A Topological Theory of Autism

by Gregory B. Yates

2002i23 Draft


Every population has a surface, however diffuse. This fact of topology firmly drives brain evolution. Autism in particular is primarily the result of adaptation to the inevitable sparseness of population outer regions. This yields not only autism, but an autism spectrum. It also gives rise to the great variety of autistic forms and genes. These conclusions sharply limit the possible clarity of autism diagnosis and preclude any simplistic “silver-bullet” response to the rigors of autism. The theory accounts for the observed correlation of autism and gender, and predicts a correlation with latitude of ancestral homeland. Autism emerges as a major feature of brain evolution: It is generally not a disease. Autism has been with humans as long as humans have been and has marked human history.


This theory is part of two successively more general theories of brain evolution, works in progress styled A-Priori Object and Autocatalytic Text Theories.

A tutorial on this theory, with an interactive autism evolution simulator, may be found at   www.AutismTheory.org

Introduction: Autism and Social Disconnectedness

Once autism had no name. It is useful to remember this when crafting a theory of autism because it underscores the obviousness of autism’s central defining feature: social disconnectedness. The name “autism” derives from the Greek word “auto” for self, and proclaims the apparent mental involution or self-absorption of autistic people. Historically the name “autism” was invoked at least three distinct times to describe clinical syndromes: Eugen Bleuler coined the term to name the sort of social disconnectedness seen in adult mental illness, including schizophrenia;1 Later, Leo Kanner and Hans Asperger used the term independently of each other to name behavioral syndromes in children.2,3 In all three cases it was the startling social disconnectedness of the people described that prompted use of the term “autism”: The three founders clearly saw other features of autism as secondary to social disconnectedness.

Because social disconnectedness is the central, eponymous feature of autism it is the primary feature for which a theory of autism must account. A good theory of autism may also account for some secondary features. Indeed, some subsidiary facts of autism, for example that it is more prevalent among males than females, beg for explanation. A sample case of mild autism illustrates typical secondary features of autism: A boy routinely rocks and bangs his head against furniture; He cries at the touch of a few water drops, stuffs paper in his ears to muffle sounds, and rises up on his toes when walking; He speaks like a little professor; Physically (as well as socially) awkward, practice at fluid motions brings him little skill; People remark at his poor eye contact; He has a focused, persevering interest in mechanisms like flying machines, electronic devices and, later, brains. This is only one example: Secondary autistic traits occur in kaleidoscopic variety. Although secondary traits like the ones given may merit explanation at some point, it is social disconnectedness that most defines autism and that demands the primary attention of a good theory of autism. Social disconnectedness is the horse of autism: Secondary features are baggage in its cart.

There is a further reason to ask that a theory of autism first address the social disconnectedness of autism, and that is in the interest of compassion. While some secondary features of autism are unpleasant, in a social world one autistic trait is truly devastating. That is autism’s defining characteristic itself -- social disconnectedness. Autistic people live like Tantalus, with the fluent social interaction of others suspended before their eyes, out of reach. This hell can extend to others, for example to an autistic person’s parents, who may surround their child with affection only to be treated in return like objects. Autistic people vary in their desire for social interaction. However, even those who do not desire social contact can be devastated by its lack, for thriving in human society depends on social ability. In tracing the origins of autistic social disconnectedness a theory of autism can suggest ways to respond to autism’s most grueling challenge. Logic and compassion alike thus dictate that a theory of autism first address autism’s central and defining feature, social disconnectedness.

Physical and Social Disconnection

Increasing the physical distance between people generally decreases their social interaction. This is particularly true in the absence of modern technology. Without technology people separated by more than a stone’s throw rarely relate socially at all. At the opposite extreme people embracing are almost certainly -- and in a basic sense absolutely are -- relating socially. Without technology people living near others may be socially disconnected, but people living far from others must be. Social and physical disconnection correlate strongly, and this correlation is profound in the bulk of human history preceding modern technology.

