UNIT 3 LOOKING AT DATA-2

Structure
Objectives
Formal Linguistics - An Introduction
Generative Grammar
3.2.1 Principal Goals
Generativists and Structuralists
3.3.1 Generativists and Bloomfieldians
Tranformational Generative Grammar
Let Us Sum Up
Key Words
Questions
Suggested Readings

3.0 OBJECTIVES

, In this unit, we will discuss

o the Generative framework of grammar

o its differences with the Structuralists
We will also briefly talk about Transformational Generative Grammar.

Some of these ideas may appear difficult but they will be clear to you as you read the
course. Don't get discouraged if some concepts appear difficult and complex. They
will be clarified as we proceed along.

3.1 FORMAL LINGLTISTICS - AN INTRODUCTION

In the early 1950fs,signs of restiveness began to disturb the calm of structuralism,

and by the end of the decahe new ideas emerged in a big way. Chomsky, a student of
Zellig Harris was concerned with discovering a general theory of grammatical
structure. He believed that an adequate grammar should provide a basis for
explaining how sentences are used and understood. He reproaches the
Bloonlfieldeans for "their satisfaction with description [and] their refusal to explain"
( 198 1:38). According to him, as other 'developing sciences, linguistics should also
endeavour to establish a more ambitious goal than mere description and
classification. Linguists should aim at developing methods not just for the description
of language but also for understanding the nature of language. And this was
possible only if one takes recourse to intuition of native speakers. Intuition had,
however, remained a source of discomfiture for Bloomfield. American structuralists
have maintained a Bloomfieldian mistrust of meaning. They held the opinion that for
describing and classifying the forms of expression in a language it was unnecessary -
nay misleading to have a knowledge of the meanings. All they considered necessary
was to know which utterances are possible, which are ambiguous and which ones are
syntactically related to each other. This, they believed, could reduce the dependence
on the intuitions of native speakers (which most of them recognized, was
unavoidable). *,

Recognition of possible utterances provides a basis for making a formal analysis of

language. Making a distinction between several aspects of this "possibility" is
necessary because an utterance may be appropriate in some circumstances and not in
others depending upon the situation. An utterance may be wellformed but not
acceptable, for acceptability involves wellformedness together with
appropriateness to context-linguistic or situational, e.g. an utterance Colorlrss
green ideas sleep furiously is completely nonsensical but grammatically correct.
Judgements on the wellformedness of expression cannot simply be based on
observable occurrences of utterances in the speech of native speakers, As Chonisky
points out, the mere occurrence of an utterance, even wellfolmed, does not make it
more representative, because there are countless other possible utterances which may
not have occurred during the observation. This makes the recourse to native speakers
indispensable. This could not be obtained through discovery procedure. Also, if one
were to look into the syntactic relatedness between two sentences, an immediate
constituent analysis would fail to tell us anything about an underlying kin relationship
between the active and passive voice : Mary sees George-George is seen by Mczry.
The flaws in the structuralist edifice brought a shift in n~ethodologyand orientation 111
linguistic theory and initiated an all-encompassing theory of language in which the
whole assumes primacy over the parts. Intuitions and judgement of native
speaker of language became pivotal for revealing the underlying relationship
between sentences and parts of sentences. The speaker of a language, not the text or a
corpus, was considered a source of all linguistic studies. This creative aspect of
language i.e. ability to produce or create 'novel' sentences which slhe might not have
heard before became a central concern of language study. Behaviourists' postulate
that la langue is a "system of habi'ts" was replaced by the idealist position of "~nnate
ideas" and the belief that human beings are born with the innate capacity to learn
languages. The theories of universal grammar were outlined to account for the
creativity and the recursive property in language.
The theory of language which came to be associated with the shift in 'orientation' has
been referred to as generative grammar.

3.2 GENERATIVE GRAMMAR -

It is the theory of language proposed by Chomsky in his Syntactic Structul-es (1 957).

It provides a set of finite rules that defines the unlimited number of sentences of the
language and associates each with an appropriate grammatical description.

