What If the DSM-5 Got Autism Wrong?

A French Canadian researcher thinks so, and takes it apart, piece by piece.

Photo by chen lei on Unsplash

Autism is once again making headlines. Here we go again: the tired claim that “autism is caused by vaccines” is back in circulation. It’s unfortunate that autism only draws media attention in such sensational ways, rarely to highlight the actual experiences of autistic people, their challenges as well as their contributions to society. Do we ever get a thank you for bringing fresh perspectives to the world? Wouldn’t it be nice, for once, if the media simply asked: “Hey, autistic people! What can we do as a society to make your life better?”

Anyway.

So it’s in this frustrating context that I came across a different and slighly thought-provoking perspective on autism: “Si l’autisme n’est pas une maladie, qu’est-ce ?” (“If autism is not an illness, what is it?”) by French Canadian researcher Laurent Mottron¹. Drawing on over forty years of research, Mottron dismantles the dominant paradigms and offers a radically new way to understand autism. In this article, I will try to convey the core of his insights.

Autism as an Asymmetrical Bifurcation

Mottron no longer supports the idea of autism as a neurodevelopmental disorder. Instead, he proposes that autism represents an alternative developmental path, a bifurcation that occurs at a crucial moment in early development. He illustrates his point by drawing parallels with other asymmetrical bifurcations in human life, such as left-handedness, homosexuality or breech birth.

These phenomena aren’t “caused” by anything. They’ve always existed as natural variations in human development. They mark a divergence from the majority path, hence the term "asymmetrical": most people follow one trajectory, while a smaller group diverges and takes a different, but equally human, path.

Social Bias and Bifurcation

According to Mottron, typical infants process information through a social bias — meaning their cognitive and emotional development is shaped by constant social interaction. They don’t perceive objects in isolation, but in reference to others: parents, siblings, peers.

This social bias underpins behaviours like joint attention, imitation, shared reference in language, and internalisation of social norms. For typical children, learning is deeply embedded in social context at least until around 18 months.

At that point, some babies, for no apparent reason, diverge. They stop processing information through the social bias. This developmental shift is a developmental bifurcation. Social cues lose their privileged status, not because they are rejected, but because they are no longer prioritised. In a sense, the child "chooses" — much like a river finds its course — a path of cognitive independence.

This shift coincides with observable signs of autism:

• reduced interest in others’ actions,

• language treated as information rather than communication,

• less responsiveness to social interaction.

The autistic child does not withdraw from others entirely but develops a self-directed mode of engagement, intensely focused on personal activities. What is interpreted as a deficit in social cognition by the typical observer, may actually reflect independence from the social bias. For instance, not responding to one’s name may simply mean that the child is already focused on something else, not that they lack understanding or awareness. For the autistic child, there is no added value in information carried by others. All information has the same weight, there is no hierarchy.

Autism Is Not an Illness

The DSM-5’s definition of Autism Spectrum Disorder (ASD) has become so broad that it borders on meaningless. Mottron argues for a return to a prototypical model that maintains conceptual clarity and biological coherence instead of lumping together unrelated profiles under a single vague label hence the idea of developmental rather than neurodevelopmental bifurcation. This bifurcation is a natural event that occurs in a small percentage of humans, where the individual takes a different path by letting go of the social bias that shapes typical human development. It’s a shift that happens without cause or pathology. It simply is.

The Annihilation of the DSM-5

Based on this framework, Mottron deconstructs the DSM-5² line by line. Let’s walk through its core diagnostic criteria and see how he reframes each point.

1. Persistent Deficits in Social Communication and Interaction

Social-Emotional Reciprocity

Once a toddler begins processing information without social bias, they typically show a reduction in joint attention, except when they need assistance. Joint attention refers to behaviours such as pointing at objects or following another person's gaze. When autistic toddlers do not engage in these behaviours, it is often interpreted as a deficit in Theory of Mind, the assumption being that they fail to understand the intentions or mental states of others.

However, according to Laurent Mottron, this interpretation misses the point. The autistic child doesn’t need to share their perceptions or actions because they have decoupled perception from social context. They act independently, engaging others only when required, for example, to reach an object they cannot access alone. This tendency to seek help only when necessary has sometimes been mischaracterised as manipulative, a term that is unjustly and frequently applied to autistic individuals.