The link of physical and social connection holds even when communication technology exists: With such technology people living together can socialize both directly and indirectly, and so are more socially connected in general than are physically isolated people, who can socialize only indirectly. In general, people spending much time at great average distance from neighbors are socially disconnected relative to others living near each other.

The argument now proceeds to the central theme of object topology.

Population Surfaces

Every object has a surface, however diffuse. That is, any object, whether individual or collective, can be substantially contained in a sufficiently large bag. This basic fact of topology gives the present theory its name: It is the simple idea at the theory’s core. The fact of object surfaces assures that the density of any object declines outward at some point. In the case of a collective object like a population this means that the density of a population always declines outward at some point.

Where population density is low, as it must be at population surfaces, mean distance to nearest neighbors is high. This shows that people are more physically disconnected in a particular place: People in the sparse outer regions of a population are more separated physically, and therefore more disconnected socially on average. Furthermore, because every population has a surface, people who are relatively disconnected socially always exist in a population. The facts of biological adaptation now add force to these conclusions.

Defining Adaptation

How do we tell that an organism is adapted to its circumstances? Initial answers to this question are likely to take the form of teleological stories like this: “A mackerel has gills suiting it to undersea life. Gills show that a mackerel is adapted to life in the sea. Roots show that an oak is adapted to life on land.” A more rigorous reply to the question follows.

An organism that perishes instantly in a particular circumstance is not adapted to that circumstance, whereas an organism that, itself or in kind, persists in a particular circumstance is adapted to it. An oak perishes almost instantly in flaming lava and so is not adapted to life there. Mackerel persisting for generations in the sea are adapted to life there. The case intermediate between longevity and instant death proves instructive. An oak in sea water and a mackerel on land do not perish instantly. They are not as poorly adapted to life in those circumstances as is an oak in lava. Regardless of any teleological stories a fish out of water and an oak at sea are adapted to their lives there for precisely as long as they survive. An organism that persists in a circumstance for a time is adapted to the circumstance for that time. Adaptation has a time scale and existence in a circumstance proves adaptation to it.

An organism generally appears not as an undifferentiated mass but as a collection of properties and behaviors -- of traits. To say that an organism is adapted to a circumstance is thus to say that its collection of traits is adapted to that circumstance: The persistence of an organism in a circumstance proves the adaptation to that circumstance not only of the organism but of its collected traits. This conclusion applies to all organisms, including to humans at their population surface.

Adaptation to Population Surfaces

What are the circumstances of humans at the surface of their population? As we have seen, a fundamental fact at a population surface is that the population density there is relatively low. (A rare exception is considered later.) Even in a phalanx the outermost soldiers border on empty space, which increases the average distance to their neighbors. The existence of humans in the sparse outer regions of a population proves that those humans are relatively adapted to large mean distance between neighbors, and their traits necessarily reflect this. Adaptation to topology relentlessly acts to create a syndrome of social disconnectedness whether or not this syndrome is called autism.

The Persistence of Autism

Existence proves adaptation, but there is also adaptive inertia: The longer a trait persists in a population the more slowly it is likely to change, even when the organisms’ circumstances change abruptly. Thus a polar bear transported to a temperate climate demonstrates adaptation to that climate as long as it survives there, and yet the bear’s progeny in temperate climes do not immediately lose the thick white fur adaptive to arctic life. Why this is so is obvious from the perspective of Charles Darwin's theory of evolution. In that theory a trait is seen as the result of randomly arising genetic fluctuations fixed in a population through natural selection: What arises in such a fashion through multiple time-consuming events does not change instantly when circumstances change abruptly.4

The roots of adaptive inertia, however, go deeper than Darwin's accounting for them. Adaptive inertia is traceable even to an inertia of phenomena in general: Phenomena that have been long-lived in the past are more likely to be long-lived in the future than those that have been short-lived. A mountain is more likely to persist than is a lightning bolt. If there were no generalized inertia the lifetimes of objects and events would change haphazardly: Some lightning bolts would last eons, and mountains would often vanish in a flash. Such generalized inertia is reflected in organism traits, both genetic and cultural, and appears there as adaptive inertia. Abrupt trait changes are possible in spite of adaptive inertia, of course: They are simply less probable.

