still associated with their original site, so the location
can be examined to establish the appropriateness of
the description. By this means, recent work has identified a transferred meaning of Swedish kil tool,
wedge to refer to a wedge-shaped piece of land in
place names such as Kila in Sweden, paralleled by a
similar development of the West Norse cognate kll in
place names such as Kildale in England.
Comparative evidence is also important. An unattested ON *strjo n occurs in Norwegian place names
designating good fishing places, while the cognate OE
(ge)stre on wealth, offspring occurs in similar locations in England. Examination of the corpus as a
whole suggests that the toponymic usage is related
to productivity. The first element of Pusk in Scotland
(Pureswic 1209) is an unattested OE *pur. Here, the
combination with OE wc specialized farm, together

with linguistic evidence such as comparison with

present-day English dialect pur, suggests an interpretation male lamb, which is supported by the situation of Pusk and other place names containing the
same qualifier within sheep-farming country.
Topographical Vocabulary Groundbreaking work
in the area of historical semantics has been carried
out by Gelling and Cole (2000), throwing new light
on Old English topographical vocabulary by demonstrating a high degree of lexicalization for terms previously regarded as synonyms. This finding is the
result of fieldwork comparing the use of terminology
for identical land formations across England. As
regards OE words for hill, they show that du n
was used for a low hill with a level summit suitable
for a settlement site, whereas beorg referred to a hill
with a rounded profile, cru c to one with an abrupt
outline, and hyll to one that was neither rounded nor
flat-topped. Similar precision characterizes terms for
other landscape features, such as clearings, floodplain, marshland, meadow, pasture, river crossings,
roads, valleys, water courses, and woodland: examples include gela d difficult river-crossing, holt
single-species wood, and wsse land by a meandering river that floods and drains quickly. Research
of this kind has so far focused on England, but the
results of a pilot study in Scotland suggest that
the approach could be applied usefully in other
parts of the British Isles and elsewhere to recover
further information on early topographical lexis.
Orthography and Morphology

Place-name spellings can reveal patterns of orthographic changes at different periods. Whitebaulks in
Scotland, from Scots quhite white and bauk
unploughed ridge, is recorded throughout the 16th
century with spellings in initial <Quh->, replaced
by <Wh-> from the mid-17th century. This change
is evidently the result of anglicization, a process that
also affects morphology in such Scottish place names
as Hangingside (hanging i.e. sloping hillside), recorded up to 1607 with the Middle Scots present
participle inflection <and> (Hingandside 1551,
Hingandsyd 1564, Hingandsyid 1607; Scott, 2003:
25, 27). Recent work on Irish and Scottish Gaelic
also emphasizes the value of place-name evidence for
noun morphology (both inflectional and derivational)
Maolalaigh,
and its interface with phonology (O
1998).
Phonology

Historical spellings of place names, especially those

recorded in local sources, also preserve primary evidence for phonological change. In a Scottish context,

Place Names 619

Nicolaisen (1993) has argued that semantically

opaque place names may be influenced less by standardized spelling conventions, and are more likely to
reflect changes in pronunciation, than vocabulary
words. Stirling is first recorded in the early 12th
century as <Striuelin>, and the majority of spellings
end in <n> or <ne> until the mid-16th century.
A few spellings in final <ng> appear alongside
them, gradually becoming more common until they
take over completely from the mid-17th century.
The same pattern appears in place names such as
Dunfermline and Tealing, and seems to reflect a phonetic change from /n/ to /N/. Similarly, toponyms from
the Gaelic generic baile homestead can help provide
dating evidence for the loss of postvocalic /l/ in words
such as ball and wall, as most spellings without <l>
appear during the late 15th16th century. Also significant are such place names as Falkirk, in which the
medial <l> is not original but first appears in the
mid-15th century, suggesting a hypercorrect spelling
introduced at a time when words were known to be
spelled with medial <l> despite the lack of a
corresponding sound in contemporary pronunciation. Some of Nicolaisens conclusions have been
challenged: for instance, whereas he regards the development from final Gaelic -ach to -o in place names
such as Aberlemno and Belmerino as evidence for

the loss of final fricative /x/ in eastern Scotland, O

Maolalaigh (1998: 3844) argues that Gaelic -ach
in such names may simply have been replaced by
the nearest equivalent as the phonological structure
of Scots did not allow velar fricatives in unstressed
position. Nonetheless, Nicolaisens approach is
clearly one that would repay further research both
in Scotland and elsewhere.
Onomastic Dialect

Although place names are to some extent formed

from lexical items, evidence is emerging of an onomastic dialect predating the splitting up of the
Germanic languages (Nicolaisen, 1995). Cognate
pairs of elements in the North and West Germanic
toponymica, such as OE e a/ON a river, OE hli/ON
hl slope, and OE widu/ON vir wood, support
the theory of a common ancestry in a Northwest
Germanic onomasticon as distinct from the lexicon.
Some 36 such pairs within Old English and Old
Norse are identified by Nicolaisen, as well as another
10 pairs within German and North-Frisian; subsequent scholarship has added OE (ge)stre on/ON *strjo n
productive (fishing) area, OE laf/ON lev inherited
property, and ON *kll/Swedish kil wedge. This
makes it possible to interpret onomastic vocabulary
by comparison with cognate place-name-forming

elements rather than with lexical items, and also

has important implications for language history.
Kitson has pointed out out that some elements within
the hydronymic system discussed above survive as
lexical items in eastern Indo-European languages
but not in European ones, indicating that the naming
system was in operation since before the eastern languages separated from the western continuum
(1996: 86).

Organizations and Place-Name Surveys

Place-name research involves both the close analysis
of individual toponyms, tracing the earliest historical spellings to establish an etymology, and the comparison of corpora from different areas and time
periods. The first requires detailed local and linguistic
knowledge; the second, collaboration on a national
and international scale. The international organization for name studies is the International Council of
Onomastic Sciences, which organizes biennial conferences and publishes the journal Onoma. NORNA,
the cooperative committee for onomastic research in
the Nordic countries of Denmark, Finland, the Faroe
Islands, Iceland, Norway, and Sweden, organizes regular congresses and symposia, the proceedings of
which are published in the series NORNA-rapporter.
In the British Isles, the Society for Name Studies in
Britain and Ireland organizes annual conferences and
publishes the journal Nomina.
On a national and local level, there are too many
individual societies and projects to list here, but mention should at least be made of the American Name
Society, which organizes annual conferences and publishes the journal Names, the Canadian Society for
the Study of Names, which holds annual meetings
and publishes the journal Onomastica Canadiana;
the English Place-Name Society, which has to date
produced some 80 volumes of the English PlaceName Survey and also publishes the Journal of
the English Place-Name Society; the Institute of
Name Research at the University of Copenhagen,
which produces Danmarks Stednavne; and the
Names Society of Southern Africa, which holds biennial congresses and publishes the journal Nomina
Africana. So much relevant material is published in
these sources that individual items have not been
listed separately in the following bibliography unless
directly cited above.
Place-name surveys for other parts of the British
Isles are less advanced than for England, but excellent
work is being carried out by amongst others the
Northern Ireland Place-Name Project; the PlaceName Research Centre, University of Wales Bangor;
and the Scottish Place-Name Society; and in Ireland

620 Place Names

by the Placename Branch of the Department of

Community, Rural and Gaeltacht Affairs.
In addition to etymological work, active steps are
being taken to recover indigenous names in areas
overlaid by foreign imports. The Australian National
Placenames Survey aims to produce an electronic
database analyzing the full corpus of both imported
and indigenous names, and part of the remit of the
New Zealand Geographic Board is to collect and
to encourage the use of original Maori place names.
See also: Personal Names; Proper Names: Linguistic Sta-

tus; Proper Names: Semantic Aspects.

Bibliography
Brainse Logainmneacha (1989). Gasaite ar na hE irean/
Gazetteer of Ireland. Baile A tha Cliath/Dublin: Placenames Office of the Ordnance Survey.
Cameron K (ed.) (1975). Place-name evidence for the
Anglo-Saxon invasion and Scandinavian settlements.
Nottingham: English Place-Name Society.
Cameron K (1996a). English place names (rev. edn.).
London: Batsford.
Cameron K (1996b). The Scandinavian element in
minor names and field-names in north-east Lincolnshire.
Nomina 19, 527.
Coates R & Breeze A (2000). Celtic voices English places:
studies of the Celtic impact on place-names in England.
Stamford: Shaun Tyas.
Debus F & Schmitz A (2001). (Mikro-) Toponyme im
slawisch-deutschen Kontaktgebiet Norddeutschlands.
Onoma 36, 5170.
Eichler E, Hilty G & Lo ffler H (eds.) (19951996).
Namenforschung: ein internationales Handbuch zur
Onomastik/Name studies: an international handbook of
onomastics/Les noms propres: manuel international
donomastique (3 vols.). Berlin & New York: Walter de
Gruyter.
Embleton S (19941995). Place names in Finland: settlement history, sociolinguistics, and the Finnish/Swedish
language boundary. Onoma 32, 124139.
Gelling M & Cole A (2000). The landscape of place-names.
Stamford: Shaun Tyas.
Kitson P R (1995). The nature of Old English dialect distributions, mainly as exhibited in charter boundaries. In
Fisiak J (ed.) Medieval dialectology. Berlin & New York:
Mouton de Gruyter. 43135.
Kitson P R (1996). British and European river-names.
Transactions of the Philological Society 94, 73118.
Kristensson G (2001). Language in contact: Old East Saxon
and East Anglian. In Fisiak J & Trudgill P (eds.) East

Anglian English. Cambridge: D.S. Brewer. 6370. [First

published in Melchers G & Warren B (eds.) (1995).
Studies in anglistics. Stockholm: Almqvist & Wiksell
259268].
Nicolaisen W F H (1993). Scottish place names as evidence
for language change. Names 41, 306313.
Nicolaisen W F H (1995). Is there a Northwest Germanic
toponymy? Some thoughts and a proposal. In Marold E
& Zimmermann C (eds.) Nordwestgermanisch. Berlin &
New York: Walter de Gruyter. 103114.
Nicolaisen W F H (2001). Scottish place-names: their study
and significance (rev. edn.). Edinburgh: John Donald.
O Maolalaigh R (1998). Place-names as a resource
for the historical linguist. In Taylor S (ed.) The uses
of place-names. Edinburgh: Scottish Cultural Press.
1253.
Scott M (2003). Scottish place-names. In Corbett J,
McClure J D & Stuart-Smith J (eds.) The Edinburgh
companion to Scots. Edinburgh: Edinburgh University
Press. 1730.
Stemshaug O (2002). Place-names and dialectology.
Onoma 37, 219247.
Vennemann T (1994). Linguistic reconstruction in the
context of European prehistory. Transactions of the
Philological Society 92, 215284.

Relevant Websites
http://www.icosweb.net International Council of Onomastic Sciences.
http://www.sofi.se NORNA.
http://www.snsbi.org.uk Society for Name Studies in
Britain and Ireland.
http://www.wtsn.binghamton.edu American Name Society.
http://geonames.nrcan.gc.ca Canadian Society for the
Study of Names.
http://www.nottingham.ac.uk English Place-Name Society.
http://www.hum.ku.dk Institute of Name Research at the
University of Copenhagen.
http://www.osu.unp.ac.za Names Society of Southern
Africa.
www.qub.ac.uk Northern Ireland Place-Name Project.
http://www.bangor.ac.uk Place-Name Research Centre,
University of Wales Bangor.
http://www.st-andrews.ac.uk Scottish Place-Name Society.
http://www.pobail.ie Placename Branch of the Department of Community, Rural and Gaeltacht Affairs.
http://www.anps.mq.edu.au Australian National Placenames Survey.
http://www.linz.govt.nz New Zealand Geographic Board.

Plagiarism 621

Plagiarism
D Woolls, CFL Software Development,
Birmingham, UK
2006 Elsevier Ltd. All rights reserved.

Introduction
Plagiarism is more often a violation of regulations
than the law, although it is clearly implicated in copyright. Plagiarism occurs mostly within educational
establishments, where the matter is normally handled
by the internal disciplinary procedure of the institution. The penalties can be severe, up to and including
the failure to be granted a degree, but it is relatively
rare for cases to be carried beyond the confines of the
establishment. Outside educational institutions, those
accused of plagiarism are most likely to be prosecuted
for breach of copyright, if the aggrieved author decides to pursue the case. Such cases are also relatively
rare; accusations and defenses are more frequently
debated in the national newspapers and journals
than in courts of law.
Most accusations of plagiarism are made in relation to student work; the subject of the plagiarism
usually knows nothing about the matter. That is why
the common reference to kidnapping a text is somewhat imprecise. The words written by the author are
not removed from their original location, and the
original author seems to experience no loss since the
author does not see the plagiarism. For the same
reason, borrowing and theft are equally imprecise
explanations of the practice. Any sense of grievance is
normally first felt by the reader, who has been misled
about the true authorship of the work in question.
Where such misleading is contrary to the regulations
governing the production of the work, this grievance
can be extended into some form of prosecution. Since
this occurs most often in predominantly academic
settings, the majority of the discussion below centers
on the academic issues.

The Law of Copyright and the Concept

of Plagiarism
A sentence such as the following is now common at
the front of books published in the United Kingdom:
The author has asserted her moral right to be identified as the author of this work in accordance with
section 77 of the Copyright, Designs and Patents Act
1988. This applies equally to fiction and nonfiction
writing, and appears because, unless and until this is
explicitly asserted, no protection for breach of this
right is available. However, whether or not such an
assertion has been made, the presence of one authors
work in that of another author without attribution is

still considered plagiarism. There is no mention of the

word plagiarism in the above Patents Act statement,
although it is related predominantly by the moral
issue of giving credit where credit is due. Copyright
is designed to protect the rights of the author. Plagiarism, while certainly breaching such rights, normally
has a greater impact on the reader, by presenting a
misleading origin for the material that he or she is
reading.
That an author has the right to be identified as such
does not imply that only the author in fact created all
the material in the book. In works of nonfiction this
will rarely be the case, and many works of fiction
contain references to material by other authors. The
claim is that the author has both created some of the
text, and selected and arranged pre-existing elements
in such a way as to form a new and unique text. In
such cases, the author of the whole work, particularly
in the area of nonfiction, is expected to alert the
reader to the presence of direct or indirect use of
external material, and to include a list of the publications to which reference has been made at the end of
the work.
A number of systems have been formalized into
style guides to assist an author in compiling a bibliography, the predominant guides being Harvard style
manual for editors, The MLA style manual, The
Chicago manual of style (University of Chicago),
and the Publication manual of the American Psychological Association style guide (APA). All these set
guidelines for the contents and style of the bibliography. To avoid repetition of the same method of
identification, and to acknowledge the utility of paraphrase in summing up a longer argument in a referenced work, while the bibliographical section
remains consistent within any one set of conventions,
the internal textual referencing is more flexible. So,
for example, inverted commas to indicate quoted
material are used for shorter extracts, indentation is
used to show a number of sentences, and simple
reference to the author or authors with no explicit
marking is acceptable where accurate paraphrase of
material is employed.
The following sentence illustrates the correct use of
many of the standard techniques by the author Alison
Johnson:
Using Hallidays (1989) terminology for the division of
vocabulary into lexical (open class) and grammatical
(closed class), he raises the case of apparent open class
lexical items which have connective properties, especially . . . of paraphrasing closed class connectors such as
subordinates and conjuncts which are [therefore better]
treated as closed class vocabulary (Winter, 1996: 46)
(Johnson, 1997: 213).

622 Plagiarism

This sentence indicates clearly which element of the

discussion originates with Halliday, and the concepts
of open and closed classes of words. Since these are
general concepts associated with Halliday, it is sufficient to identify him as their originator in the body of
the text. How Winter discusses them within his text is
indicated by the double quotes surrounding the part
of the sentence so used, referenced by page and dated
publication at the end. In the quote itself, two other
features are used. Ellipsis indicates that material was
omitted, between especially and of, so that this was
not all that Winter said, and the square brackets
indicate Johnsons interpolation of an explanatory
segment into the original quote.
If we make only slight alterations to the text we can
illustrate how plagiarism can occur.
Using Hallidays (1989) terminology for the division
of vocabulary into lexical (open class) and
grammatical (closed class), it is possible to identify
apparently open class lexical items which have
connective properties, especially of paraphrasing
closed class connectors such as subordinates and
conjuncts which are therefore better treated as
closed class vocabulary.

In the italicized sentence, all but one of the features of

correct citation have been removed. If this was presented in a paper, Halliday would still be duly credited, but the origin of the ideas encapsulated in the
correctly cited quote would be credited to the putative author. In such a case, both Johnson and Winter
would be plagiarized, since neither are credited, although the main victim would be Winter, whose ideas
have been absorbed entirely by the plagiarist, along
with Johnsons glosses on them.
This illustrates that the rules are cast so that in the
absence of any such internal indication, the reader of
a work may safely assume that the remainder of the
work is that of the stated author. Correct attribution
assists the reader, the author, and the cited authors.
A reader of a text cannot necessarily be expected to be
able to identify material taken from other sources,
even if he or she has read the publications from
where the material was taken, and so needs the assistance of the author in separating the sources to better
evaluate the material being presented. The author
benefits from this convention by having his or her
distinct contribution clearly identified. The quoted
author benefits from the level of exactitude of quoted
material being indicated, and is so protected from
potential misrepresentation. A reader can refer to
the quoted source in cases of paraphrase to assess
whether the summary presented is in fact something
like what the original author wrote. If these conventions are not followed, the reader may be misled

by their absence into believing that the main author

is the original source for the work of others, and
accusations of plagiarism can arise.

How Different Is Work on the Same

Subject?
The objection is frequently raised that work on similar topics, particularly essay questions, is bound to
look more similar than different, because the students
use the same reference material, and might be expected to reach similar conclusions. However, it is
possible to show clear distinctions between work
that is dependent on another students work and
work that is independently produced, using computer
analysis of the content of student essays.
The two pie charts below show the distinctive difference in word use between two independently produced
texts (Figure 1) and two dependent texts (Figure 2).
In each case, the figures represent two essays,
which each answer the same question, that have been
compared for content by a computer program. Figure
1 shows what a normal comparison produces. The
bulk of the words appear only in one text or another;
the slice indicating vocabulary that is shared between
the two texts more than once is smaller than the
distinctive vocabulary slices, and the slice indicating
words used only once is very small. Figure 2 shows
what happens when two texts are dependent on each
other, or make heavy use of a shared source. The
pattern is completely reversed, with the vocabulary
shared just once forming a substantial slice of the pie.
Readers are alerted to the feature in quite a different

Figure 1 Independently produced essays answering the same

question.

Plagiarism 623

way, of course. In the independently produced texts,

they read different sentences nearly all the time. In a
dependent text, sentences such as those below are
presented, and the reader will almost certainly remember seeing something like the sentence of Text 3
before, when encountering the sentence in Text 4,
even if some time has elapsed since reading it. For
example, consider the two sentences below.
Text 3: British history suggests that the
phenomenons of a multiracial and multicultural
nation has always existed and that immigration
was not just post World War Two theories.
Text 4: British history reveals that it has always been
a multiracial and multicultural nation and that
immigration was not just a Post World War Two
phenomenon.

Text 3 is exactly what the author wrote, including the

ungrammatical phenomenons. The underlined words
indicate the differences between the two sentences.
What the reader will almost certainly not be aware
of is that a number of words in these sentences
occur only once in each of the texts, as shown by
the boldface words below.
Text 3: British history suggests that the
phenomenons of a multiracial and multicultural
nation has always existed and that immigration
was not just post World War Two theories.
Text 4: British history reveals that it has always been
a multiracial and multicultural nation and that
immigration was not just a Post World War Two
phenomenon.

Figure 2 Two closely related essays answering the same

question.

Further discussion of this case, with illustration of the

use of computer assistance to analyze the data, can be
found in Johnson (1997) and Coulthard (2004).

