What is the definition of autism (ASD)?

Autism — formally autism spectrum disorder, or ASD — is a lifelong neurodevelopmental diagnosis describing differences in social communication and interaction, alongside focused interests, a need for sameness, and differences in sensory processing. It is one of the most discussed diagnoses of the age and one of the most misunderstood, and the word itself has changed meaning more than once since it was coined.

The Directory reads autism with The Neurodiversity Book: a real and often profound difference in how a nervous system senses, attends and relates — attention cast narrow and deep (monotropism), a low optimal rate of change, overload met by withdrawal rather than flooding outward — sitting toward one end of axes every human being runs. The difference is real. What is abstract is the norm it is measured against; the difficulties autistic people face are, to a large degree, the cost of running from a different baseline in conditions set for an average nobody is. See also autism’s narrow interests and autistic pattern recognition.

Frequently asked questions about the term autism

What’s the difference between autism and Asperger’s syndrome?

Asperger’s syndrome was a separate diagnosis, introduced into English-language psychiatry by Lorna Wing in 1981 and added to the manuals in the 1990s, for autistic people without early language delay or intellectual disability. In 2013 the American manual (DSM-5) folded it, with other subtypes, into the single diagnosis of autism spectrum disorder, and the international classification (ICD-11) followed. Some people diagnosed with Asperger’s still use the term for themselves, and that is their choice. Others have dropped it, partly because the research of Herwig Czech and others has documented Hans Asperger’s part in the child-euthanasia system of Nazi-era Vienna. The underlying reason for the merger was that the subtypes could not be reliably told apart: they described positions on one spectrum, not separate conditions. For anyone diagnosed with Asperger’s before 2013, nothing about the diagnosis has been taken away: services in the UK treat it as an autism diagnosis. The change is one of vocabulary and of understanding, from separate boxes to one spectrum of positions, which also makes it easier to see that the old line between “Asperger’s” and “autism” was often drawn by speech in early childhood rather than by anything that mattered later in life.

Can you be “a little bit autistic”?

In one sense yes, and in another no, and it helps to hold both. The traits autism describes — narrow, deep attention, a preference for sameness, sensitivity to sensory input — sit on dimensions every human being has, which is why many people recognise some of themselves in descriptions of autism. A diagnosis marks a position far enough out, on enough of those dimensions, to shape a life substantially. So “we’re all a little autistic” is true about the machinery and misleading about the experience: it can erase the weight of being autistic in a world that does not fit, and it can erase the smaller group whose autism is profound. The spectrum describes positions, not a single line from mild to severe. Clinicians sometimes use the phrase “broader autism phenotype” for relatives of autistic people who show some traits without meeting diagnostic criteria, which is the scientific version of the same observation: the traits run through families and populations, and the diagnosis marks where they become large enough to shape a life.

Should I say “autistic person” or “person with autism”?

Surveys of autistic adults in the UK and elsewhere consistently find most prefer identity-first language — “autistic person” — because autism is experienced as part of who they are rather than something they carry. Many clinicians, researchers and parents were trained to use person-first language — “person with autism” — to put the person before the condition. The respectful default is identity-first in general writing, and whatever a particular person prefers when writing about them. The Directory uses “autistic person”. Community spaces, many of them listed under neurodiversity communities, are a good place to hear the preference argued directly. The debate matters because language carries a model. Person-first language was introduced to resist people being reduced to a condition; identity-first language resists the idea that autism is something separable from the person, like an illness. Both come from respect. When in doubt, ask, and follow the answer.

What causes autism, and can it be prevented?

Autism is strongly heritable: twin studies estimate that genetic factors account for a large majority of the variation, and hundreds of genes are involved, each contributing a small effect. Some environmental influences in pregnancy, such as prematurity and certain exposures, modestly raise likelihood. Vaccines do not cause autism, and neither does paracetamol taken as directed in pregnancy; both claims have been tested at scale and not supported — see the Directory’s piece on the paracetamol claims. The Neurodiversity Book adds a harder question that research is only starting to address: whether some presentations are shaped by early environments, such as heavy screen exposure in infancy, which clinicians have begun, awkwardly, to call “virtual autism”. The honest answer to prevention is that autism is not a disease to prevent; the useful question is what helps an autistic child develop well. Research on causes is also shaped by what it is used for. Some genetic research is welcomed by autistic people for what it can explain about co-occurring conditions such as epilepsy; some is opposed because of fears that it could be used to screen autistic people out. Those concerns are part of why autistic people increasingly insist on a voice in the research agenda.

