The Role of Innate Knowledge in First and Second
Language Acquisition
by Hasanbey Ellidokuzošlu
I. INTRODUCTION
That some kinds of migratory birds navigate thousands of miles toward their
destination by calibrating the positions of stars against time of day and year,
poses no serious problem for many scientists, who can easily attribute this
amazing success to the birds' instinctive behavior (Pinker, 1994, p.19). It is
apparent, after all, that these animals cannot learn such complicated
astronomical facts through a trial and error fashion; they neither have enough
time nor necessary cognitive capacity.
The same scientists, however, including some professional linguists, are quite
reluctant to attribute any form of instinct to human infant, who arrives at
complex linguistic knowledge within a remarkably short period of time
(Karmiloff-Smith, 1992, p.1). The infant's is no less a complicated task than
that of the bird's as the linguists themselves have spent decades (or even
centuries) to discover the intricacies of the very same system and with no final
theory. Infants, on the other hand, not only arrives at an almost complete
knowledge of grammar in their brinds (brain+mind) but also accomplish this task
within less than a decade.
Although a human infant and a migratory bird are essentially alike in terms of
the complexity of the task to be accomplished and their inability to handle the
task with their current cognitive capacity, only the latter is believed to rely
on its instincts.
There are, of course, some differences between an animal and a human baby; it
would be unwise to equate the cognitive capacities of the two. And it is also
impossible to underscore the importance of environmental factors in child
language acquisition. After all, thousands of hours of exposure is required in
order for a child to acquire his mother tongue, whereas animals like sonar-using
bats or web-building spiders seem to be ready to use their instinctive knowledge
with minimum, if any, learning experience. It is equally unwise, however, to
suggest that a cognitively immature child can accomplish a task which has yet to
be accomplished by professional linguists.
[A] child may well not have grasped the property of conservation of volume nor
be able to perform but the most rudimentary arithmetic calculations, yet will
have the knowledge linguists formulate as the binding principles, none of which
has been explicitly taught. (Carston, 1988, p. 41)
The amazing success of children in picking up their mother tongue is no recent
discovery. Slobin (1979) quotes Rene Descartes commenting on human beings'
disctinctive ability to formulate a linguistic system:
...[E]ven those man born deaf and dumb, lacking the organs which others make use
of in speaking, and at least as badly off as the animals in this respect,
usually invent for themselves some signs by which they make themselves
understood by those who are with them enough to learn their language (p. 113)
In the literature of child language acquisition there are cases in which
infants, deprived of linguistic input, invent a rudimentary grammar not
attributable only to the external factors. Children are also known to build a
natural language when exposed to unsystematic pidgin data (Bickerton, 1981,
1983). The resulting creole is almost as systematic and sophisticated as any
natural human language and more interestingly contain rules that are not
attributable to the languages forming the pidgin, out of which the creole is
driven.
II.
PLATO'S PROBLEM AND CHOMSKY'S SOLUTION
This imbalance between the external input--linguistic data-- and the
output--complex linguistic knowledge-- is called Plato's problem. Chomsky's
solution to the Plato's problem is to seek the richness in the
processor--infant's brind--rather than in external stimuli.
Our knowledge of language is complex and abstract; the experience of language we
receive is limited. Our minds could not create such complex knowledge on the
basis of such sparse information.It must therefore come from somewhere other
than the evidence we encounter; Plato's solution is from memories of prior
existence, Chomsky's from innate properties of the mind. (Cook, 1988, p. 55)
Chomsky believes that child's brind is equipped with the principles and
parameters of the Universal Grammar (UG) which underlies the grammar of any
human language. With the help of this language-specific knowledge children can
figure out roughly what the shape of his or her mother tongue is like. That is,
UG provides a skeleta knowledge upon which the child is supposed to dress the
flesh. To make an analogy, UG is similar to the genetic information in the seed
of a flower and the external linguistic input is similar to the water which
activates this latent information. It is vain to attribute the beauty of a
flower only to the minerals in water, and the complexity of child's grammar to
the external data only. Both water and input act as a trigger working on a rich
genetic blueprint.
III.
INNATE KNOWLEDGE:DOMAIN-SPECIFIC OR GENERAL?
Many scientists, especially the psychologists, however, are hesitant to
attribute a domain-specific innate linguistic knowledge to the human infant.
