Autism: a Mind-Body Problem

Dinah KC Murray, 1993. From Biological Perspectives in Autism: Proceedings from the 1993 conference of the Autism Research Unit of the University of Sunderland.

• Editor’s note: this was Dinah’s second Durham talk, one year after she first used the term ‘monotropism’ and two years after she started working with autistic people. It is presented here more-or-less as it appeared in the conference proceedings, aside from reformatting to take advantage of the technology of the web, and a few typo corrections. Feel free to email with any further corrections, or if you are keen to see scans of the original paper.

Fifty years after Kanner (1943) identified and named the syndrome, autism is still largely an enigma (see Frith 1989). There is a consensus that it has biological origins, but the biological abnormalities which have been found in people with autism are so many and various, and human neurochemistry is so complex, that a unitary explanation looks radically unlikely, and even a set of distinct explanations looks remote. Huge strides have been made but it must be years before practicable biological solutions to this baffling problem will be available¹.

Luckily, biological systems are not closed systems, and biological events are affected by environmental ones². In practical terms, that means that we can think it worthwhile to concern ourselves with its psychological realities. A considerable consensus about what those are has evolved over the years, as people with many different points of view have added their observations to a common pool. A relative newcomer to autism, I have drawn deeply on that pool³. What can I bring to the area is yet another point of view, amplified by a mathematical model – see Table 1-and guided by a particular understanding of meaning⁴. During the rest ofthis paper,I speculate uninhibitedly in applying those ideas to autism.

Despite the psychological consensus, there is little agreement about any underlying psychological cause, nor about to. relation between any such cause or causes and the observed biological abnormalities. How could a biological dysfunction underlie the psychological dysfunctions we observe ? My basic idea⁵ is that the autistic syndrome makes sense if we assume these people have monotropic⁶ interest systems; where most of us have many simultaneous interest active and alert (polytropism) – see Figures 1 and 2 – people with autism can’t do that but are perpetually attention-tunnelled, disconnected, contextless, and unable therefore to model other people as ordinary people crucially do, unable to develop rich networks, unable to strive for central coherence or to maintain a grasp of time or an awareness of alternatives. (Thin networks may eventually be developed, which modifies this picture – they might mitigate the severity of the presented symptoms; see discussion below “The Older Child”).

Naively, it seems obvious to me that in some sense the distribution of attention in any individual’s interest system must register chemically as well as psychologically.Certain coincidences between features of our mathematical model and what we know about biological systems and autism are noted in Table 1. I don’t know what to make of them, am not qualified to judge, and welcome any better informed speculation. Whether or not there are clues to be found in our model to specific biochemical abnormalities, the model certainly has a value in stimulating the perception of psychological analogies, and my discussion is restricted to those.

In the latter part of this paper I use those analogies as a basis for further speculation, and ask what must it be like to have autism, if I’m right that (rigid) monotropism is its essential feature. Before that I’m going to explore the implications of the model for our understanding of the autistic syndrome as a developmental disorder. In doing so, I will at the same time be giving an account into which a number of apparently conflicting theoretical positions can be integrated; it is an attachment problem (Kanner, Hobson), it is a dysfunction in the “social acquisition device”(Jordan, these proceedings), it is a “social coding” dysfunction (Newson), it is “encapsulation” (Tustin), it is an absence of “theory of mind” (Frith, Leslie, Morton), it is an absence of any drive for central coherence” (Frith).

It is my view, based on personal experience of babies, that they all start monotropic. Polytropism can be seen as a skill which most people acquire and maintain most of the time (it may break down), while for some reason people with autism just don’t acquire it. Why not? and how come the rest of us do ? The answer may lie in the relative proportion of N (the scarce attentional resource) bound up in the monotrope, to the amount of N outside the monotrope, to the amount of N elsewhere. In even the most extreme case, there will be some N outside the monotrope, just not enough for detectable activation to co-occur. If there is quite a plentiful supply of N to handle exotropic input, then it will be a relatively small step to maintain dual arousal.