Normal Text Composition

This relative infrequency of word use arises from the
normal distribution of words within a text. Separating content from function words allows texts of
very different lengths to be compared. Figure 3
shows the relative proportions of content vocabulary
occurrence in an 18th-century essay of 13 074 words.
Figure 4 shows a very closely matching pattern in a
20th-century student essay of 3888 words. As can be
seen, the majority of content words will only occur
once (they are generally referred to as hapax legomena), and account for approximately 65 percent of
all the content words used. The proportion of vocabulary occurring just twice (dis legomena) is remarkably
stable over even very long texts, with the 16 percent
shown on the graph varying only slightly with text
length. The implications of this are that words new to
the reader in the surrounding context occur throughout a text, and it is this turnover of words that keeps
readers reading.
The figures in Table 1 illustrate that this is indeed
broadly the case. The figure shows how many of each
successive 350 words in the running text are either
content or function words that occur just once in the
full text. The number of content words in each segment is fairly stable, and this stability can be found in
the majority of texts across groups of approximately
20 sentences such as these. At a lower level of word

Figure 3 Eighteenth-century essay showing content word

frequency of occurrence.

624 Plagiarism
Table 1 Text 3 divided into successive 350-word segments and
analyzed for the Hapax occurrence of content and function words

Figure 4 Twentieth-century student essay showing content of

word frequency of occurrence.

sequence, local variation can be present. An extended

noun group, through a multiplicity of nouns or additional adjectives, would push up both the content
hapax and the content count in a segment, for example. What is also of interest is that some function
words are also introduced throughout the text and
not reused, but obviously in much lower quantities
per segment.

Communicative Function of the Hapax

Legomena and Dis Legomena
The hapax legomena and dis legomena also represent
the original contribution of an author to addressing a
common problem or presenting an argument. The
texts may be different responses to an essay title, or
may be two entire books on the same broad topic.
When reading successive related texts, in whatever
order they are read, the reader is continually encountering new words because each pair can be expected
to show the pattern of shared hapax legomena (see
Figure 1 above). So, when confronted with a text
where a number of the words appear familiar, a reader is alerted to the oddity of that experience. This will
be even more noticeable when, as in the sentences
above, the words appear in the same sentences and
in approximately the same order. Just one such sentence can alert a reader to the problem; where, as
might be expected from the size of the hapax legomena slice in the Text 3 and Text 4 comparison in
Figure 2, several sentences are involved, serious
doubts as to the originality of one of the documents
can arise, and collusion or plagiarism investigation
is started.

Content hapax
per 350 words

Content count
per 350 words

Function hapax per

350 words

49
52
46
50
54
69
43
55
60
50

158
158
156
177
165
168
143
166
181
156

6
7
6
10
12
5
13
9
5
6

How Much Unattributed Sharing Is

Allowable?
In pure terms, any unattributed borrowing must be
considered plagiarism, and this may be as little as one
word. In considering how plagiarism comes about in
the material produced by students, one observed
strategy could be described as equivalent to the construction of a patchwork quilt, sewn together from
fragments of material to make a new artifact. If this
article now includes the sentence I call this patchwriting, with no further reference, then the editors
will correctly accuse this author of plagiarism, because the term was coined by Rebecca Moore
Howard (Howard, 1993). An attempt to avoid the
problem might make an author consider using the
image of the practice as being like that of using
the technique of collage, to create a picture from
fragments of paper and fabric, and call it collagism.
The probability is that this would be considered the
plagiarism of the patchwriting idea, because it is a
representation of a similar practice with no acknowledgment of the idea. The only way in which this
might be defended is if the author could demonstrate
that there was no prior knowledge of the concept at
the time of writing.
Such a case illustrates the problem of quantifying
two texts to judge on similarity. Once plagiarism has
been suspected, it is not uncommon for the amount
of shared material to be identified, and, the more
there is, the worse the plagiarism is considered to
be. But, especially in areas where there are clear
guidelines as to expected practice, such as universities, any occurrence that can be identified is generally
considered sufficient to constitute a problem. Given
that this is the case, questions relating to how much of
a relationship there must be between the sentences for
the source and the work under review, and how many
such sentences are required to constitute plagiarism,

Plagiarism 625

are very difficult to provide satisfactory answers to.

While acknowledging that such questions are valid on
a case-by-case basis, the general principle adopted by
writers of plagiarism legislation is that case-by-case
evaluation can only take place under a general acceptance that any uncredited work that can be discernibly traced back to a source available to a writer
requires an explanation.

Detection of Plagiarism
Until the end of the 20th century, plagiarism has been
detected largely by human (as opposed to computerassisted) readers familiar with the subject area, the
likely sources, and, in the case of assessing the work
of students, with the expected capabilities of the writers. In such cases, the triggers for recognition can
be specific vocabulary items or phrases associated
with particular sources or recognition of substantial
portions inserted. The most common reported alert
of suspicion is of a sudden change of style within
the flow of the text. This can be an unexpected improvement in clarity, the use of terminology outside
the expected boundaries of reference, or the use
of material that is not as fully integrated with the
surrounding text as the rest of the material. The sentences below, for example, are incorrectly punctuated
while containing two highly distinctive and wellcontrasted sets of adjectives, marked in bold type.
In this text, the role of the leading man Okonkwo
seems to show that he values masculinity and he is
rebellious, the fact that his father was lazy makes
Okokwo more determined to make his way in the
world that values manliness. He rejects everything
that his father stood for it seems to show that he
believed his father to be idle, poor, cowardly, gentle
and interested in music and conversation. So he
decides to do the opposite of everything his father
was, Okonwo becomes productive, rich, brave,
violent, and adamantly opposed to music and
anything else that seems to be soft.

In the eventually acknowledged source reproduced

below the overlaps have been marked. The underlined words are retained intact, with modifications
of the intervening words, e.g., seems to value being
transformed into values.
Analysis of Important Characters
Okonkwo
Okonkwo, the son of the effeminate and lazy Unoka,
strives to make his way in a world that seems to
value manliness. In so doing, he rejects everything
for which he believes his father stood.

The second part of the essay extract above, shows

retention of some adjectives, marked in bold type,

and the replacement or omission of others; wealthy

has been replaced by rich and thrifty omitted. Further rewriting is present in this section, with the
underlined elements representing the exactly borrowed material; e.g. he perceives becoming seems.
Unoka was idle, poor, profligate, cowardly, gentle,
and interested in music and conversation.
Okonkwo consciously adopts opposite ideals and
becomes productive, wealthy, thrifty, brave,
violent, and adamantly opposed to music and
anything else that he perceives to be soft, such as
conversation and emotion.

As might be expected from the presence of the header,

this is from a set of notes on a book, in this case
Things fall apart by Achebe. The source is Spark
Notes, a set of materials published on the Internet,
primarily aimed at secondary education, so an inappropriate authority for a third-year degree-level
student to be employing, even if referenced, which it
was not. The major transformation is at the start of
the second sentence, where sentences two and three
of the source are conflated in a less than clear or
comfortable fashion.
He rejects everything that his father stood for it
seems to show that he believed his father...

This form of transformation is not uncommon,

although many examples are less directly related
than this.

Identification of Sources
Once suspicions have been aroused about the stylistic
integrity of the essay, the reader then has the problem
of attempting to identify the source. This is generally
a requirement in disciplinary proceedings, because it
is necessary to show that the material substantially
exists in the source and to indicate the extent and the
manner of the incorporation of this material in unreferenced form. In cases where the reader is familiar
with the relevant literature, this can be a relatively
simple task, but many cases of suspected plagiarism
cannot be pursued to a conclusion because the source
data cannot be readily traced. It is possible to make
comparisons with earlier work, but there is as yet no
reliable stylistic indicator that can demonstrate that
an author could not have written the words under
scrutiny. Some research into provision of computerassisted style checking is reported in Woolls (2003).
Computer assistance is also available in the form of
Internet search engines, where the material under
suspicion can be checked for occurrence on any
open Internet site. Since students are often in need
of a rapid answer, it is not uncommon for the source

626 Plagiarism

or sources to be readily found, since in-depth research

into potential sources revealed by an initial enquiry
based on the question is unlikely to have occurred.
So, the most likely candidate sources are those at the
top of the search list.

Native and Nonnative Speaker

Differences
It should be noted that, one side effect of the stylistic
issue is that nonnative speakers of the language in
which the work has been presented, are frequently
perceived to be at a disadvantage over native speakers
employing similar incorporation strategies. This is
because the improvement in expression and confidence in language use is extremely apparent, and
they are most likely to use the standard sources.
As such a group is also potentially one most likely
to have a problem with either comprehension or
transformation of the source data, this places a particular burden of responsibility on those teaching
and marking to ascertain the nature of the problem
before assuming deliberate plagiarism. The question of whether plagiarism can in fact be other than
deliberate is addressed below.

Electronic Detection
A number of factors have resulted in the development
of electronic plagiarism detection tools. The growth
in class sizes has made knowledge of student style
much less common, and crosscomparison much
more difficult. Source material is increasingly available both within institutions, and freely published on
the Internet. And more work is being produced electronically and is increasingly being collected electronically. So it has become practical to compare student
work with that of their peers, and to search the Internet for material incorporated from open websites or
relevant data held by the web search services.
It is impossible to be definitive about search methodologies since they are not generally revealed by
their constructors. However, in the unstructured
space that the Internet is, as opposed to the indexed
databases employed in document retrieval systems,
for example, a successive word methodology is likely to be the preferred method of search. World Wide
Web searching is generally performed by looking for
consecutive strings of words that have been transformed electronically into distinctive patterns that
can be found in other texts treated in the same way.
As few as 6 to 8 words can be used in such pattern
building for identification across a very large number
of potential sources. If those making use of the web
sources are aware of this, it is possible to attempt to

defeat the search mechanism by systematically changing words throughout a text within the presumed
successive word boundary. The precise nature of the
operation of the algorithms is not known, however, so
this may not be a fruitful concealment attempt. It is
also a considerable task to prepare oneself to make
amendments that are coherent and still answer the
question. In any extended source use, it is probably
highly likely that a consecutive word string will be left
intact, and this gives a search algorithm a chance of
identifying the source. What such search engines cannot deal with is that not all material available from
the Internet is immediately available. Some can only
be obtained on payment of a fee, either for prewritten
material or to have an essay written for that fee, so the
essay has no obvious antecedents.
Related problems are growing class sizes with students sharing work by email attachments, and large
distance-taught groups where the electronic sources
in learning material are often simply reproduced in
the answers of the student. To address these issues,
other software programs have been developed to
compare all the work on a given task in a student
group with all the others. This allows comprehensive
comparison of the level of similarity not available to
generalized web searchers. These programs operate
using either using successive strings of words, or individual word similarity within a document. This provides assurance with regard to the independence of
student work from their peers, but cannot assist in the
identification of external material.

Can Plagiarism Occur and Not Be

Deliberate?
Within the academic area, the relationship between
plagiarism and learning how to write has been explored by Rebecca Moore Howard (1993, 1995,
1999). She argues that many students find the
requirements of the academic referencing system too
difficult to apply in their first years in higher education, because they have not learned to write in an
academic style. This is not ignorance of the underlying expectations, but a perceived or actual inability
to meet those expectations. Although able to find
the sources for the task given them, they are unable
to process the material sufficiently well to produce an
accurate or acceptable summary to include in their
essays. To reduce the amount of direct quotation
that would be necessary if all their sources were correctly cited, some students adopt the policy of selecting complete sentences or sections from different
sources and stitching them together with their own
material. This she has called patchwriting, as noted
earlier. Since this leaves clear traces of the original

Plagiarism 627

material, particularly in the first years in higher education, their teachers will have great familiarity with
the source data, and they are frequently accused and
found guilty of plagiarism for adopting this practice.
Howard argues that this should be seen rather as a
stage on the path to achieving academic writing standards, and is more accurately described as inadequate
paraphrase, or an inappropriate use of sources. She
further contends that students can be encouraged to
build their way out of this practice during their
course, and that it is not an acceptable part of final
work assessed for a degree.
Diane Pecorari examined this view sympathetically
in a Ph.D. thesis (Pecorari, 2002) and subsequent
paper (Pecorari, 2003). The thesis traced the progress
of nine postgraduate masters degree students through
their course, with the consent and cooperation of
both students and supervisors. She found that in all
cases patchwriting was present to a greater or lesser
extent, even though the students knew that they
were being monitored and were in fact involved in
the discussion process during their course. In addition, Pecorari looked at eight completed Ph.D.s selected for their similarity in provenance to the M.A.
students, and discovered that in all but one case,
patchwriting was present at different levels. This
implies that the practice persists beyond first-degree
level. In her conclusion Pecorari (2002) comments,
What can be concluded is that between performance and
expectations a wide gap exists, one which presents a
danger for every student and a disaster for the few who
err, and whose errors are detected. It is in the interest of
every member of the academic community to mind the
gap and try to close it.

An educational establishment has the role of student

development, of which writing in an academic style
is but one aspect. All higher education institutions
have explicit statements of the severity with which
plagiarism is viewed by the academic community,
most will have examples of what is deemed to
be plagiarism available for student reference, and
almost all will retain the option of failure of a complete course element, or ultimately expulsion from
the institution. Both Howard and Pecorari believe
that this stance is unhelpful and unduly severe for
institutions to take while existing to develop their
studentsminds. Neither denies the existence of what
is defined clearly as plagiarism in much student work,
but they regard the issue as one of education and
prevention rather than as a disciplinary matter.
The view taken by Howard and Pecorari almost
certainly represents a minority one. The mainstream
position is that the rules of accurate referencing
are relatively few and are clearly set down from the

beginning of a students career. Observation of these

rules is a fundamental requirement to the awarding of
a degree. Every student benefits at the institution,
both while taking the degree course and in later life.
Due to the evaluative nature of the writings of both
authors, the bibliographies attached to their work can
be recommended as rich sources of further reading in
the field.

Conclusion
While plagiarism can be clearly defined as a departure
from the convention of attribution of sources, in
practice, identifying that plagiarism is present is not
always an easy task, and the identification of the
actual sources that have been plagiarized can be
even more difficult. The growth of electronically
available data, much of which will not be known to
either the specialized or general reader, makes the
problem of identification even greater. Electronic detection tools are only a partial solution that have
limitations on the amount of available data that can
be checked, and therefore the overall accuracy of a
negative report. Classical plagiarism the use of material from books is still a practice that is employed
and detecting this continues to be a problem only
really accessible to a reader (human, not computer)
with a broad knowledge of the topic. This problem
may in the near future lend itself to computer-assisted
detection as more publications become available online. At the time of writing, a combination of detection and prevention methodologies within education
is the favored method for encouraging good practice.
See also: Applied Forensic Linguistics; Authorship Attribution: Statistical and Computational Methods; Computational Stylistics; Computers in the Linguistic Humanities:
Overview; Corpus Linguistics; Linguistic Features.

Bibliography
Coulthard M (2004). Author identification, idiolect, and
linguistic uniqueness. Applied Linguistics 25, 43447.
Gibaldi J (1988). The MLA style manual and guide
to scholarly publishing (2nd edn.). New York: Modern
Language Association of America.
Howard R M (1993). A plagiarism pentimento. Journal of
Teaching Writing (Summer), 233245.
Howard R M (1995). Plagiarisms, authorships, and the
academic death penalty. College English 57, 708736.
Howard R M (1999). Perspectives on writing: theory,
research, practice, vol. 2: Standing in the shadows of
giants. Stamford, CT: Ablex Publishing.
Johnson A J (1997). Textual kidnapping a case of plagiarism among three student texts? Forensic Linguistics.

628 Plagiarism
The International Journal of Speech, Language and the
Law 4(2), 210216.
Pecorari D E (2002). Original reproductions. Ph.D. diss.,
University of Birmingham, UK.
Pecorari D E (2003). Good and original: plagiarism and
patchwriting in academic second-language writing. Journal of Second Language Writing 12, 317345.
Woolls D (2003). Better tools for the trade and how to use
them. Forensic Linguistics. The International Journal of
Speech, Language and the Law 10(1), 102112.
Woolls D & Coulthard M (1998). Tools for the trade.
Forensic Linguistics. The International Journal of
Speech, Language and the Law 5(1), 3357.

Yancey K B & Huot B (eds.) (19972000). Perspectives on

writing: theory, research, practice (5 vols). Stamford, CT:
Ablex Publishing.
(2001). Publication manual of the American Psychological
Association (5th edn.). Washington, DC: American
Psychological Association.
(2002). Style manual for authors, editors and printers
(6th edn.). Sydney: John Wiley & Sons.
(1993). The Chicago manual of style (14th rev., edn.).
Chicago: University of Chicago Press.

Planning Strategies and Production of Spoken Discourse

E Milano, Universita` di Napoli, Napoli, Italy
2006 Elsevier Ltd. All rights reserved.

Introduction
It is not surprising that, even if the primary goal of all
research focused on speech planning and production
is to account for spontaneous spoken language, the
relationship between research on speech planning and
production and the study of spoken discourse has
often been difficult. This has much to do with the
impossibility of direct observation of conceptualization processes and with the difficulties in achieving
control over the production of speech. The creativity
inherently involved in speech planning sets two counteropposing effects in motion: the less creativity is
associated with a production task, the less interesting are the insights into the language production
processes; the more creativity is associated with a
production task, the wider is the variability in speech
productions and the more difficult it is to draw
conclusions from the data. As such, the historical
relationship between speech production and planning
research and the study of spoken discourse has
been strongly affected by the choice of methodology and data in language planning and production
studies.
The two research fields appeared strongly connected in the 1970s, when the analysis of error
corpora collected from daily conversations (over
many years) gave a large impetus to language production research. The underlying idea was that
errors are not random but seem to be connected to
fundamental characteristics of the speech production
process (see Fromkin, 1971, 1973; Garrett, 1980;
MacKay, 1972): errors shed light on the underlying

units of linguistic performance and on the production

of speech (Fromkin, 1971: 29). Thus speech error
data opened a window into linguistic mental processes, letting the analyst observe the processes between thought generation and its verbal articulation
(Fromkin, 1973: 43).
Together with speech errors, insights on the interior
mental process during the production of spontaneous
spoken discourse came from garnering information
about hesitations and their role in the decisionmaking process, primarily as markers for when
planning and decision making are being made by the
speaker (see, among others, Beattie, 1979).
The emphasis on dysfluencies in the early research
on speech production and planning eventually was
recognized as paradoxical in the effort to define the
properly functioning processes that allow smooth
production of speech (Ferriera, 1996: 724). This, as
well as the difficulty in managing data from spontaneous spoken speech, distanced the study of speech
production from spontaneous spoken discourse:
Thus the ultimate models of normal speech production will try to not lie in how they account for infrequent derailments of the process but rather must lie in
how they deal with the normal process itself (Levelt
et al., 1999: 2). Notwithstanding the above trend, the
study of errors and dysfunction in language remains
even today an invaluable tool for understanding the
mechanisms of language production (see, among
others, Meijer, 1997).

The Speaker as Information Processor

Leading the just-mentioned trend, Levelt (1989) provides an excellent overview of major advances in the
language planning and production field up to the end

628 Plagiarism
The International Journal of Speech, Language and the
Law 4(2), 210216.
Pecorari D E (2002). Original reproductions. Ph.D. diss.,
University of Birmingham, UK.
Pecorari D E (2003). Good and original: plagiarism and
patchwriting in academic second-language writing. Journal of Second Language Writing 12, 317345.
Woolls D (2003). Better tools for the trade and how to use
them. Forensic Linguistics. The International Journal of
Speech, Language and the Law 10(1), 102112.
Woolls D & Coulthard M (1998). Tools for the trade.
Forensic Linguistics. The International Journal of
Speech, Language and the Law 5(1), 3357.

Yancey K B & Huot B (eds.) (19972000). Perspectives on

writing: theory, research, practice (5 vols). Stamford, CT:
Ablex Publishing.
(2001). Publication manual of the American Psychological
Association (5th edn.). Washington, DC: American
Psychological Association.
(2002). Style manual for authors, editors and printers
(6th edn.). Sydney: John Wiley & Sons.
(1993). The Chicago manual of style (14th rev., edn.).
Chicago: University of Chicago Press.