How do I know if my child is autistic?

Signs vary with age and with the child. In young children, parents often notice differences in shared attention (pointing, showing, following a gaze), delayed or unusual speech, strong reactions to sounds, textures or changes, intense interests, and repetitive play or movement such as stimming. In older children, especially girls, the signs can be quieter: exhaustion after school, difficulty with friendships despite wanting them, meltdowns at home after holding it together all day. If you are concerned, speak to your GP or health visitor and ask about referral; the Directory’s guide to getting an autism diagnosis in the UK explains the routes. Research on early support also suggests that the years before five are a period of real opportunity. Waiting times for assessment are long in much of the UK, and support should not wait for a diagnosis. Schools can put adjustments in place on the basis of need, and understanding your child’s sensory and communication preferences at home helps whatever the eventual outcome. Many parents recognise themselves in the process, and adult assessment is increasingly common among parents of autistic children.

What is autistic burnout, and how is it different from depression?

Autistic burnout is a state of profound exhaustion, loss of skills and reduced tolerance for sensory and social input, following long periods of strain — often masking, overload and life changes. Research led by Dora Raymaker and autistic colleagues (2020) described it as the result of chronic life stress and a mismatch between expectations and abilities without adequate support. It overlaps with depression and can coexist with it, but the two differ: burnout centres on exhaustion and withdrawal of capacity, and people often still want to do things they cannot, whereas depression centres on low mood and loss of interest. Recovery usually needs reduced demands, sensory rest and time, rather than activation. Misreading burnout as depression, or as laziness, is common and costly. The distinction matters for treatment. Approaches built for depression, such as pushing to increase activity, can deepen burnout, while approaches that reduce load, remove masking pressure and restore sensory rest usually help. Many autistic adults describe repeated cycles of burnout before diagnosis, and part of what diagnosis offers is the chance to recognise the pattern early and change the conditions that produce it.

Can autistic people have relationships, live independently, and work?

Yes — autistic people do all of these, across the full range of lives. Many form lasting partnerships and friendships, raise families, and work in every field. At the same time, employment and independence outcomes for autistic people as a group remain poor, and many need significant support. The difference between those outcomes is usually not ability but fit: predictable environments, clear communication, manageable sensory conditions and understanding employers change what is possible. And a smaller group of autistic people — those whose autism is profound — need care throughout their lives, which is not a failure of anyone. In the UK, only a minority of autistic adults are in paid work, one of the largest employment gaps of any group, and the reasons are mostly about recruitment and workplaces rather than ability: interviews that test social performance rather than skill, open-plan offices, and vague expectations. Adjustments such as written questions in advance, quiet workspaces and clear instructions are often inexpensive and highly effective. Relationships, too, tend to go best where differences are named rather than hidden: many autistic people describe partnerships that work well once both people understand what the other needs, whether that is explicit communication, time alone to recover, or a different shape of shared life.

What’s the best treatment or therapy for autism?

There is no single best approach, because autism is not an illness with a cure. Good support depends on the person and targets what they want help with: speech and language therapy, occupational therapy for sensory needs, support for anxiety (often adapted for autistic people), and help with daily living, education and work. Applied Behaviour Analysis (ABA) is the most contested approach: some modern forms focus on building skills, but many autistic adults describe traditional ABA as training compliance and suppressing autistic behaviour at great personal cost, and the Directory does not list ABA providers. The best support treats the person as the authority on their own experience. The counselling and therapy category lists neurodivergent-affirming practitioners. For co-occurring conditions — anxiety, depression, sleep problems, epilepsy, gastrointestinal difficulties — ordinary treatments apply, often adapted. The key question for any approach is whether it helps the person live the life they want, or mainly makes them easier for others to manage. Autistic-led peer support, from people who have lived the same things, is one of the most valued forms of help and one of the least funded. Families often find that the most useful early step is simply understanding their child’s sensory and communication profile, which changes everyday life more than any single programme.