The discontinuity between the monotrope and its surround will tend to be less steep in ordinary infants than it is in those with autism. But that is itself a graded difference,and not an absolute one (see Table 1) – and its varies from occasion the occasion in individuals. I’m postulating that the steeper the discontinuity (a) the harder it is to stimulate interest, or draw attention, exotropically (this is demonstrable in the mathematical model), and (b) the more uncomfortable – the more like being forced – it is when something succeeds in yanking a monotropic person out of an attention tunnel.

Variety among children with autism

According to our model, there should be a variety in aetiology, in severity, and in overall capacity – as well as in individual interests.

Variable aetiology : Although as babies, infants with autism are typically either dozy or hyper⁷, as they get older their monotropic development and steep discontinuities generally interfere with normal social and cognitive development in essentially the same ways. Perhaps the hyper type will go on being more likely to go suddenly out of control than the dozy sort, but in other respects their similarities should be much more important it is to adopt the attitude proposed : restrict stimuli, maximise predictability, start where the child is. The less severe it is, the more likely it is that the child may find other people and their meanings acceptable if those same precepts are observed; also the more quickly some sort of network between interests will be developed (the less rigid their boundaries)

Variable overall capacity : If attention (N) is a scarce resource which may vary in total from interest system to interest system (as well as fluctuating regularly and irregularly in individuals), then functional capacity shouldco-vary. Very low N may itself promote monotropism:we need to do more experiments with our model to find this out). At the low end of this range the afflicted children will show very little interest in the world, will not even move about much, but will show some simple stereotypies such as head-banging, finger-shaking etc. At the high end, we should find overt activity,including elaborate stereotypies, successful acquisition of a range of skills,some islands of super ability. These children should pick up receptive language quite fast and develop a consumerate degree of cognitive working. In some cases, there may even be so much N free outside the overexcited monotropes that learning can take place more efficiently exotropically once an adequate cognitive network has been developed.

TABLE 1

My previous writing about autism has focused on the cognitive and communicative consequences of monotropism, emphasising (a) how hard it is to stimulate interest exotropically, but it hasn’t really dealt with core social meaning problems. Since those are unmistakably fundamental difficulties for people with autism, I have been wanting to understand how they fitted into my otherwise rather coherent account of the syndrome. I now believe that (b) how uncomfortable it is when someone gets pulled out of a monotrope – clearly related to Kanner’s observation of an obsessive desire for sameness – is the key to the social dysfunction⁸.

Language is typically used by Human Beings to manipulate each other’s interest systems. One way or another speech expresses a person’s interests in such a way that it simultaneously reaches out and lines up the hearer’s interest system with the utterer’s. As far as that goes, language is a survival tool, a way of getting communities of interest to work for you. The absence of responsive Other-models in their interest systems has an enormous range of consequences for people with autism⁹, including direct effects on their use of language¹⁰, But what interests me here is rather what its like to be on the receiving end of other people’s meanings, linguistic and otherwise.

About Meaning

Meaning always involves activity in interest systems, it is not primarily linguistic – though linguistic meaning is a particulary interesting special case.Expressions – bodily, facial, or spoken – are actions of interest systems, which can be perceived: their meaning depends on the interests they express. Recognising their meaning entails having the like interests activated in your own interest system; a language is a very finely tuned tool for this purpose. Gaze and touch prepare you by engaging your attention-opening your interest system for input from some Other; that seems to be the universal meaning¹¹. Without that engagement the Other can’t be sure you’ll even take the other meaning in: the relevant activity may not occur at all.If there is relevant activity it will take the form of cognitive effects – ie changes – in your interest system¹², generally including a redistribution of attention. Some of those effects will be common access all hearers within any language community,namely those cognitive effects given by the syntax and lexicon of that language.

Accordingly when babies learn a language, they learn both a way of creating specific cognitive effects in other people, and a way of other people getting into their heads and doing the same. Now recall the crucial feature in autism that I believe may be detectable very young. Those are very steep discontinuities, and the steeper the discontinuity : (a) the harder it is to stimulate interest exotropically , and (b) the more uncomfortable it is when someone gets pulled out of a monotrope. When parents teach babies to talk, they typically begin the process by observing what the baby’s interested in, and naming it – usually with a strongly marked intonation pattern. The exploit the infant’s spells of unmistakable monotropism.

