The development of language The ultimate objective of transformational- generative linguistics is to understand what language is and how it operates. 3 In this sense the study of language is closely associated with the field of psychology. As Lenneberg (1964, p. 78) asks, "Might it not be possible that language ability—instead of being the consequence of intelligence—is its cause?" If the goal is to comprehend language more fully, the ideal place to begin is with children's acquisition of their mother tongue. It is this desire to understand language itself better that has fostered increased attention among psycholinguists on first-language learning.
Former studies of language learning stressed word counts and pronunciation. The object was to describe the language as spoken. The emphasis was on phonology, morphology, and semantics. Language learning was viewed as a product of imitation and reinforcement. However, stimulated by Chomsky's theory of language and language learning, newer studies have focused on the acquisition of syntax and the development of competence. 4 McNeill (1966, p. 17) states, "It is possible to describe performance without explaining it, but if we wish to explain performance, we must show how it derives from competence. . . ." In other words, the crucial aspect of comprehending how language and mind interact is to determine the role of language in the process of converting thought to expression and vice versa. However, before performance can be explained, competence must be understood.
As the psycholinguists have studied first-language learning from this new perspective of syntax, a new viewpoint has developed. 5 In the first place, these studies reveal that elementary competence underlying functional language performance is acquired in a relatively short period of time. Speech in a grammatical sense usually occurs around the time the child is one-and-one- half years old. By the time she is three-and-one-half, she already has the elementary syntactical patterns of the language. The accomplishment of this tremendously complex task in a period of two years, compared with the child's cognitive achievements in other areas, is little short of amazing.
The fact that children can handle elementary syntactical patterns at age three-and-one-half does not mean that they have reached the adult language level, however. In fact, the more researchers explore first-language acquisition, the more they are impressed by the length and the complexity of the
process. Contrary to earlier belief that the child has mastered the language syntactical system by the age of five or six, Chomsky (1969) found that the grammar of a child of five is still significantly different from that of an adult and that it is not until she is around ten that her language is essentially that of the adult.
Theoreticians and researchers in the area of first-language learning now seem relatively certain that children's utterances are not a reduced imitation of adult speech. Close examination of the words children use and the way they put them together indicates that certain combinations and forms occur that are not to be found in adult speech. However, the combinations the children use do have a definite pattern, even though it does not correspond to the adult system of combining words. Therefore, the conclusion is that children develop their own simple linguistic competence, i.e., grammar rules, which makes their speech patterns possible.
It has been found that young children usually do not imitate a new pattern prior to incorporating that pattern into their own language system. For example, a child will not repeat, "Daddy is going to work," until he has reached a certain level of linguistic and cognitive maturity. Only after he has begun to use the -/ngform of the verb will he repeat the sentence as he hears it. Ervin and Brown found that approximately 10 percent of all child utterances were imitations (McNeill, 1966).
The stand has also been taken that language is not learned by means of repetition and practice. Patterns within the language system are much more influential in language acquisition. This conclusion is based on the fact that children first learn irregular verb forms. The irregular verbs are the most common and are, therefore, the verbs with which children come in contact first and with which they have the most practice. Yet, as soon as they acquire a regular pattern, such as the past tense of regular verbs in English, they extend the pattern to include the irregular verbs that they have practiced repeatedly and that, up to this point, they have handled correctly. For some time thereafter, they continue to say "goed" and "seed," etc. McNeill (1966, p. 71) states, ". . . the amount of practice given to a feature is less relevant to language acquisition than the ability of a child to notice that a feature is part of a pattern."
Language is not learned by analogy. In order to acquire language children must develop hierarchical structures in their language system. Yet, these hierarchical structures are not apparent in overt, adult speech which children receive from their surroundings. The necessary knowledge is hidden within the deep structure of the language. 6 Therefore, it would be impossible for them to analogize that which is not available for analogy. For example, in the
First-Language Learning
sentences: "I like raising flowers/' "I like amusing stories/' and "I like entertaining guests/' there is no surface indication that the sentences are different. However, when the native speaker of English attempts to insert the word the in the sentences, the fact that they are not the same is readily apparent (Saporta, 1966).