Planning Strategies and Production of Spoken Discourse

E Milano, Universita` di Napoli, Napoli, Italy
2006 Elsevier Ltd. All rights reserved.

Introduction
It is not surprising that, even if the primary goal of all
research focused on speech planning and production
is to account for spontaneous spoken language, the
relationship between research on speech planning and
production and the study of spoken discourse has
often been difficult. This has much to do with the
impossibility of direct observation of conceptualization processes and with the difficulties in achieving
control over the production of speech. The creativity
inherently involved in speech planning sets two counteropposing effects in motion: the less creativity is
associated with a production task, the less interesting are the insights into the language production
processes; the more creativity is associated with a
production task, the wider is the variability in speech
productions and the more difficult it is to draw
conclusions from the data. As such, the historical
relationship between speech production and planning
research and the study of spoken discourse has
been strongly affected by the choice of methodology and data in language planning and production
studies.
The two research fields appeared strongly connected in the 1970s, when the analysis of error
corpora collected from daily conversations (over
many years) gave a large impetus to language production research. The underlying idea was that
errors are not random but seem to be connected to
fundamental characteristics of the speech production
process (see Fromkin, 1971, 1973; Garrett, 1980;
MacKay, 1972): errors shed light on the underlying

units of linguistic performance and on the production

of speech (Fromkin, 1971: 29). Thus speech error
data opened a window into linguistic mental processes, letting the analyst observe the processes between thought generation and its verbal articulation
(Fromkin, 1973: 43).
Together with speech errors, insights on the interior
mental process during the production of spontaneous
spoken discourse came from garnering information
about hesitations and their role in the decisionmaking process, primarily as markers for when
planning and decision making are being made by the
speaker (see, among others, Beattie, 1979).
The emphasis on dysfluencies in the early research
on speech production and planning eventually was
recognized as paradoxical in the effort to define the
properly functioning processes that allow smooth
production of speech (Ferriera, 1996: 724). This, as
well as the difficulty in managing data from spontaneous spoken speech, distanced the study of speech
production from spontaneous spoken discourse:
Thus the ultimate models of normal speech production will try to not lie in how they account for infrequent derailments of the process but rather must lie in
how they deal with the normal process itself (Levelt
et al., 1999: 2). Notwithstanding the above trend, the
study of errors and dysfunction in language remains
even today an invaluable tool for understanding the
mechanisms of language production (see, among
others, Meijer, 1997).

The Speaker as Information Processor

Leading the just-mentioned trend, Levelt (1989) provides an excellent overview of major advances in the
language planning and production field up to the end

Planning Strategies and Production of Spoken Discourse 629

of the 1980s. Moreover, he presents a very detailed

model of speaking, from planning to articulation.
According to Levelt, studying speech functions and
patterns of spoken interaction between speakers,
although of crucial importance (in fact he starts
with a qualitative analysis of a case study taken
from the English conversation corpus of Svartvik
and Quirk (1980)), is not enough. The primary objective is to develop a theory of speaking that can
be broken down from system into subsystems or processing components with characterizations of the
representations they compute.
The model is made up of four components: a
Conceptualizer, a Formulator, an Articulator, and a
Speech-Comprehension System. The Conceptualizer
generates preverbal messages. It consists of two steps,
macroplanning and microplanning. The former elaborates the communicative intention and determines
the content of the speech act. The latter gives the
expression to each speech act, supplying the message
with an information structure that helps the addressee to infer the communicative intention. The
Formulator is made up of two subcomponents: the
Grammatical Encoder, which retrieves lemmas from
the lexicon and generates grammatical relations,
and the Phonological Encoder, which builds a phonological plan. The Articulator executes the phonetic
plan. The Speech-Comprehension System allows the
speaker to monitor his own productions.
Each of the above components is an autonomous
specialist that transforms its particular input into its
particular output. The processing is incremental. It
includes both serial and parallel processing. All components can work in parallel on different bits and
pieces of the utterance. Each processing component
will be set into motion by a minimal amount of its
characteristic input. An interesting, and still debated,
issue is how small that minimal amount can be.
A central thesis in Levelts theory is that the lexicon
is a mediator between conceptualization and grammatical and phonological encoding (lexical hypothesis). The preverbal message activates lexical items.
The syntactic, morphological, and phonological properties of an activated lexical item in turn trigger the
grammatical, morphological, and phonological encoding procedures in charge of the generation of an
utterance.
In the last 20 years, research by Levelt and his
colleagues has generated a lively debate focused on
different aspects of the theory. The more the discussion has delved into details, the stronger has been the
necessity of achieving experimental control over
speech production, particularly at the higher levels
of sentence production. As a result, scholars have
conducted research mainly using experimentally

induced speech error and reaction time experiments

in picture description tasks, sentence production
tasks, and sentence completion tasks. In this framework, these mainly quantitative studies, taking into
account different variables, were able to addressee
very detailed features of the general theory.
As far as speech production is concerned, some of
the more debated points at the higher levels of grammar are the incremental vs. competitive nature of
grammatical encoding, the functioning of syntactic
encoding concerning hierarchical relations and word
order in sentence production, the nature of combinatorial information in grammatical encoding, and
phrasal ordering constraints in sentence production
(see Ferreira, 1996; Hartsuiker and Westenberg,
2000; Vigliocco and Nicol, 1998; Bock, 1986;
Pickering and Branigan, 1998; Stallings et al., 1998;
Yamashita and Chang, 2001; etc.).
As far as speech planning is concerned, a muchdiscussed topic is how far ahead speakers plan before
initiating utterances. Even if it is generally accepted
that speakers prepare utterances at different planning
levels in parallel while at the same time speaking, the
sizes of phonological and grammatical planning units
are still largely indefinite. Speech errors and hesitations provide some evidence that the clause is probably a key planning unit at the grammatical level, and
that the phonological planning units probably are
made up of only one or two words within a phrase.
In fact, word exchange typically switches words from
different phrases within the same clause (i.e., pauses
are more frequent between, as opposed to within,
clauses), while sound exchange typically affects
words from the same phrase, often adjacent words.
Yet much remains to be resolved: studies using various techniques to elicit utterances with dissimilar
syntactic structures reach different results (for an
overview of the problem, see Meyer, 1996: 477
481). Speakers probably employ planning units of
differing sizes according to the above circumstances
(see Schriefers, 1992). Therefore, the inconsistencies
among the results of different research could be
attributed to differences in the planning strategy of
the speakers.
In the same framework, another debated issue
concerns the cost of speech planning. Syntactic
planning should sustain processing costs in speech
production (just like any other process), but it is has
been difficult to empirically observe such costs. More
recently, besides hesitation in speech, this issue has
been approached using syntactic priming research.
The role of the conceptual planning stage has been
emphasized as a source of structural priming (Heydel
and Murray, 2000). Benefits in cost reduction to
the speaker have been hypothesized to underlie the

630 Planning Strategies and Production of Spoken Discourse

motivation for avoidance of syntactic planning and

encourage syntactic persistency. In the study by Smith
and Wheeldon (2001), for example, an on-line picture description task is used to investigate structural
priming, with the objective being to analyze the scope
and cost of syntactic structure generation.

Combining Perspectives
Up to this point in all the referenced works, the strict
control of planning (visually presenting the words to
be used) and production (time pressure during articulation) renders generalization to natural communicative situations tenuous: the production tasks used are
distinctly distant from natural production. Therefore,
the experimental conclusions would be reinforced
by more evidence from spontaneous natural language production. Combining observations taken
from natural situations and artificial controls applied
in a laboratory remains of the utmost importance to
advancing production theory.
Such a combination can be seen in the paper by
Branigan et al. (2000) that aims to connect the cost
reduction hypothesis to syntactic priming by examining coordination in dialogue. Do speakers in dialogue, regardless of lexical and semantic content,
have a tendency to coordinate the syntactic structures
of their contributions? Branigan et al. have established an original experimental technique in this
field called confederate scripting, which allows the
study of syntactic structure under controlled conditions in dialogue. Pairs of speakers took turns describing pictures to each other. One speaker (an accomplice
of the experimenter) produced pre-scripted descriptions designed to systematically vary in syntactic
structure. The natural production of the true experimental subject was analyzed to determine if it produced matching syntactic forms, the objective being
to investigate not only syntactic persistency in dialogue but also issues such as syntactic planning. Such
syntactic representations might be encoded as a component of lexical entries that are accessed during both
production and comprehension.
Of certain interest to this discussion are some crosslinguistic language production studies prompted by
the observation that most of what we know about the
cognitive processes underlying the production of spoken utterances has come from the study of English
speakers. Among them are Bates and Devescovi
(1989), who contrast the structural complexity of
Italian and English spoken utterances, and Holmes
(1995), who compares some of the message packaging devices used by speakers of two different
languages, specifically French and English. Direct
comparisons of production processes between English

and other languages are not common, although they

are critical for models of utterance production, i.e.,
allowing evaluation of language-specific or languagedependent characteristics. These studies include
simple comparisons of the production frequency of
distinct syntactic constructions in spontaneous speech
(Holmes, 1995; Bates and Devescovi, 1989) and the
rates of various hesitations produced by speakers
(Holmes, 1995). Three of the more interesting aspects
of these studies are the choice of data-spontaneous
spoken discourse, the choice of French, a presumably
topic-comment prominent language, and the choice
of the phenomena analyzed, such as topicalization,
cleft sentence, and so forth.

From Spontaneous Spoken Discourse to

Planning Strategies and Speech
Production
These studies are reminiscent of the more qualitatively inclined studies begun at the end of the 1970s in the
area of American functionalism that were focused on
message packaging and the flow of language and
thought (Chafe, 1976, 1979, 1994, 1998; Levy,
1979).
The objective of this research is to understand
how a speakers stream of thought is followed and
influenced by language and how language sheds light
on the nature of the thought process that lies behind
it. To this end, Chafe (1979) showed silent 7-minute
films to groups of different people and then interviewed them individually, asking them to describe
what happened in the film. He collected a large sample of quite natural verbalizations. A qualitative analysis of the interviews gave the author the opportunity
to present a model for the verbalization of recalled
experience: the Flow Model. In this model, a previous
vertical, hierarchical organization of thought and language is replaced and enriched by increased attention
to the horizontal aspect. The role of the speaker is
emphasized as being engaged in a real-time process of
focusing on sequential ideas and converting these
ideas, one after another, into language (see Chafe,
1979: 160166). Chafe (1976) takes into account
topicalization phenomena while introducing the notion of the topic as a premature subject based on
hypotheses of how speakers organize blocks of knowledge into sentences. Chafe has continued to further
refine his hypotheses (Chafe, 1994; 1998), placing
even more emphasis on the role of consciousness in
thought, language, and behavior. In the same framework, the fundamental thesis of Levy (1979) is that
the study of discourse must be approached in terms of
the mental activity of the speaker, referring to the
speaking process in which the speaker is engaged.

Planning Strategies and Production of Spoken Discourse 631

To gather data, Levy used techniques like approaching students just as they had finished class registration
to question them on why they decided to order their
schedules in the manner they did. Most the cohesive
properties of discourse, realized through the speakers
language resources, are best described by taking into
account the speakers mental states and process rather
than the structured text: in other words, the flow of
the speakers thought process is central to understanding and explaining the flow of the discourse.
Some syntactic ambiguities are not easily resolved
without accounting for the role of the mental structure involved in the speakers production of the
clause. For example, reference should be seen as a
strategic process in which the speaker does not
merely identify a particular object but rather constructs that object by choosing among a selection of
properties that are momentarily relevant (see Levy,
1979: 184190).
Other frameworks for considering speech planning
and production originate in the study of spontaneous
spoken language, which shows a large amount of
phenomena readily understandable when taking into
account the processes involved in the generation of
real texts (see, among others, Sornicola, 1979, 1981;
Milano, 2004). Sornicola (1979, 1981) analyses
large series of phenomena in spoken text, where real
speakers in actual dialogue present deviances from
that expected by ideal users, demonstrating how
such an approach can contribute to further modifying
the concept of appropriateness. The analysis of spoken discourse gives the author the opportunity to
discuss in detail some remarkable concerns of linguistic theory. Milano (2004) focuses on processes of
topicalization in spontaneous speech. The analysis
(at different levels of abstraction) provides some
insights on syntactic variation, topicalization phenomena, and characteristics of the text in which
they occur.
Further integration of research on spoken discourse
and speech planning and production, as well as research conducted with a shared viewpoint (the marriage of qualitative and quantitative approaches),
could only serve to greatly benefit that analysis and
improve the techniques employed to that end.
See also: Cognitive Science: Overview; Consciousness,
Thought and Language; Discourse Processing; Hesitation
Phenomena and Pauses.; Language of Thought; Pauses
and Hesitations: Psycholinguistic Approach; Phonology in
the Production of Words; Psycholinguistic Research Methods; Psycholinguistics: Overview; Speech Errors as Evidence in Phonology; Speech Errors: Psycholinguistic
Approach; Speech Production; Spoken Language Production: Psycholinguistic Approach.

Bibliography
Bates E & Devescovi A (1989). Crosslinguistic studies
of sentence production. In MacWhinney B & Bates E
(eds.) The crosslinguistic study of sentence processing.
Cambridge: Cambridge University Press. 225253.
Beattie G W (1979). Planning units in spontaneous
speech: some evidence from hesitation in speech and
speaker gaze direction in conversation. Linguistics 17,
61181.
Bock J K (1986). Syntactic persistence in language production. Cognitive Psychology 18, 355387.
Branigan H P, Pickering M J & Cleland A A (2000).
Syntactic co-ordination in dialogue. Cognition 75(2),
1325.
Chafe L (1976). Giviness, contrastiveness, definiteness,
subjects, topics and point of view. In Li C N (ed.) Subject
and topic. New York: Academic Press. 2556.
Chafe L (1979). The flow of language and thought. In
Givo n T (ed.) Discourse and syntax. New York: Academic
Press. 160181.
Chafe L (1994). Discourse, consciousness, and time: the
flow and displacement of conscious experience in
speaking and writing. Chicago: University of Chicago
Press.
Chafe L (1998). Language and the flow of thought. In
Tomasello M (ed.) The new psychology of language:
cognitive and functional approaches to language structure. Mahwah, NJ: Lawrence Erlbaum. 93111.
Ferreira V S (1996). Is it better to give than to donate?
Syntactic flexibility in language production. Journal of
Memory and Language 35(5), 724755.
Fromkin V A (1971). The non-anomalous nature of
anomalous utterances. Language 47(1), 2752.
Fromkin V A (1973). Speech errors as linguistic evidence.
The Hague: Mouton.
Garrett M (1980). The limits of accommodation: argument for independent processing levels in sentence production. In Fromkin V A (ed.) Errors in linguistic
performance: slips of the tongue, ear, pen, and hand.
New York: Academic Press.
Hartsuiker R J & Westenberg C (2000). Word order
priming in written and spoken sentence production.
Cognition 75(2), 2739.
Heydel M & Murray W (2000). Conceptual effects in
sentence priming: a cross-linguistic perspective. In De
Vincenzi M & Lombardo V (eds.) Cross-linguistic perspectives on language processing. Dordrecht: Kluwer
Academic. 227255.
Holmes V M (1995). A crosslinguistic comparison of the
production of utterances in discourse. Cognition 54(2),
169207.
Levelt W J M (1989). Speaking: from intention to articulation. Cambridge: MIT Press.
Levelt W J M, Roelofs A & Meyer A S (1999). A theory of
lexical access in speech production. Behavioral and
Brain Sciences 22(1), 138.
Levy D M (1979). Communicative goals and strategies:
between discourse and syntax. In Givo n T (ed.) Discourse and syntax. New York: Academic Press. 182195.

632 Planning Strategies and Production of Spoken Discourse

MacKay D G (1972). The structure of words and syllables:
evidence from errors in speech. Cognitive Psychology 3,
210227.
Meijer P J (1997). What speech errors can tell us about wordform generation: the roles of constraint and opportunity.
Journal of Psycholinguistic Research 26(1), 141158.
Meyer A S (1996). Lexical access in phrase and sentence
production: results from picture-word interference
experiments. Journal of Memory and Language 35(4),
477496.
Milano E (2004). Sulla variazione sintattica: invarianza
e variabilita` dei processi di topicalizzazione in italiano
parlato. Bollettino Linguistico Campano 3/4, 153176.
Pickering M J & Branigan H P (1998). The representation
of verbs: evidence from syntactic priming in language
production. Journal of Memory and Language 39(4),
633651.
Schriefers S H (1992). Lexical access in the production of
noun phrases. Cognition 45, 3354.

Smith M & Wheeldon L (2001). Syntactic priming in spoken sentence production an online study. Cognition
78(2), 123164.
Sornicola R (1979). Egocentric reference as a problem for
the theory of communication. Journal of Italian Linguistics 4, 764.
Sornicola R (1981). Il parlato. Bologna: Il Mulino.
Stallings L M, MacDonald M C & OSeaghdha P G (1998).
Phrasal ordering constraints in sentence production:
phrase length and verb disposition in heavy-NP shift
Journal of Memory and Language 39(3), 329417.
Svartvik J & Quirk R (eds.) (1980). A corpus of English
conversation. Lund: CWK Gleerup.
Vigliocco G & Nicol J (1998). Separating hierarchical relations and word order in language production: is proximity
concord syntactic or linear? Cognition 68(1), 1329.
Yamashita H & Chang F (2001). Long before short
preference in the production of a head-final language.
Cognition 81(2), 4555.

Planudes, Maximus (12601310)

S Matthaios, University of Cyprus, Nicosia, Cyprus
2006 Elsevier Ltd. All rights reserved.

Maximus Planudes was one of the most important

figures in Byzantine philology, grammar, and scholarship and was one of the main representatives of the
so-called Palaiologan Renaissance, the prosperous
period of the last two centuries of the Byzantine era.
According to the latest estimates, Planudes most likely lived between 1260 and 1310. He was born in
Nikomedia of Bithynia, Manuel Maximus was the
name he took when he became a monk and was
highly educated in Constantinople; his educational
status was reflected in his political efficiency and,
most of all, in his activity as a teacher, writer, and
scholar. Before becoming a monk the terminus post
quem for this event is the year 1283 he served as a
manuscript copyist and secretary in the imperial administration. Planudes remained involved with politics even during his monastic life. Due to his vast
knowledge of Latin, he participated in political missions; for example, he went on an embassy in Venice
in 1296 under the order of the emperor. As a monk
he became soon Father Superior (Z< gouB menoB) at the
Monastery of Five Saints at the foot of mountain
Auxentios. Initially at the Monastery of Chora, and
from the beginning of the 14th century on at the
Monastery of Akataleptos, Planudes became significantly involved in teaching. His school was like a
university in modern times. It was at first in the

Monastery of Chora and then in the Monastery of

Akataleptos.
Planudes left behind a rich and prodigious collection of writings (see Wendel, 1950 for an overview
and a description of his writings). Apart from his
theological works, the content of which was influenced by the unstable relationship of his period with
the Roman Catholic Church, and his literary writings
(poetry and prose), Planudes is also distinguished
for his grammar monographs and his extensive philological studies. The latter include a series of editions
and annotations of ancient Greek writers, anthologies, and collections of epigrams, proverbs, and rhetorical and scientific pamphlets (for an evaluation
of Planudes scholarly work, see Wilson, 1996: 230
241), as well as translations of Latin religious and
nonreligious authors into Greek (on the importance
of Planudes translations, see Schmitt, 1968). The
main body of his writings on grammar consists of
the following works: (1) Pee` geammatiknB dialogoB;
(Dialogue on grammar; edited by Bachmann, 1828:
II, 1101), a conversation between a teacher and his
pupil regarding the definition and scope of grammar
that is, of philology in ancient terms as well as
topics on morphology, orthography, syntax, and semantics in relation to certain parts of speech and
specific words. (2) The treatise PeeiA suntaxeoB toCn
tou
C luBgou meroCn (On the syntax of the parts of speech;
edited by Bachmann, 1828: II, 103166), which, following an introduction on the subject matter of syntax, deals with the syntax of the article, the pronoun,

632 Planning Strategies and Production of Spoken Discourse

MacKay D G (1972). The structure of words and syllables:
evidence from errors in speech. Cognitive Psychology 3,
210227.
Meijer P J (1997). What speech errors can tell us about wordform generation: the roles of constraint and opportunity.
Journal of Psycholinguistic Research 26(1), 141158.
Meyer A S (1996). Lexical access in phrase and sentence
production: results from picture-word interference
experiments. Journal of Memory and Language 35(4),
477496.
Milano E (2004). Sulla variazione sintattica: invarianza
e variabilita` dei processi di topicalizzazione in italiano
parlato. Bollettino Linguistico Campano 3/4, 153176.
Pickering M J & Branigan H P (1998). The representation
of verbs: evidence from syntactic priming in language
production. Journal of Memory and Language 39(4),
633651.
Schriefers S H (1992). Lexical access in the production of
noun phrases. Cognition 45, 3354.

