Dame Professor Emeritus Uta Frith has recently suggested in an article in the Times Education Supplement that autism should no longer be understood as a spectrum. In her view the category has expanded so far that it has “widened to the point of collapse”.
I think she is wrong. However, I am not going to attempt a line-by-line refutation of her argument. Several other writers have already done that very effectively. I just want to offer some more situated insights.
Speaking from the inside, the expansion makes perfect sense.
I say that both as an autistic person (diagnosed or, as I prefer, “professionally identified” as such by psychologists) and as someone who has spent much of his working life thinking about how different minds organise perception, attention and creativity.
The idea that autism is a spectrum is not an abstract theory to me. It simply and effectively describes the sheer range and variety of autistic people I have known, worked with and read over many years.
For most of the twentieth century autism was defined quite narrowly. It was primarily associated with children who had very visible developmental differences and who were usually diagnosed early in life. Many other people who would now be recognised as autistic simply passed through life under different descriptions: shy, eccentric, obsessive, socially awkward, gifted, difficult. Some struggled quietly, while others built lives that accommodated their differences in various ways.
What has changed in the past two decades is that these experiences have begun to coalesce into a shared language. Adults have started recognising themselves in descriptions of autism. Women who were overlooked in childhood have begun receiving diagnoses later in life. Autistic people themselves have increasingly written about their experiences in ways that were largely absent from earlier research.
To some observers this widening of recognition looks like diagnostic inflation. I believe instead that we have simply become better at recognising autism.
Part of the tension here is historical. Much of modern autism research developed around a set of cognitive theories that attempted to explain autistic behaviour in terms of deficits. The most influential of these was the idea that autistic people struggle with “Theory of Mind”: the ability to infer other people’s thoughts and intentions. This idea, which was partly developed by Uta Frith herself, has been enormously influential and helped shape research agendas for decades. But it has also caused great harm to autistic people.
Once autism is framed primarily as a deficit in understanding other minds, it becomes easy to make assumptions that autistic people lack empathy, lack imagination, or lack insight into their own experience. Those assumptions have travelled far beyond academic psychology to become part of the cultural story about autism.
What has changed in recent years is the arrival of autistic writers, researchers and advocates into a space that was previously populated by non-autistic psychologists. Their presence alone contradicts the idea that autistic people lack a theory of mind.
These autistic writers and researchers have increasingly challenged the notion that social differences can be explained simply as a one-sided deficit. Damian Milton’s “double empathy problem”, for example, suggests that misunderstandings between autistic and non-autistic people arise from differences in perspective on both sides. Communication difficulties may reflect a mismatch between neurotypes rather than a failure on one side alone.
I should add a personal observation here, although it is not a scientific one. In my own experience I often find that I “click” more quickly and more comfortably with autistic people who have significant learning disabilities and little or no spoken language than I do with many neurotypical people. Communication takes a different form, of course, but the sense of mutual understanding can be surprisingly immediate. Experiences like this make it difficult for me to see autism as a collection of unrelated conditions. There is often a recognisable affinity across the spectrum itself. If the spectrum had truly “collapsed”, as Frith suggests, that affinity would be much harder to explain.
The growing diversity of people identifying as autistic is not evidence that the concept has broken down. It reflects the fact that autism was never a single uniform condition to begin with. Even among those diagnosed in early childhood there has always been enormous variation: in language, cognition, sensory experience, interests and ways of engaging with other people. The word “spectrum” was introduced precisely to capture that diversity.
To be clear, I am not especially attached to the term “spectrum” itself. It is often misunderstood as a simple linear scale running from “mild” to “severe”, which is not a very good description of autistic variation at all. A more accurate image might be something closer to a circle, or perhaps a landscape, in which different traits appear in different combinations. Two autistic people may have very little in common on the surface and yet still recognise something of themselves in each other.
It is also not possible to infer very much about someone’s intellectual abilities simply by looking at them. Some autistic people who speak little or not at all have extremely rich inner lives. Equally, people who appear highly articulate, successful and independent may still find themselves overwhelmed in particular situations and, at those moments, require a great deal of support.
My disagreement with Frith is therefore not really about terminology. It is about the argument that the growing diversity of autistic people somehow invalidates the concept altogether. In my view the opposite is closer to the truth.
What has expanded is not the category of autism, but our recognition of how many different forms it can take. My own experience illustrates the point. Once I had the diagnosis, many aspects of my own history suddenly became easier to understand.
Much of my music is built from rule systems, permutations and constraint-based processes derived from areas of focus that are often non-musical. This is a way of working that many autistic people immediately recognise. What I have described as a “stubborn literalism” often gives rise to abrupt formal changes in focus, rather than smooth and continuous transitions.
Only since my diagnosis did I begin to recognise how closely these habits of thought aligned with descriptions of autistic cognition. Ideas such as monotropism, intense focus or Flow and pattern-based reasoning are highly appropriate. The autism was not new. What was new was finally having a name for it.
When I speak with other autistic people, I often hear similar descriptions of attention and perception: a pleasure in pattern, an attraction to detail, an instinct to explore structures very thoroughly. These ways of thinking can sometimes create friction in social environments that rely heavily on implicit expectations and rapid intuitive signalling. But they can also be powerful creative resources.
None of this means that autism brings no difficulties. For many people it involves real challenges, and some autistic individuals require substantial support. But it also involves distinctive patterns of attention and perception that are part of the richness of human cognitive diversity.
From that perspective, the claim that the autism spectrum has “collapsed” looks less like a discovery than a reaction to a changing conversation.
For many years autism research was shaped almost entirely by clinicians and psychologists observing autistic people from the outside. Today autistic people themselves are participating in the discussion about what autism means, as writers, artists, scholars and researchers. What Frith describes as the “collapse” of the spectrum looks rather different from the inside. What has expanded is not autism itself, but our ability to recognise it.
That shift inevitably unsettles some older frameworks.
Autism may not be a straight line running from “mild” to “severe”. It may be closer to the circle or landscape I described earlier: a space in which different patterns of perception, attention and communication appear in many different combinations.
The autism spectrum has not collapsed: it is the people describing it who have changed.