Crucially, Mottron emphasises that this bifurcation does not impair emotional connection or attachment. Autistic children still engage in behaviours that nurture strong emotional bonds: they express joy during play with their parents and seek comfort when distressed. The absence of joint attention is therefore not a sign of emotional detachment, but rather the result of a different cognitive prioritisation.

Nonverbal Communication

In typical development, the social bias naturally sustains nonverbal communication, through gestures, facial mimicry, eye contact, and vocal intonation. These cues are constantly attuned to social interaction and are used almost effortlessly. In typical children, they are in synch. In autistic children, however, these nonverbal expressions often appear disconnected or even absent. The child may seem serious, unexpressive, or out of sync with others. Yet, as Laurent Mottron points out, this is not a deficit in nonverbal communication. Rather, the autistic child is typically immersed in their own activities, following ‘their own agenda’.

While the typical child processes information in constant relation to others, the autistic child does not. Because their attention is so deeply focused elsewhere, their nonverbal functions are often in the background and remain inactive unless specifically needed. Their nonverbal behaviours, or the apparent lack thereof, are not dysfunctional; they are consistent with their way of relating to the world in a direct, non-socially mediated manner.

Understanding and Maintaining Relationships

Autistic toddlers often interact with others in ways that differ from typical social expectations, but this doesn't mean they’re incapable of understanding and forming meaningful relationships. In fact, they often form deep and enduring bonds with their parents. While they might not show clear preferences for other children before school age, or sometimes later, they typically do form connections in time.

They may not follow conventional social codes, for example, they often dislike physical contact with strangers. Or, they may ignore strangers altogether or be upset in their presence. But who is to say that this is a lack of understanding of relationships? Maybe soon after the bifurcation, the child learns that people, adults and peers, are likely to interrupt their activities and therefore are terribly annoying…

But this changes over time. At some point, especially after starting school, autistic children begin to show preferences for certain peers, with whom they may develop deep and lasting connections — probably with those who are less annoying!

A side note: in autism, regardless of the domain being studied, one recurring pattern stands out — quality is consistently preferred over quantity. For example, autistic teenagers rarely engage in stereotypical flirting, yet they often build stable and meaningful romantic relationships, and statistically, they tend to have fewer partners.

2. Restricted, Repetitive Patterns of Behaviour, Interests, or Activities

Following the logic of a developmental bifurcation that favours non-social information processing, Laurent Mottron proposes replacing the term “repetitive behaviours” with “perceptual overfunctioning.”

Stereotyped or repetitive motor movements, use of objects, or speech

Autistic children often spend extended periods exploring objects in depth, for instance, rotating them in three-dimensional space to examine all possible movements, lighting effects, or positions. These activities tend to focus on objects belonging to the same category (e.g., items that can be aligned, that share colour or shape), and include experimenting with their physical properties and transformations. They may involve repetition, but how about we call this curiosity and deep focus instead of dysfunction? In the same logic, autistic toddlers may be drawn to objects that can be sorted, aligned, or categorised by shape, colour, or physical transformation. Is this a “restrictive pattern of behaviour” or detailed and thorough investigation?

Behaviours like hand flapping or head banging are often labelled as pathological and grouped under the same category as object exploration. But Mottron questions this interpretation. These behaviours do not belong to the domain of object or environmental exploration. They are expressions of emotional states, something the DSM fails to recognise. Hand flapping can express joy, excitement, or even euphoria. Biting one’s hand, slapping one’s face, or hitting one’s head against the wall expresses frustration, distress, or anger. It is true that these behaviours can occur regularly and may appear repetitive. But would anyone describe a non-autistic child as having a "restricted behaviour" simply because they always laugh or cry in the same way?

Finally, the author suggests that actions like rocking or pacing often emerge from informational voids — moments where the child lacks new input to process, either because nothing new has occurred or because their exploration has been interrupted or blocked by those around them.

Insistence on sameness, inflexible routines, or ritualised behaviour

Ever since Kanner’s earliest descriptions, it has been taken for granted that autistic individuals display atypical outbursts of anger. This anger has rarely been questioned; it is usually interpreted as a pathological aversion to change. But Mottron proposes a different explanation. These reactions usually occur when a child is interrupted while focused on an activity. As he puts it:

“We cannot invoke intolerance to change as an explanatory principle when we interfere with an activity the child is in the middle of, or when we alter the outcome of that activity.”