Smith M & Wheeldon L (2001). Syntactic priming in spoken sentence production an online study. Cognition
78(2), 123164.
Sornicola R (1979). Egocentric reference as a problem for
the theory of communication. Journal of Italian Linguistics 4, 764.
Sornicola R (1981). Il parlato. Bologna: Il Mulino.
Stallings L M, MacDonald M C & OSeaghdha P G (1998).
Phrasal ordering constraints in sentence production:
phrase length and verb disposition in heavy-NP shift
Journal of Memory and Language 39(3), 329417.
Svartvik J & Quirk R (eds.) (1980). A corpus of English
conversation. Lund: CWK Gleerup.
Vigliocco G & Nicol J (1998). Separating hierarchical relations and word order in language production: is proximity
concord syntactic or linear? Cognition 68(1), 1329.
Yamashita H & Chang F (2001). Long before short
preference in the production of a head-final language.
Cognition 81(2), 4555.

Planudes, Maximus (12601310)

S Matthaios, University of Cyprus, Nicosia, Cyprus
2006 Elsevier Ltd. All rights reserved.

Maximus Planudes was one of the most important

figures in Byzantine philology, grammar, and scholarship and was one of the main representatives of the
so-called Palaiologan Renaissance, the prosperous
period of the last two centuries of the Byzantine era.
According to the latest estimates, Planudes most likely lived between 1260 and 1310. He was born in
Nikomedia of Bithynia, Manuel Maximus was the
name he took when he became a monk and was
highly educated in Constantinople; his educational
status was reflected in his political efficiency and,
most of all, in his activity as a teacher, writer, and
scholar. Before becoming a monk the terminus post
quem for this event is the year 1283 he served as a
manuscript copyist and secretary in the imperial administration. Planudes remained involved with politics even during his monastic life. Due to his vast
knowledge of Latin, he participated in political missions; for example, he went on an embassy in Venice
in 1296 under the order of the emperor. As a monk
he became soon Father Superior (Z< gouB menoB) at the
Monastery of Five Saints at the foot of mountain
Auxentios. Initially at the Monastery of Chora, and
from the beginning of the 14th century on at the
Monastery of Akataleptos, Planudes became significantly involved in teaching. His school was like a
university in modern times. It was at first in the

Monastery of Chora and then in the Monastery of

Akataleptos.
Planudes left behind a rich and prodigious collection of writings (see Wendel, 1950 for an overview
and a description of his writings). Apart from his
theological works, the content of which was influenced by the unstable relationship of his period with
the Roman Catholic Church, and his literary writings
(poetry and prose), Planudes is also distinguished
for his grammar monographs and his extensive philological studies. The latter include a series of editions
and annotations of ancient Greek writers, anthologies, and collections of epigrams, proverbs, and rhetorical and scientific pamphlets (for an evaluation
of Planudes scholarly work, see Wilson, 1996: 230
241), as well as translations of Latin religious and
nonreligious authors into Greek (on the importance
of Planudes translations, see Schmitt, 1968). The
main body of his writings on grammar consists of
the following works: (1) Pee` geammatiknB dialogoB;
(Dialogue on grammar; edited by Bachmann, 1828:
II, 1101), a conversation between a teacher and his
pupil regarding the definition and scope of grammar
that is, of philology in ancient terms as well as
topics on morphology, orthography, syntax, and semantics in relation to certain parts of speech and
specific words. (2) The treatise PeeiA suntaxeoB toCn
tou
C luBgou meroCn (On the syntax of the parts of speech;
edited by Bachmann, 1828: II, 103166), which, following an introduction on the subject matter of syntax, deals with the syntax of the article, the pronoun,

Plato and His Predecessors 633

the conjunction, and the preposition. This treatise is

modeled on and has as its sources Apollonius Dyscolus work PeeiA sunta xeoB (2nd century A.D.) and Priscianus Institutiones grammaticae (6th century A.D.)
(on the issue of sources, see Murru, 1979a, 1979b).
Despite his direct relationship with and attachment
to his ancient models, Planudes succeeded in overcoming the descriptive and normative character of
his contemporary grammaticography. Furthermore,
he made the theoretical discussion and argumentation
accessible to his Byzantine audience in a brief, precise,
and contemporary form and offered, as well as promoted, a deeper study of certain issues, such as case
and tense (for an evaluation of Planudes linguistic
contribution, see Robins, 1993: 201233). With his
prodigious profile and breadth of his writing, including his translations, Planudes played a crucial role in
and contributed to the strengthening of the cultural
and intellectual relations between the East and the
West. He was a pioneer of the spirit of the humanistic
movement and the Renaissance from the 15th century
onwards.

Bibliography
Bachmann L (ed.) (1828). Anecdota Graeca (2 vols).
Leipzig: Henrichs. (Reprinted Hildesheim: G. Olms,
1965).
Hunger H (1978). Die hochsprachliche profane Literatur
der Byzantiner (2 vols) (Handbuch der Altertumswissenschaft XII 5.12). Mu nchen: Beck.
Murru F (1979a). Planudea. Indogermanische Forschungen 84, 120131.
Murru F (1979b). Sullorigine della teoria localista di
Massimo Planude. LAntiquite Classique 48, 8297.
Robins R H (1993). The Byzantine grammarians. Their
place in history (Trends in Linguistics. Studies and Monographs 70). Berlin, New York: Mouton de Gruyter.
Schmitt W O (1968). Lateinische Literatur in Byzanz.
sterreichischen Byzantinischen GesellJahrbuch der O
schaft 17, 127147.
Wendel C (1950). Planudes, Maximos. In Paulys Realencyclopa die der classichen Altertumswissenschaft 20/2.
22022253.
Wilson N G (1996). Scholars of Byzantium (rev., edn.).
London: Gerald Duckworth & Co. Ltd.

See also: Classical Antiquity: Language Study; Grammar.

Plato and His Predecessors

R Bett, Johns Hopkins University, Baltimore, MD, USA
2006 Elsevier Ltd. All rights reserved.

Predecessors
The first attempts at explicit theorizing about language in ancient Greece occurred among the itinerant
teachers known as the Sophists (late 5th century
B.C.E.). Ideas about language are, however, at least
implicit in some texts prior to this. Parmenides (late
6thmid-5th century B.C.E.) famously declared the impossibility of speaking or thinking of what is not,
which seems to presuppose a view of meaning as
consisting in, or at least requiring, reference to something in the world. And a generation or so earlier
Heraclitus drew attention to the seeming paradox
that one of the words for a bow, a deadly weapon,
was bios, which is identical in spelling with the Greek
word for life (DK 22B48). This remark appears to
trade on the notion that one would expect a certain
fitness between words and their objects. And a similar
notion is apparent in the etymologizing that occurs
periodically in Homer and Hesiod; for instance,
Odysseus was named by his grandfather, who was

angry (odussamenos) towards many (Od. 19.407),

and also himself suffers the effects of the gods anger
(Od. 1.62).
It is, though, in the Sophistic period that language
first becomes the object of sustained reflection. Several texts from this period address the question of
the origins of language, sometimes in the context
of a broader account of the origins of society.
A consistent picture emerging from these texts is
that language was devised by humans themselves, in
response to the needs of early communities. It came
into being through a set of decisions about what
names should apply to what objects, and in that
sense is conventional rather than natural. These decisions were not necessarily seen as entirely arbitrary;
the Sophist Protagoras, in Platos dialogue of that
name (322a6), speaks of language as the product of
a techne , skill, analogous to building or agriculture
which implies that one set of names may be better
crafted (whatever that means) than another. But still,
according to this view, there were clearly many different possible and acceptable sets of names. The picture
is hardly sophisticated. However, it is the first attempt
in Greek thought to understand, in something like

Plato and His Predecessors 633

the conjunction, and the preposition. This treatise is

modeled on and has as its sources Apollonius Dyscolus work PeeiA suntaxeoB (2nd century A.D.) and Priscianus Institutiones grammaticae (6th century A.D.)
(on the issue of sources, see Murru, 1979a, 1979b).
Despite his direct relationship with and attachment
to his ancient models, Planudes succeeded in overcoming the descriptive and normative character of
his contemporary grammaticography. Furthermore,
he made the theoretical discussion and argumentation
accessible to his Byzantine audience in a brief, precise,
and contemporary form and offered, as well as promoted, a deeper study of certain issues, such as case
and tense (for an evaluation of Planudes linguistic
contribution, see Robins, 1993: 201233). With his
prodigious profile and breadth of his writing, including his translations, Planudes played a crucial role in
and contributed to the strengthening of the cultural
and intellectual relations between the East and the
West. He was a pioneer of the spirit of the humanistic
movement and the Renaissance from the 15th century
onwards.

Bibliography
Bachmann L (ed.) (1828). Anecdota Graeca (2 vols).
Leipzig: Henrichs. (Reprinted Hildesheim: G. Olms,
1965).
Hunger H (1978). Die hochsprachliche profane Literatur
der Byzantiner (2 vols) (Handbuch der Altertumswissenschaft XII 5.12). Munchen: Beck.
Murru F (1979a). Planudea. Indogermanische Forschungen 84, 120131.
Murru F (1979b). Sullorigine della teoria localista di
Massimo Planude. LAntiquite Classique 48, 8297.
Robins R H (1993). The Byzantine grammarians. Their
place in history (Trends in Linguistics. Studies and Monographs 70). Berlin, New York: Mouton de Gruyter.
Schmitt W O (1968). Lateinische Literatur in Byzanz.
sterreichischen Byzantinischen GesellJahrbuch der O
schaft 17, 127147.
Wendel C (1950). Planudes, Maximos. In Paulys Realencyclopadie der classichen Altertumswissenschaft 20/2.
22022253.
Wilson N G (1996). Scholars of Byzantium (rev., edn.).
London: Gerald Duckworth & Co. Ltd.

See also: Classical Antiquity: Language Study; Grammar.

Plato and His Predecessors

R Bett, Johns Hopkins University, Baltimore, MD, USA
2006 Elsevier Ltd. All rights reserved.

Predecessors
The first attempts at explicit theorizing about language in ancient Greece occurred among the itinerant
teachers known as the Sophists (late 5th century
B.C.E.). Ideas about language are, however, at least
implicit in some texts prior to this. Parmenides (late
6thmid-5th century B.C.E.) famously declared the impossibility of speaking or thinking of what is not,
which seems to presuppose a view of meaning as
consisting in, or at least requiring, reference to something in the world. And a generation or so earlier
Heraclitus drew attention to the seeming paradox
that one of the words for a bow, a deadly weapon,
was bios, which is identical in spelling with the Greek
word for life (DK 22B48). This remark appears to
trade on the notion that one would expect a certain
fitness between words and their objects. And a similar
notion is apparent in the etymologizing that occurs
periodically in Homer and Hesiod; for instance,
Odysseus was named by his grandfather, who was

angry (odussamenos) towards many (Od. 19.407),

and also himself suffers the effects of the gods anger
(Od. 1.62).
It is, though, in the Sophistic period that language
first becomes the object of sustained reflection. Several texts from this period address the question of
the origins of language, sometimes in the context
of a broader account of the origins of society.
A consistent picture emerging from these texts is
that language was devised by humans themselves, in
response to the needs of early communities. It came
into being through a set of decisions about what
names should apply to what objects, and in that
sense is conventional rather than natural. These decisions were not necessarily seen as entirely arbitrary;
the Sophist Protagoras, in Platos dialogue of that
name (322a6), speaks of language as the product of
a techne, skill, analogous to building or agriculture
which implies that one set of names may be better
crafted (whatever that means) than another. But still,
according to this view, there were clearly many different possible and acceptable sets of names. The picture
is hardly sophisticated. However, it is the first attempt
in Greek thought to understand, in something like

634 Plato and His Predecessors

the spirit of a social scientist as opposed, say, to

appealing to divine fiat how language could have
come about.
The question of constraints on the naming of
objects was further pursued by the Sophists in a
field of inquiry known as orthote s onomato n, correctness of names. It is hard to reconstruct precisely what this involved, but it seems to have
been concerned, in some sense, with making the fit
between words and the world as good as possible;
it thus shares something with the pretheoretical
concerns implied by Homer and Heraclitus. Two
Sophists, Protagoras and Prodicus, were especially
associated with this subject. Protagoras is said
to have distinguished the grammatical genders of
words, and to have been willing to criticize existing
genders of words as not appropriate to the nature of
their objects; pe le x helmet, for example, should be
masculine, not feminine. He also drew distinctions
among what would now be called speech acts entreaty, question, answer, and command and was
again ready to correct peoples choices among these
modes of discourse. Prodicus, on the other hand,
was especially concerned with precise definitions of
words, including precise distinctions among nearly
synonymous words; reports (and parodies) of this
are common in Plato and elsewhere.
Despite Protagorass anticipation of the notion of
speech acts, the overriding conception of language in
this period was as consisting purely of names applying to objects. And this led to some extraordinary
claims about language. Several Sophists and their
contemporaries are reported to have argued that
falsehood is impossible, that there is no such thing
as contradiction, or both. The essential point of these
arguments is that there are only two possible options:
to speak the truth or not to speak at all. The connection with the naming conception of language is clear.
If all language does is name, then truth can only
consist of successfully naming an object or state
of affairs. But in that case, there is no prospect of
successfully referring to something, yet saying something false about it, or of two people successfully
referring to the same object, yet saying contradictory
things about it. The only alternative to successful
naming that is, truth on this conception is unsuccessful naming; but if naming is all language does,
unsuccessful naming is not genuine language at all.
One other text from this era deserves mention.
The Sophist Gorgias wrote a treatise entitled On
What Is Not (Peri tou me ontos), in which, again,
extraordinary conclusions were argued for, namely:
(1) there is nothing, (2) even if there were anything,
it could not be known about, and (3) even if it could
be known about, this knowledge could not be

communicated to anyone else. There is much dispute

about the purpose of this production, although
Parmenides is clearly in some way the target. But the
third part, on the impossibility of communication, is
of particular interest for the study of language. The
main arguments in this part were, first, that words are
wholly different types of entity from the objects that
they name, and second, that the contents of any two
peoples minds are necessarily different. We need not
suppose that Gorgias actually believed communication
was impossible. However, it is clear that he was raising
(without attempting to answer) the deep question of
how words manage to refer to objects at all.

Plato
The Sophists ideas on language form the context for
the reflections on this subject by Plato (mid-420s347
B.C.E.). In a number of dialogues, widely (though by no
means universally) regarded as from early in Platos
career, Socrates and his interlocutors are shown
trying to answer questions of the form What is F?,
where F stands for some significant ethical characteristic such as courage, piety, or virtue in general. This
enterprise is not fundamentally linguistic; even
though it is often described as a search for definitions,
Socrates is not interested primarily in the meanings of
terms, but in the real natures of the items that those
terms refer to. Nonetheless, the search is understood
to require sensitivity to how the terms in question are
actually used, and in this respect it seems to owe
something to Prodicuss interest in precise distinctions
of meaning. Socrates in these dialogues occasionally
professes allegiance to Prodicus; the professions are
never entirely serious, but they are not without some
basis in his actual procedure.
Plato takes up the issue of correctness of names in
his dialogue Cratylus. Here Socrates is made to examine two opposing (and extreme) views. One view,
that of Hermogenes, is that linguistic correctness is
purely a matter of convention. Indeed, Hermogenes
even ignores the connotations of society-wide agreement normally present in the word convention
(nomos), suggesting that each individual is free to
use words in whatever manner he or she decides.
Against this, Socrates argues that if one accepts a
view of reality as fixed to which Hermogenes readily assents one cannot regard correctness in language
as a matter of merely arbitrary choice. Cratylus, on
the other hand, holds the view that some names are
naturally correct and some are not; the correct ones
fit the nature of the things they refer to, while
the incorrect ones do not. A trivial example is that
Hermogenes is not Hermogenes true name, because
he is not in fact the son of the god Hermes. But it is of

Plato and His Predecessors 635

course a much more difficult question what, in general, the natural fitness of names might consist in. The
view explored by Socrates and Cratylus is that most
names can be analyzed etymologically into a small
number of primary names, and that these primary
names are correct in virtue of a natural resemblance
between their sounds and their objects. It is not clear
how far Plato means us to be attracted by this theory.
Socrates finds much to admire in it. But he argues that
a role for convention in language cannot be altogether
excluded. He is also dissatisfied with the fact that, in
the version of the theory promoted by Cratylus, the
originators of these names were Heracliteans who
held that everything was in a state of constant change
a view that he, Socrates, cannot accept. Whatever
we are ultimately supposed to think about the theory,
however, the examination of the purported etymologies occupies a large proportion of the dialogue, suggesting that Plato thinks it deserves very serious
consideration. Given the history of etymologizing
mentioned earlier a history to which the discussion
seems sometimes to allude this is perhaps not
surprising, however frustrating and alien this portion
of the dialogue may seem to us.
Another significant point that Socrates makes in
Cratylus is that an understanding of things is more
important than, and indeed indispensable for, the
optimal naming of those things. And, as Cratylus
and other dialogues make clear, the most important
things, in the mature Platonic conception, are the
unchanging, purely intelligible Forms entities such
as Beauty itself or Goodness itself, as opposed to any
of the particular items we might call beautiful or
good. Now these Forms are not, of course, linguistic
entities. But it is plausible to suppose that one of the
reasons Plato had for believing in Forms was a linguistic reason. Given the difficulty of the Socratic
search for answers to What is F? questions about
the virtues, it is natural to wonder how we manage to
understand and use terms such as piety, courage, or
virtue at all. Even though we seem to have no difficulty employing these terms in ordinary discourse,
Socratess attempts to pin down what exactly they
refer to consistently end in failure. It might well
have seemed an attractive solution to this puzzle to
suppose that our understanding of such terms derives
from our (no doubt incomplete) grasp of the
corresponding Forms, which are only imperfectly
exemplified in the world around us; if Forms are
what the terms really refer to, then the Socratic
search, focused as it is on ordinary instances, is
bound to fail. Of course, the project of understanding
the Forms is hardly less ambitious, as Plato readily
concedes.

Plato also tackles the question of how falsehood

(along with contradiction) is possible. And here
he makes what should be considered a definitive
breakthrough. In the Theaetetus the characters are
shown struggling unsuccessfully with the question
of how false judgment is possible. Socrates also
proposes a model of language that involves names
being woven together into an account (logos,
202b), and this seems to promise an advance beyond
the Sophistic conception of language as doing nothing
but naming. But the promise is not fulfilled; no linguistic items are recognized other than names, and
the implications of structure present in the notion of
a weaving together (sumploke ) are not followed
through. In the Sophist, however, Plato has the central
character (a visitor from Parmenidess hometown,
Elea) draw a crucial distinction between a name
(onoma) and a thing said (re ma): the name identifies
the thing being spoken of, and the thing said delivers
information about it. Once this distinction is in place,
it is easy to see how both true and false statements
may be made about the same objects. The dialogue
involves much broader discussion about the legitimacy of thinking or speaking of what is not; and the term
speaking of what is not regularly refers, among
other things, to the speaking of falsehoods. But once
the possibility of this is established (against Parmenides) in general terms, the specifically linguistic
details of how falsehood is possible are established
relatively quickly. The importance of moving beyond
the naming conception of language can hardly be
overestimated; until it was clearly understood that
language does more than just naming, the study of
grammar, syntax, or anything else to do with the
structure of language was out of the question.
See also: Classical Antiquity: Language Study; Greek
Lexicography, Classical.

