A Few Observations on Autism, Language, and the Brain
by Mark English
Areas of the brain used in mathematics have been identified. These areas, which are associated with our sense of number and with spacial reasoning, are used not only for basic arithmetic and to process our ordinary perceptions and intuitions about shape and form and space but also, it now appears, for advanced math. (1)
High levels of mathematical ability come at a price, however, and often at the expense of cognitive ability in other areas, especially areas related to the social world. There is abundant anecdotal evidence for this and there is also a growing body of solid research on the topic.
Social skills draw on many different brain areas: some of these skills (and the respective areas of the brain where they are processed) relate specifically to language and some don’t. Ordinary language is obviously a powerful and crucially important social tool, but it generally works in concert with other perceptual and behavioral functions. Neither body language nor facial recognition capacities, for example, are related directly to language or draw on the linguistic (grammatical) processing centers of the brain, but they too are important elements in our day-to-day interactions. (2)
As it happens, facial recognition deficits in mathematicians have been demonstrated. (3) This finding shows that mathematical processing power comes at the expense of processing power in at least one area associated with social functioning. It also brings autism to mind.
One of the standard tests designed to diagnose autism involves showing the subject photographs of human faces and asking them to select for each the best match from a short list of words descriptive of particular emotions or attitudes. If you are on the autism spectrum, you typically do badly in such tests.
The research referred to above has shown that mathematical reasoning occurs in different parts of the brain, not only from those used for processing the grammar of natural language but also from those associated more generally with language-based reasoning – specifically, in this case, historical reasoning. Autism is a condition which is associated with cognitive deficits in social domains, usually including language, and it often manifests itself in a formal educational context by weakness in subjects, like history or English, which rely predominantly on language-based reasoning.
Where does music fit in here? Is it processed quite separately from language and mathematics or is there some overlap? Our brains are not modular in any absolute sense, and there are many possible links between all three areas. I have the sense that the links between music and language processing are sometimes overstated, however. (4)
Language – understood in terms of core grammatical processes – is quite distinct from music, I would say. The very existence of medical conditions which involve dramatic deficits in linguistic ability but which don’t affect other abilities such as the processing of music demonstrates that very specific areas of the brain are responsible for dealing with language’s basic morphosyntactic, phonological and semantic mechanisms.
One aspect where there is overlap between linguistic and musical processing relates to pitch and tone. The main language centers are on the left side of the brain, but speakers of tone languages like Mandarin make much greater use of sections of the right hemisphere, specifically those involved in the processing of music and melody.
We all use tone in speaking, of course. In languages like English or French, while a difference in tone may affect the pragmatics of communication, it does not change the basic meaning of words. Tone in these languages is often seen, with other pragmatic elements, as paralinguistic because it does not contribute to core grammatical processes. In tone languages, by contrast, different tones or tone patterns are built into basic grammar and so have direct and immediate semantic relevance.
While tone languages are in no way more primitive than other languages – in fact they often developed from non-tone-based languages – there is another form of speech which some believe may provide some clues about the early development of the human language capacity, and which also has links to music. I am talking about infant-directed speech, or motherese, where tone and pitch-based vocalizations combine with repetitive structures and rhythmical patterns – all music-related elements.
This form of communication is usually classed as nonlinguistic, even though it may utilize words, because it depends primarily on rhythm, tone and pitch to convey meaning. Its features generally include simplified grammar, exaggerated speech melody, diminutive forms of words, and a highly repetitive style. It is sometimes suggested that motherese – or rather something like it – preceded the development of the human language capacity and was used in many contexts, not just with infants.
It’s fascinating to watch the way language develops in infants. What is most striking is the apparent ease and speed of acquisition. Moreover it is based on random, disorganized data. Language acquisition is a large – and extremely contentious – field of study, and I don’t want to make too many very specific claims here. I will, however, make a few points which tie in with some of the themes raised above.
Firstly, it’s clear that a period of free experimentation with a wide range of sounds precedes and is a prerequisite for normal language development. Very soon, however, a baby’s babbling starts to reflect the particular sounds of the language (or languages) he or she is hearing. Each language utilizes only a subset of possible sounds, and the infant starts to imitate these building-blocks (or phonemes) well before she starts imitating full words or phrases. (5)
This process is strangely dependent on other aspects of social interaction. For example, autistic infants (whose linguistic development is usually delayed) will typically not give special attention to human speech or to faces, preferring to focus on the sights and sounds (and tactile sensations) associated with the non-social world. This applies equally to natural and artificial sounds and objects. A lack of early babbling is often a sign that a child is autistic. Likewise, a lack of attention to, or interest in, faces is an early indicator of autism.
It goes without saying that many mathematicians and scientists are on the autism spectrum. There are a lot of myths about autism, both positive (based on popular films and/or famous people) and negative (based on the stereotype of the disruptive child or the person who is incapable of sympathy). The positive myths may lead us to overlook the fact that full-fledged autism is in most cases associated with a lower-than-average IQ. For every high-functioning autistic person, there are many who remain seriously disabled and dependent throughout their lives.
Further evidence for language processing being to a large degree separate from, say, mathematical thinking and musical perceptions and abilities comes from the cognitive testing of children. The cognitive deficits of many autistic children are restricted to – or at least concentrated in – areas associated with language and communication.
Whereas normal (or neurotypical) children learn the basics of the syntax of their native language very quickly, without specific instruction and seemingly without effort, autistic children are typically not only slow to develop linguistically, they also often have to make a conscious effort to learn correct syntactic forms. It’s as if they are using general learning mechanisms rather than the specific areas of the brain that are primed (in neurotypicals) for language processing. (6)
- Since we associate certain familiar faces with names, some facial recognition processing will also activate the areas associated with the relevant lexical items.
- See link at note 1.
- For example, by Anirruddh D. Patel in his book, Music, Language and the Brain (OUP, 2007).
- Phonemes are the smallest sound units capable of distinguishing meaning in any given language.
- There are many ways of conceptualizing this and, as I indicated, I don’t want to get into language acquisition controversies here. But it certainly does appear that the specific areas of the brain devoted to grammar are open to a large (but finite) set of possible linguistic systems. If they are functioning properly and are fed appropriate input within a given timeframe, they allow the child very quickly and easily to lock into the syntactic and phonological conventions of the particular language(s) to which he or she has been exposed.