Sunteți pe pagina 1din 2

LINGUISTIC THEORY, NEUROLINGUISTICS

AND SECOND LANGUAGE ACQUISITION*

Lisa Travis
McGill University

I. INTRODUCTION

Several papers within this volume have raised the question of whether the
adult second language learner has access to the principles and parameters
of universal grammar (UG). Flynn, Mazurkewich, and White all suggest that
the principles involved in first language acquisition are indeed at work in
second language learning. Clahsen, on the other hand. argues that adults
learners resort to other learning strategies in an effort to systematize
the data of the second language. In this paper I suggest that a very
different type of evidence may be brought to bear on this issue. The claim
is that agrammatic speech of a Broca's aphasic who has learned a second
language as an adult may help tease apart the relevant issues. The paper is
intended not so much as an answer to the question of the rule of UG in
second language acquisition as much as a justification for a possible line
of inquiry.
The organization of the paper is as follows. Since the theory which I
will be assuming is that of Government and Binding (Chomsky. 1981 and
references therein), in Section 2 I give a brief overview of the theoreti-
cal notions that are needed for an understanding of the issues which I
will raise. In Section 3 I take a detour through certain neurolinguistic
literature in order to justify the stance that I take towards neurolinguis-
tic data. It is crucial to my argument concerning the problem of second
language acquisition that neurolinguistic data be used as evidence for the
mental representation of linguistic knowledge. Finally in Section 4, I
raise the question again of whether any language specific learning compon-
ent is responsible for the acquisition of a second language in the adult
learner and show how neurolinguistic data may provide an answer.

2. THEORETICAL BACKGROUND

A recent shift in the goal of generative grammar has moved the focus of the
linguist's task from the formulation of the rules to formulation of the
principles and parameters necessary to account for the mental represent-
ation of natural language. For instance. powerful transformational rule
systems such as those described in Chomsky (1957) have been replaced by

90
S. Flynn and W. O'Neil (eds.), Linguistic Theory in Second Language Acquisition, 90-108.
r&l1988 by Kluwer ACfldemic Publishers.
NEUROLINGUISTICS 91

interacting modules which contain cel1ain language universal principles and


language specific parameters. Movement now is contrained through conditions
on empty categories and binding. Just as transformational rules receive a
different handling in this new framework, so do phrase structure rules.
Phrase markers are represented through the interaction of principles and
parameters of the theory rather than through a context free production
grammar.
This shift has an effect on the linguist's view of word order. In the
EST framework, word order was represented through Phrase Structure rules of
the type given in (I) creating the structures given in (2).

(I)
a. VP ---> V NP
b. VP ---> NP V

(2)
a. VP b. VP
1\ I \
V NP NP V

However, since emphasis has been shifted from rules to principles and
parameters, the precedence and dominance relations previously encoded in P~
Rules are encoded within GB theory in (i) principles (such as X'-theory,
the Principle of Full Interpretation, the Theta Criterion, and the Rule of
Predicate Linking), (ii) parameters (such as headed ness , direction of
theta-role assignment, case assignment, and predication). and (iii) lexical
properties (such as argument structure and complementation requirements).
With the interaction of principles, parameters, and the lexicon, the phrase
structure rules of (I) are no longer needed. The fact that a VP has a
daughter V is dictated by X'-theory (every maximal projection must have a
head). The presence of the NP will be insured by the argument structure of
the V (devour will require an NP. die will not).
Language specific word order will be determined by the setting of
parameters such as headed ness , direction of Case assignment. direction of
theta-role assignment. and direction of predication (see Travis 1984, for a
slightly different account. Koopman 1984). With these parameters. we may
describe three different domains within the VP as shown in (3) below.

(3)
V NP PPI PP2 PPI = argument PP
~ PP2 = non-argument PP
Case
"-.r-'
theta-role
~
headedness

In this account of word order, if all the elements of the VP precede


the verb as in (4a) , the only relevant parameter is the headedness para-
meter: the VP is head-final. If, however, the object NP follows the V while

S-ar putea să vă placă și