LANGUAGE ACQUISITION IN AUTISTIC CHILDREN: WHAT DOES IT TELL US?

James R. Cash (1989, 24 novembre)
jcash@lifestories.com
testo in italiano

The purpose of this paper is to gain theoretical insight into language acquisition in humans by examining the characteristics of language in the autistic child. There have been many theories put forth to help us understand how language may be learned. Especially interesting is the question raised by treatments designed by Lovass (1966) to teach language to autistic children. Lovass employed a behaviour modification (therapy) procedure utilizing reinforcement learning theory and shaping techniques to develop a "program for the establishment of speech in psychotic children" (Wing, 1966, p.115). The main questions arising from his program are: Were the children actually learning ie. acquiring language; that is, were they acquiring the facility to extract linguistic rules in order to produce sentences appropriate to situational changes? Or, were the children merely learning (assembling) a highly complex "phrase book" in which they had compiled a repertoire of sentences for all occasions? Put simply, do these autistic children understand their newly learned speech in a way that is at all similar to the understanding of a normal child of equal linguistic development? Let us first examine some characteristics of autistic language as reported by Aug (1974). Then, after a brief overview of the Lovass method of behaviour therapy, we will attempt to determine the effectiveness of his therapy. Aug (1974) outlines nine characteristics of language that are common in autistic children. Later we shall see how Lovass applied reinforcement theory to modify some of these so that speech can be learned. Aug begins by pointing out a failure in most autistic children to communicate a suitable affect in their speech due to an impairment of qualities such as intonation, pitch, emphasis and timbre. He calls this sing-song or monotonous speech. He also found that these children would repeatedly recite stereotyped words and phrases in a ritualistic way, such as the slogan in a television commercial, or a specific question that may be endlessly repeated and asked of everyone. Yet, often was the case where autistic speech failed to address anyone specifically. Even so, these children had special difficulty with personal pronouns even when directing their speech to another person. Especially problematic are the pronouns "I" and "you". And if these pronouns were ever acquired, they were misused. Aug indicates that "yes" is often a difficult word for these children to use. Instead the child who wishes to respond in the affirmative will repeat the question that was just asked. This Aug calls affirmation by repetition. He also found that these children would prompt the speech of the other person in order to strictly adhere to the script of a previous conversation, or the child may directly prescribe the other person to say something by commanding, "Say _________". Autistic children also typically employ extreme literalness in communication as illustrated by the example of one child who said that a picture is not "on" the wall but "next to" the wall. Idiosyncratic metaphorical speech is another common characteristic in these children. Aug notes an observation by Kanner in 1946 where a seven-year old boy said, "Annette and Cecile make purple." This was an enigma until the original situation was revealed to Kanner: the boy had five bottles of paint which were named after the Dionne quintuplets--Annette was blue and Cecile was red. The metaphors are rigidly bound to the some initial situation which makes the metaphor's frame of reference incoherent. Finally, echolalia is the term used to describe the duplication of speech the child hears. For example, a child may hear the teacher say, "What is this? ...cookie?" to which that child will respond with same intonation and rhythm, "What is this? ...cookie?". Echolalia is characteristic of (but not restricted to) the speech of autistic children. Lovass understood this to be a fundamental aspect of autistic speech and its abnormalities. Aug (1974) concludes that "autistic children fail to use words flexibly as vehicles of general meaning, and rather experience them as undifferentiated parts of some unique original situation." (p.165) This concept is elemental to understanding many of these characteristics and the learning problems that disrupt language acquisition in autistic children. It is interesting that Lovass has entitled his 1966 paper "A Program for the Establishment of Speech in Psychotic Children" rather than "A Program for the Establishment of Language in Psychotic Children" (Wing, 1966, p.115). Perhaps he was aware of the formidable abyss that may very well separate the two. Yet in the relatively brief description of his program below, one can conclude that he has considered the challenge and nature of the abyss very well. In brief, his program is highly dependent upon the specific severity of the psychotic child involved and the characteristics of that child's language development. Initially, before any sort of speech reinforcement therapy can begin, any and all psychotic behaviours that may interfere with this program of language behaviour modification must be extinguished. Echolalia is an important consideration because there is value to any actual verbalization in this program and it was found that previously autistic, echolalic children progressed at much faster rate than previously autistic, mute children (the difference is between one and a half years and eight months of training; in addition, the previously echolalic children's language was higher quality across the board). Yet, uncontrolled echolalia can and did interfere with this program and it had to carefully managed. Violent self-destructive acts common in more severe psychotic children obviously needed to be suppressed before any attempt of communication or therapy can be made. If the child is left unrestrained extinction of these destructive acts does take place, however, it is slow and therefore very painful. In one case, the child did not stop until he had given himself over 10,000 blows. Electric shock therapy contingent upon self- destructive acts was immediately effective (although somewhat controversial). After self-destructive and self-stimulating acts are suppressed, a program of shaping can begin in which the process of acquiring speech is broken down into many, very simple behaviours. The sequence of behaviours begins and, bit by bit, each behaviour is learned. Primary reinforcers (those that attend a physiological need, such as food attends to hunger) are used initially because of their power and effectiveness. First the child learns to be attentive: eye to face contact is prompted and reinforced. In the use of prompts, given a certain stimulus situation, one wants a