In the case of sparseness-generated autism adaptive inertia causes autism-producing genes (and memes) to remain autism-producing even when they move from the sparse population surface to the dense population interior. The expected lifetime of these traits is related to the time scale of their tenure at the population surface. Because of adaptive inertia autistic traits observed in cities today can easily be the result of adaptations that persisted for thousands and even millions of years at the sparse frontiers of ancient human and hominid populations.

Charles Darwin’s theory of evolution suggests how organism adaptations can arise. The Topological Theory of Autism is more concerned with the fact of adaptation than with its underlying mechanism. As a result the Topological Theory of Autism is compatible with Darwin’s theory but does not rely on it.

The shift in emphasis from the mechanism to the fact and persistence of adaptation permits the Topological Theory of Autism to apply on a wide range of time scales, including the scales of both genetic and cultural change. Thus the theory provides an alternative to debate about whether a trait is genetic or cultural in origin. Within the Topological Theory of Autism the expected lifetime of a trait matters more than how in detail the trait arises.

Trait Exchange in Autism

When an organism’s circumstances change so that some of its traits are no longer adaptive, those traits can sometimes find other uses. The flippers of whales are apparently the former feet of ancient terrestrial mammals, for example.5,6 Such trait recycling is in fact to be expected in evolution because it allows adaptation to proceed without gross overriding of adaptive inertia. When a trait that is no longer adaptive forms the basis of new adaptation the resulting change can destroy the earlier function. Whale flippers make poor feet, for example. Such loss of function is to be expected in time even without trait recycling because a trait usually requires resources to sustain -- resources wasted when its function is no longer adaptive.

Human social interaction is complex and doubtless needs many brain resources to sustain. Social interaction is, however, largely absent at the sparse population frontier. For this reason, among outwardly migrating humans it is precisely those brain resources used in social interaction that would be expected to recycle over generations, and to come to serve the more pressing needs of surface dwellers -- for example, the need to cope with sparseness by using available objects. Surface-dwellers thus not only gain objective ability, they lose prior social function. The effect is that there is trait exchange in the history of autism: social ability is effectively exchanged for ability with objects. The syndrome of social disconnectedness at the human population surface is reinforced by trait recycling and trait exchange. Social ineptitude and unusual objective ability are often found together in autistic people today.7

A “topological force” drives human adaptation to sparseness. The result is outlying individuals suited to life at some remove from most neighbors -- people we would not hesitate to call socially disconnected, and whose traits may be expected to resemble those we call autistic today. These traits, born in sparseness, are expected to persist even when transplanted from wilderness to city.


This summarizes the argument thus far:

  1. Social disconnectedness is the defining and most difficult feature of autism and is thus the first concern of an autism theory.

  2. Because physical and social connection are linked people in sparse outer population areas are relatively disconnected socially.

  3. Adaptation to the topological fact of sparse outer areas generates persistent social disconnectedness, whether or not this is called autism.

2 and 3 are two bridges to a single result: At the sparse population surface people are generally more autistic both because to be socially disconnected is what it functionally means to be autistic and because persistent social disconnectedness is unavoidably expected from adaptation to sparseness.

It is easy to make up Just So stories about evolutionary adaptations. An imaginative person can invent "adaptive reasons" for almost any trait. For example, one can find in noses an adaptation to party masks and in elbows an adaptation to dinner tables. What makes theTopological Theory of Autism different from these is the stark unavoidability of the fact that every population always has a surface. An ancient person may or may not have had party masks or a dinner table, but the person's population (and all ancestor populations) always had a surface. This is not a matter of fancy, but of fact. It is like an x-ray into all of past history. For the reason of its unrelenting certainty, the fact of population surfaces is strongly expected to leave a genetic imprint. Few Just So stories can claim this degree of certainty in their premises, and it is this concrete certainty that sets the Topological Theory of Autism apart from them.