Bibliography
Ackrill J L (1999). Language and reality in Platos Cratylus. In Fine G (ed.) Plato 1: metaphysics and epistemology. Oxford: Oxford University Press. 125142.
Barnes J (1982). The Presocratic philosophers. (rev. edn.).
London: Routledge & Kegan Paul.
Bostock D (1994). Plato on understanding language. In
Everson (ed.). 1027.
Broadie S (2003). The sophists and Socrates. In Sedley D
(ed.) The Cambridge companion to Greek and Roman
philosophy. Cambridge: Cambridge University Press.
7397.
Denyer N (1991). Language, thought and falsehood in
ancient Greek philosophy. London: Routledge.

636 Plato and His Predecessors

Everson S (ed.) (1994). Language. Cambridge: Cambridge
University Press.
Frede M (1992). Platos Sophist on false statements. In
Kraut R (ed.) The Cambridge companion to Plato. Cambridge: Cambridge University Press. 397424.
Dillon J & Gergel T (trans.). (2003). The Greek Sophists.
London: Penguin.
Kerferd G B (1981). The Sophistic movement. Cambridge:
Cambridge University Press.
Mourelatos A (1987). Gorgias on the function of language. Philosophical Topics 15, 135170.
Plato (1993). Sophist. White N P (trans.). Indianapolis:
Hackett.

Plato (1998). Cratylus. Reeve C D C (trans.). Indianapolis:

Hackett.
Sedley D (2003). Platos Cratylus. Cambridge: Cambridge
University Press.
Sluiter I (1997). The Greek tradition. In Van Bekkum W,
Houben J, Sluiter I & Versteegh K (eds.) The emergence
of semantics in four linguistic traditions: Hebrew,
Sanskrit, Greek, Arabic. Amsterdam/Philadelphia: John
Benjamins. 147224.
Sprague R K (ed.) (2001). The older Sophists. Indianapolis:
Hackett.
Williams B (1994). Cratylus theory of names and its
refutation. In Everson (ed.). 2836.

Platos Cratylus and Its Legacy

J E Joseph, University of Edinburgh, Edinburgh, UK
2006 Elsevier Ltd. All rights reserved.

Platos Cratylus is the first complete surviving philosophical inquiry into the arbitrariness of language,
and arguably the greatest. Although it was long
thought to have been an early work of Platos on
account of being merely about language, heightened
appreciation of its importance over recent decades
has prompted scholars to redate it to his great middle period. It takes the form of a debate among three
of Platos teachers, Cratylus, Hermogenes, and
Socrates, about the correctness of words (the subject
taught by Cratylus). Cratylus and Hermogenes are
arguing the question within the Sophists staple pedagogical dichotomy of physis nature versus nomos
convention. Socrates joins them, and they invite
him to adjudicate as to who is right: Cratylus, who
holds that a word is correct only if naturally
connected to its meaning, or Hermogenes, who thinks
that any word can designate anything just as well as
any other.
They begin by considering proper names, which
seem obviously conventional since chosen willfully
for individuals, usually by their parents although,
significantly, Plato had changed his own name (originally Aristocles) and then proceed to common
nouns, the choice of which is lost in prehistory. In
both cases, Hermogenes takes the conventionalist
view and Cratylus the naturalist one. But Cratylus
sounds absurd in the proper-name context when he
holds that Hermogenes isnt really his interlocutors
name because he isnt actually born of Hermes
(which would imply that he is lucky and eloquent,
when in fact he is neither). And Hermogenes sounds
equally absurd in the common-noun context when he

holds that it would make no difference if someone

referred to a man as horse and to a horse as man.
Socrates raises the objection that, were this latter
argument the case, there would be no way of distinguishing truth from falsehood, a position taken by
relativists like Protagoras but anathema to Socrates
(and Plato), for whom things have a natural reality of
their own that does not depend on the perception of
individuals. Socrates lays out the characteristically
Platonic view that the things we perceive in the
world around us are not really real, in the sense
that they are not permanent but in flux, are not the
wholes we perceive but conglomerations of atoms,
and are not perceived in the same way by all of us.
The table at which I am now writing was not a table
30 years ago but has been made up from what was
then parts of trees, which themselves did not exist 100
years ago, and in time the table will burn or decay, its
substance taking on yet another form. The meaning
of the word table, the knowledge I have of what a
table is, is not this transient thing. On the contrary,
actual tables are made in accordance with the function of a table, which requires it to have a particular
shape, that of a flat top surface supported by legs.
This is what Plato calls the idea of a table, what
defines the ideal form of a table. That ideal form is the
true, unchanging reality. Any individual table is merely an attempt to realize that ideal form in transient
matter. The meaning of the word table is that idea,
not any particular material instantiation of it. Moreover, ideas are all that we can know, since knowledge must be of permanent, unchanging things, as
opposed to the perception or opinion we can have
of material things. We may disagree over whether a
particular table is beautiful, but we cannot know that
it is or is not beautiful in the same way that we can
know what a table is and what beauty is although,

636 Plato and His Predecessors

Everson S (ed.) (1994). Language. Cambridge: Cambridge
University Press.
Frede M (1992). Platos Sophist on false statements. In
Kraut R (ed.) The Cambridge companion to Plato. Cambridge: Cambridge University Press. 397424.
Dillon J & Gergel T (trans.). (2003). The Greek Sophists.
London: Penguin.
Kerferd G B (1981). The Sophistic movement. Cambridge:
Cambridge University Press.
Mourelatos A (1987). Gorgias on the function of language. Philosophical Topics 15, 135170.
Plato (1993). Sophist. White N P (trans.). Indianapolis:
Hackett.

Plato (1998). Cratylus. Reeve C D C (trans.). Indianapolis:

Hackett.
Sedley D (2003). Platos Cratylus. Cambridge: Cambridge
University Press.
Sluiter I (1997). The Greek tradition. In Van Bekkum W,
Houben J, Sluiter I & Versteegh K (eds.) The emergence
of semantics in four linguistic traditions: Hebrew,
Sanskrit, Greek, Arabic. Amsterdam/Philadelphia: John
Benjamins. 147224.
Sprague R K (ed.) (2001). The older Sophists. Indianapolis:
Hackett.
Williams B (1994). Cratylus theory of names and its
refutation. In Everson (ed.). 2836.

Platos Cratylus and Its Legacy

J E Joseph, University of Edinburgh, Edinburgh, UK
2006 Elsevier Ltd. All rights reserved.

Platos Cratylus is the first complete surviving philosophical inquiry into the arbitrariness of language,
and arguably the greatest. Although it was long
thought to have been an early work of Platos on
account of being merely about language, heightened
appreciation of its importance over recent decades
has prompted scholars to redate it to his great middle period. It takes the form of a debate among three
of Platos teachers, Cratylus, Hermogenes, and
Socrates, about the correctness of words (the subject
taught by Cratylus). Cratylus and Hermogenes are
arguing the question within the Sophists staple pedagogical dichotomy of physis nature versus nomos
convention. Socrates joins them, and they invite
him to adjudicate as to who is right: Cratylus, who
holds that a word is correct only if naturally
connected to its meaning, or Hermogenes, who thinks
that any word can designate anything just as well as
any other.
They begin by considering proper names, which
seem obviously conventional since chosen willfully
for individuals, usually by their parents although,
significantly, Plato had changed his own name (originally Aristocles) and then proceed to common
nouns, the choice of which is lost in prehistory. In
both cases, Hermogenes takes the conventionalist
view and Cratylus the naturalist one. But Cratylus
sounds absurd in the proper-name context when he
holds that Hermogenes isnt really his interlocutors
name because he isnt actually born of Hermes
(which would imply that he is lucky and eloquent,
when in fact he is neither). And Hermogenes sounds
equally absurd in the common-noun context when he

holds that it would make no difference if someone

referred to a man as horse and to a horse as man.
Socrates raises the objection that, were this latter
argument the case, there would be no way of distinguishing truth from falsehood, a position taken by
relativists like Protagoras but anathema to Socrates
(and Plato), for whom things have a natural reality of
their own that does not depend on the perception of
individuals. Socrates lays out the characteristically
Platonic view that the things we perceive in the
world around us are not really real, in the sense
that they are not permanent but in flux, are not the
wholes we perceive but conglomerations of atoms,
and are not perceived in the same way by all of us.
The table at which I am now writing was not a table
30 years ago but has been made up from what was
then parts of trees, which themselves did not exist 100
years ago, and in time the table will burn or decay, its
substance taking on yet another form. The meaning
of the word table, the knowledge I have of what a
table is, is not this transient thing. On the contrary,
actual tables are made in accordance with the function of a table, which requires it to have a particular
shape, that of a flat top surface supported by legs.
This is what Plato calls the idea of a table, what
defines the ideal form of a table. That ideal form is the
true, unchanging reality. Any individual table is merely an attempt to realize that ideal form in transient
matter. The meaning of the word table is that idea,
not any particular material instantiation of it. Moreover, ideas are all that we can know, since knowledge must be of permanent, unchanging things, as
opposed to the perception or opinion we can have
of material things. We may disagree over whether a
particular table is beautiful, but we cannot know that
it is or is not beautiful in the same way that we can
know what a table is and what beauty is although,

Platos Cratylus and Its Legacy 637

actually, such knowledge is not accessible to just anyone. The Ideal Forms inhabit a heaven into which
only the philosopher, the wisest of men, can see
(hence the political stance of Platos Republic, where
the ideal ruler is the philosopher-king).
From this point on, Socrates directs all his questions to Hermogenes, with Cratylus silent until quite
late in the dialogue. In response to Socratess point
about truth, Hermogenes raises a powerful objection:
If truth depends on some kind of natural relationship between word and thing, how is it possible for
different languages to exist? Socrates does not attempt to answer this directly (a sign of how seriously
Plato took the question), but steers the dialogue off in
the direction that the question demands. He asks
Hermogenes about the purpose of words, and they
conclude that words exist for two reasons: to discriminate among things, i.e., to pick out their ousia, the
true essence that belongs to them alone; and to transmit that knowledge from the few who can perceive it
directly to the many who cannot.
This leads Socrates to ask about the origins of the
words we use, paving the way to the etymological
inquiry that will form the great central bulk of the
dialogue. Etymology, which in Greek means the study
of truth, was one of the sciences based on language
in which instruction was offered by Socratess contemporaries the Sophists. The most successful and
widely sought after of these was rhetoric, the art of
using language in order to persuade, persuasion being
the ultimate political commodity in the democracy of
Athens. Just as they had no faith in democracy,
Socrates and Plato disdained rhetoric as mere wordplay, not at all concerned with real knowledge. Etymology, however, had more of an appeal. It not only
claimed to be the study of truth, but did offer real
insights into what words meant in an earlier time,
closer to the moment of their creation. The great
question, Socrates says, is whether whoever made
the words we use the semimythical nomothetes,
which also means lawgiver really perceived the
true essence of the thing he was naming, and if
so, whether he succeeded in the word makers craft
of mimesis, imitation of that essence in the sounds
of language. This provides, in theory at least, the
answer to Hermogeness question about how it is
possible for different languages to exist unless words
are purely conventional: any number of correct
words are conceivable to designate a given meaning
as long as they capture its essence and make it plain.
As Socrates proceeds through various classes of
words and their etymologies, the thesis emerges
that the creation of the Greek lexicon took place
under the influence of Heraclituss doctrine that
everything flows. Accounts are given whereby

words of a positive moral sense are traced back to

roots and sounds indicating motion, while those of a
negative moral sense have roots expressing immobility. But all sorts of problems ensue. Socrates points
out how easy it is, when no such account suggests
itself, to have recourse to some ad hoc explanation,
such as attributing a word to foreign origin. When it
comes to mimesis, he notes that r, a trilled consonant
in Greek, seems naturally to denote motion by the
very way it is produced, and indeed appears in many
of the words with positive qualities that he traces
back to the idea of motion yet it does not appear
in the word for motion itself, kinesis. Socrates also
cites cases of words known to have undergone sound
change, which he puts down to people who care
nothing about the truth, only about how they shape
their mouths. This is a swipe not just at the hoi polloi
in general but especially at rhetoricians and the poets
who along with them are banned from Platos ideal
Republic.
He finally turns back to Cratylus to discuss another
fundamental problem with words being naturally
connected to their meaning. Cratylus has insisted
that the meaning of a correct word must not only
be embodied directly in the sounds of the word but
also be indistinguishable from the idea of what it
designates. Socrates leads him to admit that this is
not in fact the nature of mimesis if the image of
something were identical to that something, it would
not be an image at all but would actually have become the thing itself. Instead, the art of mimesis in
language is to capture and reproduce some part of
the essence of things; indeed, Socrates affirms, it is
precisely in this way that a language should be
constructed and used. But he admits, once constructed, the language passes into the hands of the
many, who care only about the vulgar function of
communicating with their fellows, for which conventional words (not correct ones in Cratyluss sense)
suffice. The investigation reaches an impasse, and
Socrates concludes by opining that the Greek language really was created under the Heraclitean
philosophy, but that the philosophy itself is misguided; that the lawgiver did not consistently
embody its principles in sound; and that, even when
he did so, words have subsequently undergone change
at the hands of the mob, who care only about
communication and sounding nice.
As a result, the dialogue ends up rejecting the possibility that the study of language opens a path to
understanding the true nature of the universe. For
each word, one would first have to decide whether it
was created with an eye to truth, whether that truth
was properly and consistently imitated in sounds, and
whether the original form of the sounds has remained

638 Platos Cratylus and Its Legacy

unaltered. But the first of these decisions already

demands that we know what the truth is independently of the word in question and if we can do that,
there is no point inquiring further. This is clearly
disappointing to Socrates, who believes that words
should be naturally and deterministically bound to
their meanings, but he is forced to admit that, at
least as they are used for the vulgar purpose of communication, they are not in fact so bound. The way to
knowledge of the truth is not therefore through the
study of etymology but through the kind of dialectic
inquiry that the Cratylus itself embodies.
This is a blinding condemnation of the futility of
any language-based science. Yet so intense and sustained is Socratess attempt to disprove Hermogeness
conventionalist stance through an examination of
the hidden etymological structure of words that,
although his attempt is ultimately given up as unsuccessful, the dialogue nevertheless has been read for
centuries as a detailed study of Greek etymology. It
was even interpreted as coming down on the side of
Cratylus, holding that language really is grounded in
nature, when in fact it makes clear that the positions
taken by both Cratylus and Hermogenes are unsustainable in isolation from one another and from deeper inquiry into the nature of word making and
mimesis.
The misinterpretations are less surprising when one
considers how powerful and universal is the impulse
to interpret language as having a meaning far deeper
than what appears on the surface. This is nowhere
more true than in European culture, formed by the
doctrine of a god who is Himself logos, the Word, the
language-based and language-like intelligence that
created the world and continues to order it. From
medieval through modern linguistic thought we find
various attempts at containing arbitrariness, for instance, by locating it in nouns only, as opposed to
verbs and other more functional parts of speech; in
the lexicon as a whole, as opposed to grammar; in
conventional signs, which are imagined as having
come about as an accretion onto more original and
basic natural signs; in a universal grammar that underlies all languages, which differ from one another
only in relatively superficial particulars.
Modern secular culture has not given up its faith in
language but rather has problematized and debated
its centrality in various guises that can be read as
reworkings of the problem outlined and investigated
in the Cratylus. Modern linguistics is predicated upon

the belief in something deeper than the actual words

we speak and hear, wherein their real structure lies,
and perhaps also their real meaning. We find this
impulse in Humboldts energeia, in Jakobsons functionalism, in Chomskys universalism, and in various
attempts to portray thought as conditioned by meaning, including the SapirWhorf hypothesis at the
micro level of words, and the regimes of language
described by present-day linguistic anthropologists
and discourse analysts (in the wake of Nietzsche
and Foucault) at the macro level of discourse. Even
Saussure, with whom the modern doctrine of the
arbitrariness of linguistic signs is most closely associated, taught that everything having to do with
the systematic nature of languages and languages
are entirely systematic by his account needs to be
approached from the point of view of limiting
the arbitrary.
The legacy of the Cratylus has yet to play itself
out. Once linguists manage to confront the powerful
challenge it puts to the whole enterprise, we can
perhaps begin to approach an understanding of
language complex enough to take account of how
Hermogenes, Cratylus, and Socrates are actually
each right and wrong, and then to formulate, for the
first time, new questions that actually pass beyond the
concerns of the ancient dialogue rather than endlessly
recycling them.

See also: Naturalism; Nominalism; Plato and His Predecessors; Realism and Antirealism.

Bibliography
Baxter T M S (1992). The Cratylus: Platos critique of
naming. Leiden: E. J. Brill.
Joseph J E (2000). Limiting the arbitrary: linguistic naturalism and its opposites in Platos Cratylus and modern
theories of language. Amsterdam & Philadelphia: John
Benjamins.
Kretzmann N (1967). History of semantics. In Edwards P
(ed.) Encyclopedia of philosophy, vol. 7. New York:
Macmillan. 358406.
Sedley D (1998). The etymologies in Platos Cratylus.
Journal of Hellenic Studies 118, 140154.
Sluiter I (1997). The Greek tradition. In ven Bekkum W,
Houben J, Sluiter I & Versteegh K (eds.) The emergence
of semantics in four linguistic traditions: Hebrew,
Sanskrit, Greek, Arabic. Amsterdam & Philadelphia:
John Benjamins. 147224.

Pliny the Elder (2379 A.D.) 639

Pliny the Elder (2379 A.D.)

T Fogen, Humboldt-Universitat, Berlin, Germany
2006 Elsevier Ltd. All rights reserved.

C. Plinius Secundus (Plinius maior), Roman equestrian, civil servant and officer, historian, rhetorician,
encyclopedist, born in Novum Comum (Como) in
23 A.D., died in 79 A.D. during the eruption of Mount
Vesuvius, as reported by his nephew Pliny the Younger (Epist. 6.16, 6.20). Under the emperor Claudius,
he was employed as an officer and financial administrator and held further administrative posts under
Vespasian (for details, see Konig and Winkler, 1979;
Serbat, 1986: 20732077; Sallmann, 2000: 1135
1136), but at the same time he was a very prolific
author of numerous treatises on history (De iaculatione equestri [On the use of the throwing-spear by
cavalrymen], Bella Germaniae in 20 books, A fine
Aufidii Bassi historiae in 31 books), grammar and
rhetoric, as well as science. Most of his works are
lost or preserved only in fragments, with the exception of the monumental and encyclopedic Naturalis
historia in 37 books (Beagon, 1992; Healy, 1999;
Sallmann, 2000: 11381140; Naas, 2002; Murphy,
2004), which had an enormous impact on later periods (Serbat, 1986: 21702183), even though Pliny
has frequently been criticized for the compilatory
nature of his work and for his lack of original
thought. Pliny the Younger, who in one of his letters
compiled a catalogue of his uncles and adoptive parents works (Epist. 3.5; cf. Suetonius, De hist. fr. 80
Reifferscheidt), described Pliny the Elder as an untiring scholar who devoted an enormous amount of time
to his multifaceted research activities.
To Pliny the Elders grammatical, rhetorical, and
biographical works belong three treatises of which
only fragments have survived or which are lost: (1)
De vita Pomponii Secundi in two books, a biography
of his friend Pomponius Secundus, military commander and dramatist; (2) Studiosus in three books
on the education of the orator, apparently soon
replaced by Quintilians Institutio oratoria; and (3)
Dubius sermo in eight books, which is referred to in
the preface to the Naturalis historia (praef. 28) and
can be dated around 67 A.D. The majority (more than
60%) of the Dubius sermo fragments, most of which
belong to Book 6, were preserved by Charisius via
Iulius Romanus. It is difficult to reconstruct the work
as a whole, but it can be said with some certainty that
it was a collection of words and word forms whose
grammatical correctness, including their orthography, was open to debate. In order to determine the
status of certain forms, Pliny applied either analogy

(ratio) or usage (consuetudo) as normative principles,

depending on the individual case; the authority of
exemplary writers (auctoritas), euphony (suavitas),
and antiquity (consuetudo veterum) played a role
for him in the discussion of linguistic norms as well
(cf. Fogen, 1998, 1999). Pliny also dealt with the
quality of sounds and questions of sound change.
The fragments demonstrate that in his own work he
often quoted passages from Varro and from C. Iulius
Caesars treatise De analogia. In general, he does not
seem to have had the highest opinion of grammarians,
as he condemned their tendency toward oversubtlety
and pedantry (Nat. hist. 35.13: perversa grammaticorum subtilitas); this judgment might have been
caused by his rejection of the all too strict linguistic
normativism of some school grammarians.
See also: Aristotle and the Stoics on Language; Classical

Antiquity: Language Study; Greece: Language Situation;

Greek, Ancient; Italy: Language Situation; Language Attitudes; Language Change and Language Contact; Language Education: Correctness and Purism; Language
Education: Grammar; Latin; Normativity; Norms and Correctness; Roman Ars Grammatica; Varro, Marcus Terentius
(11627 B.C.).