Lenneberg (1966) argues that language is not learned due to a psychological need. He cites the cases of deaf-mute children who seem to be as happy psychologically as those who have learned to speak. Nor does he feel that language is developed by means of "training procedures and extrinsic response-shaping." Instead, he maintains that language is acquired as a normal corrolary of the typical maturation process. When children are ready, they will learn the language, and neither their parents nor anyone else can speed up this developmental sequence. He gives the example of a father attempting to get a child to say, "Daddy, bye-bye." The child until a certain age will not join the two words into a single phrase, even though she may "babble" much longer utterances. The block apparently does not result from a lack of motor development but of cognitive development. In the average child, language is not conditioned or hastened, but ". . . emerges by an interaction of maturation and self-programmed learning" (Lenneberg, 1966, p. 239).
An internal, mentalistic model of first-language acquisition If children do not learn to speak by imitation, practice, and reinforcement, how do they learn the language? What is the process whereby children learn to connect words in order to express meaning?
As children's cognitive functions mature, they begin to use one-word sentences, i.e., holophrastic sentences. Wah-wah may mean "I want a drink," "I had a drink," "Let's go play in the pool," or any number of things connected with water. It has been suggested that children quickly abandon this holophrastic means of communication for two reasons: (1) the mind would shortly be filled with a "cognitive clutter," which would overtax its capabilities, and (2) the communications of such a system are ambiguous and imprecise.
Therefore, for the sake of efficiency and exactness children progress as they mature to the formation of a grammar that allows them to produce the exact expressions they need. At first this grammar is quite simple. The first sentences are made up of two-word expressions, such as the following examples given by McNeill (1965):
|
allgone |
boy |
|
byebye |
sock |
|
big |
boat |
|
more |
milk |
|
pretty |
vitamins |
my plane
see hot
The words listed above fall into two classes, an open class and a pivot class. Words in the open class correspond to naming words (i.e., nouns) in the language. Words in the pivot class correspond to function words (i.e., verbs) in the language. By combining pivot words and open words or open words with other open words, the child achieves an elementary, but satisfactory, manner of self-expression. The important factor in this development of grammar is that learning a system enables the child to spin out needed language in its infinite variety from a limited corpus of material.
How children learn language Gradually, the child's simple grammar acquires the complexity of the adult model. But a description of the process does not explain how it all takes place. The answer of the mentalists is that the child is born with an innate capability for language acquisition. As children mature cognitively and linguistically, this innate predisposition for language learning enables them to compare their language system with the adult system. The deep structure that they produce at first is modified by means of rules of transformation to enable them to produce the surface structure that they hear. 7 It is the children's "innate linguistic endowment" that enables them to formulate "linguistic hypotheses," and it is by means of these hypotheses compared with the language they receive that they manage to conform to adult speech. In McNeill's (1965, p. 32) words, "Languages have deep features, unmarked in overt speech, precisely because children possess the specific linguistic capacities that correspond to them. Otherwise, deep features would be unlearnable."
THE N.ATMSTIC EXPLANATION
It is on this point of the learner's innate capability for language acquisition that nativistic theory and cognitive theory disagree. Those who hold to the nativistic theory of first-language learning maintain that the child is born with an indwelling predisposition to learn a language. They refer to this innate component as a language acquisition device (LAD). They base their conclusions on the following observations:
1. Only humans have the physiological and psychological means to participate in sustained speech.
2. Given a typical environment, all children learn language. In fact, suppression of language learning would be almost impossible.
3. The sequence in which grammatical structures are acquired seems to be
First-Language Learning
practically the same for all people, and the beginning of language usage and the minimal achievement in language skills seem to be relatively unaffected by cultural or linguistic variations.