3.2.1 Principal goals

There are two principal goals which underline this theory. These are :

(a) The universal features (i.e. features which are intrinsic to language as a whole)
which constitute grammars of individual language should be characterized III
formal terms.
(b) Formal statements should be provided for characterizing the graminars of
individual languages. This goal is equated with characterizing the tacit
knowledge or competence which native speakers have about syntactic,
phonological, morphological and semantic patterning in their language.
Generative grammar sees the theory of Competence as forming a central
component of language which interacts with principles from cognition.
neurology, physiology and other domains to give language its overall character,

Generative grammar has its roots firmly grounded in the structuralist tradition.
Generativists share with structuralists the idea that "the grammar of a la~iguageis a
 statement of'the systematic structural interrelationships holding between linguistic Looking at Data-2
(Newmeyer 1992 : 46). Even Chomsky's notion's of 'competence' and
are in many ways modem reinterpretations of Saussure's classic

t
dis 'nction between 'langue' and 'parole'. However, there are differences between
gen rativists and structuralists-the most significant being Chomsky's reinterpretation
of the goals of linguistic theory. He proposed a novel conception of what a linguistic
theory actually addresses. While the structuralists' goal of linguistics was to
construct inventories of the linguistic elements in particular languages, alongwith
statements of their distributions, Chomsky believed that the goal of linguistics has to
be redefined in order to provide a rigorous and formal characterization of a "possible
human language" i.e. specification of a universal grammar (UG). This UG is innate
to human mind. He has gone to the extent of characterizing linguistics as a branch of
cognitive psychology.

3.3.1 Generativists and Bloomfieldians

Generativists had certain differences with Bloomfieldians. Chomsky reacted against
the (post-) Bloomfieldian's "separation of level" principle (phonemic, morphemic,
etc.), that is, an analytical, structural description that breaks the utterance down into
watertight layers. This principle gives functional independence to each level i.e. one
need not refer to morphology if one is conducting a phonemic study. Chomsky,
however, believes that the opposite is true. Further, Chomsky believes that the post-
Bloon~fieldianlinguistics does not "take the speaker and his role in the constitution of
the utterance into account; instead it proposes an empirical description, which claims
to be 'neutral' and 'objective', of the spoken chain in itself' (Kristeva 1989: 253).

Nevertheless, generative grammar owes a considerable historical debt to post-

Bloornfieldians. Chomsky has been a student of Zellig Harris, one of the leading
post-Rloomfieldian, and this certainly had an impact on him. He remained faithful to
r.l~t:post-Bloomfieldian demands of rigor, neutral and formal descriptions, as well as
to their mistrust of the meaning (which has led to an emergence of a doctrine that has
come to be h o w n as the 'autonomy of syntax'). He took up certain concepts given
by his teacher, Harris, his inspired precursor-and gave them a new interpretation (for
example, the notion of transformation, which in the sense of Harris is an inter-
sentential phenomenon, while in Chomskian sense it is an intrasentential
phenomenon). He also interpreted many features of the American structuralism in a
new fashion , as for example his notion of deep structure could be traced back to
Sapir's inner-form. Despite the resemblances to his predecessors, there was an
element of novelty in Chomskian theory.

As opposed to post-Bloomiieldian's analytical approach to structure, Chomsky

proposed a synthetic description. He believed that instead of breaking down the
sentence into immediate constituents, "one should follow the synthetic process that
leads these constituents to a syntagmatic structure, or transfonns this structure into
another one" (Kristeva 1989 : 253-54).

In this operation, the implicit linguistic intuition of a speaker becomes, according to

Chomsky, the only criterion for the grammaticality or ungrammaticality of a
sentence. To quote Chomsky: "The fundamental aim in .the linguistic analysis of a
language L is to separate the grammatical sequences which are the sentences of L
froin the ungrammatical sequences which are not the sentences of L and to study the
structure of the granlmatical sequences"(Chomsky 1957:13). Chomsky noticed that
the notion of grammaticality could not be identified with that of "meaningful" or
"significant" in any semantic sense, as we can see in the following two sentences:

I. Colorless green ideas sleep furiously

2. Furiously sleep ideas green colorless
is For an English speaker, while both (1) and (2) are nonsensical, the first is
grammatical and the second is not. Chomsky did not suggest "meaningful" to be the
criterion for identification because he believed that "semantic theories have not been
made sufficiently exact to justify proposing a notion of meaning that could give us a
practical way of deciding on the grammatical status of pairs of sentences [given
above]" (Dinneen 1967:362). From such considerations Chomslcy notes that "we are
forced to conclude that grammar is autonomous and independent of meaning"
(1957:15 and 19).

