Beruflich Dokumente
Kultur Dokumente
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The New School is collaborating with JSTOR to digitize, preserve and extend access to Social Research.
http://www.jstor.org
Noam Chomsky's/
/
Linguistic
T IGO ry
/BY JAMESHIGGINBOTHAM
A WENTY-FiVE
of Noam
yearshavepassedsincethepublication
x
a slendervolumethatis widely
Structures,
Syntactic
Chomsky's
concededto have inaugurateda revolutionin linguistics.
The
period has witnessedmanytrendsand tidesof thought,and
ownviewshave notstoodstill.In booksand articles
Chomsky's
he has expoundedand defendedhis ideas on mattersof general scientific
and philosophicalinterestno less than on particularquestionsof linguisticanalysis.His mostrecentbook,
Lectures
on Government
andBinding,2
is thebroadestin scope of
on linguistics
sincethe 1965 Aspects
any of his writings
ofthe
3
4
Theoryof Syntax. The title essay of Rules and Representations,
144
SOCIAL
RESEARCH
State
Languagea Cognitive
this
a personwho knowsa languagehiscompetence,
contrasting
to verbalbehavior,or
withhisverbalbehaviorand dispositions
here seems
In retrospectChomsky'sterminology
performance.
The term"competence"suggests
to have been unfortunate.
thatthepossessorofcompetencepossessesa skillof somesort;
and "performance"
suggestsa domain of actual
correlatively
behaviorthatfallsshortin variousrespectsof being ideally
"competent."Both suggestionsare misleading.The contrast
betweencompetenceand performanceis a contrastbetween
CHOMSKY
145
146
SOCIAL
RESEARCH
to a more nearly
as we may see by switching
terminological,
It
is
essential
to
neutralinitialposition.
Chomsky'sresearch
programthat it be true or false to attributegrammarsto
personsas descriptionsof theircognitivestates.Simplifying
is thatpersonsare in thestateswe
theassumption
onlyslightly,
commonlycall "knowingEnglish,"or "knowingChinese,"etc.
G are to be attributed
to them.To
just whencertaingrammars
characterize
competenceis to solveforG. We mayat
linguistic
thispointleave open the questionjust whatrelationa person
withgrammarG standsin to G itself;but it mustbe trueor
falsetosaythathe hasG. The philosophical
questionis whether
as a case of knowledge.
thisrelationis interpretable
whether
the
grammarsare in somesense
Although question
it is not as significantfor the
knownis not insubstantial,
the
as
of
practice linguistics
priorquestionwhetherattributionsof grammarsare a corrector fruitful
way to describe
cognitivestatesin the firstplace. The questionof fruitfulness
but it will be
can certainlybe answeredin the affirmative;
usefulto describethe workingsof grammarsin more detail
beforecontrasting
Chomsky'sapproachwithothersthathave
been suggested.
to the varietyof
I willfollowcustomary
usage in referring
thatemergedchieflyin consequenceof Chomsky's
linguistics
work as "generativegrammar."What does a generative
grammarof a languagesay about it,and how do generative
grammarsdifferfrom traditionalgrammars?Quite apart
and specifically
or metatheory,
fromquestionsof methodology
of
own
from
interpretation his enterprise,
Chomsky's
apart
thereis an importantrespectin whichgenerativegrammaris
an intellectualnovelty:it is the firsttypeof linguistictheory
whose avowed aim is to make grammaticaldescription
fully
that
in
observation
the
There
is
generative
explicit.
justice
CHOMSKY
147
148
SOCIAL
RESEARCH
CHOMSKY
149
The descriptions
providedbytheascriptionof grammarsto
of cogpersonsare, we have seen,to be abstractdescriptions
nitivestates of those persons. Now, these cognitivestates
doubtlessadmit of physicaldescriptionsas well, and surely
mustbe countedas havingthe cognitivepowersthattheydo
in virtueof theirphysicalorganization.I say"surely,"thereby
a positionthatis now as formerly
acquiescingin physicalism,
doubt,and controversy.
subjectto interpretation,
Supposing,
that
is in some sense true,we can bring
however,
physicalism
out a distinctive
featureof Chomsky'stypeof inquiry,a featurethatis bornealso byFreud'saccountsof mentallife.This
featureis the thesisthat the theoryof mind can fruitfully
proceedin theabsenceof all butthemosttenuousconnections
betweenits typeof descriptionsof cognitivestatesand their
in
physicalembodiments.
