The Nature of Language Acquisition
By: Ronald W. Langacker
From: Language and its Structures, Harcourt Brace Jovanovich, 1967
J. Geffen
5
10
15
20
25
30
1.
The problem of understanding language acquisition is that of finding out how
the child comes to possess the abstract linguistic system that is largely responsible
for shaping his verbal activity. We have no adequate theory to explain, stage by
stage, the child’s discovery of the structure of his language. Consequently, we will
focus our attention on a more restricted question: How much of the linguistic
structure is innately specified? To put it another way, how much of the lingustic
system that he ultimately commands is the child born with, and how much must he
discover on the basis of his linguistic experience?
2.
In the empiricist view, the human child is born with no special capacity for
language, only a general ability to learn. A linguistic system is built up through
experience by means of the same basic intellectual ability (association, for example)
that is responsible for shaping all other aspects of cognitive structure. In the
rationalist view, on the other hand, there is, in addition to general intelligence, a
special inborn capacity for language. Above and beyond our natural ability to form
and manipulate concepts, we are endowed with an innately specified predisposition
to acquire a linguistic system having certain properties as opposed to all other
conceivable ones.
3.
The evidence for the rationalist claim is very strong. Consider first the
uniformity of language acquisition throughout the human race. Every human child
learns a language unless he is the victim of extreme mental deficiency or isolation
from language use. There are all sorts of physical and intellectual skills that children
can fail to master despite a considerable amount of instruction, but talking is not
among them. This is precisely what one would expect on the basis of the assumption
that language is innately specified almost fully, with linguistic experience serving
mainly to activate the genetically specified system.
4.
While the species’ uniformity of language acquisition fits in perfectly with the
rationalist position, it conflicts with what we would expect on the basis of the
empiricist viewpoint. If the acquisition of language depended mainly on the training
the child receives, we would expect differences in training to correlate directly with
differences in language acquisition (if general intelligence is held constant). In fact,
however, this expectation turns out to be false. A child learns to talk regardless of
whether or not his parents constantly pursue him, correct him, and put him through
linguistic drills. Some parents do this, others do not, and some children don’t even
Problems of Socioeconomic Organization / 2
35
40
45
50
55
60
65
70
have parents – but they all learn to talk. Despite wide variation in the amount of
speech they are exposed to, all children acquire a full-blown linguistic system. There
are no cases on record of children who have only learned half a language, who have
failed to master any syntactic rules, who lack underlying phonological
representation, or who have not picked up any complex lexical items. The
vicissitudes of early linguistic experience are not matched by any comparable
variations in linguistic structure.
5.
A second argument in favor of the rationalist position is provided by the fact
that only human beings learn to talk. The most likely nonhuman candidates, of
course, would be the higher apes – chimpanzees, for example. They are anatomically
similar to humans and are also reasonably intelligent; they can learn to use tools and
to solve simple problems. The difference in intelligence between apes and human
beings is thus not absolute, but only a matter of degree. When we consider language,
however, we find an absolute distinction. The progress that apes can make toward
mastering a human language is not proportional to their intelligence – in fact, they
can make no progress whatsoever. Experience has shown that a chimpanzee, even
when raised exactly like a child, acquires nothing that bears even the faintest
resemblance to the linguistic systems that human children learn so easily. Language
is therefore peculiar to our species. Moreover, it is not directly tied to intelligence.
These observations are perfectly compatible with the view that language develops in
the human child because of a special, inborn linguistic capacity. Apes cannot learn to
talk because they do not possess this innate structure.
6.
This simple and natural explanation is not available if one adheres to the
empiricist position. If language is a function of general intelligence and not of any
special linguistic capacity, then other animals should, given proper training, succeed
in acquiring language to a degree proportional to their intelligence. Experiment has
shown that this is not the case. There is absolutely no evidence to indicate that
anything even remotely resembling the complex system of rules and abstract
underlying representations of a human language can arise in other species.
7.
The relative perfection of language acquisition is a third argument for innate
specification. If language reflected general intelligence and not a special linguistic
capacity, we would expect differences in intelligence to correlate directly with
differences in language acquisition (if training is held constant). We would expect
bright children to do better than stupid ones in mastering a linguistic system. We
would also expect some children to fail miserably at acquiring language, just as
many children fail to learn geography or the procedure for extracting square roots.
We would expect some children to wind up with linguistic systems so deficient or so
distorted as to be unrecognizable.
8.
These expectations are not borne out, however. Bright children, average
children, and stupid children all learn to talk. They are all successful at mastering a
linguistic system that is virtually identical to that of their models, one which is
Problems of Socioeconomic Organization / 3
75
80
85
90
neither distorted nor deficient. Regardless of general intelligence, a child succeeds in
mastering a complex system of rules and underlying representation that specifies an
infinite set of sentences. Children may vary on minor points such as volubility or size
of vocabulary, but they do not vary with respect to the significant structural features
of linguistic organization.
9.
The abstractness and structural complexity of language is a fourth strong
argument in favor of the rationalist view…. The child masters the entire set of lexical
items and structural principles that constitutes a linguistic system. He does this at an
age when he is not yet capable of logical, analytical thought. This remarkable
phenomenon can be explained in terms of the rationalist view, but hardly in terms of
the empiricist position. If linguistic structure is genetically transmitted, its
tremendous complexity does not pose a problem for the child. It is not a matter of
building up the entire linguistic system from scratch, but only of filling in some
peripheral details.
10. With no inborn structural principles to guide him, the child could hardly be
expected to discover the nature of the linguistic system. He would have no basis even
for suspecting the existence of this system. Furthermore, its abstractness and
complexity put it well beyond the possibility of being discovered by trial and error.
Problems of Socioeconomic Organization / 4
Questions should be answered in your own words, in English, unless otherwise
indicated.
Answer the question below in Hebrew.
1.
2.
What is the underlying assumption from which the author proceeds in his
discussion of the various aspects of language acquisition? (paragraph 1)
Answer: ____________________________________________________________
Answer the question below in English.
What is the essence of the empiricist view on language acquisition?
Answer: ____________________________________________________________
Answer the question below in Hebrew.
3.
What arguments does the author provide – paragraph 3 – in support of the
rationalist view?
Answer: ____________________________________________________________
4.
5.
Answer the question below in English.
In what circumstances will a child fail to acquire language?
Answer: ____________________________________________________________
Quote at least one of the statements that suggests that for children language
acquisition is not a function of intellect.
Answer: ____________________________________________________________
Answer the question below in Hebrew.
6.
What is the function of exposure to language – or as one might call it linguistic
experience – in the child’s process of language acquisition?
Answer: ____________________________________________________________
Problems of Socioeconomic Organization / 5
8.
Choose the best answer.
7.
What is the main idea in paragraph 4?
a.
Language acquisition is a function of linguistic experience.
b.
The dull rarely acquire language.
c.
Excessive tutoring will damage a child’s innate capacity for
language.
d.
It is the quality of tutoring rather than the amount of exposure that
will decide the pace of language acquisition.
e.
Children will acquire language with very little regard to the quality
of tutoring they are exposed to.
Complete the sentence below.
The qualification exclusively possessed by human beings is ___________________
Answer the question below in English.
In what context and to what end are chimpanzees mentioned in this
article?
Answer: ____________________________________________________________
9.
10.
Answer the question below in English.
What does the very fact that both bright and stupid children successfully master
the linguistic system of their models indicate?
Answer: ____________________________________________________________