Chomsky also devoted himself to looking into the highly abstract theory capable of
finding universal formalism valid for all languages "with no specific reference to
particular languages" (1957 : 11). He looked at grammar more as a theory of la

believed tbat the "grammar of a given language must be considered in accordance

with the specific theory oflinguistic structure in which terms like 'phoneme' and
'syntagm' are defined independently of any particular language" (Kristeva 1989:
i
langue leading to a "condition of generality" and less as an eimpirical description. I e

255).

How dld Chomsky establish the rules of his theory ? He went about doing it by
examining two models of grammatical description namely, finite state grammars
and phrase structures grammars. The former bore a close resemblance to the type
of device promoted by communication theorists. It was rejected because it could not
explain the speaker's ability to produce and understand new utterances. The kind of
descriptions which phrase structure grammars provided were identical to of the post -
Bloomfieldians' procedures (resembling IC analysis)-the way sentences are broken
into parts. The tree diagram is used, but turned upside down, since the progression IS
from the sentence to the parts, as in case of the following sentences : the malt hi! !he
ball shown in Figure-1 .

Fig. 1
Step-I Sentence (S)

nzan hit Art N

t Ithe I
ball
Here the step (I) basically gives a summary diagram. There are two other steps also
which can be applied on the same sentence.

(i) Sentence >NP+VP

(ii) NP >(Art)+ N
(iii)W 'v + (NP)
(iv) Art (article) >the
(v) N (noun) . >man, ball
(vi) V (verb) >hit

Sentence
NP+w (9
Art+N+VP (ii)
 I
Art+N+V+NP (iii) Looking at Data-2
The +N+V+NP (14
The+man+V+NP (v)
The+man+hit+NF' (vi)
The+man+hit+Art+N (vii)
The+man+hit+the+N (viii)
The+man+hit+the+ball (ix)

Here the steps (11) and(111) respectively, provide a grammatical analysis and
information regarding derivation.

Chomsky himself rejected this type of grammatical description, as he realized that it

was inadequate for describing the structure of English sentences. By means of several
examples he showed the limitations of phrase structure grammars.
The first example of the weakness inherent in phrase structure grammars can be seen
in the ordinary formation of new sentence by conjunctions. If one had two sentences
Z+X+W and Z+Y+W, and if X and Y are the "constituents" of these sentences, one
could in principle form a new sentence, such as :

e.g. 1 (i) (a) The scene-of the movie-was in Chicago

(b) The scene-of the play-was in Chicago
(ii)(c) The scene-of the movie and of the play-was
in Chicago.

If, however, X and Y are not the constituents, then the formula cannot be applied, for
it would produce, for example :

2 (i) (a) The -liner sailed down the-river

(b) The-tugboat chugged up the-river
(ii)(c) The-liner sailed down the and the tugboat chugged up the -river.
It is clear from the above examples (i) and (ii) that in order to apply syntagmatic rules
to a language like English, one needs to know not only the final shape of sentences
but also the structure of their constituents at the time of application i.e. the "history of
derivationV(p37).
Similar weaknesses in phrase structure grammar may be seen when attempting to
deal with fonns of the verbs other than the simple fonns produced by rewriting rules
l~keVerb----> hit or take, since we may be required to have in other contexts, forms
like takes, has+taken, willi-take, and so on. Limitations may also be seen in the
t treatment of the active-passive relation, statements and questions, emphatic
utterances (Idid tell them) and unemphatic ones ( I told them) and so on.
?'he inadequacies of the phrase structure grammars for a language like English
demand incoporating new rules into it. But doing that completely changes the
conception of the linguistic structure. Henbe the concept of a "grammatical
transformation" was proposed by Chomsky, which he formulated as : "a grammatical
trasformation T operates on a given string (or. . . on a set of strings) with a given
constituent structure and converts it into a new sting with a new derived constituent
structure" (1957:44). This introduction of the concept of a transformation has led to
referring to the entire formal approach as Transformational Generative Grammar.
i

I 3.4 TRANSFORMATIONAL GENERATIVE GRAMMAR-

1957 MODEL
A transformatiin is a way of specifying, by rule, the relationships between sentence
I types. Transfomiatians perform two types of functions: 29

I
 What is Lnnguage?
(a) They change underlying grammatical relations, as in the case of passives
which are derived from actives, for example. John saw Jill-->Jill 1vu.s .see,,
by John. Since the relationship applies to the syntactic elements, not just to
these particular words, it may be algebraically expressed as :

This may be verbalized as : The two noun phrases exchange places, with by placed
before the one that now comes last. The tense (Aux) remains thc same (past sa1.r:
matches past saw), but a form of be is inserted and the verb takes its past participle
(en) form : was seen by applying the A@ hopping rules, under which the affix
moves to a place immediately after the V.

