?tekauer Pavol, Lieber Rochelle. Handbook of Word-Formation

Подождите немного. Документ загружается.

ONOMASIOLOGICAL APPROACH TO WORD

FORMATION

229

3.6

ete

r K

och

hile Koch’s model (2001) is based on similar principles it is more elaborate

and more comprehensive – it is a lexicological rather than purely word-formation

odel. Koch presents it in the form of a three-dimensional ‘grid’.

Its horizontal axis represents cognitive-associative relations which underlie new

igns. These include (a) identity; (b) contiguity (relations within a conceptual frame,

he above-mentioned engynomies; (c) metaphorical similarity, (d) co-taxonomic

imilarity between the concepts of the same level of hierarchy, e.g.,

and

;

(

e) taxonomic superordination; (f) taxonomic subordination; (

) co-taxonomic

ontrast, e.

GOO

D – BA

; and

(

)

conce

tual contrast between more or less

ncompatible concepts, e.

., PRI

N – H

TEL

The vertical axis re

resents formal

eans of expressing new concepts. They

nclude (i) zero = semantic change (diachronically) or polysemy (synchronically));

(

ii) gender change (

wood

–

woods

); (iii) genus change; (iv) diathesis change; (v)

onversion; (vi) mutation (change of word-class by replacing a word-class specific

bound grameme, e.g., fr.

manquer

‘t

obes

f’

–

e manque ‘shortage’;

(

vii)suffixation; (viii) prefixation; (ix)

ompounding; (x) lexicalized syntagma, e.g.,

red

wine

red wine

;

(

)

idiom.

thir

‘

utochtonous-borrowed’ opposition. Then,

ndividual words result from the combination of the elements re

resented on the

former two axes, for exam

le, the relation between

peer

and

eer tre

ontiguity relation expressed by compoundi

g in English, but by suffixation in

French (

poire – poirier

(

). Obviously, not all theoretical combinations actually occu

n the individual languages.

avol Štekauer

Department o

British and American Studies

Facult

Arts

ov University

7. novemb

080 78

ešov

lovakia

e-mail: stekpal

unipo.s

EFERENCES

Alinei, Mario. 1995. “Theoretical Aspects of Lexical Motivation.”

venska Landsmål och

venskt Folkliv

118, 321, 1

Aronoff

Mark. 1976.

ord Formation in

enerative

rammar. Linguistic Inquiry Monograph 1.

ambridge, MA: MIT Press.

Brammesber

er, Albert and Grze

a, Joachim. 2001. “ModE

girl

and Other Terms for ‘Young Female

Person’ in En

lish Lan

e Histor

nomasiolo

Online 2 [www.onomasiolo

.de].

auer, Laurie. 2001.

orphological Productivit

Cambridge Studies in Linguistics, 95. Cambridge:

ambridge University Press.

230

TEKA

eard

Robert E. 1995. Lexeme-Morpheme Base Morpholo

. A Ge

eral Theor

lection and Word

mation

. SUNY Series in Linguistics. State

niversity of New York Press.

lank

Andreas. 1997. Prinzipien des lexikalischen Bedeutungswandels am Beis

iel der romanischen

rachen. Tübingen: Niemeyer.

lank, Andreas. 1998a. “Outlines of a Cognitive Approach to Word-formation.” In: Proceedings of the

International Congress of Linguists, Paper No. 0291. Oxford: Pergamon.

lank, Andreas. 1998b. “Cognitive Italienische Wortbildungslehre.”

talienische Studien 19

5-27.

lank, Andreas. 2001. “Words and Concepts in Time.” In: metaphorik.de.

lank, Andreas and Koch, Peter

(

eds.

)

. 2003. Ko

nitive romanische Onomasiolo

ie und Semasiolo

Tübin

en: Max Nieme

er Verla

alton-Puffer, Christiane. 1997. “Developinm

a meani

oriented theor

of En

lish word formation.”

View

[

]

–

Vienna En

lish Workin

Papers

4-18.

iez

Friedrich. 1875.

omanische Wortschöp

. Anhan

zur Grammatik der Romanischen S

rachen.

irven, René and Verspoor, Marjolijn. 1998. Cognitive Exploration of Language and Linguistics

msterdam/Philadelphia: John Benjamins.

okulil

Miloš. 1962.