Bibliography
Barwick K (1922). Remmius Palaemon und die romische
ars grammatica. Leipzig: Dieterich.
Beagon M (1992). Roman nature. The thought of Pliny the
Elder. Oxford: Clarendon.
Beck J W (ed.) (1894). C. Plinii Secundi librorum dubii
sermonis VIII reliquiae. Leipzig: Teubner.
Citroni Marchetti S (1991). Plinio il Vecchio e la tradizione
del moralismo romano. Pisa: Giardini.
Della Casa A (1969). Il Dubius sermo di Plinio. Genova:
Istituto di filologia classica e medioevale.
Della Casa A (1982). Plinio grammatico. In Plinio il Vecchio sotto il profilo storico e letterario. Atti del Convegno
di Como 5/6/7 ottobre 1979. Atti della Tavola rotonda
nella ricorrenza centenaria della morte di Plinio il Vecchio, Bologna, 16 dicembre 1979. Como. 109115.
Detlefsen D (1867). Zur Flexionslehre des alteren Plinius.
In Symbola Philologorum Bonnensium in honorem
Friderici Ritschelii collecta, vol. 2. Leipzig: Teubner.
695714.
Fogen T (1998). Bezuge zwischen antiker und moderner
Sprachnormentheorie. Listy filologicke 121, 199219.
Fogen T (1999). Spracheinstellungen und Sprachnormbewutsein bei Cicero. Glotta 75, 133.
Healy J F (1999). Pliny the Elder on science and technology.
Oxford: Oxford University Press.
Holtz L (1987). Pline et les grammairiens: Le Dubius
sermo dans le haut moyen age. In Pigeaud J & Oroz J
(eds.) Pline lAncien temoin de son temps. Conventus

640 Pliny the Elder (2379 A.D.)

Pliniani internationalis Namneti 2226 Oct. 1985 habiti
acta. Salamanca/Nantes. 549570.
Jan L & Mayhoff K (eds.) (18751906). C. Plini Secundi
Naturalis historiae libri XXXVII (6 vols). Leipzig:
Teubner.
Ko nig R, Winkler G et al. (eds.) (1973). C. Plinii Secundi
Naturalis historiae libri XXXVII (Lateinisch-deutsch).
Munich: Heimeran (for later volumes: Zu rich: Artemis
& Winkler; not yet completed).
Ko nig R & Winkler G (1979). Plinius der A ltere: Leben und
Werk eines antiken Naturforschers. Munich: Heimeran.
Mazzarino A (ed.) (1955). Grammaticae Romanae fragmenta aetatis Caesareae (Vol. 1). Turin: Loescher.
214331.
Murphy T (2004). Pliny the Elders natural history. The
empire in the encyclopedia. Oxford: Oxford University
Press.

Naas V (2002). Le projet encyclope dique de Pline lAncien.

Rome: E cole franc aise de Rome.
Rackham H, Jones W H S & Eichholz D E (eds.)
(19381962). Pliny: Natural history. With an English translation (10 vols). Cambridge: Harvard University Press.
Sallmann K (1975). Plinius der A ltere 19381970.
Lustrum 18, 5299.
ltere). Der Neue
Sallmann K (2000). Plinius (1) (der A
Pauly 9, 11351141.
Schottmu ller A (1858). De C. Plini Secundi libris grammaticis particula prima. Leipzig: Teubner.
Serbat G (1986). Pline lAncien. E tat pre sent des e tudes sur
sa vie, son uvre et son influence. Aufstieg und Niedergang der ro mischen Welt II 32(4), 20692200.
ltere. Paulys
Ziegler K et al. (1951). Plinius Secundus der A
Real-Encyclopadie der classischen Altertumswissenschaft 41, 271439.

Pluractionals (Distributives)
P Newman, Indiana University, Bloomington, IN, USA
2006 Elsevier Ltd. All rights reserved.

Many of the worlds languages have a derived pluractional verb form that indicates that the verbal
action is characterized by one or another kind of
multiplicity: it can happen habitually; it can be executed by a certain number of subjects; it can be applied to a certain number of objects; it can continue
over a longer period of time; or it can be performed at
different places (Gerhardt, 1984: 12), e.g., Gaanda
(Chadic) [lax] tear (something), [lelax] tear many
things, shred. The wide semantic range of pluractionals, which varies in detail from language to
language, is evidenced in Parsonss (1981: 206) description of Hausa, namely one actor, or a number of
actors doing the same thing to a number of objects,
either simultaneously or in succession; or a number of
actors doing the same thing to the same object severally and/or in succession; or else one actor doing the
same thing to the same object several times over . . ..
With intransitive verbs it adds a notion of multitude
and/or succession . . . or sometimes of distribution in
space. In the linguistic literature, these verb forms
are variously referred to as intensives, which is semantically inexact, or plural verbs, which leads to
confusion with inflectional verb forms that agree in
number with the subject (e.g., Dutch: run loop/loopt
[sg.] vs. lopen [pl.]). These forms have less often, but
more appropriately, been described in terms of plurality of action or simply as plural action verbs. In
recent years, the term pluractional, a lexical blend

coined by Newman (1980: 13) has been increasingly

adopted by Africanists, and more gradually by scholars working in other language areas. The exact distribution/prevalence of pluractional verbs is hard to
determine since language-specific descriptions and
terminology vary so greatly; however, it is probably
fair to say that pluractionals are very common and
quite widespread. For example, they are well attested
in all four major phyla in Africa (Brooks, 1991), in
Native American languages (Sapir and Swadesh,
1946), in Dravidian languages (Steever, 1987), in
Australian languages (Dixon, 1972), and in the
now-extinct Sumerian language (Steinkeller, 1979).
Generally speaking, use of pluractionals represents
a free choice to express semantic plurality rather than
being determined by agreement rules (see above), e.g.,
Bachama (Chadic) (Carnochan, 1970) [nda mbura
Fiye] he extinguish fire he put out the fire (single
action) vs. [nda mbara Fiye] he extinguish.PLURACT
fire he put out the fire (by going around and beating
it out). Nevertheless, there is a natural tendency for
ergative-type collocations, that is to say, pluractionals
commonly co-occur with (but are not necessarily required by) plural objects of transitive verbs or plural
subjects of intransitive verbs, for example:
(1) Bachama
/nda pir
vuney/
he
thatch hut
he thatched the hut
/nda pyer
vonye/
he
thatch.PLURACT huts
he thatched the huts

640 Pluractionals (Distributives)

Pliniani internationalis Namneti 2226 Oct. 1985 habiti
acta. Salamanca/Nantes. 549570.
Jan L & Mayhoff K (eds.) (18751906). C. Plini Secundi
Naturalis historiae libri XXXVII (6 vols). Leipzig:
Teubner.
Konig R, Winkler G et al. (eds.) (1973). C. Plinii Secundi
Naturalis historiae libri XXXVII (Lateinisch-deutsch).
Munich: Heimeran (for later volumes: Zurich: Artemis
& Winkler; not yet completed).
Konig R & Winkler G (1979). Plinius der Altere: Leben und
Werk eines antiken Naturforschers. Munich: Heimeran.
Mazzarino A (ed.) (1955). Grammaticae Romanae fragmenta aetatis Caesareae (Vol. 1). Turin: Loescher.
214331.
Murphy T (2004). Pliny the Elders natural history. The
empire in the encyclopedia. Oxford: Oxford University
Press.

Naas V (2002). Le projet encyclopedique de Pline lAncien.

Rome: Ecole francaise de Rome.
Rackham H, Jones W H S & Eichholz D E (eds.)
(19381962). Pliny: Natural history. With an English translation (10 vols). Cambridge: Harvard University Press.
Sallmann K (1975). Plinius der Altere 19381970.
Lustrum 18, 5299.
ltere). Der Neue
Sallmann K (2000). Plinius (1) (der A
Pauly 9, 11351141.
Schottmuller A (1858). De C. Plini Secundi libris grammaticis particula prima. Leipzig: Teubner.
Serbat G (1986). Pline lAncien. Etat present des etudes sur
sa vie, son uvre et son influence. Aufstieg und Niedergang der romischen Welt II 32(4), 20692200.
ltere. Paulys
Ziegler K et al. (1951). Plinius Secundus der A
Real-Encyclopadie der classischen Altertumswissenschaft 41, 271439.

Pluractionals (Distributives)
P Newman, Indiana University, Bloomington, IN, USA
2006 Elsevier Ltd. All rights reserved.

Many of the worlds languages have a derived pluractional verb form that indicates that the verbal
action is characterized by one or another kind of
multiplicity: it can happen habitually; it can be executed by a certain number of subjects; it can be applied to a certain number of objects; it can continue
over a longer period of time; or it can be performed at
different places (Gerhardt, 1984: 12), e.g., Gaanda
(Chadic) [lax] tear (something), [lelax] tear many
things, shred. The wide semantic range of pluractionals, which varies in detail from language to
language, is evidenced in Parsonss (1981: 206) description of Hausa, namely one actor, or a number of
actors doing the same thing to a number of objects,
either simultaneously or in succession; or a number of
actors doing the same thing to the same object severally and/or in succession; or else one actor doing the
same thing to the same object several times over . . ..
With intransitive verbs it adds a notion of multitude
and/or succession . . . or sometimes of distribution in
space. In the linguistic literature, these verb forms
are variously referred to as intensives, which is semantically inexact, or plural verbs, which leads to
confusion with inflectional verb forms that agree in
number with the subject (e.g., Dutch: run loop/loopt
[sg.] vs. lopen [pl.]). These forms have less often, but
more appropriately, been described in terms of plurality of action or simply as plural action verbs. In
recent years, the term pluractional, a lexical blend

coined by Newman (1980: 13) has been increasingly

adopted by Africanists, and more gradually by scholars working in other language areas. The exact distribution/prevalence of pluractional verbs is hard to
determine since language-specific descriptions and
terminology vary so greatly; however, it is probably
fair to say that pluractionals are very common and
quite widespread. For example, they are well attested
in all four major phyla in Africa (Brooks, 1991), in
Native American languages (Sapir and Swadesh,
1946), in Dravidian languages (Steever, 1987), in
Australian languages (Dixon, 1972), and in the
now-extinct Sumerian language (Steinkeller, 1979).
Generally speaking, use of pluractionals represents
a free choice to express semantic plurality rather than
being determined by agreement rules (see above), e.g.,
Bachama (Chadic) (Carnochan, 1970) [nda mbura
Fiye] he extinguish fire he put out the fire (single
action) vs. [nda mbara Fiye] he extinguish.PLURACT
fire he put out the fire (by going around and beating
it out). Nevertheless, there is a natural tendency for
ergative-type collocations, that is to say, pluractionals
commonly co-occur with (but are not necessarily required by) plural objects of transitive verbs or plural
subjects of intransitive verbs, for example:
(1) Bachama
/nda pir
vuney/
he
thatch hut
he thatched the hut
/nda pyer
vonye/
he
thatch.PLURACT huts
he thatched the huts

Pluractionals (Distributives) 641

(2) Bachama
/keembeto a
canoe
PAST
the canoe sank
/keembyee a
PAST
canoes
the canoes sank

Fimo/
sink
Fyemo/
sink.PLURACT

There are occasional instances, however, where

pluractional usage has become grammaticalized as
an agreement feature. In Kanakuru (Chadic), a handful of verbs obligatorily require pluractional agreement with plural objects (of transitive verbs) or plural
subjects (of intransitives), for example:
(3) Kanakuru
/na
Fowe
I
tie
I tied the horse
/na
Fope
I
tie.PLURACT
I tied the horses

dow-i/
horse-the
donjin-i/
horse.PL-the

(4) Kanakuru
/dow-i
a
Fowe-ni/
horse-the
PAST
tie-it(ICP)
the horse is tied
/donjin-i
wu
Fope-wu/
horse.PL-the they tie.PLURACT.-them(ICP)
the horses are tied
(ICP intransitive copy pronoun)

A phenomenon that is prevalent in the Chadic family is the existence of frozen pluractionals, namely
verbs that are pluractional in form but for which the
nonpluractional counterpart no longer exists. In some
cases the original pluractional meaning is still evident
to some degree; in others the special semantics of the
pluractional has been bleached out, e.g., Hausa furfura barter, cf. kirkira call many or often, pluractional of kira call; sansana smell, cf. tuntuna
remind many or often, pluractional of tuna remind; sassabe clear a farm (with plural action implied even though the simple stem *sabe doesnt
exist); yagalgala tear to pieces (with plural action
evident even though *yagala doesnt exist).
As with other morphological formations, pluractional derivation is accomplished by a wide variety of
means; nevertheless, some patterns stand out. First
pluractionals are commonly formed in iconic fashion
by means of reduplication. In the Chadic language
family, partial reduplication (sometimes accompanied by vowel modification) is the norm, but full
reduplication is also attested. Here are some examples (with the pluractional form to the right):
Bole [looFu]/[lolooFu] ask; Gaanda [fel]/[fefal]
break; Sha [mot]/[motot] die; Lamang [sula]/
[sulala] fry; Hausa [tuna]/[tuntuna] remind,

[dagura]/[dagurgura] gnaw; Daba [pez]/[pezpez]

sharpen; Margi [kutsa]/[kutsakutsa] shake. Possibly related to reduplication is the use of gemination to
form pluractionals, e.g., Pero [lofo]/[loffo] beat,
[liguno]/[ligguno] answer; Bole [poFu]/[poFFu]
take out, [salu]/[sasallu] slash. In Kanakuru, the
geminate has been reduced to a single consonant,
leaving the surface pluractional formation as one
of consonant hardening, e.g., [muri] (<[*muti])/
[mute] (<[*mutte]) die, [Fowe] (<[*(Fope])/[Fope]
(<[*Foppe]) tie. In Kwang, similar (but independent) degemination has resulted in pluractional
formation by devoicing, e.g., [beje]/[peje] wash
(something), [geFe]/[keFe] disguise oneself. Finally,
languages with pluractionals often have a small number of suppletive forms, e.g., Kaje (Benue-Congo)
[wrak]/[ban] mount, climb, [byin]/[tyey] give
birth; Mupun (Chadic) [Fen]/[le] put, [cit]/[nas]
beat, [ta]/[FoN] fall down.
See also: Intensifiers; Reduplication; Iconicity; Natural
Semantic Metalanguage; Plurality.

Bibliography
Brooks B (1991). Pluractional verbs in African languages.
Afrikanistische Arbeitspapiere 28, 157168.
Carnochan J (1970). Categories of the verbal piece in
Bachama. African Language Studies 11, 81112.
Cusic D D (1981). Verbal plurality and aspect. Ph.D.
dissertation, Stanford University.
Dixon R M W (1972). The Dyirbal language of North
Queensland. London: Cambridge University Press.
Dressler W (1968). Studien zur Verbalen Pluralita t. Vienna:
Hermann Bo hlaus.
Durie M (1986). The grammaticalization of number as a
verbal category. Proceedings of the Berkeley Linguistics
Society 12, 355370.
Gerhardt L (1984). More on the verbal system of Zarek
bersee 67, 1130.
(Northern Nigeria). Afrika und U
Moravcsik E A (1978). Reduplicative constructions. In
Greenberg J H (ed.) Universals of human language 3:
Word structure. Palo Alto, CA: Stanford University
Press. 297334.
Newman P (1980). The classification of Chadic within
Afroasiatic. Leiden: Universitaire Pers.
Newman P (1990). Nominal and verbal plurality in Chadic.
Dordrecht: Foris.
Parsons F W (1981). Writings on Hausa grammar: the
collected papers of F. W. Parsons. Ann Arbor, MI: UMI
Books on Demand.
Sapir E & Swadesh M (1946). American Indian grammatical categories. Word 2, 103112.
Steever S B (1987). The roots of the plural action verb in the
Dravidian languages. Journal of the American Oriental
Society 107, 581604.
Steinkeller P (1979). Notes on Sumerian plural verbs.
Orientalia 48, 5467.

642 Plurality

Plurality
P Lasersohn, University of Illinois at
Urbana-Champaign, Urbana, IL, USA
2006 Elsevier Ltd. All rights reserved.

Plural expressions may be intuitively characterized

as those involving reference to multiple objects. Semantic theories differ, however, in how this intuition
is worked out formally.
The most popular approach is to treat plural
expressions as referring to some sort of plural object, or group. That is, alongside individual objects
such as people, tables, chairs, etc., we assume there
are groups of such objects, and that plural expressions refer to these groups in much the same way as
singular expressions refer to individuals. Just as singular noun phrases denote and quantify over individuals, plural noun phrases denote and quantify over
groups; just as singular predicates hold true or false
of individuals, plural predicates hold true or false of
groups.
An alternative, advanced by Schein (1993) (building on earlier logical work by George Boolos) is to
regard a plural term as denoting each of several individuals directly, rather than denoting the group containing these individuals. This gives the effect of
treating denotation for plural terms as a relation rather than a function; it is the denoting relation itself,
rather than the denoted object, which is plural. The
primary advantage of this technique is that it allows
a reasonable treatment of noun phrases such as the
sets that do not contain themselves, which potentially
give rise to Russells paradox in more conventional
treatments.
Among analyses that do regard plural expressions
as referring to groups, the main options are to identify
groups with sets, or with mereological sums. The
latter choice is favored especially by those who regard
sets as abstract, mathematical objects existing outside
of space and time. As Link (1998) puts it, If my kids
turn the living room into a mess I find it hard to
believe that a set has been at work. However, not
all authors share the intuition that sets of concrete
objects are themselves abstract, and in any case the
issue seems more philosophical than linguistic (but
see the discussion of conjoined noun phrases below).
No matter which approach is adopted, a central
problem in the semantics of plurality is determining
the range and distribution of readings available to sentences containing plural expressions. Sentence (1a),
for example, is intuitively interpreted as predicating
the property of being numerous to our problems
collectively; no individual problem can be numerous.

Sentence (1b), in contrast, requires all (or nearly all)

the individual children to be asleep; the predicate
applies distributively. Sentence (1c) seems ambiguous
between collective and distributive readings, meaning
either that the T.A.s together earned exactly $20 000,
or that they each did:
(1a) Our problems are numerous.
(1b) The children are asleep.
(1c) The T.A.s earned exactly $20 000.

Such examples raise several issues: Are plural

expressions authentically ambiguous between collective and distributive readings, or are both interpretations covered under a single, very general meaning? If
there is an authentic ambiguity, is it a simple two-way
ambiguity between collective and distributive readings, or are there other possibilities? What is the locus
of the ambiguity: in the noun phrase, the predicate, or
both?
In favor of the view that there is an authentic ambiguity, consider a situation in which there are two
T.A.s, and each of them earned exactly $10 000.
In this case, sentence (1c) is true. Sentence (2) is also
true in this situation:
(2) The T.A.s earned exactly $10 000.