Autism (ASD), defined

Autism spectrum disorder is diagnosed on two groups of characteristics: persistent differences in social communication and interaction, and restricted, repetitive patterns of behaviour, interests or activities, including sensory differences. They must be present from early development and cause significant difficulty in daily life. Around one to two in every hundred people are autistic by most current estimates, though rates have risen sharply over two decades as criteria widened and recognition improved; see the Directory’s neurodiversity statistics. Many autistic people are also ADHD, dyspraxic or dyslexic, or have anxiety, epilepsy or gastrointestinal conditions.

The word has an unusual history. The Swiss psychiatrist Eugen Bleuler coined it in 1911 from the Greek autos, the self, to describe something he saw in patients with schizophrenia: a turning inward, the private world preferred to the shared one. It is the only diagnostic word with the self built into it, and it began by filing inwardness itself as a symptom. In 1943 Leo Kanner borrowed it for a group of children no existing category could hold, and in 1944 Hans Asperger described a similar group in Vienna. The word left schizophrenia and moved into childhood, where it has lived ever since, carrying its original sense with it: attention pointed somewhere the observer could not see.

In 2013 the manual dissolved the subtypes and the word became a spectrum. The grammar of that change did a great deal of work. A category has an outside; a spectrum has only positions. Once autism named a spectrum, recognition widened enormously — to women and girls, to adults, to people who had spent their lives being told they were odd or difficult — and diagnosis rose accordingly. The same grammar also produced the phrase “we’re all a bit autistic”, and it left the most profoundly affected autistic people harder to see inside a much larger population. In 2021 a Lancet commission proposed the term “profound autism” for that group, whose needs have not changed while the word around them grew.

The Neurodiversity Book reads the underlying mechanisms in the same way it reads ADHD. Lay the autism criteria beside the ADHD criteria and look past the labels: “restricted, intense focus” and “easily distracted” are the same axis — how wide attention is cast — read at opposite ends; “insistence on sameness” and “craving novelty” are one axis of optimal rate of change; withdrawal under overload and outward flooding are two directions of one response. Autism names sitting far toward one end of several of these axes, far enough to cross the threshold the manuals draw. It is not a different kind of brain, but it is a genuinely different tuning, and a life is shaped by it.

That reading changes where the difficulty is located without denying it. Much of what disables autistic people is a mismatch: sensory environments too loud and bright, communication conventions built around implicit cues, pace and change imposed from outside, social expectations enforced by penalty. Masking — performing typicality to avoid those penalties — is one of the heaviest costs autistic people carry. Change the conditions and the same person often functions very differently. For others, especially those with profound autism, the difficulties remain substantial in any arrangement, and good support is lifelong. Both are true, and a definition that holds only one of them fails someone.

Autism is also increasingly recognised later in life, and the Directory’s own writing on late autism diagnosis argues that it is more than delayed recognition: it often arrives when a person’s compensations finally fail under a load they can no longer carry. Research also suggests that when autism becomes visible is itself partly written in genetics.

Autism is also a word in motion. In speech, “autistic” and “ADHD” are blending: people diagnosed with one increasingly identify with the other, and the manuals, which kept the two strictly apart for decades, now permit both. At the same time, the word is being carved at its edges — “profound autism” for those whose needs are highest, “virtual autism” for presentations shaped by early environments — while its middle widens to include people who arrive at it in adulthood, often with relief, sometimes as an identity as much as a diagnosis. The Directory’s view is that all of these are real, that they are not the same, and that support has to follow the actual person rather than the word.

Two further points complete the definition. First, autism is a developmental description, not a fixed portrait: the same autistic person at four, fourteen and forty may look very different, because they develop, compensate, find their people and choose their conditions. Second, autism is not only a set of difficulties. Many autistic people describe it as the source of what they value most in themselves — the depth of their interests, their honesty, their attention to detail and fairness, their sensory joys — and a definition that lists only deficits describes them as badly as one that lists only gifts.