The next natural stage is to start using those words to attract the baby’s attention to whatever the carer expects the baby to be interested in. In this way,infants begin to learn what to care about – for an interest system is a value system as well as an information store (what you value is demonstrated by what you pay attention to). Most babies find this an agreeable and worthwhile experience; ones with autism I suspect find it too disorienting and disagreeable ever to discover it could be worthwhile. If I’m right, its because the discontinuities are so steep between monotrope and a background that some infants develop an aversion to meanings¹³.

What is more, precisely because of those discontinuities, exotropic meanings which are not high profile will stimulate too little interest to have any effect : infants with autism will tend to pick up only those meanings that belong with their monotropes. So, as the carer’s language passes from 100% co-tropical towards its use as a way of getting the infant to “join in”, so it will become trebly disadvantaged.The infant will tend not to be affected by it, but if s/he is, will tend to be averse to it and what succeeds in getting through exoptopically will be restricted to the highly emphatic, thus compounding the aversion.

All joining in entails engagement with Others’ meanings, entails being stimulated by an Other’s interest. Only if the Other’s interest and the infant’s are the same with the Other’s meanings be acceptable to a steeply monotropic infant. So long as the Other does all the joining, some of those infants may go on finding other people and their meanings quite acceptable. Some minority might even find them so acceptable that they would not after all go on to develop the autistic syndrome, but would acquire a motivation to model Others sufficient for that first crucial step out of monotropism to occur. If, for the sake of argument, we take Other people to be a sort of fuel for our interest systems, we could accommodate it inour model. Though local in immediate effect of input of Other’s meaning could be to increase the general level of arousal in an interest system. It would thus reduce the discontinuities, and take the system closer to a polytropic mode of action. But infants with autism will only get that benefit from co-tropical meanings.

If I’m right about both autism and meaning, then much follows. Autism involves recoil from a particular aspect of social life, not from social life itself. It means that attachment will become progressively less attractive as the baby discovers that it pulls both ways. If we want these infants to develop attachments and create Other models then our best chance will involve maximising their control of events, minimising exotropic events, and giving them as much attention co-tropically as possible. This approach should be beneficial to some degree in all but the most severe cases, from the earliest months especially, but also for older children and adults with autism (I expand on this below, and see Tables 2 and 3).

One real source of concern does occur to me. These children are exceptionally unbiddable, if we put them in control so young, isn’t there a danger of them becoming tyrants? May they not learn just enough about some other people to achieve control over them, rather than forming reciprocal attachments? That’s always a risk with autism, and perhaps my approach would make a higher risk? – my guess is that it would have the reverse effect. These infants would not be encountering people control, and so would not need to find ways of preventing it – taking control yourself being a most effective way of doing that if need be¹⁴. And, the more you see people as friends, as co-members of communities of interest, the less likely it is that you will feel an urge to control them. Compare you feelings towards, eg., friends and fellow supporters avidly watching a match with you on the TV, versus your feelings towards someone ringing up while you’re so engrossed.

In an ideal world, it would be possible to diagnose autism by around six months¹⁵. Until a reliable method of early diagnosis has been devised, no amount of advice on early treatment will do any good. So I shal turn my attention now to the further implications of the model for autism in older children. For context, see Table 2.

Older Children

There are some further points to make about the use of language after its foundations have been laid. Wherever possible, a carer’s utterances should “start where the child is”, or as I have been putting it, they should be co-tropical. According to Howlin and Rutter (1987; 57), and in accordance with what we would predict,expansions of the child’s speech are particularly helpful to language development,whilst “high rates of imperatives, demands, or negative remarks” slow it down. But the exigencies of real life are such that attempts to arouse exotropic interests are sometimes unavoidable.