The Autism Spectrum and Autistic Variety

From the results given above two of this theory’s main conclusions follow directly: They are that autism necessarily exists in a spectrum of intensities and also in a variety of forms. If autism is a reflection of and adaptation to sparseness at population surfaces it follows that where population is less sparse less autism is expected. An autism spectrum is thus expected to extend from relatively low incidence and intensity where population has long been dense, to relatively high incidence and intensity at sparse population surfaces. This means that systematic average differences of human temperament are probable. Because a population density gradient always extends from where a population is to where it is not, the forge of autism and of an autism spectrum has never cooled. Every body being physically distinct, all partake to some degree of an essential autism. Beyond this, however, there must be an autism spectrum. The Topological Theory of Autism shows not that there is an autism spectrum but that there must be one.

Autism must also exist in a variety of forms. That is because the topology mandating social disconnectedness does not mandate what traits underlie it. A huge variety of traits can underlie social disconnectedness. A person may lack or fail to develop any of myriad social instincts transparent to those who have them and opaque to those who do not. An anomaly in almost any sensory or motor pathway, an unusual gait or peculiar manner of eye contact -- almost any oddity of behavior or appearance -- can promote social disconnection. Topology mandates the disconnection but not the specific traits. This is comparable to a guitar string, where length fixes the fundamental frequency of its vibration but allows great freedom in the specific pattern of harmonics that plays upon it. There is similar freedom in how social disconnection is realized at population surfaces. The expectation of autism itself is as certain as the fact that a population has a surface: Any constraint on autistic variety is unlikely to match this degree of certainty. The result is that autism necessarily arises in a variety of forms. Autism is not a single condition: Its inherent variety precludes any simplistic “silver-bullet” response to its rigors.

If autism is strongly expected to arise in a spectrum of intensities and with a variety of forms it follows that there can be no clear diagnosis of autism. For any person with a given degree of autism there are people slightly more and people slightly less autistic; For any autistic person with a given set of traits there are others equally autistic but with widely differing traits. There may be modest clusters of autistic types but intermediate forms are strongly expected, and a person quite able to manipulate text and verbal symbols can be fully as autistic as a person spinning mutely in a corner. There is no pure autism: All are hybrids, not only of autistic traits, but of autistic and non-autistic traits as well. The medical love of tidy diagnosis must be disappointed in the case of autism.

Sometimes the full adaptive value of a trait depends on the presence of other traits. A finger is of less use without a hand or arm, for example. The same is true of adaptations to sparseness, with the result that autistic traits can appear as fragments whose adaptive value is not immediately obvious. The adaptive value of autistic rocking is not immediately obvious, for example, but it may be that in concert with other traits it translates to a taste for the rhythmic action of walking, which is adaptive to food gathering in sparse regions.8 This is merely speculative, of course, but it illustrates how autistic traits that are in fact part of adaptations to sparseness can appear nonfunctional.

Maladaptive Autism

Surely, however, many autistic people could not survive at rugged frontiers. Profound autism appears to be, if anything, maladaptive. Doesn't this contradict the Topological Theory of Autism?

Profound autism is the variety likely to be noticed first. This biases a view of autism in general, and it probably favored the historical interpretation of autism as a pathology. The autism revealed in the theory is likely to be so ordinary in many forms that it is hardly noticed. Consider for example traits described by words like shy, aloof, introverted, "nerd", loner, awkward, and eccentric. The Topological Theory of Autism does expect cases of profound autism, just in smaller numbers than moderate cases. The autism spectrum provides that even exceptional cases do not lie far from the rest of humanity and that, however maladaptive those cases may be, they can well arise from the combining of traits that are common and otherwise adaptive.

Social ability, while transparent to those who have it in good measure, is very complex, with an underlying mechanism that can be damaged. Doubtless some autism is the result of such damage and may be maladaptive. However, the theory underscores that lack of social ability is not in itself evidence of brain damage, and can easily be adaptive. Before concluding that some form of autism is maladaptive it is well to recall that it is sometimes possible to survive in sparse areas where people often perish, if one lacks socially conventional food prejudice and eats bugs.

