The Articulate Mammal (28 page)

Chomsky himself is somewhat vague about how many switches there are, and what they switch. But one possible switch, he has suggested, is ‘head position’ (
Chapter 5
). Children might know in advance that linguistic structures have a head (key word), and that languages tend to put the modifiers (words relating to the head) consistently either before or after it. So, to repeat the English example given on p. 95, English children might say:

THE DOG DROPPED(H) THAT SLIPPER(M) DOWN(H) THE DRAIN(M)

with heads (H) preceding modifiers (M), while Turkish children would reverse this order, and say the equivalent of:

THE DOG THE DRAIN(M) DOWN(H) THE SLIPPER(M) DROPPED(H)

with modifiers preceding heads. Does this suggestion work?

At first sight, this is a plausible idea. Children are on the whole consistent in their treatment of heads and modifiers. But on second thoughts, this may be because youngsters are sensitive to the order of the words they hear. There is no need to assume an English child has ‘set a parameter’ when it says WANT MILK, rather than *MILK WANT. It just listened to its mother saying: ‘Do you want some milk?’ and remembered the word order. Furthermore, if a switch had been set, we would expect children to iron out various inconsistencies. They should say *AGO TWO WEEKS instead of TWO WEEKS AGO, where the
modifier occurs (exceptionally) after the words it modifies. But they show no real signs of behaving like this.

Perhaps the biggest weakness of the switch-setting theory is that no one can agree how many switches there are, nor how exactly they are set (Roeper and Williams 1987). This may be because language acquisition is just too messy a process to be explained by the flick of a switch.

Let us now summarize our conclusions about Content Cuthbert. This approach does not seem to be borne out by the evidence. Children do not appear to have firm advance expectations about language. They do not necessarily steer clear of sentences which are prohibited by language universals. They do not acquire chunks of language by flicking a switch. Of course, Chomskyan language universals may still exist. But they are not there ‘ready to go’ at a relatively early stage, triggered by simple data, and requiring very little effort on the child’s part, as Chomsky has suggested. Let us now consider whether Process Peggy provides a better explanation for language acquisition.

IS PROCESS PEGGY A GENERAL PROBLEM SOLVER?

The most general process approach proposes that Process Peggy simply makes use of a wider set of puzzle-solving abilities which she brings to bear on the world as a whole. Proponents of this viewpoint put forward various nonlinguistic factors which they consider to be critical for guiding the child forward through the thickets of language. We shall consider two of these: children’s needs and their general mental development.

According to an ‘everyday needs’ approach, children are by nature sociable little animals who need to interact with other humans. They also have certain material needs, such as MILK or JUICE. They are therefore concerned primarily with interacting with other people, and with getting what they want. They acquire speech in order to help them in this quest (e.g. Donaldson 1978). Within a particular culture, there is relatively little variation in the interests and requirements of different children. Therefore, it is not surprising that children develop language in a parallel fashion, even though they have never met one another.

This viewpoint is certainly borne out by children in the very early stages of development. As we noted in
Chapter 6
, children all over the world seem to talk about very similar things at the two-word stage. We find requests such as WANT MILK, rejections such as NO WASH, questions such as WHERE DADDY?, and so on, in widely separated children. Some researchers have suggested that this state of affairs lasts throughout the language learning period. They argue that children are concerned primarily with the external world, both with finding out about it, and with getting what they want. As youngsters attempt to learn about and manipulate some aspect of their
environment, they look for ways to talk about it. Language, therefore, mirrors the preoccupations of the child at each stage.

In a trivial way, this is undoubtedly true, in that children talk about the things which concern them. But it cannot explain why children proceed to further stages of language development when their own primitive structures have the desired effect. For example, if a child says WHERE KITTY?, she is likely to be told what she wants to know – where the cat is. Why, therefore, should she, and most other children, proceed to (probably) WHERE KITTY GO?, then some weeks or months later to WHERE KITTY HAS GONE? and finally to WHERE HAS KITTY GONE? In brief, the argument that the child learns language in order to help her to manipulate the world does not explain why she does not stop learning as soon as she starts obtaining what she wants, nor why we find similar structural developments in different children.

A child called John provides further problems for the notion that children develop language in order to cope with everyday needs (Blank
et al.
1979). John used language creatively, and had a firm grasp of linguistic structures – but he did not use language to communicate. He disliked interacting with others so much that he never spoke directly to anyone, even his parents. He simply talked to himself as he played with his toys: ‘Let’s go shopping. Where’s the money? OK here’s the change. Open the door. Pretend it’s a shopping centre. OK get elevator. Push button.’ John provides evidence against the view that children are sociable beings who cater for their needs by communicating with others.

Let us, therefore, look at another factor, which may be important in understanding the stages by which children acquire language. This is general mental development, or rather, general cognitive development as it is more usually expressed. Some people have suggested that language acquisition is both dependent on it, and caused by it. Like the ‘everyday needs’ view, such a belief is obviously justified to a limited extent, since ‘It is tautological that linguistic development presupposes cognitive development in the uninteresting sense that one cannot express a concept that one doesn’t have’ (Fodor
et al.
1974: 463). Certain concepts seem to be easier for children to grasp than others. For example, English, Italian, Turkish and Serbo-Croatian children were asked to describe where an object such as a nut was placed in relation to one or more other items, such as plates or glasses (Johnston and Slobin 1979; Slobin 1982). They could all cope with the nut being IN, ON, BESIDE or UNDER a plate before they could describe it as being BETWEEN two plates.