Basic Symptoms	Speculation about how it feels for people with autism	Appropriate Responses
extreme responses/lack of response	Events can be inside or outside their attention Inside everything is strong, very vibrant, but sometimes too much and they overload; Outside everything seems random, senseless.	Restricted stimuli
	Also events can let them stay in theirattention tunnels or force them out – allnew events that need coping with (demand attention) do that. Other people’s meanings do that.	reassurance calm environs
rigidity aversion to change	Being forced to leave an attention tunnel is never pleasent and always confusing: they long to get back in.	gentleness
meaning-aversion gaze-avoidance speech lack	To avoid other people’s meanings is therefore highly desirable – that’s especially easy for individuals with autism because they are so attention-tunnelled. Much of the time it’s not that hard anywayto avoid communication, to avoid gaze, to keep from speaking.	cotropicality empathy
objects preferred to animates	Other people are often very alarming, sudden, unpredictable, indeed all animate beings do the the unexpected a lot. They don’t like this.	structure & sequence
agitation anxiety repetitions/stereotypes obsessed with predictability	They are always being taken by suprise, evenwhen people leave them alone. New events whichthey have ne way anticipated keep shockinglyhappening. At those times they often make somethingpredictable happen instead, it comforts them tobe in conrol and to know what’s going to happen next.	reassurance child in control
	They are happy when people let them stay in their attention tunnels, they are accepting of meanings which belong in there, they are comfortable to have their interests shared. they like to see people appreciating what they appreciate – it makes them feel good.	cotropicality empathy

Some of those necessary interventions can be time-tabled, and become acceptable as part of a structures, predictable day, but some just crop up and need dealing with ad hoc. At such times, there are ways of tackling this which should help get your meaning across with minimum stress to both parties. First, say the individual’s name and await a response, that response should indicate a relative readiness to receive new input. Then, using some such phrase as “about ….”, locate your utterance in that individual’s interest system, thus tying it firmly to something known. By now the individual will be as ready as s/he will ever be to absorb and make sense of what you have to say. In my observation, some version of this procedure is very often naturally employed by the caregivers of people with autism.

Meeting the educational needs of children with autism is a recurrent challenge.Hoe do you get them to accept the idea of being taught ? Firstly, of course, if you address them solely in terms of their current manifest interests and build up from there, you stand the best chance of achieving uptake: you’ll be telling them something they want to know, or at least something they don’t mind hearing. Effective interest systems are multiply connected, so its important to create as many branches out of each monotrope as possible, thus reducing their modularity, and maximising the range of potential cognitive effects. In this way, cognitive networks could gradually become quite rich, even in naturally monotropic people. Once again, the exigencies of life-and particularly the National Curriculum-ensure that some teaching will have to be more directive than that. Here the alternative route to acceptability must be followed:maximise predicability.

Making people and events as predictable as possible is not just a matter of kindness. At times of stress there is an increase in autistic withdrawal andthe appearance of stereotypies which is incompatible with successful teaching. It is therefore of real practical value to reduce the high levels of stress in the lives of these children: maximising predictability is an important and achievement way of doing so.

It is cheering to find that all the points I have been making about how to interact successfully with its sufferers can be found among the Approaches to Autism discussed in the NAS booklet of that name. In Table 3 I’ve summarised what I consider the most vital attributes of any approach to autism and used them to analyse the various approaches listed. It is my belief that the life skills and the motivation to communicate which are the practical aims of anyone teaching people with autism will be most effectively learnt when they are delivered in a calm, predictable environment and when the child’s control over events – including the activation on interests -is maximised.

In sum, I believe autism is a highly variable, biologically caused dysfunction, the severity of which can be significantly affected by environmental factors.

Thanks are due to so many people, I fear I’m sure to leave someone out. My immediate colleague, Mike Lasser, Peter Allen and Grant Worrell, have in their several ways greatly furthered my understanding of the mathemetical model. All the staff and children at Harborough School, especially the Red Class, all the helpers and the infants with autism at the NAS workshops, everyone who has commented on any of my papers on this topic (including Frith, Happé, Hobson, Jordan, Morton, Newson, Powell, Shettock and Sigman) – all of these have greatly furthered my understanding of autism. I am immensely grateful to all concerned. In the last year, in particular, critical comments trom E. Newson, and reading and re-reading Donna Williams’ brave book, have significantly influenced my thinking.