There is more to some autism than the theory can explain. The theory shows only one important part of autism's origins: It does not show all the parts. Profound, seemingly maladaptive autism merits careful and caring attention, and it is consistent with the Topological Theory of Autism. The theory suggests looking for the roots of profound autism not only in damage to genes and brains, but also in combinations of commonly occurring traits.

A Direction of Evolution

The Topological Theory of Autism predicts a direction of evolution. It is a common misunderstanding of Darwin’s theory of evolution that it specifies a direction of evolution. It does not. Darwin’s theory does not predict that monkeys will give rise to humans, nor that some humans will be autistic. Darwin’s theory is a rudderless boat.9 The Topological Theory of Autism theory predicts minimally that wherever brains arise there also will arise autism. Because of the iron influence of population topology autism emerges not as a minor pathology, but as a major feature of brain evolution.

Autism, Gender and Race

The Topological Theory of Autism predicts that autism will correlate with gender and race. This is, of course, a very sensitive subject so I present the basic facts for the reader’s own assessment. The facts are simple. Population density gradients exist on a variety of scales, including tribal and global scales. These density gradients correlate with differences in tendency to move about. The sparse outer regions of a tribe are generally populated by the people who are able and inclined to travel into and throughout the outer regions. For the vast bulk of human history one population within most tribes has been systematically more mobile than another: Men have been generally more mobile than women for the simple reason that bearing and nursing children -- the biological legacy of women -- makes rambling widely difficult. A mother’s movement also requires energy that is effectively stolen from fetus and infant. Simple mechanics and bioenergetics thus populate the sparse outer regions of early tribes disproportionately with males, with correspondingly simple adaptive consequences: Human males have lived in surface sparseness and their genes have come to reflect it. By most accounts autism is about three times more prevalent among human males than females.10 Adaptation to topology plausibly accounts for the far greater incidence of autism among males than females.

The same argument applies on the global scale. Current evidence -- both archaeological and genetic -- suggests Africa as the cradle of the human species or of its ancestors.11 From there migration proceeded outward into sparse regions. The sparsest regions were those with the fewest nutrients to support human life, and those farthest away from Africa. Deserts and mountains cannot support dense human populations and so would be expected to leave an adaptive imprint on the incidence of autism. However, there is on Earth a nutrient gradient that dwarfs those formed by deserts and mountains: It is the gradient that results from the fact that life-giving sun shines more brightly at the equator than the poles. As a result a great nutrient gradient extends from equator to pole on the terrestrial sphere. The poles simply cannot sustain dense human population. Any humans approaching the poles -- or approached by polar ice during ice ages -- must have been living sparsely to say the least. Adaptation to sparseness should render autism more prevalent among people with ancestral homelands near the poles. There has long been a latitude-correlated racial spectrum on Earth. However unpleasant it is to contemplate human differences, some correlation of the autism and racial spectrums seems inevitable.

It is possible but not necessary that the incidence of autism be high among present-day Arctic peoples. The expected incidence of autism depends not just on present-day homelands, but on ancestral homelands, which is to say on entire evolutionary histories. These histories must be considered in assessing expected autism rates. Where tenure in sparse regions has been relatively brief, the expected impact on autistic traits is not as great. DNA analysis is beginning to make studies of human population histories easier.


In two exceptional cases an autism spectrum may not form: They are the case of a spherical population (in any dimension) and the case of a perfectly stirred population. In the first case there is little or no population density gradient, and in the second the average exposure of each individual to different population densities, and hence the average autistic adaptation, does not vary. It is unlikely that either case has persisted long in a pure form in human history. There has, of course, been sufficient mixing of human populations that the species has remained a single species. On the other hand, the fact of geographic variation in skin color and other features shows that stirring has been far from perfect. Even at a tribal scale, where stirring would seem inevitable, the symbiotic arrangement between the genders, with males devoting energy to risk-facing motion and females to reproduction, has prevented perfect population mixing. The ultimate effect of mixing is to force the evolution of brains capable of both autistic and gregarious behavior, and it is fairly obvious that to some degree this has happened. Brain evolution doubtless traces a path intermediate between zero and full effect of the two exceptions. Study of the exceptions should help gauge their relative influence, but in any case this influence has been only partial.