Tvo

ení slov v

ĜĜ

eštin

þþ

I. Teorie odvozování slov.

Prague: Nakladatelství

okulil, Miloš. 1964. “Zum wechselseitigen Verhältnis zwischen Wortbildung und Syntax.” In:

avaux

inguistiques de Prague 1. Prague: Academia, 369-376.

okulil, Miloš. 1968a. “Zur Fra

e der Konver

on und verwandter

tbildun

svor

än

e und -

beziehun

”

ravaux lin

uistiques de Pra

ue 3. Pra

ue: Academia, 215-239.

okulil

Miloš. 1968b. “Zur Theorie der Wortbildun

slehre.” Wissenscha

tliche Zeitschri

t der Karl-

arx-Universität Leipzig. Gesselschafts- und S

rachwiss

nschaftliche Reihe 17

203-211.

okulil, Miloš. 1968c. “Zur Frage d

sog. Nullableitung.” In: H. E. Brekle and L. Li

(

eds.

)

Wortbildung, Syntax und Morphologie. Festschrift zum 60. Geburtstag von Hans Marchand am 1.

ktober 1967. The Ha

ue: Mouton de Gru

ter, 55-64.

okulil, Miloš. 1968d. “Zu

Frage der Stelle der Wortbildung im Sprachsystem.” Slovo a slovesnos

29,

-16.

Dokulil

Miloš. 1997. “Th

Prague School‘s Theoretical and Methodological Contribution to ‚Word-

formation‘ (Derivology).” In: Obsah – výraz – význam. Miloši Dokulilovi k 85. narozeninám

Praha

F UK

210

ressler, Wolfgang U.; Mayerthaler, Willi; Panagl

Oskar; and Wurzel, Wolfgang U. (eds.). 1987.

Leitmoti

s in Natural Morpholo

Amsterdam: John Ben

amins.

leischer, Wolf

. 1969. Wortbildun

der Deutschen Ge

enwartssprache

Leipzi

: VEB

ibliohraphisches Institut.

Grzega, Joachim. 2001. “On the Names for Wednesday i

Germanic Dialects with S

ecial Reference to

est Germanic.”

nomasiologiy Online 2 [www.onomasiology.de].

Grzega, Joachim. 2002. Some Thoughts on a Cognitive Onomasiological Approach to Word-Formation

with Special Reference to English

Onomasiology Online 3

1-29

Grzegorczykowa, Renata. 1979.

Zarys slowotwórstwa polskie

go. Slowotwórstwo opisowe

Warszawa:

PWN.

ansen

Barbara

;

Hansen

Klaus

;

Neubert

Albrecht; and Schentke, Manfred. 1982

ed. 1990

)

nglische Lexikologie. Einführung in Wortbildung und lexikalische Semantik. Leipzig: VEB Verlag

Enzyklopädie.

oreck

, Ján. 1983. V

vin a teória

ka.

ratislava:

PN.

oreck

, Ján; Buzáss

ová Klára; Bosák Ján et al. 1989.

namika slovne

zásob

sú

asnej sloven

þþ

iny

ratislava:

eda.

oreck

, Ján. 1994.

emantics o

Derived Words

Prešov: Acta Facultatis Philoso

hicae Universitatis

afarikanae.

oreck

, Ján. 1999. “Onomasiolo

ická interpetácia tvorenia slov.”

Slovo a slovesnost

60, 6-12.

uke

Ivana. 1977. Die Wortbildun

stheorie von Miloš Dokulil

Inau

ural-Dissertation Giesen.

üllen, Werner. 1999a. “A plea for onomasiology.” In: W. Falkner and H.-J. Schmid

(

eds.

)

, Words

exemes, conce

ts – A

roaches to the l

xicon: Studies in

onour of Leonhard Lipka

Tübingen:

Narr

343-352.

üllen

Werner. 1999b.

lish Dictionaries 800-1700. The Topical Tradition

xford:

xford

niversit

Press.

ONOMASIOLOGICAL APPROACH TO WORD

ORMATION

231

Kastovsky, Dieter. 1968.

ld English Deverbal Substantives Deri

ed by Means of a Zero Morpheme.

Esslin

en/N.: Lan

er.

Kastovsk

, Dieter. 1969. “Wortbildun

und Nullmorphem.” Lin

uistische Berichte

2, 1

Kastovsk

, Dieter. 1982. Wortbildun

und Semanti

Tübin

en/Düsseldorf: Francke/Ba

el.