These psychologists view the human brind as a homogeneous computational system
which analyze varying types of data using general information processing
principles. Postulating a language-specific mechanism within such a
general-purpose computational system is considered to be a violation of Occam's
principle which favors minimum amount of principles to account for maximum
amount of data rather than ad hoc explanations restricted to specific phenomena.
Piaget, being a typical representative of this reductionist paradigm, views
language acquisition as an instance of general human learning with no appeal to
domain-specific innate knowledge. He asks, 'If one wants to introduce innateness
into language, why not introduce it into the symbolic function in its totality,
and finally into anything that is general' (Piaget, 1980, p.167). It is believed
that linguistic concepts are reducible to general cognitive terms:
...Piagetians seek precursors of all aspects of language in the child's
sensorimotor interaction with the environment ....Playing with
containers--embedding objects one into another--is considered a necessary
precursor to the embedding of clauses....Notions such as noun phrase, verb
phrase, subject and clause are ... said not to be available to the young child's
linguistic computations before the acquisition of elaborate cognitive
structures. (Karmiloff-Smith, 1992, p.34)
IV.
PARALLELISM BETWEEN PIAGET AND SKINNER
A profound analysis of Piaget's ideas will reveal the basic similarity between
his reductionist view and behaviorist paradigm. A version of "tabula rasa" can
easily be detected in this modern version of behaviorism. Here are some
interesting remarks of a former Piagetian:
At this juncture I shall risk some of my colleagues at the Geneva University by
suggesting that Piaget and behaviorism have much in common....Neither the
Piagetian nor the behaviorist theory grants the infant any innate structures or
domain-specific knowledge. Each grants only domain-general, biologically
specified processes ....These domain-general learning processes are held to
apply across all areas of linguistic and nonlinguistic cognition. Piaget and the
behaviorists thus concur on a number of conceptions about the initial state of
the infant mind. The behaviorists saw the infant as a tabula rasa with no
built-in knowledge...; Piaget's view of the young infant as assailed by
'undifferentiated and choatic' inputs is substantially the same.
(Karmiloff-Smith, 1992, p.7)
V.
NATIVISM AND OCCAM'S RAZOR
Chomskyan linguist themselves are willing to follow the principle of Occam's
Razor in their theorizing. Such an attitude can easily be seen in their
eagerness to reduce the number of linguistic rules to a minimal set of UG
principles and parameters (Chomsky's latest theory is called "the Minimalist
Program", for instance). But such a minimalist attitude is different from
psychologists' reductionism in that the former seeks the Razor within the
boundaries of language whereas the latter tries to apply it to whole human
cognition. It would, of course, be more economic thus desirable to have minimum
amount of principles capable of accounting for every kind of cognitive
phenomenon, but as Jackendoff (1993) puts it, accepting language-specific,
innate knowledge in the brind of a human neonate is a desparate move:
[W]e're struck between a rock and a hard place. On the one hand, the expressive
variety of language demands a complex mental grammar that linguists can't figure
out. But on the other hand, children can manage to acquire this grammar. Thus,
in a sense the Genetic Hypothesis is a move of desperation. (p.33)
There is some point which should be underlined here: The problem will not be
solved some decades later when linguists are able to figure out the grammar in
its entirety, as some theorists speculate. When we arrive at a super theory with
all its glamour, our desperation will probably be even bigger than it is now.
Because reaching at such a theory will be accomplished after decades of
systematic and collaborative efforts of cognitively mature professional
linguists: after thousands of researches, conferences, discussions, books, etc.
But our curious child will continue acquiring the same grammar within a few
years. It seems Plato will not lose his popularity even in the third millennium.
As for the reducibility of linguistic phenomenon to general cognitive terms, one
should not forget that just like vision or music, language has its own life with
a set of domain-specific concepts:
[D]omain-general sensorimotor development alone cannot explain the acquisition
of language. Syntax does not simply derive from exploratory problem solving with
toys, as some Piagetians claim. Lining up objects does not form the basis for
ord order. Trying to fit one toy inside another has nothing to do with embedded
clauses. General sensorimotor activity alone cannot account for specifically
linguistic constraints; if it could, then it would be difficult to see why
chimpanzees, which manifest rich sensorimotor and representational abilities, do
not acquire anything remotely resembling human language despite extensive
training. (Karmiloff-Smith, 1992, p.11)
VI.