Chomsky'stheoryis thusmentalistic
a doublesense,abstracting
bothfromthedirectexplanationof
behaviorand fromthe physicalunderpinningsof the states
thatit is the theory'saim to describe.
It is interesting
to contrastChomsky'sadvocacyof mentalistic linguistics
witha view thathas been workedout in some
detail,namelythatof W. V. Quine as expressedin
systematic
severalof hiswritings
overtheyears.Quine's view,as I understandit, is thatexplanationof the growthof human knowl-
150
SOCIAL
RESEARCH
as Theory-Construction
LanguageAcquisition
CHOMSKY
151
If grammaris viewedas the outcomeof linguisticexperience, the experienceservingto convertthe stateof the child
fromignoranceto knowledgeof language,then thereis an
initialstateon whichthisexperienceactsin some determinate
Univergrammar.
way.The initialstateChomskycallsuniversal
sal grammarrepresentsthe contribution
of the child to the
state
attained
on
the
basis
of
cognitive
experience;it is, by
about both
definition,
innate,and mustincludeinformation
whatgrammarsare possiblefor human languagesand how
grammarsfromamong the possibleones are to be selected.
Chomsky'sprogramof research,fullyrealized,would characterizeboththe grammarsof humanlanguagesand the selection to be ascribedto universalgrammar.
Withinlinguistics
itself,whatnow principallydistinguishes
Chomsky'spositionand type of researchis the thesisjust
outlined,thatthe fundamentalaim of linguistictheoryis to
explain(insofaras explanationis possiblewiththetoolsavailable,and withinthelimitsof thedegreeof abstraction
presently
required)the acquisitionof language by normalhuman beings. From this perspective,the descriptionof featuresof
languageis neveran end in itself;rather,it is at bestpreliminaryto thetaskof deducingthosefeaturesfromthestructure
ascribedto universalgrammar,underthe conditionsof exposure to language thatchildrentypically
undergo.
The term"universalgrammar"misleadingly
suggeststhat
the studyof universalgrammarwould intimately,
or perhaps
exclusively,involvetaking a principledinventoryof those
featuresthatgrammarsof humanlanguageshave in common.
It may thereforebe worth stressingthat, understood in
Chomsky'ssense,universalgrammaris nothingelse but the
initialstateof the human language-learner.
This initialstate
well
involve
that
factors
determine
universal
featuresof
may
consistin princilanguage,but mayalso,and even principally,
that
select
of
forms
that
are verydifples
among
grammars
ferentfromeach other.To put the point anotherway,the
cogencyof Chomsky'sprogramis not underminedby obser-
152
SOCIAL
RESEARCH
CHOMSKY
153
154
SOCIAL
RESEARCH
a theory
questionof acquisitionmoreclosely,by formulating
of comparative
syntax thatis, a theoryof the wayslanguages
in theirsyntactic
and
do
differ
organization.
may
Given the psychologicalorientationof Chomsky'stheory,
comparativesyntaxfor him becomesprimarilythe studyof
availainformation
howthechild,on thebasisof thelinguistic
the syntaxof the languageto which
ble to him,distinguishes
A good theoryin
he is exposedfromotheradmissiblesystems.
thata
thisdomain,Chomskyargues,shouldhave theproperty
fewdetectablefeaturesof a languageshouldsufficeto fixthe
rules.A simple
of a hostof grammatical
formand functioning
image may help to conveyhow such a theorymightwork.