Koch

Peter. 1999a. “TREE and FRUIT: A

nitive-onomasiolo

ical approach.” In:

tudi di Lin

uistica

Teorica ed A

licata XXVIII

emantica lessical

331-347.

Koch, Peter. 1999b. “Frame and Conti

uit

. On the

nitive Bases of Meton

and Certain T

pes of

ord Formation.” In: K.-U. Panther and G. Radden

(

eds.

)

etonymy in Language and Though

Amsterdam/Philadelphia: John Benjamins, 139-167.

Koch, Peter. 2001. “Bedeutungswandel und Bezeic

nungswandel: Von der kognitiven Semasiologie zu

ognitiven Onomasiologie.” Zeitschrift für Literaturwissenschaft und Linguisti

121,

Krefeld, Thomas. 1999. “Cognitive ease and lexical borrowing: the recategorization of body parts in

Romance.” In: A. Blank and E. Koch

(

eds.

)

istorical semantics and cognition

rlin

Lieber

Rochelle. 1981.

n the Or

anization o

the Lexicon

Doctoral dissertation, MIT, Bloomin

ton,

N: Indiana Universit

Lin

uistics Club.

Lipka, Leonhard. 2002.

lish Lexicolo

Tübin

en: Narr.

Marchand

Hans. 1960.

he Categories and Types of Present-Day English Word-Formation.

iesbaden:

tto Harrassowitz.

Marchand

Hans. 1964a. “A Set of Criteria for the Establishin

of Derivati

nal Relationshi

between

ords Unmarked by Derivational Morphemes.” Indogermanische Forschungen 69, 10-19.

Marchand, Hans. 1964b. “Die Ableitung desubstanti

scher Verben mit Nullmorphem im Englischen,

ranz

ös

” Die Neueren S

rachen 10

105-118.

Marchand, Hans. 1965. “The Analysis

erbal Nexus

ubstantives.”

ndogermanische Forschungen

70,

7-7

Marchand

Hans. 1969.

he Categories and Types of Present-Day English Word-Formation

sed

edition. München:

. H. Beck.

Neš

menko

Galina Parfenevna. 1963. “Zakonomerno

ti slovoobrazovanija, semantiki i upotreblenija

suš

yx s suffiksami subjektivnoj ocenki v sovremenno

þeskom jazyke.” In:

ssledovanija

ešskomu jazyku: Voprosy slovoobrazovanija i gramatiki

þþ

. Moscow

105-158.

Neš

menko

G. P. 1968.

stori

a imenno

o slovoobrazovani

a v

ešskom literaturnom jazyke konca

þþ

III-XX vekov. Moscow.

Polenz, Peter von. 1973. “Synpleremik I. Wortbildung.” In: Lexikon der germanistichen Linguistik.

Tübin

en, 145-163.

Puz

nina, Jadwi

a. 1969. Nazw

ci we ws

pó

czesnym j

zyku polskim (slowotwórstwo, semantyka,

ĊĊ

kladnia

)

. W

d. 1. Rozpraw

Uniwes

tetu Warszawskie

o Nr. 40. Warszawa: W

dawnictwo UW.

uadri, Bruno. Aufgaben und Methoden der onomasiologischen Forschung.

Eine

ntwicklungsgeschichtliche Darstellung

Bern: A Francke Verlag.

roat, Richard W. 1988. “Brack

ting Paradoxes, Cliticization and Ot

her Topics: The Mapping between

Syntactic and Phonological Structure.” In: M. Everaert, A. Evers, R. Huybregts, and M. Trommelen

(

eds.

)

orpholo

and Modularit

. In Honour of Henk

chultink

Publications in Lan

Sciences 29. Dordrecht: Foris

339-360.

Štekauer, Pavol. 1992. “On Some Issues of Zero Morpheme in English.” Linguistica Pragensia 2

73-87.

Štekauer

Pavol. 1996. A Theory of Conversion in English.

rankfurt am Main: Peter Lang.

Štekauer

Pavol. 1997. “On the Semiotics

f Pro

er Names and Their Conversion.” AAA – Ar

beiten aus

nglistik und Amerikanisti

Band 22/1

27-36.

Štekauer, Pavol. 1998. An Onomasiological Theory of English Word-Formation

Philadelphia: John Benjamins.

