Integrationist Notes and Papers. No.18 © Roy Harris 2007

Integrating Autism


   •Roy Harris

What was wrong with so many 20th-century approaches to language was the divorce between language and communication. It is almost as if linguistic theorists (Chomsky being a notable example) had deliberately set out to formalize a view of language that is ‘natural’ only to autistic children. What is characteristic of various forms of autism (and the root of all ensuing social and developmental problems) is the failure, or inability, to contextualize what other people say as something anticipating or requiring a response.

This failure is immediately interpreted as indicating one or other of various negative responses; e.g. inability to understand, indifference to what has been said, refusal to take part in the conversation, etc. Thus the autistic non-response is treated as evidence branding the non-responder as an ‘outsider’ – someone not a full member of the linguistic community.

Various issues about our conceptualization of ‘language’ arise here. It seems important to distinguish between e.g. saying nothing because you did not hear what was said, saying nothing because what was said did not seem to make sense, saying nothing because you did not realize that what was said was addressed to you, saying nothing because you did not know what to say, saying nothing because you did not wish to commit yourself to a response, saying nothing because you thought it was not the moment to say what you wanted to say, etc. In practice, the difference between some of these alternatives is difficult to draw. From an integrational point of view, however, such distinctions cannot be ignored, because they have implications for the principle of cotemporality, which treats speech acts as events. Saying nothing is itself a speech act. If proof were needed, one need only look at the way society brands those who are too often silent as ‘autistic’.

In his autobiography, Daniel Tammet, subsequently diagnosed as suffering from ‘Asperger’s syndrome’, recounts how at school he would at first fail to respond to the teacher’s simple arithmetic questions like ‘seven times nine’, not because he did not know the expected answer (‘sixty-three’), nor because he did not realize that the question was addressed to him, but because ‘I did not realise that I was expected to say the answer out loud to the class’ (Tammet 2006: 85).

This excuse speaks with the voice of unquestionable honesty, and its honesty drives a cart and horses through simplistic theoretical dichotomies like langue and parole, ‘competence’ and ‘performance’, etc. In Saussurean terms, this can be neither a failure of langue (since Daniel both understood the question and knew the ‘right’ answer), nor a failure of parole (since he could quite easily have uttered the syllables ‘sixty-three’: there was no physiological or neurological impediment). From a Saussurean perspective, then, there was no communicational breakdown at all: the sounds uttered by the teacher had ‘triggered’ the right ‘concepts’ in Daniel’s brain, and Daniel’s brain had come up with the appropriate reply (‘sixty-three’). Ergo, whatever went wrong must have had nothing at all to do with language. Here we can put a finger on the precise point at which orthodox linguistics compartmentalizes between ‘language’ and ‘the world’. (‘Sorry, Daniel. There’s no linguistic problem. You’re just ‘autistic’. Nothing we can do for you.’)

Current medical thinking about autism endorses this diagnosis. But that – an integrationist might suggest – is doubtless because medical authorities have been induced to subscribe to the same language myth as is propagated by orthodox linguists. The core of this myth is the proposition that what underwrites speech (and all other ‘cognitive’ forms of human communication) is the transference of identical ideas from sender to receiver. That is why the theory cannot cope with cases where the supposedly ‘identical ideas’ are intact, along with the physiological apparatus and the physics of transmission, but the ‘speech circuit’ breaks down all the same.

The last thing that orthodox theorists want to do is revise their simplistic conception of the linguistic sign as a binary pairing of ‘form’ and ‘meaning’. That is why they are so reluctant to entertain the semantic possibility that propositions like ‘seven times nine is sixty-three’ are radically unlike ‘beer is made from fermented hops’; or that perhaps the reason why so many of those suffering from autistic disorders are surprisingly ‘good at numbers’ but ‘bad at communication’ is that numerals are not typical linguistic signs. Being ‘good at numbers’ is essentially a matter of grasping the connexions between numerals and other numerals, and these connexions are context-neutral. Seven times nine is still sixty-three, regardless of whether we are counting days of the week or buttons in a box.

True arithmetical propositions, philosophers tell us, are ‘analytic truths’; unlike ‘beer is made from fermented hops’. And the characteristic of analytic truths is that they yield no information about the world. But ‘2+2=4’ is a special kind of analytic statement: it is ‘language-neutral’ in the sense that it stands you in just as good stead in an English classroom as in a Chinese classroom (although the equation will be ‘read’ differently in the two languages). That puts it in a different class from ‘Anything white is white’.

So is the autistic mind one that somehow privileges language-neutral analytic truths to the detriment of language-based non-analytic truths? That does not quite fit Daniel Tammet’s account either. He describes his difficulty in following a narrative as an inability to ‘connect the different statements and get the overall picture’. And, more picturesquely: ‘It is like joining the dots in a children’s colouring book and seeing every dot but not what they create when joined together.’ In other words, not a problem of relating language to the world, but of integrating various parts to form a whole.

Tammet, D. (2006), Born on a Blue Day, London, Hodder & Stoughton.

INP 19 - The Grammar of Numbers


© Roy Harris, Emeritus Professor of General Linguistics, Oxford, 2010