These transformations are referred to as singularly (simple) transformations. They

include transfomlations like Passive, Auxilliary, Negative.

(b) As a second function, transfonnations create corrlplex sentences out of simple

ones, as in the case of embedding, e.g. the sentence John t~ziizl&that Bill will
leave was derived by an embedding transfom~ationthat combined Johrr
thiriks and Bill will leave.

These transfonnations are referred to as generalized (double-based)

transformations.

In formulating the principle of transformational grammar, Chomsky clarified

essential properties of transformations, for example, the order in which these
transformations are applied. Two types of rule ordering were identified - Extriruic-
ordering and Intrinsic ordering. The foniier explicitly specifies the order in which
the rules must be applied, while in the latter the ordering is a consequence of the
way in which the rules are stated. Chomsky further argued that at least some
transformations had to be strictly ordered with respect to each other in the grammar
i.e, transformation rules are lined up in a certain order. For example, reilexivization
must apply before 'you' deletion:

(a) Reflexivization
(b) 'you' deletion

Where there are identical NPs in the same sentence, the second NP is
changed into reflexive by applying rule (a):

NP, NP,
You - hang YOU You hang yourself
Then rule (b) is applied to arrive at
Hang yourself

If the order is changed, it will result in giving a sentence 'hang you' and then
there is no way of reflexivizing 'you'.

He also pointed out that some transformations are 'obligatory while others are
optional. The Auxiliary transformation and the Do transforlllation are obligatory
while passives, negatives, imperatives, Wh-and yes-no cluestions are optlonal.
Sentences that are produced by applylng obligatory transformations are called kernel
strings while the sentences obtained by applying optlonal transformations are called
derived.

-
With regard to the question of grammar and meaning, Chomsky felt that "grammar 1s
30 autonomous and independent of meaning" (1957: 17). However, his insistence on the
independence of grammar of meaning is not in tune with post-Bloomfieldian I

I
structuralism. He was clear that the question of the relation of grammar and meaning Looking at Data-2
is an empirical one and he gave many examples to illustrate his position.
lvevertheless, the role of syntax remains crucial in determining the meaning. This can
be seen in case of handling ambiguity and paraphrases, which are semantic notions.
The ambiguity of the sentences, according to him, can easily be resolved by
transformational description by merely reestablishing the transformational rules that
produced it. To quote Chomsky : ". . . if a certain sentence S is ambiguous, we can
test the adequacy of a given linguistic theory by asking whether or not the simplest
grammar constructible in terms of this theory for the language in question
automatically provides distinct ways of generating the sentence S" (1957 :123),

Thus, the Chomskian approach offered a dynamic vision of syntagmatic structure that
was missing in structural grammar. It also eliminated the atomization of la langue
that accoinpanied post -Bloomfieldian methods. Instead, it suggested a processual
conception of la langue in which "each sequence of rules stems fiom a coherent
whole centered on the consciousness of the subject - locuter whose freedom consists
of submitting to the norms of grammaticality" (Kristeva 1989:259).

You will read more on Chomskyan linguistics in Block 5 .

3.5 LET US SUM UP

In this unit, we gave you another point of view fiom that of the structuralist -
behaviourist. .We gave you the reasons for the dissatisfaction with the structuralists.
We discussed the main points of generative framework, and its differences from the
structuralists.

We have given you several examples and used some technical terminology. Do not
be intimidated by it. You will understand all that is discussed by the time you finish
the course.

3.6 KEY WORDS

Intuition: A judgement *hich you make about your

own language (whether something is
grammatical or not), what it means, whether
it is ambiguous or not, how it is related to
something else, and so on.

The state of being grammatical. A well-

formed sentence in a language is a sentence
which is consistent with all the grammatical
rules of that language. However well-
formedness includes acceptability and
appropriacy.

An approach to psychology which holds that

psychologists should study only observable
and measurable phenomena, and should not
appeal to unobservable things like 'minds'
and 'intentions'. Leonard Bloomfield and
the American structuralists were much
influenced by behaviourist ideas, but ever
since Noam Chomsky's devastating critique
 of B.F. Skinner's behaviourist account of
What is Litngsnge? language acquisition, linguists have
generally rejected behaviourism in favour of
mentalism.