It is also true that certain cognitive abilities and language structures tend to emerge at similar times. For example, one researcher claimed that the development of comparative constructions (I AM BIGGER THAN YOU) occurs at a time when the child can recognize that a pint of milk remains the same whether it is poured into a long thin container or a short fat one (Sinclairde-Zwart 1969). However, the simultaneous development of different
abilities does not prove that one is dependent on the other, since in the normal child, many aspects of growth take place at around the same time. As one researcher noted ‘Hair growth and language development might be positively correlated, but few psycholinguists would wish to posit interesting links between the two’ (Curtiss 1981).

Perhaps the best way to test whether language acquisition and cognitive development are inextricably linked is to search for children who show some discrepancy between cognitive and linguistic abilities. If such a discrepancy can be found, then clearly the link is not an inevitable one. And there are reports of several children whose general cognitive development is unrelated to their grasp of language structure.

Consider Laura, earlier known under the pseudonym Marta (Yamada 1988, 1990). Laura had been a limp, floppy infant. In spite of coming from a loving, supportive home, her general development was delayed. She could not sit alone until she was 15 months old. She was also severely mentally retarded, and as a teenager was unable to perform tasks which even normal 2-year-olds can carry out successfully. When she was given a stack of pictures to sort, she did not separate humans from objects, as normal children tend to do. She did not understand numbers, and did not know her age. Her short-term memory was limited, and she could not repeat back sequences of more than three unrelated items.

In contrast, her speech was fluent and sometimes richly structured, and had apparently been so since around the age of 9. When her speech was studied in her teens, she produced sentences such as:

SHE DOES PAINTINGS, THIS REALLY GOOD FRIEND OF THE KIDS WHO I WENT TO SCHOOL WITH LAST YEAR, AND REALLY LOVED.

She used syntactic structures which are acquired relatively late in normal development, such as ‘full’ passives as in:

I GOT IT CUT ALREADY BY A MAID (when talking about her hair).

I DON’T WANT TO GET EATEN BY ONE (about crocodiles at the zoo).

Laura was not just repeating back sentences she’d heard, as shown by occasional errors, as in:

WHEN I FIRST WENT THERE THREE TICKETS WERE GAVE OUT BY A POLICE LAST YEAR.

She could also repeat back correctly a sentence such as:

AN APPLE WAS EATEN BY JENI.

This construction is difficult for children who have not acquired the passive.

But the passive was not the only advanced construction she had acquired. Consider:

I SHOULD’VE BROUGHT IT BACK.

I DON’T LIKE HIM PUTTIN’ PAPER TOWELS IN MY MOUTH.

DID YOU HEAR ABOUT ME NOT GOING TO THIS SCHOOL?

HE WAS SAYING THAT I LOST MY BATTERY-POWERED WATCH THAT I LOVED.

These all show a considerable degree of linguistic sophistication as far as syntax is concerned.

But Laura’s speech was by no means ‘normal’. Her utterances were often semantically odd, or inappropriate, as in:

I WAS 16 LAST YEAR AND NOW I’M 19 THIS YEAR.

I WAS LIKE 15 OR 19 WHEN I STARTED MOVING OUT O’ HOME.

SHE WAS THINKING THAT IT’S NO REGULAR SCHOOL, IT’S JUST PLAIN OLD NO BUSES.

WELL, WE WERE TAKING A WALK, MY MOM, AND THERE WAS THIS GIANT, LIKE MY MOTHER THREW A STICK.

In brief, she was able to deal with the structure of language to a perhaps surprising extent, but found it difficult to cope with the type of concepts which language normally expresses.

Genie, the Californian girl whose development was outlined in
Chapter 4
, illustrates the reverse situation (Curtiss 1977). Genie was able to cope with complex feelings and concepts, but her ability to deal with language structure was minimal. She expressed herself mainly by means of content words strung together with little syntactic structure as in: THINK ABOUT MAMA LOVE GENIE, or DENTIST SAY DRINK WATER. Her utterances were appropriate, and often conceptually sophisticated, even though telegraphic, as in:

As Curtiss noted: ‘Genie’s semantic sophistication suggests a conceptual level far surpassing what one would imagine from her otherwise rather primitive utterances’ (1981: 21) – an impression borne out by her relatively good performance in a wide range of intelligence tests.

Laura, therefore, showed that severe conceptual deficits can exist alongside a surprisingly developed language ability, while Genie illustrated the opposite – that conceptual ability can outstrip language structure. These case studies suggest that cognitive development cannot provide the definitive key to the acquisition of language structure – even though it is clearly important for meaningful communication.

But Laura and Genie are not the only ones who showed a bizarre mismatch between linguistic and general cognitive abilities. Christopher, a multilingual savant, and Kate, a savant poet, are two gifted but disadvantaged individuals whose use of language massively outweighs their other mental abilities. A savant, incidentally, is someone who has exceptional skills in one particular area (Treffert 1989/1990). Judging from the literature, most savants have either musical or mathematical skills. But Christopher and Kate are both savant linguists.

Christopher is a man, now over 40, who is unable to look after himself. Yet he can speak English perfectly well. In addition, he is obsessed by languages other than his own, and can translate fluently from over a dozen. Here is a translation by Christopher from Polish, followed by an accurate version (Smith and Tsimpli 1995: 15):

Kate is over 40, but her case is equally strange. Her mental age has been assessed at around that of a 7-year-old. She cannot solve even the simplest verbal intelligence test problems, and has huge difficulties using language in everyday situations. Yet she is judged by professionals in English literature to be a highly gifted poet. Over half her poems describe her own problems (Dowker
et al.
1996):