Autism Thresholds

Having shown that autism exists as a seamless spectrum and so cannot be clearly diagnosed I will now seem to contradict myself by highlighting two thresholds along the spectrum. The first of these I term the autism horizon and is essentially absolute. A person has crossed the autism horizon when he becomes so socially disconnected that his genetic and cultural information perishes with him. Extreme loners at population frontiers cross this horizon. Populations constantly expand to their autism horizon: Beyond this they cannot move far because the genetic and cultural information needed to do so is lost. From a genetic standpoint the minimum sustainable behavior near the autism horizon is that an autistic person mates successfully at least once: Otherwise the autistic genetic information is lost. From a cultural standpoint less contact is required -- a discarded tool or hastily scrawled note can transmit information enabling future pioneers to survive in sparse regions. The autism horizon is diagnostically almost useless because one must die to cross it with certainty.

The second threshold along the autism spectrum is relative and may be termed the threshold of self-evident pervasive autism. After years of earnest but failed attempts to form comfortable social relationships it can become self-evident to the affected person that his radical social disconnection is not the result of any moral failing or lack of effort. To such a person autism is a fact of life as stark as a wall. The threshold of self-evident pervasive autism is entirely subjective and depends on myriad variable conditions, including in particular local population density. Being private and subjective, this autism threshold is also diagnostically useless. Autism can be overwhelmingly self-evident: Sometimes things are obvious, as when a trout finds itself in the milk.

Autistic Compensation

Autistic people often compensate for their autism. This is true both for individuals who mask their autism or simulate conventional behavior, and for entire societies which adapt culturally to offset social disconnectedness. The need to compensate for autism arises when an autistic person comes into contact with other people -- in other words when there is a local increase in the density of a population containing relatively autistic people. There are as many kinds of compensation as there are varieties of autism. For example, a person whose autism stems largely from nervous sensitivity so extreme as to preclude most human contact may compensate by meeting others in quiet places with easy exits. Where the autism stems largely from a congenital bias of instinct toward object-manipulating ability an autistic person may compensate by methodically regarding social interactions as a form of intellectual puzzle involving warm objects.

A compensation strategy possible in most cases is the approach of practice-effort: It is a rare deficit that cannot be at least partly relieved through practice. Practice at compensating for autism can be emotionally demanding because even heroic effort may yield social abilities far below those common in people who make essentially no effort. The problem is compounded by the general incomprehension of the socially able who, never having experienced radical social difficulty themselves assume it must be like some episode of awkwardness within their experience and thus attribute the autistic person’s difficulties to laziness or moral turpitude. People who through cerebrovascular accidents lose the capacity to speak or walk often sense the incredible complexity underlying commonly effortless acts. In spite of comparable obstacles many autistic people compensate quite successfully for their autism when circumstances require it.

Autistic Societies

Societies forming near the sparse frontier of the human species are likely to contain a disproportionate number of relatively autistic people. Such societies become possible when a shift in climate or technology allows a population to grow in a previously sparse region, or when outliers migrate back into fertile zones. These societies must compensate culturally for their elevated tendency to social disconnection if they are to thrive. The topological view of autism thus anticipates systematic geographic variation in the cultural concerns of societies. Because the autistic tendency is so fundamental societies are unlikely to compensate simply by denying or reviling it. Rather, a successful autistic society is likely to acknowledge and even to exalt autistic traits, but also to encourage weaving them into a social fabric.

An autistic person is the archetype of the individual. Thus in relatively autistic societies one may expect an emphasis on individual rights and liberties and on the role of the individual in general. The reader may therefore test the present theory in part by considering where historically the greatest cultural emphasis has been placed on the role and rights of the individual and what have been the dominant latitudes of ancestral homeland where the individual has been culturally most promoted.