Bloomfield, Leonard: A distinguished and influential American

linguist (1887-1949). Trained in historical
linguistics, Bloomfield dld important work
on Germanic and Austroneasian languages.
and his famous work on Algonqu~an
languages pioneered the application of the
comparative method to native-American
languages. His 1933 book Language '
revolutionized linguistic thinking and his
ideas were developed by his colleagues and
successors into American structul.alism,
which donlinated the Amer~canl~nguistic
scene until about 1960.

Generative grammar: 1. A particular grammar of a pai-ticular

language which, in a purely mechanical way,
is capable of enumerating all and only the
grammatical sentences of that language.
Generative gralnar in this sense was
introduced by Noam Chomsky in the 1950s.

2. Any theory of grammar which has as its

goal the construction of such gl-amliiars.

3. The enterprise of constructing s ~ ~ c l i

theories of grammar.

Universal language: Any property which is present in all

languages or in nearly & languages
l
example: all langu'ages distinguish nouns &
-
verbs.

Competence: An idealisation of a speaker's knowledge 01'

herhis language, excluding s'ucli tictors as
slips of tongue, memory limitations or
distractions.

Phonology: Pertaining to the sound system.

Morphology: Pertaining to word-formation '

Semantics: Pertaining to meaning

Performance: The actual linguistic behaviour of particular

individuals on particular occasions, .
including any hesitations, memory lapses, !.
'
slips of the tongue or processing difficulties
arising from long,or complty 'structures.

Universal Grammar (UG): The hypothetical structural properties which

are necessarily common to all human I

languages, both real and possible,

' 32 presumably because these properties are part
of the human language faculty. The tenn is
, ,
 particularly associated with the work of Loolung at Data-2
Noam Chomsky and his followers; not all
linguists are convinced that UG exists.

Transformation: In a formal grammar, a type of

grammatical rule which has the power to
change the structure of a sentence which is
being generated by the grammar, for
example by deleting something or by
moving it to a different position.

Transformational grammar (TG): A theory of grammar developed by Noam

Chomsky in the 1950s and extensively
modified by Chomsky and others in the
succeeding decades.

3.7 QUESTIONS

1. What is the difference between the Generativists and the Structuralists? In

what ways have the Generativists made advancements on the ~trucbralists?

2. "The relationship between the signifier and the signified is arbitrary". Are
there any exceptions to this rule? Think of some words in English and your
mother tongue in which the relationship between the signifier and the
signified is not arbitrary, but is based an some similarity between them,

3. 'Noun is the name of a person, place or thing' Do you think that this definition
is adequate? What about the words like investigation, division,
congratulation? Are they the name sf a person, place or thing? The
Struturalist approach to language provides a better alternative definition. Can
you define noun using its distributional pattern in language?

3.8 SUGGESTED READINGS

Benveniste, E. 1971 Problems in General Linguistics, by Mary E. Meek, University

of Miami Press.
.
Hloon~field,L. 1933 Language, New York: Henry Holt
Chomsky, N. 1957 Syntactic Structures, The Hague: Mouton.
Dinneen, F. P. 1967 An Introduction to General Linguistics, New York: Holt,
Rinehart and Winston:.
'
Hymes, D. and J, Fought 1975 American Structuralism, The Hague: Mouton.
Ivic, M. 1965 Trends in Linguistics, The Hague: Mouton
Kristeva, J. 1989 Language- The Unknown: Aiz Initiation into Linguistics. Tr. by
Anne M. Meaker, London, Sydney and Tokyo: Harvaster Wheatsheaf.
1-epschy,G. 1992 : "Early Structuralism", in W, Bright (ed) International
Encyclopedia of Linguistics, Oxford: Oxford University Press.
Newmeyer, F. J. 1986 Linguistic theory in America, New York: Academic Press, Inc.
, --_-_____________
1992 " Generative grammar" in Bright (ed)
Rastogi, K. 1997 Structural Linguistics: Its Origin and Development, Delhi: Penman
1
Publishers.
Saussure, F. de 1 959 Course in General Linguistics, McGraw Hill.
Thibaul t7P.J, 1998 "Structuralism" in P. Bouissac (ed) Encyclopedia of Semiotics,
Oxford: Oxford University Press,
I 33