Autism, masking and the limits of observable behaviour
Earlier this year, TES published an interview with Dame Uta Frith under the headline Why I no longer think autism is a spectrum. Frith has been one of the most influential figures in autism research for decades, and her change of position is significant. She argues that the autism spectrum has become too broad, bringing together people whose difficulties may have very different origins, and that the autism label should be reserved more narrowly for those whose differences are apparent early in childhood.
There are parts of Frith’s argument that deserve serious consideration. Autism is extraordinarily heterogeneous. People currently sharing the same diagnosis can have very different intellectual, language, sensory and support profiles, and there are legitimate questions about how useful one diagnostic category can be when it encompasses such a wide range of experience. Frith also acknowledges how little we still know about the underlying biology of autism. There is no single biomarker, and she suggests that future research may reveal many different kinds of autism rather than one underlying condition.
However, some of the conclusions she draws raise more difficult questions, particularly for those of us working in education. In the TES interview, Frith suggests that smooth conversational interaction may be a “contraindicator” for autism. Being able to understand irony, humour or what is implied in conversation may count against a diagnosis too. She is also highly sceptical of masking, particularly as an explanation for why autistic girls and women may remain unidentified until adolescence or adulthood.
The difficulty is that these arguments rely heavily on what another person can observe. A child communicates fluently. A young person understands a joke. A pupil gets through the school day. From the outside, each of those things can look like evidence of capacity. What they cannot tell us is how much effort that performance required, what else the pupil had to sacrifice in order to achieve it, or whether it can be sustained.
That distinction has implications far beyond the current debate about autism diagnosis.
What can we actually see?
Teachers necessarily work with observable behaviour. We see whether a pupil enters the classroom, answers a question, completes the task, talks to classmates or copes with an unexpected change. We make judgments from those observations because, much of the time, they are the information available to us.
The difficulty comes when successful performance is treated as evidence that the task or environment is unproblematic. A pupil who sits through a noisy lesson has demonstrated that they were able to remain in that lesson. They have not demonstrated that the noise was comfortable, easily processed or harmless. A pupil who holds an articulate conversation has demonstrated the ability to participate in that particular interaction. We still know very little about how consciously they were managing eye contact, facial expression, turn-taking, tone of voice or the effort involved in deciding what to say.
The same problem appears in academic assessment. A learner may successfully complete a piece of work while simultaneously managing dense instructions, working-memory demands, sensory distraction, handwriting difficulties, anxiety and uncertainty about what the question requires. Looking only at the finished answer can conceal a considerable amount of additional cognitive work.
This is one reason apparently capable autistic pupils can be so easily misunderstood. Their success becomes evidence against their difficulty. The better they become at meeting the expectations around them, the harder it can become for adults to recognise the effort involved.
The problem with “everyone masks”
Frith is particularly critical of masking. In her TES interview, she describes autistic people reporting that they imitate neurotypical behaviour and feel exhausted afterwards, but questions whether exhaustion can be attributed to masking when tiredness can arise for many other reasons. She also suggests that, in some sense, everybody masks because all of us adjust our behaviour to social expectations.
There is some truth in that observation. Most people alter their behaviour according to context. We speak differently in a job interview from the way we speak with family, suppress irritation in situations where expressing it would be inappropriate, and learn social conventions that do not necessarily come naturally.
The existence of a behaviour across the population, however, tells us little about whether its frequency, intensity or cost is the same for everyone. Everyone experiences noise, but that does not mean everyone experiences sound in the same way. Everyone becomes tired, but that does not mean all forms of fatigue are equivalent. Most people modify their social behaviour, but that does not establish that monitoring facial expression, gesture, eye contact, conversational timing and appropriate responses carries the same cognitive load for everyone.
There is also now a substantial body of research examining autistic masking and social camouflaging. Studies by Laura Hull and colleagues, Amy Pearson and Kieran Rose, and others have explored both the experience of masking and its relationship with stigma, identity and exhaustion. The concepts and measurements used in this research can certainly be questioned and refined, as they can in any developing field. Describing masking as an idea with no scientific basis seems harder to reconcile with the literature that now exists.
More importantly for schools, we do not actually have to settle every disagreement about the terminology before responding to the underlying phenomenon. If a pupil tells us that maintaining apparently typical behaviour throughout the school day requires intense conscious effort and leaves them exhausted afterwards, that information deserves investigation. Whether we eventually call the process masking, camouflaging, compensation, adaptation or something else does not remove the experience we are trying to understand.
Capacity is not the same as sustainability
A recent paper by Lori Hogenkamp, Dhwani Sanghavi and Heini Natri offers another way of thinking about this. Their Evolutionary Stress Framework brings together biological, psychological, social and neurodiversity perspectives and asks how neurodevelopmental differences interact with stress and environment over time.
The framework was developed principally in relation to autistic adulthood, so applying it directly to schools requires some caution. Its central idea is nevertheless useful for thinking about children and young people. Hogenkamp and colleagues describe autistic traits and responses in terms of stress–energy regulation, and introduce the idea of “stress incoherence” for situations in which ways of adapting that may previously have worked become unsustainable under current conditions.
That word – sustainable – is useful.
A strategy can work today and still be too expensive to keep using tomorrow. A pupil can get through a lesson, assembly or school day by investing enormous regulatory effort. If we judge the success of the arrangement solely by whether the pupil remained in the room, completed the activity or appeared calm, we may miss the accumulating cost.