Štekauer, Pavol. 2001a. “Beheading the Word? Please, Stop the Execution.” F

olia

Linguistica 34/3-4

333

Štekauer, Pavol. 2001b. “Fundamental Principles of an Onomasiological Theory of English Word-

rmati

” Onomasiology Onlin

2, 1-42 [www.onomasiology.de].

Stepanova, Maria D. 1973.

ethoden der s

nchron

Wortschatzanal

Hall

manek, Bo

dan. 1988. Cate

ories and Cate

orization in Morpholo

Lublin: Katolicki Uniwers

tet

Lube

agliavini, Carlo. 1949. “Di alcuni denominazioni della <

illa>

(

studi

di onomasiologia, con speciale

iguardo alle lingue camito-semitiche e negro-africane).” Scritti minori

Bologna, 529-568.

2 PA

TEKA

Tappolet, Ernst. 1895. Die romanischen Verwandtscha

tsnamen

it besonderer Berücksichti

der

französischen und italienischen Mundarten. Ei

Beitrag zur vergleichenden Lexikologie. Strassbourg:

Karl.

. Trübner.

aszakowa, Krystyna (1994).

lowotwórstwo ws

pó

czesnego j

áá

yzka polskiego – rzeczowniki sufiksalne

ĊĊ

obce

Warszawa: Wydawnictwa Uniwersytetu Warszawskiego.

auner, Adolf. 1902. Die romanischen Namen der Kör

rteile. Eine onomasiologische Studie. Erlangen.

233

tekauer, P. and R. Lieber (eds.), Handbook of Word-Formation, 233—265

2005 Sprin

er. Printed in the Netherlands.

COGNITIVE APPROACH TO WORD-FORMATION

ID T

UGG

1. BASIC NOTIONS OF COGNITIVE GRAMMAR

(

)

.1 The grammar of a language under CG

ognitive grammar (CG), as developed by Ronald W. Langacker (1987, 1991a,

2000), Ta

lor (2002) and others, is one of the most influential streams in the

eneral

ovement known as Co

nitive lin

uistics. CG holds that a lan

e (or its

grammar) “can be characterized as a struct

ured inventory of c

nventional lin

uistic

nits.” (Langacker 1987: 57) Most of the ter

s in this definition are technical terms

but here are some of the im

lications of them.

•

rammar is an

nventor

not a machine. It holds resources that s

eakers use to

onstruct utterances; the grammar itself does not construct utterances

automatically and mindlessly.

•

This inventory is not haphazard, but

uctu

the parts of it are related to each

ther in important ways.

•

The inventory consists of

units

. A ‘unit’ is a cognitive routine

hat a speaker has mastered quite thorou

, to the extent that he can emplo

it in

automatic fashion, without havin

o focus his attention specificall

on its

individual parts or their arrangement. … It is effectively simple, since it does not

demand the constructive effort required for the creation of novel structures.

cholo

ists would speak of a ‘habit’, o

that ‘automatization’ has occurred

(Lan

acker 1987: 57).

•

a r

hip bet

lf a

ves u

s, the whole complex

becomes ‘effectively simple’, without the parts thereby losing their unit status.

n other words, speakers can wield eithe

he parts or the whole with ease. Units

an thus become quite hi

complex.

•

These units m

st be lin

uistic, defined as either semantic (constitutin

or bein

part of a meanin

) or phonolo

ical (in a broad sense encompassin

estures and

writing or other signals as well as spoken sounds), or symbolic. Symbolic

uctu

are bipolar; that is, they involve the pairing of a semantic with a

phonological structure. For the word

cat

for instance, the cognitive routines

onstituting the meaning

AT ar

nit

h i

emantic pol

the

articulatory and auditory cognitive

outines involved in pronouncing and

ecognizing [

æt

]

form a unit which is the phonological pol

and the pairing of

234

ID T

UGG

the two poles to form the symbolic structure

cat

is also a unit.

does not

allow you to posit any other kinds of structures in a language, beyond these

ee.

•

The units of a language are

conventional

That is, they are established by usage as

hared by a community of people. All of language is in this sense usage-based,

and usage is a central, not a peripheral concern of linguistics. Usage is generally

ot 100% predictable, but it tends to be reasonable, and the units it produces,

thou

h often motivated in some de

ree, are not predictable either.