WANNA CONTRACTION
A typical langauge-specific innate constraint is found in the use of "wanna"
contraction. Native speaker of English use "want to" and "wanna" interchangeably
in following sentences.
1)
a. He wants to watch TV but I don't want to.
b. He wants to watch TV but I don't wanna.
2)
a. Who do they want to see?
b. Who do they wanna see?
A 'seemingly' natural conclusion that a child could arrive at, after being
exposed to many of 'b' type sentences would be that 'wanna' contraction is
optional in English, if he were to rely merely on general learning strategies.
But native speakers of English do know that 'wanna' contraction is not always
possible:
3)
a. Who do you want to feed the cat?
b.* Who do you wanna feed the cat?
Apparently external input does not give any cues as to why (3b) is
ungrammatical, as in other superficially similar sentences the contraction is
possible. The relevant cue comes from UG, which presupposes that a movement rule
operates between two sentence structures, i.e. d-structure and s-structure
(Cook, 1988, p. 93). Wh-questions are obtained by the movement of the wh-word
from its original position in the underlying d-structure to its new position in
the s-structure, leaving behind a coindexed trace in its former location. This
trace is preserved in s-structure though it is phonologically null. So the
s-structures of(2a) and (3a) are as follows:
4)
Who(i) do they want to see t(i)?
5)
Who(i) do you want t(i) to feed the cat?
Native speakers of English know that when a wh-trace intervenes between 'want'
and 'to', the contraction is not possible. But neither the existence of this
trace nor the operation of the movement rule is derivable from the surface
structure of sentences, nor from the sematics as all of these sentences are
meaningful with wanna. UG theorist believes that this is a typical manifestation
of innately specified linguistic constraints.
A non-UG explanation might be that native speakers judge this type of sentences
as ungrammatical simply because they never hear them. Such an explanation is not
plausible as it presents a vicious cycle: why don't they hear such sentences at
the first place. If there is no UG constraints working on these sentences, why
don't native speakers of English produce such sentences? Secondly, absence of
such sentences in input does not justify their ungrammaticality. That is one
should not judge a unheard utterance as ungrammatical as the majority of
sentences we are exposed to are heard for the first time in our lives. There are
potentially infinite number of utterances which are perfectly grammatical. How
do the native speakers differentiate these grammatically correct unheard
utterances from their ungrammatical counterparts?
Again a non-UG response to such a question might be that on the basis of their
repeated exposure to the pairs of sentences like 1a and 1b and hearing only 3a
and not 3b, native speakers conclude that sentences like 3b are not possible
(Haiman, 1997). One problem with such an explanation is that native speakers
don't hear sentences in such identical pairs. Suppose that an English child
heard a sentence like
6)
Never will he want to do the same thing again once he is punished.
It is quite likely that the child will never hear the WANNA counterpart of this
sentence. If he is to follow the inductive repeated-exposure principle, then he
need to formulate a rule for the absence of wanna in such a sentence whereas it
is perfectly grammatical to use WANNA in this context. Such an inductive
learning also requires quite a great amount of memory load (Gleitman and
Gleitman, 1997).
VII.
UNIVERSAL GRAMMAR (UG) IN SECOND LANGUAGE ACQUISITION (SLA)
It is quite interesting that even the fervent advocates of UG turn out to be
hardline reductionists, when it comes to the role of UG in SLA: Pinker and
Jackendoff are typical examples. Even Chomsky in his earlier theorizing
speculated that innate linguistic knowledge is not accessible to L2 learners.
Still for many UG theorist innate mechanism for language acquisition atrophies
especially after puberty which is generally assumed to be the critical period
for natural language acquisition (Gregg, 1984; Bley-Vroman, 1988; Clashen and
Muysken, 1989)
However, the argument in favor of UG in first language is almost equally valid
for L2 learners who can attain high levels of linguistic knowledge which cannot
be attributed to input or instruction alone. Felix (1988), for instance, shows
that L2 learners do have access to UG principles like Superiority Effects,
Parasitic Gaps etc. "which are neither learnable on positive evidence nor
transferable from corresponding structures of the learner's mother tongue" (pp.
286-7). Similarly Bulut (1996) and Cem (1996), reports that advanced Turkish
learners of English are able to acquire the L2 reflexive system which cannot be
attributed to grammar instruction nor to input alone.