That cost may become visible somewhere else. It may appear as exhaustion after school, increasing anxiety on Sunday evenings, difficulty getting out of bed in the morning, headaches or physical complaints, reduced tolerance at home, withdrawal from activities that were previously manageable, or eventually an inability to attend school at all. None of those outcomes can automatically be attributed to autism or masking, and each needs to be considered individually. They do, however, remind us that a snapshot of successful performance gives us only part of the picture.
Frith herself recognises part of this problem. In the TES interview, she describes a group of later-diagnosed pupils who may be highly anxious and acknowledges that some may become unable to attend school. She argues that schools need to build trust with these young people. Where I think the more difficult question arises is in explaining what happens before the point at which attendance breaks down.
If a pupil was apparently coping for several years and then suddenly cannot come into school, it is tempting to assume that something new has gone wrong. Sometimes it has. In other cases, what looks sudden from the outside may represent the point at which a long-standing way of coping has finally become unsustainable.
Sensory adjustments and whose evidence counts
A similar issue arises in Frith’s discussion of sensory adjustments. She questions the scientific grounding for measures such as ear defenders or reducing classroom lighting and says that something feeling good to a child does not necessarily mean that it is good for them.
It is reasonable to ask whether interventions are effective. Reasonable adjustments should not become immune from scrutiny simply because they sound intuitively helpful, and teachers need evidence about the approaches they use. There is also a difference between a temporary strategy that helps a pupil regulate and an intervention that genuinely improves long-term participation.
At the same time, subjective experience is itself a form of information. If a child consistently reports that a particular sound causes distress and that ear defenders make it possible to remain in a lesson, their report cannot establish every possible consequence of using ear defenders. It can tell us something important about what the environment feels like to that child.
This is where medical and social understandings of disability do not have to be treated as competing explanations. Biology may help explain why a child processes sensory information differently. Environment determines how frequently that child encounters overwhelming sensory demands and whether they have any control over them. The child’s own experience tells us something neither an observer nor a brain scan can fully capture.
Hogenkamp and colleagues make a similar argument at a broader level. Their framework shifts attention towards environmental design, co-regulation and support that responds to individual regulatory needs. They argue that wellbeing needs to be defined with autistic people rather than entirely for them. In schools, that suggests a process in which observation, professional knowledge and the pupil’s own account are considered together.
Looking capable can create its own disadvantage
There is a particular risk here for pupils who communicate well, achieve academically or appear socially competent. These are precisely the characteristics that can reassure adults that support is unnecessary.
A pupil who says the right things may be assumed to understand the social situation easily. A pupil achieving expected grades may be assumed to be accessing the curriculum without difficulty. A child who waits until reaching home before becoming distressed may appear to school staff to have had a successful day.
None of these conclusions is necessarily wrong. Some pupils really are coping well, and it would be equally unhelpful to interpret every sign of competence as evidence of concealed distress. The point is that competence alone cannot answer the question.
The same caution should apply in the other direction. A diagnostic label cannot tell us how much difficulty an individual pupil experiences in a particular classroom, and two autistic children may require entirely different forms of support. Frith is right to resist the idea that a single label provides a complete description of a person.
Where I differ is in what follows from that uncertainty. Heterogeneity makes careful individual assessment more important. It does not make the pupil’s account of their own experience less relevant.
What happens afterwards?
Holden Thorp, editor-in-chief of Science, returned to some of these questions after meeting Frith at her home this summer. Thorp is autistic and describes having a profile similar in some respects to Chris Packham’s. Frith told him that she would probably have diagnosed him with Asperger syndrome in the 1990s, but that she is no longer sure she would regard him as autistic today.
Their conversation reveals an important point of agreement. Both accept that autistic traits exist across a continuum and that science cannot currently identify a clean biological line showing where autism ends and personality begins. They disagree about what should happen in the absence of that line.
That diagnostic debate will continue. For teachers, however, there is a more immediate question that does not depend on resolving it.
When a pupil appears to be coping, we can ask what happens afterwards.
Do they have enough energy left to learn in the next lesson? Can they manage the same demand repeatedly across a week? Are they increasingly avoiding particular environments? Is home carrying the consequences of the effort required at school? Does a reasonable adjustment reduce the amount of effort required without preventing participation or learning?
These questions do not require us to presume hidden distress where none exists. They simply prevent successful performance from becoming the only evidence we consider.
Schools understandably want children to become more independent, more resilient and increasingly able to navigate difficult situations. Genuine resilience, however, cannot mean repeatedly absorbing demands that exceed a person’s regulatory capacity until they can no longer continue. Sometimes support enables independence precisely because it reduces the unnecessary effort being spent on coping with the environment.
Perhaps, then, one of the most useful questions we can add to our observations of autistic and neurodivergent pupils is a very simple one. When a child demonstrates that they can do something, we also need to understand what it costs them to do it, what happens afterwards, and whether that cost can be sustained.
Related posts and writing
- No Mask for Us: Why Recognition Matters for Autistic Girls
- Inclusive Education for Autistic Students: Evidence, Gaps, and Urgent Questions
- Rethinking Behaviour and Belonging
- New Publication: 6 Ways to Support Autistic Girls in Your Classroom
Written by Kate Coldrick, an educator and writer based in Woodbury near Exeter.

You must be logged in to post a comment.