•

Conventionalit

is enou

h to

uarantee a

nit a place in the lan

e. It ma

not

be excluded even if it is in some degree redundant or predictable, as long as it is

n fa

t ma

nit an

dco

ven

naliz

ed w

ithin th

speech community.

irtuall

all the concepts involved in th

s definition are matters of de

ree. (1)

Status as a unit is not a plus-or-minus quality: it takes time and repetitive usage to

establish or ‘entrench’ a habit of any ki

d, including those that constitute a language,

and there are notable differences of entrenchment even among those routines that

learly have ‘unit’ status. Catapul

is a unit for many speakers of English, but not

early so well-entre

nit f

r m

f th

m a

cat

. (2) Conventionality is

radual alon

several parameters, such as how well the conventionalit

established, for what sub

roup(s) of speakers, in what socio-historical contexts, and

o forth. Amon

sailors of the 1

centur

the meanin

s of

cat

as a ‘whip with nine

ords, cat-o’-nine-tails’ or as a ‘catboat’ were more likely to be conventional than

among American construction workers of the late 2

century, for whom the

eaning ‘bulldozer, Caterpillar tractor’ was

uite certain to be conventional.

Most categories in CG are defined in terms of gradual parameters, that is, they

are matters of more-or-less rather than absolute, plus-or-minus dichotomies. This fits

n with the cognitive model of categorization around pro

ypes or ‘best examples’

r ‘central members’ rather than with a model that emphasizes boundaries between

ate

ories and assumes that all members of the cate

are of equal status. In this

ase, there is not a plus-or-minus distinction between what forms part of a lan

and what lies outside it. Some things (e.g.

cat

with the meaning ‘domestic feline’, or

he ordering of an adjective before the noun it modifies) are very centrally part of

English, whereas others (the ordering of galor

after the noun it modifies

or the

Freddag

, which a friend used in reference to typically hard spikes in

volleyball by my brother Fred), while qualif

ying as English, are relatively peripheral

o the category.

Small caps are used, as is traditional, to represent

eanin

s. For CG, this is a shorthand for a t

picall

ery complex collection of cognitive routines (section 2.3). Phonological forms are enclosed [in

quare brackets], though the level of phonetic detail we will be dea

ing with is minimal, and the

lashes (e.g.

/) traditional for phonemic representations might have been used instead. They

epresent an acceptable (rather than ‘the right’) pronunciation. Familiarity with basic phonetic

ymbols (IPA) is assumed. Words and other bipolar,

ymbolic structures are

epresented by standard

English orthographical forms in italics

Besides being central for conventionality, usage is cr

ucial for the establishment of units in individuals’

ds.

OGNITIVE APPROACH TO WOR

FORMATION

235

.2 Lexicon and s

ntax

Man

influential lin

uistic theories have made a stron

dichotom

between the

exicon, where words and morphemes reside, and the syntax or grammar, where

phrases, clauses, and larger structures ar

ed.

a ha

sbee

escribed as constituting a ‘building bl

ck’ model

(

Lakoff and

ohnson 1980: 202-

203, Langacker 1987: 452-457), where the pieces

f language that are stored in the

exicon are like bricks, and the grammar is the mortar that binds those simple and

solid pieces together to form larger structures.

ord-formation patterns are awkward to handle under such theories. The

are

like s

ntactic rules in involvin

ran

ements of meanin

ful pieces, but the

her-order words or stems the

describe often seem intuitivel

to be part of the

exicon, mentally stored and accessed from memory as wholes rather than

ecessarily constructed on the fly. And the apparently solid bricks themselves often

prove, upon reflection, to be internally structured to some extent.

nder the CG definition, particular words and their meanings and phonological

(

or written or signed) forms are, to the extent that speakers master an

onventionalize them, included as part of the grammar, and the patterns of thei

formation are

rammatical under exactl

the

eco

iti

ue o

particular phrases, clauses, and so forth, and the patterns of their formation. Once

ain CG is positin

cate

ories that do not have strict boundaries; the differences

between lexicon, morphology (word-formati

n), and syntax are all matters of degree

ather than strict differences in kind (cf. section 4.2). Bricks and mortar and large

hings built out of them are all building materials, and not all that different fro

Thi

ill

beco

e clearer in later discussion, but one important

mplication of it is the view of syntactic and morphological patterns, not

ust lexicon,

as s

mbolic, with phonolo

ical specifications (includin

word order) at one pole

mbolizin

meanin

ful (semantic) specifications at the other. As an important part

f the picture, the small word-pieces o

ten known as ‘grammatical morphemes’ are

meaningful, and function as they do in larger

syntactic constructions because of their

eanings.