Another innate constraint which is neither applicable to Turkish nor can be
found in grammar instruction, concerns the use of WANNA. So if Turkish learners
of English can judge "3b" type of sentences as ungrammatical then this can only
be attributed to an active UG in SLA. In this paper the results of a small-scale
research is discussed.
VIII.
WANNA AND TURKISH LEARNERS OF ENGLISH
L2 Data concerning the wanna contraction is particularly valuable as Turkish
learners learning English in an EFL context are exposed to minimum amount of
wanna sentences. This is basically because of the scarcity of authentic input:
it is hard to find instances of wanna in bookish EFL input. So if these learner
are successful in a grammaticality judgement (GJ) task with wanna statements,
this can hardly be attributed to their repeated exposure to such sentences but
to their ever-active UG-based knowledge.
For this purpose, a GJ task is designed in which there are six cases of wanna
and four gonna contraction (see APPENDIX). The gonna sentences have nothing to
do with UG constraints: they are included as distractors so that the subjects
(Ss) cannot easily discover the underlying rule using their conscious
processing. Unlike other typical written GJ tasks, a the target sentences are
also recorded on a tape because it is basically during listening that their
intuition will be more evident. The wanna and gonna sentences are presented in a
context of dialogue to remove any extraneous factors due to difficulty in
understanding. For the same reason grammatical wanna sentences are deliberately
made longer than the ungrammatical ones. To remove the frequency of exposure
factor, infrequent structures are used in grammatical wanna sentences like
inversion, future tense, third person pronoun (see items 1 and 4). Only the
sentence having the wanna and gonna contraction are recorded on the tape by a
native speaker. Each sentence is read twice: first in noncontracted form (want
to and going to) and then with wanna and gonna. During the design phase native
speakers are asked to confirm the validity of the dialogues.
The Ss (30 in number) are university-level Turkish learners of English, who have
been learning English for the last six years. Their age span is between 18-20
and they started to learn English (in an extensive sense) during Lycee prep,
roughly corresponding to the critical period which is assumed to be the end of
active UG.
The test is administered in a laboratory context. Having a sense of the
dialogue, Ss are told to listen to the target sentences each of which is read
twice. They are also told to close their eyes for better concentration and
indicate when the second reading (with the contracted form) sounds awkward. The
number of Ss judging each sentence awkward are as follows:
ITEM NO------------>1--2--3--4--5--6--7--8--9--1O
NUMBER OF
SUBJECTS--------->6--4--13--5--4--9--8--2--5--11
Items 3, 8, 10 include ungrammatical wanna contractions and item 7 the
ungrammatical gonna. Of the three target items the two--items 3 and 10-- are
scored highest in terms of ungrammaticality, even higher than item 7 where gonna
sounds real awkward. Item 2, however, could not be detected as ungrammatical by
a great majority of the subjects: it is the lowest on the scale of
ungrammaticality. When Ss are asked at the end of the test they could not state
the real rule why wanna is not possible and some of them stated false rules
related to the negativity or phonological aspects of the sentences.
IX.
CONCLUSION AND SUMMARY
The results did not show a clear effect of UG in SLA as item 8 could not be
detected and items 3 and 10 are only marginally higher than the others. However,
the fact that 3 and 10 are scored highest among others without being able to
articulate the relevant rule shows that they relied on their subconscious
intuitions which are somewhat affected by UG-based knowledge.
The reason why their intuitions are not clear enough might be due to their prior
conscious learning experience in a formal context which is hypothesized to
hinder naturalistic language acquisition. As Felix and Weigl (1991) points out
classroom L2 learners are greatly constrained in their formal learning context.
They draw a distinction between a language specific (LS) system, which
corresponds to Chomsky's LAD, and a general problem-solving (PS) system to
account for the success and failure of naturalistic and classroom learners. So,
the degree of access to UG is determined by the type of learning mechanism that
is involved in SLA. The more the LS-system is involved, the more direct access
there will be, and the more the PS- system is involved, the more indirect access
(or even no access) will take place. Definitely there must be more studies
carried out in this area to arrive at a conclusive decision. But a tentative
conclusion might be that UG is still alive for those who knows how to activate
it.
APPENDIX
1.
A: Hey! Look at the accident outside. A man is lying dead on the street.