particular response to occur in its presence so that it can be reinforced. The prompt cues the correct response prior to training, or with minimal training. Later the prompt is fading so that the control of a response is shifted from one stimulus to another. This method is used often throughout the stages of this program. Next, the frequency of spontaneous vocalizations must be increased. Therefore, any vocalization will be reinforced. Incompatible responses must not be reinforced, however, so attentiveness and vocalization must both occur together to be reinforced. Then, single phonemes are vocalized by the teacher (beginning with some vowels and bilabial stop consonants such as /b/ and then progressing on to more difficult spirants) to which closer and closer approximations are reinforced. Often, a manual prompt may be used to form the lips into the correct shape, then slowly this prompt is faded. The repertoire of sounds is slowly increased. Now words can be formed by juxtaposing phonemes in the child's repertoire. The data that generates the learning curves with these children is accelerated positively. They are acquiring a discrimination in which the response resembles its stimulus. These data are characteristic of successful imitation training. Yet, these children still do not know the meanings of the words they imitate. Next, Lovass begins a program of language training to teach such meaningfulness. Lovass explains (Wing, 1966, p.128) that practically all language training following verbal imitation training can be understood as the establishment of three basic discriminations: In the first discrimination, the stimulus is nonverbal and the response is verbal. The stimuli can be objects, symbols, behaviour, etc. A good example of this discrimination is the labelling of objects or the description a situation. The second discrimination has a verbal stimulus which could arise from oneself or others; the response is nonverbal. When a child obeys an order, for example, the discrimination indicates comprehension. In the third discrimination, both the stimulus and response are verbal. A conversation or question and answer discussions are examples indicating this discrimination. Language training begins simply, by training a labelling vocabulary. This training will involve the acquisition of both the first and second discriminations outlined above. It is argued that mastery of a particular label occurs when it is generalized to all classes of that object. For example, when a member of a class objects is presented, such as a chair, and it is correctly labelled as a chair when it is first presented, then the concept of "chair" is understood. Through the use of all three discriminations, training can now extend into the following areas in the order listed: appropriate use of and response to prepositions and pronouns, preposition training, pronoun training (eg. I-mine, you-yours, he-his ie. nominative and possessive cases), discrimination of the personal case, combining discriminations of pronouns and prepositions together (additive, more and more complex discriminations must be made to a stimulus, there is also a savings across tasks). Also, after three months time, there is shift from primary to secondary (learned) reinforcers. After considerable time, these children were trained to make relatively complex discriminations. Yet, Lovass was concerned that these children were over trained--only RARELY did they volunteer to speak. They seemed to be strongly responsively oriented to the attending adult trainers. A program of training spontaneous speech then begins. The program phases were overlapping: the institution of demands, and the development of commentary and story telling including the recall of past events. An example of the latter: a story may be read by the teacher and then the child is asked to retell it, or he may be prompted to recall it with questions that reflect a comprehension of the story. Lovass also reports (Wing, 1966, p.140) that he obtained extra elaborations not expected in the program design. These he cites as being good evidence that his program was indeed having some effect on language development. He uses the example of Ricky who once made a comment on growth during the later phases of the program. During language training, Ricky learned about the concept of size and many things that were large and small were discussed. In one discussion, Ricky was told that small plants grow to larger size by pouring water on them. Ricky thought about that for a moment and then he said, "put some water on my head". Lovass feels that distinctively human contributions by these children, such as this one, defends his program against objections that it would create "trained seal" or mechanical qualities in the subjects. What insight, then, does this program by Lovass give us into the concept of language and it acquisition? A number of things must be remembered. We have seen in the Lavass study a reverse of the process exhibited by normal children. That is, in normal children, comprehension usually precedes production of a form, and comprehension is usually the immediate state after productive training. Also, from later reports of Lovass's treatment (Lovass, 1973), one finds that the initial improvement in the children's language has been quickly setback as soon as their environment failed to maintain the same demands on their behaviour as the training environment. Once the maintenance of the social interaction breaks down, so too will much of an autistic child's learned language function. The question remains, then, whether there can be a transferal of language from a treatment situation to a natural one in which the subject must "fend for him or herself". A context must be created in which meaningful and necessary speech and language must be used both in normal children and autistic children. And serving as the ultimate reinforcer must be the effectiveness of the child's communication in the child own environment.

References 

Aug, Robert C. (1974)
The Language of the Autistic Child.
In E. W. Straus (Ed.), Language and Language Disturbances.
Pittsburgh: Duquesne University Press. (pp. 155-172) 

Clark, H. H., Clark E. H. (1977).
Psychology and Language.
New York: Harcourt Brace Jovanovich, Inc. 

de Villiers, J. G., de Villiers, P. A. (1978)
Language Acquisition.
Cambridge: Harvard University Press. 

Lovass, O. I. (1966)
A Program for the Establishment of Speech in Psychotic Children.
In J. K. Wing (Ed.), Early Childhood Autism.
New York: Pergamon Press. 

Rosenhan, D. L., Seligman, M. E. P. (1989)
Abnormal Psychology.
New York: W. W. Norton and Company.

Straddon, J. E. R., Ettinger, R. H. (1989)
Learning: An Introduction to the Principles of Adaptive Behaviour.
San Diego: Harcourt Brace Janovich, Publishers. 

Wing, J. K. (1966)
Early Childhood Autism.
New York: Pergamon Press.

indietro | inizio