2. SCHEMAS AND PROTOTYPES

.1 Schemas and elaborations

ives

reat importance to schematicit

, that is to the relationship between a

ma an

rati

It i

eof

the major kinds of relationships which

nit

int

tructured inventor

f th

ede

finiti

n in

sec

n 1

s a pattern, a rough outline, a coarse-gr

ned, less-fully-specified version of a

once

t which the elaborations render, each in a different way, in finer, more

elaborate detail. All of the schema’s s

ecifications are true of its elaborations, but

each elaboration of a schema specifie

tail

h th

esc

does

236

ID T

UGG

Our thinking is shot through with relationships of this kind. For instance, the

once

NIMAL i

h in

udes

rati

MAMMAL

INSEC

f th

ese

in t

rn ha

rati

f its own, down as many levels as one

ares to analyze. Or, MOVE is a schema which can be elaborated by LOCOMOTE

(

i.e.

HANGE LOCATION), which in turn can be elaborated by R

and so forth. The

chematicity relationship is, by convention

thin CG, represented graphically by an

arr

ma t

rati

NIMAL

RTHR

SOC

IAL

ANT

IRE

N THE 1

METER

IN 1

PRIN

LOCO

(

read

‘

ANIMAL i

ssc

mati

RTHR

…

’

r ‘

SPRIN

rat

LOCOMOTE …’

f a particular pairing of a schema with an elaboration is repeated often and

trongly enough to become established (a habit), it achieves unit status and is

eligible to be, and will be if conve

ionalized, part of the language. Thus

chematicity relations are at the same time relationships between units (‘structuring’

he ‘inventory’) and, if such be the case, themselves units in the inventory.

G posits (what is eminentl

reasonable, if

ou think about it) that all lin

uistic

tructures are schemas,

atterns character

zin

families of more detailed co

nitive

events, events ultimatel

too detailed for us to talk about usefull

. So we are not

ealing with a difference in kind between schemas and non-schematic units, but

ather between more highly and less highly schematic units. In fact, schematicity is

another of the gradual parameters

G teaches one to see everywhere.

ll categories, in CG, are sets of units related by relations of schematicity,

whether full schematicity or partial schematicity (section 2.2).

.2 Partial schematicity and the growth of schematic networks

Schematicity relationships arise when a person compares two concepts and

otices similarities.

ur brains seem to be so wired that the cognitive system gets

excited when that happens. There are four possibilities (ultimately differing only in

egree, of course) when you compare a ‘ta

et’ concept, T, with a ‘standard’ of

omparison, S. Either (i) S is fully identical to T or (ii) S is schematic for T (S

)

r (iii) some of S’s specifications are true of T, but there is distortion of them, or (iv)

the distortion is such that you can’t really

recognize S in T at all. Case (iii) is

partial schematicit

and represented

raphicall

a dashed arrow (S

- -

)

you compare

APPLES an

ORATORY you will (unless you try pretty hard)

ecide that this is case (iv) and there isn’t much point in thinking about it any more.

you compare

PPLES

ith

PPLES you of course have case (i). If you compare

PPLES

ith

you have case (ii), where you simply recognize that

IT i

mati

PPLES

or as we normally express it in English, ‘Apples are (a kind

)

fruit’.

Case (iii) is particularly interesting, where you compare APPLES

ith

RAN

. The

are certainl

not identical, and

PPLE

are not

(

or at least are not

Note that the English ‘equative’

phrase normally has the same asymmetr

y as a schematicity statement. T

S does not imply, and only allows in the case of identity, T

S, and similarly if it is true that

Apples are (a kind of) fruit

it is predictably untrue that

Fruit is (a kind of) apples

OGNITIVE APPROACH TO WORD

ORMATION

asily seen as) a kind of

RAN

nor are

RAN

a kin

PPLE

eve

rth

ess

they are similar in many ways. Those ways in which they are similar constitute a

potential or nascent schema, and if the comparison is made often enough it can

become entrenched. This is illustrated in Fi

ure 1. Note that in Fi

ure 1 the

(putativel

) new schema (1.c) is not the same thin

as the alread

established

(

1.d

)

; it contains a number of s

ecifications which are common to

APPLES an

ORANGES

hara

riz

r kin

ds o

f FR

such as

BANANAS

ATERMELONS

RASPBERRIES

s this sort of process continues, mor

and more complex categories, consisting

f units related in full or partial schematicity, are built up in people’s minds. As they

are conventionalized through linguistic usage, they become part of the language in

question.

