AutismologyThe Science of Autism
Introduction
In August 2026, Uta Frith, a historic figure in cognitive autism research, published an editorial in Psychological Medicine asking whether “autism spectrum disorder” has lost part of its meaning and whether it can lead to misdiagnosis. Her article appears in a context in which the number of diagnoses has risen sharply, many adults are being identified late, diagnoses in women are increasing, and new genetic and developmental data highlight genuine heterogeneity among people currently diagnosed. https://doi.org/10.1017/S0033291726105376 ; https://doi.org/10.1038/s41586-025-09542-6 ; https://doi.org/10.1136/bmj-2025-084164
The question deserves examination without oversimplification. It would be insufficient either to reject Frith’s proposal because it challenges some contemporary conceptions of autism, or to assume that a statistical difference between people diagnosed early and late is enough to demonstrate the existence of two distinct realities. The data must be distinguished from their interpretation; the diagnostic category must be distinguished from what it seeks to identify; and the conceptual framework being used must be made explicit.
For this reason, this publication is organized into three autonomous but related sub-publications. Part A presents as faithfully as possible what Frith argues, the data she relies on, and the scientific context of her argument. Part B then examines the logical and methodological chain without relying on concepts specific to Autismology: selection bias, the status of age at diagnosis, genetic gradients, unstable thresholds, co-occurring conditions, social phenomena, causality, and the scope of the conclusions. Part C explicitly changes the frame of reference and revisits the question using notions developed in Autismology: autistic nature, the distinction between autism and ASD, downstream manifestations, the environment, sociogenic harms, the Natural Reference Framework, adaptation, and assimilation.
In brief: what the three parts show
A — What Uta Frith proposes and what the studies show Uta Frith considers that the current ASD category has become extremely heterogeneous and asks whether people diagnosed early should be more clearly distinguished from those diagnosed later. She relies in particular on changes in diagnosis, differences in developmental trajectories, recent genetic data, the increase in diagnoses among women, and questions of overdiagnosis, masking, and social influence.
B — Critical examination of the reasoning Several observations invoked by Frith are robust, but they do not necessarily suffice to demonstrate the existence of two distinct categories. Age at diagnosis depends in part on the detection system; early- and late-diagnosed groups are therefore already selected according to characteristics that may explain some of their differences. The available studies show heterogeneity and gradients, but not yet a clearly established natural diagnostic boundary.
C — Analysis according to the Autismology framework Autismology proposes distinguishing autistic nature from the diagnosis of ASD and from visible manifestations. The observed differences may result from the intensity and configuration of autistic characteristics, co-occurring conditions, adaptive capacities, the environment, and sociogenic harms. Before dividing autism into new categories, this approach therefore proposes looking further upstream for what constitutes the common core of autistic nature.
In summary: Uta Frith’s warning about diagnostic heterogeneity deserves to be taken seriously, but the essential question remains whether this heterogeneity reveals several fundamental forms of autism, or mainly several ways in which the same autistic nature can be detected, expressed, and affected by its environment.
This separation is not merely decorative. It allows readers to accept or contest each level independently. One can recognize that Part A accurately presents Frith without agreeing with the analysis in Part B; one can consider Part B methodologically sound without adopting the hypotheses of Autismology; and Part C is explicitly presented as the application of a specific conceptual framework whose hypotheses must themselves be tested against evidence.
The document takes up and reorganizes the most recent preparatory work carried out around this controversy. It therefore does not aim to reduce it to a few paragraphs: it preserves the arguments, distinctions, examples, and objections developed in the longer versions while removing unnecessary repetition and clarifying the function of each part.
This version is not limited to placing the two preparatory analyses side by side. It uses them as a foundation but broadens the documentation by integrating more explicitly the literature on the definition of “late diagnosis,” detection biases concerning girls and women, psychiatric co-occurring conditions, and the methodological criteria that would need to be met before a statistical contrast could be transformed into a new diagnostic category. Documentary additions are distinguished from proposals specific to Autismology.