The key concepts in autism

The spectrum concept and the inadequacy of functioning labels

“Spectrum” was adopted to capture how varied autism is, but it is often misread as a line from “mildly” to “severely” autistic. A better picture is a set of dimensions — attention, sensory sensitivity, communication, need for sameness, motor planning — on which each autistic person has their own profile, often a spiky one. Functioning labels (“high-functioning”, “low-functioning”) flatten that profile into a single judgement, usually based on speech and apparent independence, and they mislead in both directions: people called high-functioning are denied support they need, and people called low-functioning have abilities overlooked. The support-needs levels in the current manual are more useful, but they still describe a moment in a changing life. What helps is describing what a particular person finds hard and easy, in which conditions — and recognising that for a minority, profound autism means high support needs in every condition. Autistic people’s profiles also change over time and with circumstances: stress, illness, burnout and sensory load can reduce what someone can do, sometimes dramatically, which is another reason a fixed label describes a person badly.

Monotropic attention and special interests

Monotropism, developed by the autistic researchers Dinah Murray, Wenn Lawson and Mike Lesser, describes autistic attention as concentrated on a few interests at a time rather than spread across many. It explains much of what the criteria list separately: absorbing special interests, difficulty switching tasks, distress at interruption, overload in busy environments, and the effort social conversation demands. Special interests, on this account, are not obsessions but depth: attention selecting something for itself and going all the way in. They are the same mechanism as ADHD hyperfocus, under a different label, and for many autistic people they are the main source of joy, expertise, calm and connection. See the Directory’s long read, Understanding monotropism. Educators and employers who work with a special interest rather than against it — using it as a route into learning, or matching roles to it — often see transformed engagement. Treating it as an obsession to be limited usually produces the opposite. Losing access to a special interest, through illness, burnout or other people’s disapproval, is often one of the first signs that an autistic person is struggling.

Masking, camouflaging, and the cost of appearing neurotypical

Masking or camouflaging means suppressing autistic behaviour and performing expected social behaviour — rehearsed small talk, forced eye contact, hidden stimming, copied expressions — to avoid being noticed or penalised. Many autistic people, especially women and girls, mask from childhood, often without realising it. It works, up to a point: it can protect jobs, friendships and safety. But it costs a great deal of energy, it delays diagnosis, and research links sustained masking to exhaustion, anxiety, depression and burnout. The Directory’s reading adds one thing. Masking is not autistic people copying a real neurotypical self; it is performing an average that nobody embodies. That is why it is so exhausting, and why reducing the need for it — not better masking — is the aim. Unmasking is not simple either: after years of performance, many autistic adults describe not knowing which parts are them, and rebuilding that knowledge slowly, often with other autistic people, is part of recovery.

Sensory processing differences and environmental accessibility

Most autistic people experience sensory processing differently: some input is overwhelming (fluorescent lighting, overlapping conversations, certain fabrics), some barely registered (hunger, pain, temperature), and some sought out for comfort (pressure, repetitive movement, a particular sound). Sensory difference was added to the diagnostic criteria only in 2013, although autistic people had described it for decades. Its consequences are practical and large: a sensory-hostile classroom or office can disable an autistic person as surely as stairs disable a wheelchair user, and meltdowns and shutdowns are often the endpoint of accumulated overload rather than behaviour to be managed. Quiet space, control over lighting, noise-cancelling headphones, predictable routines and sensory tools are accessibility, not preference — and they help many non-autistic people too. Sensory needs also interact with everything else: a person who is already overloaded has less capacity for social communication, change or demands, which is why the same autistic person can cope well in one setting and poorly in another on the same day. Asking an autistic person what they notice in a room is often the quickest way to find what needs changing.

The double empathy problem and autistic communication

The “double empathy problem”, proposed by the autistic researcher Damian Milton in 2012, challenges the long-standing claim that autistic people lack empathy or social understanding. It argues that communication breaks down between people with different ways of experiencing the world, in both directions, and that non-autistic people misread autistic people as much as the reverse — but only one side is labelled deficient. Research since has supported it: autistic people often communicate effectively with each other, and non-autistic observers form quick negative judgements of autistic people that do not match those people’s actual abilities (see, for example, this 2023 discussion in Disability & Society). The practical implication is that social skills training aimed only at autistic people addresses half the problem; understanding has to be built from both sides. It also explains why many autistic people find relief in autistic company, and why communities run by autistic people have been so important to wellbeing.