Dinah Murray, June 1993

Bibliography

• Allen, Peter M. and Lesser Michael J. (1989) Evolution : Cognition, Ignorance and Selection, available from International Ecotechnology Research Centre, Cranfield,Bedford, MK43 OAL, UK.
• Allen Peter M and McGlade, J.M. (1987) “Evolutionary drive, the effect of microscope diversity, error making and noise” Foundations of Physics, Vol 17, No.7.
• Argyle, Michael. (1975) Bodily Communication. London: Routledge.
• Austin, John I. (1962) How to do things with words. Oxford: OUP.
• Dawson, Geraldine ed. (1989) Nature, Diagnosis and Treatment London : Guilford Press.
• Dawson, Geraldine and Lewy, Arthur (1989) “Arousal, attention, and the socioemotional impairments of children with autism ” in ed. Dawson.
• Dawson, Geraldine and Lewy, Arthur (1989) “Reciprocal subcortical-cortical influences in autism : the role of attentional mechanisms ” in ed. Dawson.
• Frith, Uta (1989) Autism: explaining the enigma. Oxford: Blackwell.
• Frith, Uta (1989b) “Autism and ‘theory of mind’ ” in ed Gillberg, C. (1989) Diagnosis and Treatment of Autism New York: Plenum Press.
• Frith, Uta (1991) “Cognitive development and cognitive deficit” in press.
• Frith, Uta., Morton, John., Leslie, Alan M. (1991) “The cognitive basis of a biological disorder : autism ” in Trends in Neurosciences 14 (433-438).
• Happé, Francesca (1991) “Relevance theory and the breakdown of communication in autism ” from her thesis. London: MRC Cognitive Development Unit.
• Hartmann, Hellmut (1992) “Does the ‘novely-familiarity relation’ of information control the information exchange between parts and processes of the brain and between persons ?” in Living with Autism: the individual, the family and the professional; proceedings of the National Autistic Society’s Conference at the University of Durham.
• Hobson, R. Peter, (1989) “Beyond cognition : a theory of autism ” in ed. Dawson.
• Hobson R. Peter, (1990) “On the origins of self, and the case of autism” in Developmental Psychopathology 2.
• Howlin, Patrica and Rutter, Michael (1989) Treatment of Autistic Children Chichester: Wiley.
• Jordan, Rita (1990) The Option Approach to Autism: Observer Project Report. Available from the National Autistic Society.
• Jordan, Rita. (1991) “A comparative study of the Option Method and Behaviourism as treatments for the fundamental difficulties of autism” unpublished paper.
• Jordan, Rita and Powell, Stuart (199) The Special Curricular Needs of Autistic Children: Learning and Thinking Skills London: Association of Head Teachers of Autistic Children and Adults.
• Klin, Ami (1991) “Young autistic children’s listening preferences in regard to speech:a possible characterization of the symptom of social withdrawal” in Journal of Autism and Developmental Disorders 21(1)29-42.
• Leslie,Alan M. and Happé Francesca (1989) “Autism and ostensive communication:the relevance of metarepresentation” in Development and Psychopathology 1, 205-212. Cambridge University Press.
• Leslie, Alan M. (1992) “Pretence and representation: the origins of ‘theory of mind’” in Psychological Review 94, 412-426.
• Matthews Pat. (1992) “Living with autism” in Living with Autism: the individual, the family and the professional, proceedings of the National Autistic Society’s conference at the University of Durham.
• Matthews, Pat (1993) This volume.
• Morton, John (1989) “The origins of autism ” in New Scientist, 9th Dec.
• Murray, Dinah, K.C., (1992) “Attention tunnelling and autism” in Living with Autism: the indivdual, the family and the professional; proceedings of the National Autistic Society’s conference at the University of Durham.
• Murray, Dinah K.C., (in press) Talking, Thinking and Wanting
• National Autistic Society (new ed 1993) Approaches to Autism: an annotated list. 276 Willesden Lan, London, NW2 5RB, UK.
• Newson, Elizabeth and Hipgrave, T. (1992) Getting Through to your Handicapped Child. Cambridge: CUP.
• Newson, Elizabeth (1992) “The barefoot play therapist: adapting skills for a time of need” in Miller A., and Lan D. (eds) Handbook of Child and Adolescent Therapy. Milton Keynes: Open University Press.
• Park, Clara C. (1990) The Siege: the first eight years of an autistic child (with an epilogue fiftenn years later). London : Little , Brown & Co.
• Powell, Stuart and Jordan, Rita (1991) “Using pedagogical principles to remediate the thinking of children with autism” in Therapeutic Approaches to Autism: Research and Practice, proceedings of the National Autistic Society’s conference at the University of Durham.
• Shattock, Paul (1990) “Some implications of basic physiological research for the behaviour and treatment of paeople with autism” in Experimental Psychology and the autistic Syndromes, proceedings of the National Autistic Society’s conference at the University of Durham.
• Shattock, Paul., Kennedy, Alan., Rowell, Frederick, and Berney, Thomas. (1990) “Role of neuropeptides in autism and their relationship with classical neurotransmitters” in Brain Dysfunction 3, 328-45.
• Shattock, Paul and Lowdon, Gillian (1992) “Proteins, peptides and autism. Part 2: Implications for the education and care of people with autism” in Brain Dysfunction 4(6) 323-334.
• Sigman, Marian and Mundy, Peter (1989) “Social attachments in autistic children”in Journal of the American Academy of Child and Adolescent Psychiatry, 28,1 74-81.
• Sperber, Dan and Wilson, Deirdrie (1986) Relevance, Communication and Cognition. Oxford: Blackwell.
• Williams, Donna (1992) Nobody Nowhere. London, Australia, New Zealand, Canada:Doubleday.
• Wing, Lorna ed (1988) Aspects of Autism: Biological Research. London: Gaskell and Royal College of Physicians.
• Wing, Lorna and Gould, Judith (1979) “Severe impairments of social interaction and associated abnormalities in children: epidemiology and classification” in Journal of Autism and Developmental Disorders 9, 11-30.