4. Certain characteristics of phonology, semantics, and syntax appear to be universal.
THE COGNITIVE EXPLANATION
On the other hand, those who hold to a cognitive theory of first-language acquisition feel that children are born with the ability to learn. They hold that children's cognitive capacities make it possible for them to acquire many types of knowledge and skills including language and language skills, but they do not feel that children are born with a mental mechanism specifically designed for language learning per se. That is, children have certain mental abilities that make all types of learning, as well as language learning, possible (Butler, 1974).
Based on the latest research findings in first-language acquisition, both nativistic and cognitive theorists conceive of first-language learning as an internal process that is creative and rule-governed. Given the fact that the words and word combinations children first use do not correspond to the adult system, they conclude that children are creating their own language. Given the fact that the child's language follows a definite pattern, they conclude that she is creating language according to some rule-governed system. This apparent existence of a child grammar implies that the mind of the child is an active agent in the language acquisition process. As she learns the language, she uses certain learning strategies to formulate rules appropriate to the level of her cognitive development. These rules then become the basis for her participation in the "creative construction" of language (Dulay & Burt, 1974).
First-language acquisition is a dynamic and self-correcting process. Given the fact that the child's grammar system changes from rather simple beginnings to a rather complex set of rules at the adult level, theorists conclude that the system undergoes an evolutionary process. Given the facts that children hear only surface structure manifestations of language meanings and that their language tends to resist attempts to change it prior to their own incorporation of a pattern, the theorists conclude that children gradually modify their language system, changing it to conform with that of the language they receive. And how is this changing process explained? The mentalistic theorists say that the language users formulate language as hypotheses which they test against the language they receive. As time passes and as their cognitive abilities mature, they gradually acquire the adult system.
Another important aspect of the process of first-language acquisition is that the learner participates in natural communication situations. Dulay and Burt (1973) define a natural communication situation as having two characteris-
tics: (1) the attention of the producer and the receiver is focused on the content of the message, not its form; and (2) the meaning of the communication is established by the existence of clearly understood referents.
Implications of the Nativistic and Cognitive Models for Second-Language Learning The nativistic and cognitive models of first-language learning imply the following tenets for second-language learning:
1. Competence must be developed first.
2. Language usage is based on the acquisition of rules.
3. Language usage is productive, i.e., the speaker can produce language appropriate to a given situation.
4. First-language learning proceeds from deep structure to surface structure. Second-language learning reverses this process. Might it be possible to provide insights into deep structure for adult learners prior to presenting them with surface structures of the language? McNeill (1965) mentions the possibility of teaching child grammar to adults in order that they first learn the deep structure.
5. Adult learners of a second language need assistance in formulating the appropriate hypotheses about language. The proper choice of explanation and exercises might assist them to internalize the necessary language competence.
6. Typically, the surface structure of the second language is imposed upon the deep structure of the native language. The teacher should expect this native-language influence and not be upset by it.
7. Language learners generate their own set of language rules. They then test the appropriateness of these rules by comparing them with the language they hear used.
8. The learner must be given opportunities to develop language hypotheses and to test them against the language model.
9. Communication does not require a perfected language system.
10. First-language learners focus on meaning rather than on syntax (Butler, 1974). Second-language teachers have tended to stress syntax rather than meaning.
11. The acquisition of irregular structures in first-language learning proceeds from (a) not using a particular structure to (b) occasional use with no errors or overgeneralizations to (c) increased use with errors and overgeneralizations to (d) correct usage (Butler, 1974). Certainly, second-language learners will pass through a period of incorrect usage or at least incomplete control of a pattern prior to total mastery of the structure.
12. Language learning and language usage are creative processes. Therefore, the second-language learner should participate in the "creative construction" use of language.
13. The teacher should concentrate on establishing natural communication situations for practice at the "real" language level. Communicative competence cannot be acquired in a vacuum, i.e., in situations devoid of meaning or meaningful purpose.