Key figures and publications in autism

Eugen Bleuler (1911)

The Swiss psychiatrist who coined “autism” in Dementia Praecox oder Gruppe der Schizophrenien, naming a withdrawal into the inner world as a symptom — the word’s origin, and the reason it still carries the self (autos) inside it.

Leo Kanner (1943)

The Austrian-American psychiatrist whose paper “Autistic disturbances of affective contact”, in The Nervous Child, described eleven children and moved the word into childhood. His later, discredited suggestion that cold parenting contributed — the “refrigerator mother” — shaped a generation of harm.

Hans Asperger (1944)

The Viennese paediatrician who described a group of children with intense interests and social differences. His work reached English-language psychiatry only in the 1980s. Herwig Czech’s 2018 study in Molecular Autism documented his cooperation with the Nazi-era child-euthanasia programme, which is one reason the eponym has fallen out of use.

Lorna Wing and Judith Gould (1979)

The British psychiatrists whose Camberwell study described the “triad of impairments” and first framed autism as a spectrum. Wing, herself the mother of an autistic daughter, introduced Asperger’s work to English readers in 1981 and co-founded what became the National Autistic Society.

Simon Baron-Cohen

The Cambridge researcher whose 1985 study with Alan Leslie and Uta Frith, “Does the autistic child have a “theory of mind”?”, shaped decades of research. His later “extreme male brain” theory has been widely criticised, not least for its part in the long under-recognition of autistic women and girls.

Dinah Murray, Wenn Lawson and Mike Lesser (2005)

The autistic researchers behind monotropism, the theory of autistic attention that many autistic people recognise most readily in themselves.

Damian Milton (2012)

The autistic sociologist whose “double empathy problem” reframed autistic social difference as a two-way mismatch rather than a one-sided deficit.

Dora Raymaker and colleagues (2020)

The community-based research on autistic burnout, published in Autism in Adulthood, that gave clinical language to an experience autistic people had long described among themselves.

Catherine Lord and colleagues (2021)

The Lancet Commission on the future of care and clinical research in autism, which proposed “profound autism” for autistic people with high support needs throughout life, and argued for a stepped, individualised approach to care.

Ronnie Cane, The Neurodiversity Book (second edition)

The reading this entry takes: autism as a real tuning toward one end of universal cognitive axes; a word that began by pathologising inwardness; a spectrum whose grammar widened recognition and blurred the profound; and difficulty understood largely as mismatch between a nervous system and its conditions.

The Neurodiversity Directory’s guides

The guide to getting an autism diagnosis in the UK, the guide to Right to Choose, Access to Work and neurodivergent coaching cover assessment, workplace support and life after diagnosis.

Temple Grandin

The autistic animal scientist whose books, from Thinking in Pictures (1995) onward, gave many readers their first account of autism from the inside, and helped move public understanding from tragedy toward difference.

Jim Sinclair (1993)

The autistic activist whose speech “Don’t Mourn for Us” argued that autism is a way of being rather than an illness hiding a “normal” child, and co-founded Autism Network International, one of the first autistic-run organisations.

Common misconceptions about autism

Do autistic people lack empathy or emotion?

No. Many autistic people describe feeling others’ emotions intensely, sometimes overwhelmingly, and caring deeply. What often differs is how emotion is shown and read: facial expression, tone and eye contact may not match what others expect, and some autistic people find it hard to identify their own feelings (alexithymia, which also occurs in the general population). The idea of autistic people lacking empathy came from research that measured empathy through conventional social cues. The double empathy problem shows the misreading runs both ways: non-autistic people are often poor at reading autistic people, but that has rarely been called a lack of empathy. Research also suggests autistic people often attend more to fairness and to the details of what someone has said than to social smoothing, which can be mistaken for coldness when it is closer to its opposite: taking the other person’s words seriously. Many autistic adults also report being deeply affected by injustice and by animals, and describe empathy that is intense but slow to find words.

Is autism caused by vaccines, parenting, or environmental toxins?