This suggests that there are two distinct figures,

both of which are the exact amount the T.A.s
earned: in one sense they earned $20 000, and in
another sense they earned $10 000 but this appeal
to multiple senses amounts to a claim of ambiguity.
An ambiguity is also suggested by patterns of
anaphora, as argued by Roberts (1991). Sentence
(3a) allows the continuation in (3b) if (3a) is interpreted as predicating lx9y[piano(y) & lift(x, y)] of
the group of students as a whole, but not if this
predicate is understood as applying to each student
separately. (A third interpretation, in which there is
some piano y such that the predicate lifted y applies to
the individual students, also allows the continuation
in (3b), but this does not affect the argument.)
(3a) The students lifted a piano.
(3b) It was heavy.

If sentence (3a) is unambiguous, with no formal

differentiation between the collective and distributive
interpretations, it is hard to see how we could capture
this difference in anaphoric potential.
If there is an ambiguity, one may ask whether the
fully collective and fully distributive readings are
the only ones available, or if instead there may be
intermediate readings. Intermediate readings appear
to be called for in examples like (4):

Plurality 643
(4) The shoes cost $75.

This sentence is most naturally interpreted as meaning that each pair of shoes costs $75, not that each
individual shoe costs that much, or that all the shoes
together cost that much.
Gillon (1987) argues that sentences with plural
subjects have as many readings as there are minimal
covers of the set denoted by the subject noun phrase,
in which a cover of a set A is a set C of nonempty
subsets of A such that their union, [ C, is equal to A,
and a cover of A is minimal iff it has no subsets that
are also covers of A; this idea is developed further by
Schwarzschild (1996) and others. Under this proposal, the pragmatic context makes a particular cover
salient, and the predicate is required to hold of each
element of the cover. The fully distributive and fully
collective readings reemerge as special cases.
However, cover-based analyses face a challenge in
dealing with examples like (1c): Suppose John, Mary,
and Bill are the T.A.s, and each of them earned
$10 000. In this case, the predicate earned exactly
$20 000 holds of each cell of the cover {{John, Mary},
{John, Bill}}, but sentence (1c) is not intuitively true
in this situation.
Whether it appeals to covers or not, an ambiguity
analysis must address the issue of where in the sentence the ambiguity is located. Early treatments often
seemed to take for granted that it was the plural noun
phrases themselves which are ambiguous, but many
more recent treatments trace the ambiguity to the
predicates with which the noun phrases combine
(Scha, 1984; Dowty, 1986; Roberts, 1991; Lasersohn,
1995). A standard argument for this view comes from
examples like (5):
(5) The students met in the bar and had a beer.

The natural interpretation is that the students met

collectively but had separate beers. This reading is
easily obtained if we treat the subject noun phrase
as unambiguously denoting the group of students as a
whole, and treat the conjunct verb phrases as predicates of groups, with the distributive predicate had
a beer holding of a group iff each of its individual
members had a beer; the denotation of the whole,
coordinate verb phrase may then be obtained by intersection. If we try to claim that the collective/
distributive ambiguity is located in the subject noun
phrase, however, it seems impossible to give a consistent answer as to which reading it takes in this
example.
A suitable ambiguity in the predicate may be
obtained by positing an implicit operator on the
predicate. Link (1998) and Roberts (1987) suggest
an optional operator on the verb phrase, notated

D, and interpreted as lPlx9y[yPx!P(y)], where

P is the relation an individual stands in to the larger
groups of which it forms a part. This gives a simple
two-way ambiguity, depending on whether the operator is present or not; intermediate readings may be
obtained using a more complex operator which quantifies over elements of a cover (Schwarzschild, 1996).
Either operator is easily generalized across types to
give distributive readings for non-subject argument
places (Lasersohn, 1998).
It should be noted that the lexical semantics of
some predicates prevent them from participating in
the collective/distributive ambiguity: a verb like sleep,
for example, cannot apply to a group without also
applying to the individual members of the group.
Adding a distributivity operator in this case is redundant, and does not result in a difference in meaning.
Conversely, certain predicates apply only to groups:
gather, for example. Such predicates do show collective/distributive ambiguities however, when applied
to arguments denoting groups of groups:
(6) The tribes gathered.

Example (6) may be used to mean that each tribe

gathered separately, or that all the tribes gathered
together.
A good deal of work has been devoted to the relation between conjunction and plurality (Link, 1998;
Hoeksema, 1983; Landman, 1989a; Lasersohn,
1995; Schwarzschild, 1996; Winter, 2001) because
conjoined noun phrases sometimes admit collective
readings, as in (7):
(7) John and Mary lifted the piano.

Such examples suggest that coordinate noun

phrases may denote groups in much the same fashion
as plural noun phrases. Such sentences also admit a
distributive reading, which may be obtained either
through the use of a distributivity operator as with
other plural noun phrases, or by reducing the coordination to propositional conjunction, via a generalized
conjunction operator or conjunction reduction transformation.
Conjoined noun phrases have sometimes been
used to argue that semantic theory must allow
reference to higher-order groups (Hoeksema, 1983;
Landman, 1989a). This is easily accomplished in set
theory, since sets may contain other sets as members:
{a, {b, c}} 6 {{a, b}, c} 6 {a, b, c}, but is less straightforward if groups are modeled as mereological sums.
(8a) Blu cher and Wellington and Napoleon fought
against each other at Waterloo.
(8b) The cards below seven and the cards from seven
up are separated.

644 Plurality

Example (8a), from Hoeksema (1983), may be

parsed in either of two ways: [[Blu cher and Wellington] and Napoleon] or [Blu cher and [Wellington and
Napoleon]]. It is intuitively true relative to the first
parse but false relative to the second, suggesting that
the two parses correspond to denotations such as
{{b, w}, n} and {b, {w, n}}. Likewise example (8b),
from Landman (1989a), is not equivalent to The
cards below 10 and the cards from 10 up are separated. But if reference to higher-order groups is disallowed, the subject noun phrases of these two
sentences would seem to refer to the same group,
namely the group containing all the individual cards
as members. The opposing view that noun phrases
need never refer to higher-order groups has been
defended in detail by Schwarzschild (1996), who
points out that the pragmatic context may make
salient a particular division of the denotation into
subgroups even if that denotation is first-order; correct truth conditions may be obtained if the semantics
is sensitive to this pragmatically supplied division.
A number of additional issues arise in the semantics
of plurality, which can only briefly be mentioned
here: Certain adverbs, such as together or separately,
seem to force a collective or distributive reading, but
the exact mechanism by which they do this is a matter
of some dispute (Lasersohn, 1995; Schwarzschild,
1994; Moltmann, 1997). Plural expressions share a
number of characteristics with mass terms, suggesting
a unified analysis (Link, 1998). Plural noun phrases
affect the aspectual class of predicates with which
they combine, suggesting a parallel in the domain of
events to the structure of groups among individuals
(Krifka, 1989) a parallel also suggested by the phenomenon of verbal plurality or pluractionality
(Lasersohn, 1995). Finally, the interpretation of bare
plurals, or plural noun phrases with no overt determiner, and particularly the alternation between the
existential, generic, and kind-level readings, illustrated in (9), has attracted enormous attention; but
much of this work belongs more properly to the study
of genericity and indefiniteness than to the semantics of plurality per se (Carlson, 1980; Carlson and
Pelletier, 1995).
(9a) Raccoons are stealing my corn.
(9b) Raccoons are sneaky.
(9c) Raccoons are widespread.
See also: Coordination; Generic Reference; Genetics and

Language; Mass Expressions; Mass Nouns, Count Nouns,

Non-count Nouns: Philosophical Aspects.

Bibliography
Carlson G (1980). Reference to kinds in English. New York:
Garland Press.
Carlson G & Pelletier F J (1995). The generic book.
Chicago: University of Chicago Press.
Dowty D (1986). Collective predicates, distributive
predicates, and all. In Marshall F (ed.) ESCOL
86: Proceedings of the Third Eastern States Conference on Linguistics. Columbus: Ohio State University.
97115.
Gillon B (1987). The readings of plural noun phrases in
English. Linguistics and Philosophy 10, 199219.
Hamm F & Hinrichs E (1998). Plurality and quantification.
Dordrecht: Kluwer Academic Publishers.
Hoeksema J (1983). Plurality and conjunction. In ter
Meulen A (ed.) Studies in model theoretic semantics.
Dordrecht: Foris Publications. 6383.
Krifka M (1989). Nominal reference, temporal constitution, and quantification in event semantics. In
Bartsch R, van Benthem J & van Emde Boas P (eds.)
Semantics and contextual expression. Dordrecht: Foris
Publications. 75115.
Landman F (1989a). Groups, I. Linguistics and Philosophy 12, 559605.
Landman F (1989b). Groups, II. Linguistics and Philosophy 12, 723744.
Landman F (2000). Events and plurality: the Jerusalem
lectures. Dordrecht: Kluwer Academic Publishers.
Lasersohn P (1995). Plurality, conjunction and events.
Dordrecht: Kluwer Academic Publishers.
Lasersohn P (1998). Generalized distributivity operators.
Linguistics and Philosophy 21, 8392.
Link G (1998). Algebraic semantics in language and
philosophy. Stanford, California: CSLI Publications.
Moltmann F (1997). Parts and wholes in semantics.
Oxford: Oxford University Press.
Roberts C (1991). Modal subordination, anaphora and
distributivity. New York: Garland Press.
Scha R (1984). Distributive, collective and cumulative
quantification. In Groenendijk J, Janssen T & Stokhof
M (eds.) Truth, interpretation and information. Dordrecht: Foris Publications. 131158.
Schein B (1993). Plurals and Events. Cambridge, Mass.:
MIT Press.
Schwarzschild R (1994). Plurals, presuppositions, and the
sources of distributivity. Natural Language Semantics 2,
201248.
Schwarzschild R (1996). Pluralities. Dordrecht: Kluwer
Academic Publishers.
Winter Y (2001). Flexibility principles in boolean
semantics. Cambridge, Mass.: MIT Press.

Poetry: Stylistic Aspects 645

Poetry: Stylistic Aspects

L Jeffries, University of Huddersfield, Huddersfield,
UK
2006 Elsevier Ltd. All rights reserved.

Introduction
This article aims to discover what is special about
poetry in the context of literature in general and
stylistic approaches to literature in particular. What
is noticeable, first of all, about the literary works that
we generally call poetry is that they are, in a sense,
archetypal literature. In other words, to the extent
that we want to distinguish literature as an identifiable group of texts (see Jeffries, 1996), we inevitably
attempt to define what makes literature different
from other forms of language. And within the context
of literature in this sense, we might conclude that
poetry is made up of the language that is least like
normal language (whatever that is) and is therefore
more typically literary than narrative fiction or dramatic texts. Jakobsons early identification of the poetic function as one of the main functions of language
underlines the importance of poetry as one of the most
distinctive genres of text.
This is not to say that there is such a thing as an
identifiable literary language, or that literature or
poetry are actually clearly identifiable sets of texts
with different stylistic norms from other texts. Most
stylisticians from Jakobson onward would argue
quite strongly, indeed, that the same techniques and
stylistic effects are used in literature as in other texttypes and genres. See Carter and Simpson (1989) for
a restating of Jakobsons (1987) call to develop a
stylistics of literature. Despite the determination, then,
that literature was not linguistically different from
other texts, the stylistic study of poetry in the early
and mid-20th century took this prototypical role of
poetry as its starting point and attempted to apply
the techniques of the emerging discipline of linguistics
to those aspects of poetic language that seemed to
mark it out from other kinds of text.
In an insight allied to Jakobsons notion of poetic
function, the Russian formalists also focused on the
idea that literature typically makes everyday scenes,
processes, and situations seem strange to the reader
by describing them in out of the ordinary ways. This,
it is argued, causes the reader to see the world afresh,
through the eyes of the poet, and thus live life
more intensely, experiencing it more vividly. There
have been challenges to this approach, to which
I will return.
Poetry, of course, has probably existed as long
as human language, and it has been studied and

discussed throughout the ages and civilizations. One

of the approaches to language that had an enormous
influence on Western civilization was the Aristotelian
approach, which Jakobson drew on in proposing a
poetic function of language one that had no other
point to it than to provide a beautiful, and preferably
insightful, description of some aspect of the world.
This aesthetic function was juxtaposed to the much
more mundane, but highly valued, functions associated with prose, including persuasion, argument,
and reasoning.
The traditional approach to poetry, then, was
to analyze those things about its style that were different from other texts and were also, presumably,
aesthetically pleasing. These included the following:
. Figurative language: the notion that there is a clear
distinction between literal language and other
forms of expression, such as metaphor
and metonymy, with the latter seen as derivative
from and dependent on the former.
. Poetic license: the idea that poets were at liberty to
alter the rules of language, including the invention
of new words and expressions more freely than
would be expected in other text-types.
. Economy of expression: poetry was traditionally
seen as having the capacity (not always taken up,
of course) to distill thoughts and ideas into fewer
words than would be the case in any prose genre.
. Foregrounding: the notion that the artistry of the
poem lies mainly in those features (often deviant
ones) that are foregrounded by virtue of differing
in some way from the text surrounding them (see
Foregrounding).
. Patterning: the unusual (foregrounded, and/or deviant) patterning of text, either through sound
(phonology) or structure (repetition and partial
repetition).
. Authorial voice: the norm that in the absence of
clues to the contrary, means that poems are written
in the authors voice, whether or not they are in the
first person. This is not true, of course, of prose
fiction where even a first person narrative is not
automatically assumed to be the authors voice.
. Formal structures: the structuring of poems by
rhyme, meter, and stanza length has traditionally
been a very tangible mark of its separateness from
other genres.
The rise and development of linguistics and all the
sub-branches of the discipline has led to a burgeoning
range of linguistic approaches to style, all of which
share the aim of making the analysis of literary style
more rigorous, clearer in its methods, and more

646 Poetry: Stylistic Aspects

explanatory in its power. New tools of analysis have

arisen in phonology, semantics, grammar, discourse
analysis, and cognitive approaches to text analysis,
and all of these have been appropriated by stylisticians looking for new insights into literary language.
This is true not only of poetry, of course, and many of
these tools of analysis have been applied widely
across many different literary and nonliterary genres.
Many of the approaches in recent years have drawn
upon the prototypical features of poetry listed above,
though in most cases they have tried to ground them
more in the linguistic descriptions that current theories and methods provide. The earliest linguistic
treatment of poetic style was by Leech (1969) and
was organized according to the themes listed above,
but with an explicit determination to use the more
rigorous and explanatory descriptive techniques of
linguistics.
Here, we will try to identify the analyses that have
been applied specifically to poetry, in order to see the
range of possibilities of stylistic analysis of poems and
to identify those features that are most typically poetic, as well as those having a wider application in
general. It is also worth pointing out that poetry is
still sometimes used as the convenient prototype of
literature in articles on stylistics. This is partly because
poems are often a convenient length for detailed
analysis, and partly because it is more likely than
narrative or dramatic fiction to have stretched
the resources of the language in a number of new
directions, giving the stylistician a lot of material to
comment on in a short extract. In such articles, it may
not be the claim that what is being pointed out is
particularly restricted to the style of poetry. Here,
then, I am attempting to restrict my scope to discussing those treatments of poetry that are concerned
with its particular characteristics, though these are
often, of course, shared with other text-types, both
literary (e.g., dramatic texts) and nonliterary (e.g.,
advertising copy).

Music The Sounds of Poetry

The music of poetry is often cited as one of its main
attractions. But what is it? Traditional approaches
would recognize the different sound patternings that
are labeled alliteration, assonance, rhyme, stress,
etc., but what linguistics adds to these labels is an
understanding of how these phonological patternings
work. It also helps in establishing that alliteration
and assonance can be linked not simply to the
spellings of words, but to their sounds.
In his poem Robbing Myself, Ted Hughes (1998:
165) uses the phrase the snow-loaded house in a
description of his visit in midwinter to the house he

shared with Sylvia Plath. Here, the same vowel is used

twice in quick succession, and though it would be a
diphthong (/eo/) in some English accents, and a long
vowel (/o:/) in others, including the poets Yorkshire
accent, the effect is one of a repeated resonance; it is
therefore musical in a very basic sense. The effect of
this repetition may be just musical though one can
often make a case for some kind of meaningful effect
too. Perhaps the repeated use of the same vowel
sound emphasizes the extreme nature of the snowfall?
The point here is that the vowels are spelled very
differently in the two words o and oa so it
takes a phonetic transcription to clearly pinpoint
their assonantal nature.
If phonemic transcription and phonetic understanding is used, we can see patterns that show
more subtle effects than simple identity of sound.
So, for example, the use of all high vowels, or plosive
consonants, or long vowels and diphthongs, could be
said to have certain effects which would be harder
to describe accurately without the use of phonetic
terminology. The description of sound symbolism
and onomatopoeia, then, may benefit from the use
of phonetic concepts and terminology, allowing the
analyst to describe the reasons why a particular concentration of sounds could reflect the sounds or
meaning being described directly. The result is a
more accurate vocabulary for describing these effects
than was previously available and used everyday
adjectives such as spiky or smooth.
Maura Dooleys poem, Up on the Moors with
Keeper (Dooley, 1991: 66), about the Bronte s,
makes a distinction between the life and energy of
the three sisters and the darker force that, it is implied, is Branwell Bronte , their brother. The distinction is captured not only by the semantics of the
poem, but also in the phonemic structure, which contrasts light, airy sounds with darker, more ominous
sounds, as in the line: Theyve kicked up their heels
at a dull brother. The strength of a linguistic approach to such a poem is that we do not have to
leave the description dependent on adjectives like
light and airy, hoping that our readers agree. Instead, we can make a case, based on the articulatory
features of the vowels involved, that these sounds are
indeed lighter (or darker) in a very tangible sense.
Thus, the vowels in kicked /kIkd/ and heels /hIelz/
(assuming a southern English accent) are front vowels
and high (or closed) vowels, though the diphthong in
heels moves toward the centre of the mouth with the
schwa. These high, front vowels tend to be higher in
pitch than back or open vowels, and are regularly
used in English to symbolize small, delicate, or light
concepts, in both literal and metaphorical senses. The
contrast here is with the dull brother, whose vowels

Poetry: Stylistic Aspects 647

are both either /V/ or /o/, depending on your accent.

The poet has a southern English accent (though
I only know this by chance), but the poem is about
Yorkshire, and so either vowel could be relevant
to a reading of the poem. The effect in both cases is
one of a heavier, darker sense of foreboding,
since both vowels are back vowels, and more open
than /I/, they both contrast with the sisters vowels.
The sound effects described so far are features of
the segmental structure of the phonology, but poetry
also draws upon the suprasegmental phonology of
patterns of stressed and unstressed syllables as well
as the global use of rhyme schemes to create what is
traditionally known as poetic form. At the most basic
level, phonology can help us understand the difficulties of recognizing stressed syllables, in order to identify the metrical patterns of poetry, such as iambic
pentameter, etc. Recent work in this area, however,
has taken steps towards a much more comprehensive
phonological account of how poetic meter works.
See Meter for a detailed exposition of the linguistic
description of meter.