B: I have never seen a dead person in my life nor do I want to see one.
2.
A: No one has come to pick up the books, and the office will be closed during
the weekend.
B: Oh really?
A: Yes. Who do you think is going to pick them up?
B: I will.
A: OK, see you.
3.
Mother: The room is real mess. Whose turn is it for cleaning? Yours or John's?
Son: We don't take turns. Just tell me: Who do you want to clean the room?
Mother: I'd like you to do that. OK?
Son: Yeah, why not.
4.
Teacher: Bob will not be able to go out this weekend.
Student: Sir, what is the purpose of this punishment?
Teacher: Never will he want to do the same thing again once he is punished.
Student: I really don't think so.
5.
Student: Sir, do you have some free time?
Teacher: I'm afraid not. Can we talk after lunch?
Student: I was just going to ask a simple question.
Teacher: OK, then. Go ahead.
6.
A: Is the oral exam finished?
B: Not yet. There are a few more students to be interviewed.
A: Who does the teacher want to see next?
B: Call Robert. He must be the next person on the list.
7.
A: Hey! Where are you going with all those books?
B: Can't you see that I am going to the library?
A: Come on. This is Friday evening. Let's go to the movies.
B: Sorry. I have to study.
8.
Student: Sir, I don't have time to go to the library.
Can you tell John to do the research?
Teacher: John is not the student who should be doing that.
You are the person who I want to do the research.
9.
A: What will you do after you graduate?
B: I really miss my family. I am definitely going to go
back home just after graduation.
10.
A: John looks quite excited about the final game.
B: Which team does he want to win the final? The Bulls or the Pistons?
A: The Bulls, of course.
REFERENCES
Bickerton, D. (1981) Roots of Language. Ann Arbor: Karoma publishers.
------------- (1983) Creole language. Scientific American. 249:116-122.
Bley-Vroman, R. (1988). The fundamental character of foreign language learning.
In W. Rutherford & M. Sharwood Smith (Eds.), Grammar and second language
teaching (pp. 19-29). Rowley, MA: Newbury House.
Bulut, T. (1996). How Turkish adults learning English interpret English
Reflexives. Paper presented at Second IATEFL Balkan ELT Conference. Bogazici
Univrstity,Istanbul. 5-7 September.
Carston, R. (1988). Language and cognition. In F. J. Newmeyer (Ed.),
Linguistics: The Cambridge Survey: Vol. III. Language: Psycological and
Biological aspects. (pp. 39-69) Cambridge: CUP.
Cem, A. (1996). Universal Grammar in Second Language Acquisition: Knowledge of
Reflexives by Turkish Learners. Unpublished Doctoral Dissertation. Hacettepe
University.
Clashen, H., & Muysken, P. (1989). The UG paradox in L2 acquisition. Second
Language Research, 5, 1-29.
Cook, V. (1988). Chomsky's Universal Grammar. Oxford: Basil Blackwell.
Felix, S. (1988). UG-generated knowledge in adult second language acquisition.
In S. Flynn and W. O'Neil (Eds.), Linguistic theory in second language
acquisition (pp. 277-294). Dordrecht: Kluwer.
Felix, S., & Weigl., W. (1991). Universal Grammar in the classroom: The effects
of formal instruction on Second Language Acquisition. Second Language Research,
7, 162-181.
Gregg, K.(1984). Krashen's Monitor and Occam's Razor. Applied Linguistics, 5,
79-100.
Gleitman, L. and Gleitman, H. (1997) What is language made out of? Lingua.
100:29-55.
Haiman, J. (1997) Repitition and identity. Lingua: 100: 57-70
Jackendoff, R. (1993) Patterns in the Mind. New York: Harvester Wheatsheaf.
Karmiloff-Smith, A. (1992) Beyond Modularity: A Developmental Perspective on
Cognitive Science. Cambridge, Mass.: MIT Press.
Piaget, J. (1980). The psychogenesis of knowlegdge and its epistemological
significance. In M. Piattelli-Palmarini (Ed.), Language and learning: The debate
between Jean Piaget and Noam Chomsky. Cambridge, MA: Harvard University Press.
Pinker, S. (1994). The Language Instinct. New York: William Morrow.
Slobin, I. (1979) Psycholinguistics. Illinois: Scott Forestman
6
Print This Article
TOP
