How convincing are the arguments for an innate language acquisition device?

The idea of an innate Language Acquisition Device, or LAD, was first proposed by the ubiquitous linguist Noam Chomsky. It was his belief that there must be some innate built-in system that allows children to acquire language. I think it is important to be clear that the LAD is not necessarily a concrete thing – it could be that it is an innate system or process. This belief in a LAD is based on a great deal of supporting evidence.


Distiguishing speech sounds

Babies have a tremendous ability to distinguish the sounds of speech. This can be tested using the “high amplitude sucking technique”. This involves a baby being played a recording of a particular phoneme and measuring the rate at which the baby sucks on a pacifier. Eventually the baby will become accustomed to the particular phoneme and be less excited by it and this corresponds with a drop in the rate of sucking. When the recorded phoneme is changed, the rate of sucking increases as the baby is excited by the new sound. This shows that a baby is able to distinguish between different phonemes and since they cannot have learnt to do so, the conclusion is that this is an innate ability.

Babies can even distinguish between two very similar phonemes, or phonemes either side of what is known as a “phoneme boundary”, where a certain physical parameter, such as voice onset time, can change the perception of what the phoneme is. An example are the phonemes /k/ and /g/. Babies can detect a difference as small as 20ms in vocalisation over the phoneme boundary in order to distinguish the two phonemes.

What is interesting is that a baby appears to be born with the ability to distinguish between different phonemes in all languages. Over time a baby will learn which phonemes are relevant to the language it is surrounded by and it loses the ability to distinguish easily between the phonemes of foreign tongues. For instance, in Japanese, the l/r boundary that is present in English does not exist, so native Japanese speakers as a consequence have difficulty in distinguishing it.

There is a counter argument to the idea that babies innately can tell the difference between phonemes and that is they learn to tell sounds apart from what they can hear while still inside their motherswomb. However, this argument may not be a particularly good one since to learn these differences would still imply some innate ability to do so.

Vocal play

Babies also engage in what linguists refer to as “vocal play”. This is when a baby spontaneously utters repeated phonemes. This is seen as a kind of practice for the baby in learning how to make the sounds of its native language for itself. Chomsky would argue that this is another innate ability since babies do not apparently learn to make sounds. They learn what sounds to make and what the phonemes sound like, but their own vocalisation abilities appear to be innate.

Neurophysiological evidence

Some of the best evidence for a LAD comes from studies on the brain itself. There are several areas of the brain that have been identified as having a role in language, among them the auditory cortex, the visual cortex, the primary motor cortex, and the arcuate fasciculus. However there are two areas that seem to be dedicated to dealing with speech. Wernicke’s Area seems to deal with the comprehension of speech and Broca’s Area the production of speech. The presence of these dedicated areas of the brain are firmly in support of an innate ability to acquire language.

The ease of language acquisition

The ease at which children are able to acquire language could also point to the existence of a LAD. Children will typically have a vocabulary of around 15,000 words by the age of six, which requires a child to have learnt vocabulary at the rate of about nine or ten words per day. Also all children seem to go though very similar stages of acquisition which implies that there is something common to all children and this could be a LAD.

Deep structures and rules of language

An interesting proposal by Chomsky is that all human language has the same “deep structure”. Chomsky noticed that every language have things in common. For instance they all have nouns, verbs, and so forth. Chomsky’s view relates these similarities due to humans all having an innate understanding of the same deep structure to language and that this is all part of the LAD. In addition, there has been found, a place where a community of segregated deaf children have developed their own unique sign language and this language displays many similarities to human language generally, thus supporting Chomsky’s idea of there being universal deep structures to language.

Further possible evidence for a LAD is that children are able to construct hypotheses about the rules of their particular language and then use these rules to create their own utterances. This goes against theories such as that of the behaviourist B.F. Skinner, who believed that children only learn language by copying what they hear. A LAD can be seen as having a creative element to it and this creativity is seen in children saying things they would never have heard an adult say. This is seen particularly in over-regularisation. An example of this is where a child will begin to see that the past tense of verbs in English generally have an “-ed” ending and so they then apply this to all verbs. This causes errors when used with irregular verbs. So a child might say “writed” instead of “written”, or “runned” instead of “ran”. Children would never have heard an adult say these words, so Skinner’s theory is ruled out, while the LAD could account for this.

Also there seems to be some kind of innate understanding of syntax. During the early stages of language acquisition a child may only be able to make utterances that are two words in length. But even so, they still somehow order the words correctly in so that their intended meaning is conveyed. For instance, a hungry child would say “me eat” rather than “eat me”.

Problems with the language acquisition device hypothesis

However, the LAD is not an infallible idea. There is an implication with Chomsky that the LAD is all that is needed to acquire language. This is not the case. It has been found that social interaction is required in order for language to develop. This is seen tragically in the case of Genie, a girl locked away in isolation until her early teens. She had no language when discovered. Plus she could not acquire it to a level of a normal person, thus demonstrating a “critical period” of development outside which the acquisition of language is very hard, if not impossible. A good supplement to the idea of the LAD is Vygotski’s Language Acquisition Support System (or LASS). This takes into account the need for social interaction. There are also theories of how cognitive development has influence. For instance, a child would have to learn the concept of “hot” and “cold”, or “long” and “short”, before they could successfully understand and use the words “hotter”, “colder”, “longer”, or “shorter”.

Conclusions

The LAD can be seen as being part of several factors that enable a child to acquire language. It would be unreasonable to say that the LAD is the reason behind every facet of language development. But certainly, it is an important part without which the acquisition of language could not happen.


Based on the lectures of Prof. Oliver Braddick, Head of Department of Experimental Psychology, University of Oxford.

Log in or registerto write something here or to contact authors.