Words
Traditionally, the choice of words in poetry has been
known as diction, and this covers everything that might
be included in a more linguistic account under the
heading of lexical semantics. With the structures and
terminology of linguistic descriptions, we can describe
the texture and significance of a poems vocabulary
in terms of interwoven lexical fields; the use and
creation of synonyms and opposites; unusual collocations; the relationship between hyponyms and superordinates; the use of polysemy and/or homonymy to
create ambiguity and puns; and the exploitation of
the readers knowledge of a words connotation.
Vicki Feaver, for example, bases her poem Ironing
(OBrien, 1998: 261) on a set of lexical fields that
together tell a story both more subtle and more revealing than the apparent surface meaning of the
poem. Superficially, she describes how her ironing
habits have changed over the years, as she first of all
irons everything, including towels, then she irons
nothing, and finally she irons only some of her personal things. What the combined force of her lexical
choices achieves is to alert the reader to the symbolic
force of her initial dedication (and slavery) to the
domestic role she had been allocated, her rebellion
against this (by not ironing) and her final reconciliation with some of the necessary chores of any human
being, when she begins to iron again but only her
own clothes. One of the sets of words she uses is that
relating to the configuration of materials. In her overzealous days, the irons flex is described as twisting

and crinkling while the fabric is pressed to a

thinness and she wants to flatten the house. Both
of these opposing processes seem to be extreme; in the
case of the irons flex, it frays, and the house is clearly
destroyed by her fantasy of flattening it with an iron
as big as a tugboat. In her rebellious stage, she converts to crumpledness, which has a more gentle,
though slightly depressing, connotation. Finally, her
reincarnation as a contented and balanced user of the
iron sees her making wrinkled blouses creaseless
and an airy shape. It is noticeable that this final
stage of her conversion does not exactly flatten
the fabric of her silk blouse, but instead, makes it
into a three-dimensional space into which she can fit
her body.
The lexical structure of poems, then, is often very
compact, and its study through the categories and
terms of lexical semantics can help us to explain our
insights and hunches about the way in which the
poem is working at levels that do not conform to
the propositional structure and literal meaning of
the sentences and clauses that make it up.
Word combinations of various kinds can also be
said to work at levels and in ways that cut across the
surface structure of the text. Take, for example, the
idea that individual verbs tend to operate in particular
kinds of context. This might be described as selectional restrictions in a Chomskyan approach, as thematic roles in a functional approach, or perhaps as
collocational tendencies in a more structuralist approach. Without going into the theoretical basis of
these terms and the differences between them, we
might be able to accept that some verbs tend to
occur with animate subjects (e.g., eat), some require
a location in their complement (e.g., put) and some
verbs have very specific requirements, such as needing
an object to be brittle (e.g., snap). One of the most
widespread tendencies of poets throughout the ages
has been to exploit the readers knowledge of these
restrictions on verbs by putting them into unacceptable contexts. The result is often what traditional
commentators would call synesthesia (the fusing of
two of the human senses), may be metaphorical, and
often causes personification. Carol Rumens, for example, in Song of the non-existent (Rumens, 1995:
46), describes dusk as the hour between dog and
wolf, and having effectively likened the daytime to
a dog and the nighttime to a wolf, she then uses the
verb walk not with an animate subject, but with an
abstract one: anxiety. The juxtaposition of the wolf
as nighttime and this collocation of anxiety as the
subject of walk may cause the reader to fuse these
two strong images into one, and we may make
the anxiety that we feel at dusk into a wolf that is
stalking us. Rumens uses a number of these unusual

648 Poetry: Stylistic Aspects

collocations in this poem, including The watery city

thickens, blackens, in which she takes a verb (thickens) with very specific requirements a liquid subject
and puts a concrete subject (the city) into its place.
The resulting image of the city as liquid (possibly
partly seen in the reflection of a river?), and as
a liquid that gets thicker as it goes from dusk to
darkness, is a powerful metaphor, achieved with the
simplest of linguistic techniques.

Grammar
The use of grammar as a tool for creating literary
meaning is perhaps more prominent in poetry than
in other genres. It is difficult to know to what extent
the use of poetic license to play grammatical games
with the reader is partly a product of our improved
knowledge about the workings of grammatical structure as a result of the headway made by linguistics in
the last century. It would certainly not be the role of
stylisticians to attribute conscious intentions to poets
in this regard. However, it is probably true to say that
the grammatical playfulness that has been the hallmark of English poetry since the late 19th and early
20th centuries has kept pace with the increasing levels
of knowledge and understanding about grammatical
structure. Stylistics, then, has been able to use the
models and theories of linguistics not only to find
new perspectives on the literature of the past, but
also to find tools for describing the inventiveness of
contemporary literature.
The invention of new words has been a recognized
part of poetic license for many centuries, and is not
strictly limited to poetry (note Shakespeares inventiveness in his plays), but might again be said to be
more typically a poetic than a prosaic technique.
Some poets, of course, use morphological inventiveness more than others. The contribution of stylistics
to the description of such neologism is to allow definition of relationship between invented words and
existing ones, and in so doing to provide an explanation for the ability of readers to decode words they
have never met before.
Gerard Manley Hopkins, for example, is known
for his inventions of compound words. Hopkins,
writing at the end of 19th century, was a religious
poet who used compounds to try to get close to
describing the glory of god. For example, in his
poem Pied Beauty (Hopkins, 1979: 68) he is praising
god for dappled things, among which he numbers
the couple-color of certain skies and rose-moles on
trout seen through the water. Another early 20th
century poet, e. e. cummings (1960), made a particular kind of poetic license his own hallmark in the

early 20th century. He often made words change

grammatical class in order to challenge our understanding of them, and in places, used our knowledge
of morphological mistakes to conjure up a childhood innocence or naivete in the description of what
it is like to be in love: love is more thicker than
begin (see Jeffries, 1993: 28).
It is relatively recently in the history of poetry
(perhaps only about a hundred years) that the potential for meaning that ambiguity allows has been so
widely exploited. While it could be said to be a part of
the potential for economy that language has, ambiguity or vagueness about grammatical structure has
been a regular feature of poetry in English only since
the modernists challenged our preconceptions about
language, reference, and meaning.
Feminist, and particularly lesbian, poets have
found this a productive area of language to explore,
not least, presumably, because the challenging of
given categories and boundaries is in the nature of
their lives work. Audre Lorde is one of those to make
frequent use of the ambiguities available in English.
One of these (for detailed discussion, see Jeffries,
1994) is the fact that an apparent list of noun phrases
can either refer to a number of different referents or
be interpreted as a set of noun phrases in apposition,
like those we come across everyday in the newspaper
(Mr. Blair, the Prime Minister or Mrs. X, the
woman at the center of the scandal). In this way,
Lorde effectively causes the reader to consider the
referents of her lists, to see whether the phrases are
indeed coreferential, or separate or perhaps both:
Without contemplating last and late
the true nature of poetry. The drive
to connect. The dream of a common language.

In Origins and History of Consciousness (Lorde,

1993: 7), we are faced with three apparently unrelated concepts: what poetry is all about, the impulse of
human beings to relate to each other, and an ideal
society in which people speak the same language. Of
course, it is easy to see that what Lorde is doing here
is setting out the three parts of a kind of trinity, which
work for her as the Christian trinity works in that
religion both three and one. It is because people
want to make contact that they need a common language, and this is best achieved through poetry, in
Lordes view.
One of the most exciting ways in which grammar
can be exploited by poems is in using the structure,
rather than the words, as an iconic representation of
the meaning. This might be seen as the grammatical
equivalent of sound symbolism, where the sounds
directly reflect the meaning, though not always in an

Poetry: Stylistic Aspects 649

echoic way. This use of the structure itself as a signifier seems to appeal to the readers emotions at a level
that is below the referential level of the vocabulary,
and depends for its success on the information structure of clauses and reader expectations of the loading
of different clause elements. The following example
comes from Pamela Gillilans poem Doorsteps
(France, 1993: 143), which describes a woman who
is clearly no longer alive (perhaps a much-loved
grandmother?) cutting bread and delivering the
resulting slices to the tea table, and the waiting child:
Finely rimmed with crust the soft
halfmoon half-slices came to the tea table
herringboned across a doylied plate.

In the context of a hungry child, the grammatical

achievement here is to make the reader wait for the
verb, which is the central pivot of the English sentence and is not normally delayed very long. The
information structure in English sentences tends to
put new, and therefore longer and more complex,
information towards the end of a clause, with references to shared knowledge normally being located in
the early part of a clause, such as the subject. Thus, a
delayed verb may result in a feeling of frustration,
anticipation, or boredom, or any one of a number of
possible emotions connected to waiting. Here, the
frustration is compounded by the delay in reaching
the head noun (half-slices) of an unusually long
subject, an effect similar to delaying the main verb.
Together, these postponements of grammatical resolution mimic the childs hungry gaze as the carefully
laid tea is brought to the table. Gillilan contrasts this
with the present-day tendency to care less about presentation than speed and nutrition, and yet we feel
some of her loss of anticipation in the structure of
the sentence where she admits, I saw away at stoneground wholemeal. The subject is short (a pronoun)
letting us get to the verb (saw) quickly: there is no
mystery involved, and none of the ritual that she is
mourning the loss of.
This iconic use of structure is just another of the
tools that poets may choose to use. Most poets have a
particular set of favorite techniques, and some use
grammatical iconicity more than others. Wordsworth
uses grammatical iconicity in the Prelude in describing winter skating on a frozen lake. He has long,
meandering sentences full of subordinate clauses
where the skaters are racing around on the ice, and
short sharp sentences where they come to an abrupt
halt. See Jeffries (1996) for a description of this extract (see also Iconicity: Literary Texts) for more on
iconicity in literature.

Text/Discourse
One of the larger shifts in linguistic emphasis in recent
years has been the move toward more contextual
approaches to language study, dealing with texts in
larger than sentence-level units, and looking at
aspects of meaning that can range over whole texts
and take into account the background knowledge,
experience, and current situation of both producer
and recipient of a text. These discourse and textoriented approaches have been taken up in a range
of ways by stylisticians keen to find new tools to
analyze their data. In the study of poetry all of them
are potentially appropriate, but in order not to overlap too much with other articles in this volume, I will
restrict myself here to those that have been exploited,
rather than trying to take in those that could be used
for poetic analysis but simply happen not to have
been taken up so far.
Jeffries (2000) and Simpson (1993) both take the
analysis of point of view from the field of critical
discourse analysis and consider the possibilities of
this approach for poetic language, Simpson asking
this question in general terms, and not only in relation
to poetry, and Jeffries asking a rather more specific
question about point of view in Carol Ann Duffys
poetry. Point of view is an analytical technique that
draws on a number of grammatical categories from
functional grammar that together demonstrate the
viewpoint from which a poem is written. This is
much more subtle than asking whether it is in the
first or second person; it involves asking questions
about the transitivity, modality, and deixis of texts,
in addition to considering the person in which it is
written. All of these topics are covered in other articles in this volume, so here I will simply demonstrate
one or two examples of the poetic use of point of
view.
Douglas Dunn uses a form of modality in his poem
The Kaleidoscope (Dunn, 1986: 238), where he
describes the habit he got into, after the premature
death of his wife, of going up to her room, almost
expecting her to still be there: Might be to find you
pillowed with your books. His use of the epistemic
modal, might, creates an interesting effect, since it is
clear from the context that she is not going to be
there, and so the use of a modal that questions the
truth of this is shockingly apposite and emotive in
showing us the narrators sense of denial over this
bereavement. In contrast, Erin Moure (1985: 43)
uses the more straightforward boulomaic modals in
her poem about the desire she feels for her lesbian
love to be accepted by society:

650 Poetry: Stylistic Aspects

I want an age where I can turn my neck
& kiss you at dinner
among real roast beef & oranges

The use of want is more direct, more outspoken,

than Dunns wistful wishful thinking.
The other main movement in stylistic analysis in
recent years has been toward cognitive approaches to
the style of texts (see Stylistics, Cognitive). Although
poetry has again been used as the prototype of literature (see Introduction, Verdonk, 1993), there are
few analyses specific to poetry that take this approach. Semino (1997) is one of the few trying to
deal directly with the cognitive aspects of poetic
style. In doing so, she drew on Cook (1994), and
others, claiming that the distinctive feature of what
we call literature is that it challenges, and potentially
changes, the mental structures, called schemata,
through which we organize our experience of the
world. This potential for changing our worldview is
one that she has ascribed to poems in particular,
through their possibility of creating text worlds
(see Text World Theory) that are at odds with our
experience of the world. This view has been
challenged (Jeffries, 2001) and defended (Semino,
2001), and the debate about the differences between
different readers experiences of reading a poem continues. In the meantime, it is probably generally
true to say that in addition to providing aesthetic
pleasure, one of the social functions of poetry is to
affect our view of the world in some way.

Conclusion
In conclusion, we may return to the features of typical
poetic language listed in the Introduction as having
some currency in prelinguistics approaches to the
style of poetry and ask to what extent these features,
perhaps with different terminology, remain at the
heart of what poetic language is and does.
. Figurative language: Certainly interest in figurative
language has grown, and not only in literary stylistics (see Figures of Speech; Metaphor: Stylistic
Approaches). One difference of approach in recent
years is to see metaphor and other tropes as much
more normal, because they are now recognized as
common in everyday language. Metaphor has also
become the center of a large research effort in
cognitive stylistics, as a particular case in
the investigation of how readers make meaning.
Though poetry has been used for such investigations, metaphor is no longer seen as exclusively in
its province.
. Poetic license: Although the term poetic license is
used more facetiously than in earnest these days,

the phenomenon remains an important part of poetic style and has been taken to new extremes in the
20th and 21st centuries. The rules of language that
can be broken by poets now include all of them,
and the results can be relatively obscure or perceived as difficult by the reader. More recently,
poets have perhaps seen this as a problem and
have begun to stretch rules in ways that are comprehensible by readers, usually by analogy with
other forms or structures in the language.
Economy of expression: The power of poetry to say
a lot in a few words has not diminished; it may even
have increased. The willingness of poets to allow
ambiguities, vagueness, lack of cohesion, multiple
referents, etc., as well as the tendency to use every
level of language from phonology to grammar in
iconic ways, allows the poet to pack meaning into
every sound, word, and sentence.
Foregrounding: The recognition and enjoyment of
outstanding features of poetic language remain an
important aspect of poetic style, and one that is less
significant in other, less economical genres. There
is, however, a recognition that the background of
literary language can also be significant, and that
the unique voice that we recognize in many poets
may be a function of smaller, repetitive features of
language as much as the foregrounded ones.
Patterning: The importance of patterning in poetry
remains strong and has become perhaps more important as the rigors of poetic form have fallen
away. Poets, then, will use the repeated clause,
phrase, or sound to produce a kind of music that
is less insistent than a regular meter or rhyme and
thus may be meaningful as well as musical, in the
sense that every time a writer has a free choice to
make, the choice he or she makes is significant.
Authorial voice: Robert Browning is famous for his
poems in others voices, and using a range of voices
is not a new phenomenon, as any dramatic verse
will testify. What is, perhaps, new is the blending
and merging of voices in a single poem, which has
become the hallmark of some very successful poets
in the late 20th and early 21st century. Carol Ann
Duffy, Tony Harrison, and others have seen the
potential for this technique, which often draws on
more narrative-type techniques such as free indirect
style (see Dialect Representations in Texts; Speech
and Thought: Representation of).
Formal structures: the structuring of poems by
rhyme and meter may have given way to more
subtle use of line lengths, line endings, stanza
lengths, and so on; but the form of poetry remains
one of its distinguishing features, and layout continues in most cases to differentiate at a basic level
between poetry and prose.

Poetry: Stylistic Aspects 651

Poetry remains, socially and culturally, the literary

genre that people both revere and to some extent fear.
The reason for these apparently contradictory emotions is the same: poetry dares to do the unthinkable
with language, and this can be both exhilarating
and daunting.
See also: Dialect Representations in Texts; Figures of
Speech; Foregrounding; Iconicity: Literary Texts; Metaphor: Stylistic Approaches; Meter; Speech and
Thought: Representation of; Stylistics, Cognitive; Text
World Theory.

Bibliography
Attridge D (2003). Maxima and beats: a response to Nigel
Fabbs reply. Language and Literature 12, 8182.
Attridge D (2003). The rules of English metre: a response
to Nigel Fabb. Language and Literature 12, 7172.
Boase-Beier J (2004). Knowing and not knowing: style,
intention and the translation of a holocaust poem.
Language and Literature 13, 2535.
Carter R & Simpson P (1989). Language, discourse and
literature: an introductory reader in discourse stylistics.
London: Unwin Hyman.
Cauldwell R (1999). Openings, rhythm and relationships:
Philip Larkin reads Mr Bleaney. Language and Literature
8, 3548.
Cook G (1994). Discourse and literature: the interplay of
form and mind. Oxford: Oxford University Press.
cummings e e (1960). Selected poems 19231958. London:
Faber.
Cummings M & Simmons S (1983). The language of literature. Oxford: Pergamon Press.
Dooley M (1991). Explaining magnetism. Newcastle:
Bloodaxe.
Duffell M J (2002). The Italian line in English after
Chaucer. Language and Literature 11, 291305.
Dunn (1986). Selected poems 19641983. London: Faber.
Fabb N (2002). The metres of Dover beach. Language and
Literature 11, 99117.
Fabb N (2003). Metrical rules and the notion of
maximum: a reply to Derek Attridge. Language and
Literature 12, 7380.
France L (1993). Sixty women poets. Newcastle: Bloodaxe.
Goodblatt C (2000). In other words: breaking the monologue in Whitman, Williams and Hughes. Language and
Literature 9, 2541.
Hall G (2000). Coining phrases: Cliche and creativity in
the poetry of Tony Harrison. In Jeffries L & Sansom P
(eds.). 3853.
Hanson K (2003). Formal variation in the rhymes of
Robert Pinskys The Inferno of Dante. Language and
Literature 12, 309337.
Hasan R (1989). Linguistics, language and verbal art (2nd
edn.). Oxford: Oxford University Press.
Hens G (2000). What drives Herbeck? Schizophrenia,
immediacy, and the poetic process. Language and
Literature 9, 4359.

Hopkins G M (1979). The major poems. London: Dent

Dutton.
Hughes T (1998). Birthday letters. London: Faber and
Faber.
Jakobson R (1987). Language in literature. Pomorska K &
Rudy S (eds.). Cambridge, MA: Harvard University
Press.
Jeffries L (1993). The language of twentieth century poetry.
Basingstoke: Macmillan.
Jeffries L (1994). Language in common: apposition in contemporary poetry by women. In Wales K (ed.) Feminist
linguistics in literary criticism. London: Boydell and
Brewer. 2150.
Jeffries L (1996). What makes language into art? In
Maybin J & Mercer N (eds.) Using English: from conversation to canon. London: Routledge. 162184.
Jeffries L (2000). Point of view and orientation in Carol
Ann Duffys poetry. In Jeffries L & Sansom P (eds.).
5468.
Jeffries L (2001). Schema affirmation and white asparagus:
cultural multilingualism among readers of texts.
Language and literature 10, 325343.
Jeffries L & Sansom P (eds.) (2000). Reading contemporary
poetry. Huddersfield: Smith Doorstop Books.
Leech G (1969). A linguistic guide to English poetry.
London: Longman.
Lorde A (1993). The dream of a common language: poems
19741977 (Reissue edition). New York: Norton.
Mackay R (2000). Lonely meanings; Seamus Heaneys The
spirit level. In Jeffries L & Sansom P (eds.). 1937.
McCully C B (2000). Writing under the influence: Milton
and Wordsworth, mind and metre. Language and Literature 9, 195214.
McCully C B (2003). Towards a theory of poetic change.
Language and Literature 12, 525.
Moure E (1985). Domestic fuel. Toronto: Anansi Press.
OBrien S (1998). The firebox. London: Picador.
Rumens C (1995). Best china sky. Newcastle: Bloodaxe.
Semino E (1997). Language and world creation in poems
and other texts. London: Longman.
Semino E (2001). On readings, literariness and schema
theory: a reply to Jeffries. Language and Literature 10,
345355.
Short M (1989). Reading, analysing and teaching literature.
London: Longman.
Short M (1996). Exploring the language of poems, plays
and prose. London: Longman.
Simpson P (1993). Language, ideology and point of view.
London: Routledge.
Sobolev D (2003). Hopkins rhetoric: between the material
and the transcendent. Language and Literature 12,
99115.
Verdonk P (1993). Twentieth century poetry: from text to
context. London: Routledge.
Wales K (ed.) (2001). A dictionary of stylistics. London:
Pearson Education.
Wojcik-Leese E (2000). Salient ordering of free verse
and its translation. Language and Literature 9,
170181.