Imaginethata grammaris selected(apartfromthe meanings
of individual words) by setting a small number of
- 20, say- either"On" or "Off."Linguisticinformaswitches
howtheseswitchesare to
tionavailableto thechilddetermines
be set. In thatcase, a huge numberof differentgrammars
availa(here,2 to thetwentieth
power)willbe prelinguistically
to fix
suffice
ble, althougha smallamountof experiencemay
one.
of themetaphoraboveare in Chomsky's
The switch-settings
the "parameters"definedby universalgrammar.
terminology
Noticethatthisimageunderscoresthesense in whichuniverneed not
theinitialstateof thelanguage-learner,
sal grammar,
- to
in
common
have
comprisean accountof whatlanguages
could give
continuethe metaphor,differentswitch-settings
rise to verydifferent
systems.
grammatical
then
thisperspective,
from
If one viewscomparative
syntax
CHOMSKY
155
grammatical
analysesthatmightbe formulatedfor,say, Ento
glishought have the propertythattheymeshwithanalyses
of similaror interestingly
different
phenomenain otherlanguages. For a concreteexample, consider that in English,
correspondingto the sentence(1), there are two formsof
directquestion,namely(2) and (3):
You boughtthe book forJohn.
(1)
Who did you buy the book for?
(2)
For whomdid you buy the book?
(3)
In French,however,only the formcorrespondingto (3) is
permitted(in otherwords,the form"Qui avez-vousachetle
livrepour?" is ungrammatical).
Any analysis,therefore,that
and
makesboth (2)
(3) routinelyavailableto the learnerof
Englishis likelyto be wrong,because it would not contribute
to the explanationof whyonly one of these formsexistsin
French.Withrespectto these forms,in fact,it appears that
Frenchis the normamonglanguages,and Englishthe exception. The problem,then,is to explain why English should
admitformslike(2). This is not the place to discusssolutions
thathave been proposed- whatis to be notedis thatthestatus
of (2) as a problem
is directlydependentupon the incorporation of the analyticaltask of linguisticdescriptionwithina
broaderprogramof the explanationof language acquisiton.
The broader programhas arguablymade some progress,
motivatedby the analysis of examples like those above.
Whetherthatprogramwillmakeprogressin itsownterms,or
indeed whetherthose termswillnot in timebe transformed
out of all presentrecognition,
remainsto be seen.
Chomskyys
Influence
156
SOCIAL
RESEARCH
CHOMSKY
157
and thatprimitive
tivebeginnings,
peoples speak moreprimia
reflex
of thisgeneralequationof
tivelythanwe, is perhaps
The
with
culture.
growthof languageseems,on this
language
and not,
view,properlyanalogousto thegrowthof civilization,
as in Chomsky'smetaphorof mentalorgans,analogousto the
growthof liver.This attitudetowardlanguage,I think,can
make Chomsky'sviewsseem bizarre.
The second point,whetherChomsky'sresearchprogramis
is muchdebated,particularly
scientific,
respectably
by philosI
on
some
features
of
have
remarked
this
debate
ophers.
above: the questionwhetherlinguisticcompetenceis knowland studiedin
edge,whethercognitivestatescan be identified
from
theirphsyicalunderpinnings
and
comparativeisolation
behavioralcorrelates,and others.Remarkably,
therehas been
littlediscussionof the details of grammaticaltheoryitself;
mostof the criticalremarkshave been externalto it, rather
than fromwithin.Not thatphilosophers,not expertsin the
sciences,should refrainfrom tryingto formulategeneral
criteriafortheevaluationof scientific
achievement
and underWe
all
about
the
sciences
standing.
speculate
anyway,and the
hope is thatphilosopherswill do it more clearlythan most.
But I am inclinedto thinkthat,until Chomsky'stheoryis
morecritically
examinedin itsown terms,whatit mayhave to
teach us, throughits successesand its failures,will not have
been taken up withinthe broader contextof our effortto
obtaina betterunderstanding
of humanknowledge,thought,
and discourse.