Autism and Spiritual Practice

Religions are a major form of cultural expression and as such may be expected to reflect the historical geography of autism. The Topological Theory of Autism expects religions to vary systematically according to the sparseness of their authors’ ancestral homelands.

Before the domestication of animals and the advent of farming the degree to which a society depended on hunting probably formed a good measure of sparseness because in sparse regions humans could not gather enough vegetables to survive. After the rise of shepherding and farming a good measure of sparseness was provided simply by the degree to which a society depended on each to survive: Shepherding was possible in sparse regions where farming was not. The Topological Theory of Autism anticipates a rough continuum of religions extending from those correlated with hunting or shepherding and reflecting autistic traits like value for individual autonomy, to religions correlated with vegetable gathering and farming and reflecting non-autistic traits like serene immersion in a sea of humanity. The emphasis on will, choice and personal responsibility seen in the religions born of Middle-Eastern deserts and European snows contrasts markedly with the emphasis on egolessness and social conformity found in the religions of India’s and China’s fertile zones. The Topological Theory of Autism encourages open inquiry into correlations of religious doctrines and the sparseness of their authors’ ancestral homelands.

Autism and spiritual practice are linked by more than religious dogma. Perhaps the most important connection between autism and spiritual practice is that the autism spectrum prohibits a one-size-fits-all approach to spiritual practice. Even where spiritual practice is a compensation for autism it has limits. A strongly autistic person seeking spiritual practice may not fit naturally in a monastery or church community, or even into the regimen of a lone teacher. This does not mean that the autistic person is an egotistical spiritual failure. It means that the Universe has produced a strongly autistic person. For such a person unconventional practice may be an obvious choice.

Population and Technology

Population and technology have exploded on Earth, changing everything. Where each human stood on Earth several millennia ago there now stand hundreds.12 A few thousand years ago a person commonly spent all of his life within a few miles of his birthplace. Now technology allows people to travel thousands of miles easily, and to interact electronically with others on the far side of the globe. These are radical changes, and because they bear directly on population density and social interaction they alter the face of autism. Where autism is concerned we live in a time radically unlike any that has preceded it on Earth.

As the human population expands over a spherical globe and technology enables free movement and intermixing, the topological forces creating an autism spectrum diminish. At the same time, however, this makes autism at least transiently more obvious. For hundreds of thousands of years human populations were surrounded by wilderness. Most of that autism-generating wilderness has disappeared in only the last few millennia. The autism present in the human population today carries with it an adaptive inertia on the scale of at least hundreds of thousands of years, and of millions when the entire history of hominid evolution is considered. The result is that autism doesn’t vanish instantly but is, rather, made more obvious by expanding population. An autistic person is much more visible today than ten thousand years ago because he has been forced by burgeoning population into contact with others, and with fewer exits. This also has the effect of highlighting secondary autistic traits like sensory overloading, verbal anomalies and lack of social instinct, because the primary trait of large interpersonal distance is buried in the recently arisen masses of people. For these reasons autism is likely to become increasingly obvious for some time.

Autistic Symbiosis

The population and technology watershed also suggests an alternate and not strictly topological theory of autism’s origins -- what might be called a symbiosis- or informational theory of autism. Some autistic people display unusual abilities like remarkable rote memory or capacity to manipulate objects (even text) in unconventional ways, and at the same time some autistic people depend on others for sustenance, including the genetic sustenance implicit in reproduction. Because of this there is the possibility of a symbiotic relationship between more and less autistic people. This symbiosis entails the effective exchange of innovative or other informational capacity for sustenance.

It is very likely that any exchange of sustenance and information has correlated strongly with the topological population density gradient. This is because sparse regions are also a population’s frontiers where capacity to confront novel objects is adaptive. Thus the loners are also the pioneers among objects and non-human species. What distinguishes the informational from the purely topological theory of autism’s origins is that a symbiotic relationship, hence an autism spectrum, can still be sustained even in the absence of a topological gradient -- even, for example, in a highly stirred population. As discussed earlier, a similar symbiosis has already existed between the human genders. There is some suggestion of autistic symbiosis in the financial and even romantic success of socially inept scientists and engineers. It is possible that the topological theory describes autism’s past, and the informational its future.