Sometimes it is useful to express identit

with a double-headed schematicit

arr

ow:

PPLE

All the specifications of a schema are fulfilled in its

elaborations, so if all the specifications of

two structures are fulfilled in each other,

have to be identical.

The meaning represented in 1.c will most naturally,

evertheless, be named b

and thus become linked

to, the phonological pole

[

frut

]

rather than to any of the others, and such a tenuous link, in the process

f becoming established, is represented in the diagram. The convention of a dashed-dotted line

eparating semantic from phonological structure

is used without labeling later diagrams.

Oran

e color whe

ripe, thic

inedible peel,

ivi

ded

sections

ed, yellow, or

green when ripe,

edible

ski

lik

Grow on trees

near round

handheld size, o

ten eaten raw,

multiple seeds in a core section ,

green when unripe, Etc.

hings that grow,

usua

lly

man

at a t

me,

n trees

bushes or

vines; detachable,

picall

uic

and

edibl

e, conta

n see

Semantic structures

Phonolo

ical Structures

rǙzbèr iz

bԥnǙnԥz

wáqoˆmèikˆz

óokˆdԥz

Ǚmiˆz

frut

Figure

Comparing oranges with apples; establishing schemas and buildin

a netwo

3"\



BOG



PBF

\

P3

P\

238

ID T

UGG

.3 Prototypicality and salience

ot all units in a schematic network like that in Figure 1 are equal. Leaving aside

he status of 1.c

most speakers of En

lish would a

ree that apples and oran

es, in

ontrast to watermelons, raspberries or bananas, are amon

the most t

pical, first-

thou

ht-of kinds of fruits. Considerable research, much of it indebted to Rosch’s

work (e.g. Rosch 1973), indicates that such ine

ualities can be im

ortant for human

ategorization, and in fact that many categories are largely organized through

esemblances (= relationships of partial schematicity) to a central member or group

f ‘best’ members of the category. In many cases this ‘center-out’ relationship may

be more important than any ‘top-down’ relationships of full schematicity fo

haracterizing the nature of a category. But, CG maintains, it is not a matter of eithe

ne or the other; it works ver

satisfactoril

to see both kinds of cate

orization as

oexistin

and cooperatin

protot

is used to desi

nate the central part of the cate

. Arrivin

at a universally satisfactory definition of the term is not easy – the category of

prototypes itself, naturally enough, shows prototype effects – but a very important,

arguably the most important, characteristic is that prototypes are cognitively more

prominent or salient than the other members of the category. This salience naturally

esults from many aspects of typical prototypes (e.g. their being the first learned and

the most often encountered), and can be reasonably suspected of causing a numbe

f others (e.

. their bein

the easiest to recall and the quickest to be assi

ned

erball

to a cate

). Such salience, apart from contextual effects, is closel

tied

o (since it both results from and encourages) frequency of activation, and statistical

ounts of occurrences in usage provid

a useful index to prototypicality.

In Figure 1, the relative prototypicality of

PPLES an

ORANGES is represented by

r-lin

ed bo

ch they are enclosed, and the same convention will be

used in other diagrams.

.4 Access to the store o

conventional knowled

e, includin

nei

hborin

structures

G holds that meanin

s are enc

clopedi

the

include, in their de

ree, all that

he users of the language know about the concept

and

know that each other knows

(i.e. all conventional knowledge related to th

concept). This includes an essentially

open-ended list of potentially relevant f

acts and relationships, of course, and

precludes our ever being able to presume that we have exhaustively defined a

eaning. What it buys is realism, and an understanding of how meanings can link

p to ‘make sense’ through highly peripher

as well as highly central and salient

eanin

specifications. Representations like those in Fi

ure 1,

ivin

raphic

hape indication and verbal indications of a few specifications, are thus not to be

aken as full

adequate characterizations, thou

h the

are a bit better than

ust a

lik

PPLES in ca

ital letters. Note that

f th

m i

link

Despite what was posited for illustrative purposes in

he previous section, I suppose that 5.c (and/o

omething closely similar that would include peaches

pears, and plums) is relatively well established

for most American English speakers.