In other words, the anger comes from being deprived of something such as their thoughts, their observation, their construction, not from an inherent fear of change.

Typical adults, influenced by social bias, often assume their interruptions are helpful. But the autistic child, who does not share this bias, sees no added value in them: they only see the interruption! This might be one of the earliest signs of the so-called double empathy problem: mutual misunderstandings between autistic and non-autistic people, rooted not in a lack of empathy but in different cognitive perspectives.

Highly Restricted, Fixated Interests

“To be intensely interested in something means being less interested in other things.”

Autistic children often develop highly individual interests that are disconnected from what is typically valued or considered “normal” within the framework of social bias. They engage deeply with these interests, perhaps precisely because they are free from this social bias, which in turn may free up cognitive resources that they can then invest more fully in their chosen activity.

These interests can absorb their full attention, which may lead typical observers to perceive their focus as “rigid” or “fixed.” However, as Mottron points out, within the context of these intense interests, autistic children often demonstrate creativity and inventiveness, something that clearly contradicts the assumption of restriction or rigidity.

Hyper- or hyporeactivity to sensory input (e.g. adverse response to sounds or textures)

Mottron also challenges the DSM-5’s treatment of sensory features. Here again, Mottron critiques the reductive nature of the DSM-5. The reference to “sensory input” suggests a focus on low-level perception— heat, light, movement, sound. Sensory hyperreactivity may explain why a child, or even an adult, covers their ears when disturbed by loud or intense noises.

However, focusing solely on sensory information overlooks another key feature of autism that plays out at the level of complex pattern perception. According to Mottron, the selection of preferred topics or activities in autism is often based on their complexity. Indeed, preschool-aged autistic children tend to be drawn to complex patterns, such as the trajectory of falling objects, the sound produced when an object is struck, original alignment criteria, or the inspection of three-dimensional rotations. These are not just sensory inputs, they reflect a sophisticated perceptual engagement with the structure and transformation of the world.

Conclusion: Reframing Autism as a bifurcation

To this day, after more than 60 years of research and over 2 million academic papers about autism, we still have no clear idea of what causes it. Even worse, we barely know how to define autism. Too many voices speaking at once sometimes produce more noise than understanding.

Mottron offers something different: a view of autism not as a disorder with a hidden cause, but as a natural variation — a causeless form of human existence, defined by a distinct way of relating to the world and to others.

Mottron’s perspective goes beyond critiquing the DSM-5. It offers an opportunity to reframe our entire approach to autism. Rather than pathologising divergence, he invites us to recognise an alternative mode of being rooted in autonomy, perceptual complexity, and cognitive independence.

By viewing autism as an asymmetrical developmental bifurcation and therefore not a disorder, we can begin to move away from deficit-based models. Perhaps we can finally end the costly and inconclusive search for a cause, and instead focus on how to help autistic people thrive in society.

The real challenge, then, is not to fix autistic people, but to fix our own understanding of what it means to develop differently. This means abandoning models that treat divergence as dysfunction, and creating frameworks that truly encompass what it means to be autistic.

1. Laurent Mottron, Si l’autism n’est pas une maladie, qu’est-ce?, France, Ed. Mardaga, 2024.

2. The DSM-5 presents Autism Spectrum Disorder (ASD) as a neurodevelopmental disorder.

According to the DSM-5, and in order to identify autism in an individual, they have to present:

1. Persistent deficits in social communication and social interaction, across multiple contexts.

• Deficits in social-emotional reciprocity (e.g. abnormal social approach, failure of back-and-forth conversation)

• Deficits in nonverbal communicative behaviours (e.g. eye contact, body language)

• Deficits in developing, maintaining, and understanding relationships

2. Restricted, repetitive patterns of behaviour, interests, or activities, such as:

• Stereotyped or repetitive motor movements, use of objects, or speech

• Insistence on sameness, inflexible routines, or ritualised behaviour

• Highly restricted, fixated interests

• Hyper- or hyporeactivity to sensory input (e.g. adverse response to sounds or textures)