Extraterrestrial Autism

Topology does not end on Earth. As the human species expands to the stars its population will continue to have a surface, but at an ever-widening scale. By the arguments of this theory the vanguard of this expansion will be in an important sense autistic. The explosion of technology, and in particular a growing understanding of how brains are built, make human manipulation of autistic traits likely, both as an aid to space travel and on Earth. The likelihood of human self-engineering -- which is actually nothing new given the fact of selective mating -- makes the future of autism difficult to foretell. Because the fact of adaptation to population surfaces is so fundamental, when humans definitively encounter extraterrestrial civilizations they will be found to share with us far more certainly than eyes or hands, a history of autism.


This concludes the Topological Theory of Autism, or rather of its hasty portrait. The theory unites disparate early uses of the term autism under the rubric of extreme social disconnectedness. The main inferences of the theory are that autism, with both a spectrum and variety, is natural, expected universally, and without a clear diagnosis. It is generally not a disease but a consequence of adaptation to the fact that populations have surfaces. Far from being a minor pathology, autism is a major feature of brain evolution. The theory yields the testable predictions that autism correlates both with gender and with latitude of ancestral homeland. The theory is consistent with Darwin’s theory of evolution but does not rely upon it, and unlike Darwin’s theory, suggests a direction of evolution. Because the theory’s arguments apply to both genetic and cultural phenomena it provides an alternative to debate about the primacy of nature or nurture in human affairs. Finally, the theory sheds some light on the varieties of human spiritual practice, both on the historical differences of religions and on the interplay of personal temperament and spiritual practice today.



Like Darwin's theory of evolution the Topological Theory of Autism is easy to misapply. Every theory needs ability and experience to apply, and the Topological Theory of Autism is no exception. I suggest caution and rigor in attempts to use it. Any attempt to use the Topological Theory of Autism to demonstrate the global superiority of a group is a patent misuse of the theory. In the long run well-meditated written discussion of the theory's merits and weaknesses is likely to be more productive than are facile conjectures and verbal debate. Where I have inadvertantly reinvented wheels I like to think it suggests that the wheels exist. When considering objections to the Topological Theory of Autism it may help to note that the theory is at root statistical and that, whereas an exception can falsify a non-statistical theory, in a statistical theory an exception can prove the rule. Whatever the fate of this one rendering I expect that the adaptive effects of population surfaces will endure.


  1. Bleuler, E., Dementia Praecox or the Group of Schizophrenias. German edition published in 1911. H. Zinken (Trans.) New York: International University Press, (1950).
  2. Kanner, L., Autistic disturbances of affective contact. Nervous Child 2, 217-250. (1943).
  3. Asperger, H., Die "aunstisehen Psychopathen" im Kindersalter. Archiv fur psychiatrie und Nervenkrankheiten 117, 76-136. (1944).
  4. Darwin, C., On the Origin of Species by Means of Natural Selection London: John Murray. (1859).
  5. Berta, A., What is a whale? Science 263, 180-181. (1994).
  6. Ursing, B.M., Arnason, U., Analyses of mitochondrial genomes strongly support a hippopotamus-whale clade. Proceedings of the Royal Society of London Series B 265, 2251-2255. (1998).
  7. Schuler, A.L., Watanabe, A., Beyond Rainman: Thinking about people versus thinking about objects. College of Education Review San Francisco State University (1995).
  8. Chatwin, B., The Songlines New York: Viking (1987). Chatwin links a taste for rocking motion with nomadic behavior.
  9. Dawkins, R., The Blind Watchmaker New York: Norton (1987).
  10. Frith, U., Autism. Scientific American 268 June, 108-114. (1993).
  11. Wilson, A.C., Cann, R.L., The Recent African Genesis of Humans. Scientific American 266 April, 68-73. (1992). Topology would constrain a more ancient diaspora and multiregional evolution as well.
  12. McEvedy, C., Jones, R. Atlas of World Population History New York: Facts On File pg.342-351 (1978).