Notes

1. See many of the other papers in this volume, and see also the 1989 volume n this series, and references in both. ↩︎
2. See Allen and McGlade, and Paul Shattock’s contribution to this volume. ↩︎
3. See Frith, Happé, Hobson, Kanner, Jordan, Matthew, Morton, Newson, Powell, Rutter, Shattock, Sigman, Tustin, Williams. ↩︎
4. The mathematical model is derived from the Allen-Lesser equation, for discussion, see Allen and Lesser. ↩︎
5. See Shattock and Lowdon. Rosemary Waring (see this volume) and Paul Shattock (ditto) have both encouraged my in personal communications to think that speculation on these lines might not be counter-productive. I have appreciated their friendly interest. ↩︎
6. Definitions: monotropic – having active attention devoted to only one interest at a time: polytropic -having several concerrently active interests; exotropic-outside the current monotrope; cotropic – pertaining to the current monotrope. ↩︎
7. Both Newson and Wing believe, roughly speaking, that infants who go on to develop the autistic syndrome tend to fall into one of these categories. ↩︎
8. of Rutter and Howlin, also Boucher this volume, for the great importance of predictability. ↩︎
9. See especially Hobson for discussion of this; from a cognitive angle,see also Baron-Cohen, Leslie, and Frith, or Morton; from a social and pragmatic viewpoint, see Newson. ↩︎
10. See Happé. ↩︎
11. See Argyle. ↩︎
12. For a detailed account of “cognitive effects” see Sperber and Wilson, and Murray, in press. ↩︎
13. This fits with the often reported withdrawal or retreat from speech after initial apparently normal linguistic development. ↩︎
14. Anecdotal evidence suggests that children who have been subjected to “Welch Holding” when they were helpless small people, like to take violent control when they’re big enough. ↩︎
15. A major step towards reliable earlier diagnosis has been made by Baron-Cohen and his research team, with their discovery of definite indicators at eighteen months. ↩︎
16. As well as the Option Method (see Jordan 1990), Gentle Teaching and other approaches scored in Table 3, see Matthews (1992, 1993), and in practice many workers in the field; also Howlin and Rutter. ↩︎