No to vaccines and parenting. The 1998 paper that claimed a link between the MMR vaccine and autism was fraudulent and retracted; its author was struck off; and studies of millions of children have found no link. The “refrigerator mother” theory that blamed cold parenting was abandoned decades ago. Environmental factors in pregnancy play a modest role in likelihood alongside strong genetic influence, and research continues. Newer claims — such as paracetamol in pregnancy — have not held up when studied carefully, as the Directory’s piece on those claims explains. The persistence of these myths does real harm: to vaccination rates, and to autistic people treated as damage to be explained away. Rising diagnosis rates are explained mostly by widened criteria, better recognition of women, girls and adults, and changes in services; they are not evidence of a toxic cause.

Can autism be cured, or will autistic children grow out of it?

Autism is lifelong; it is a way a nervous system develops, not an illness that resolves. Autistic children change a great deal as they grow — many develop speech, skills and strategies, and some become much harder to identify from outside — but that is usually development and compensation, not autism leaving. Some autistic adults describe masking so effectively that they were never recognised as children. The more useful aims are wellbeing, communication, independence where possible, and conditions that fit, rather than making an autistic person appear non-autistic, which research links to exhaustion and poor mental health. Promises of a cure — through diets, supplements, chelation or other treatments — have repeatedly been shown to be ineffective and sometimes dangerous, and families should be wary of anyone selling one.

Are all autistic people savants or intellectually disabled?

No. Autistic people have the full range of intellectual ability. A significant minority also have an intellectual disability, and they have long been under-represented in research and public discussion. Savant skills — exceptional ability in a specific area such as calculation, music or drawing — are real but rare, occurring in a small proportion of autistic people. The two stereotypes, genius and incapable, come from films and old case studies, and both erase the ordinary majority: autistic people with uneven, often spiky profiles, strong in some areas and struggling in others, and shaped heavily by the conditions they live in. Intelligence tests themselves can mislead with autistic people, because verbal instructions, time pressure and unfamiliar testers all affect performance; results often understate what a person can do.

Is autism just extreme male brain or only affecting boys?

No. Autism is diagnosed in boys more often than girls — historically by around four to one, now closer to three to one in studies that screen whole populations — but the gap is widely thought to reflect under-recognition as much as real difference. Criteria and assessments were built largely on observations of boys; girls more often mask, have socially acceptable interests, and are misdiagnosed with anxiety or eating disorders first. The “extreme male brain” theory has been criticised on scientific grounds and for reinforcing that under-recognition. Autistic people of all genders exist, and many autistic people are transgender or non-binary, a finding research has repeatedly confirmed. The number of women diagnosed in adulthood has risen sharply, and many describe a lifetime of being told they were too sensitive, too intense or not trying hard enough before anyone considered autism. Better recognition across genders is one of the clearest reasons diagnosis rates have risen, and it is still far from complete, particularly for autistic people of colour and older adults.

Related terms and concepts

Monotropism

Monotropism is the theory of autistic attention as concentrated on a few interests at a time — the narrow end of the attentional axis.

Sensory processing

Sensory processing differences are part of the diagnostic criteria and one of the largest practical sources of difficulty for autistic people.

Masking

Masking is the suppression of autistic traits to appear typical, with high costs to energy and mental health.

Stimming

Stimming is repetitive movement or sound used for regulation, comfort or expression — not a behaviour to be trained out.

Executive function

Executive function differences — switching, planning, starting — are common in autism and often follow from monotropic attention.

ADHD

ADHD sits toward the opposite ends of several of the same axes; being both autistic and ADHD is common. See ADHD and autism as archetypes.

Spiky profile

A spiky profile is the uneven pattern of strengths and difficulties common in autism, and the reason functioning labels mislead.

Neurotypical and neurodivergent

Neurotypical names the average autism is measured against; neurodivergent is the umbrella term autism most often sits under.

Dyslexia

Dyslexia and other specific learning differences often co-occur with autism.

Nervous system

The nervous system is where autism lives: a whole system of sensing, attending and regulating, tuned differently from the average.

AuDHD

AuDHD is being both autistic and ADHD, a common combination the manuals only began permitting in 2013.

Theory of mind

Theory of mind is the ability to infer others’ mental states; once central to autism research, it is now read alongside the double empathy problem.

Alexithymia

Alexithymia is difficulty identifying and describing one’s own emotions; it is common in autistic people, and not the same as a lack of feeling.