AutismologyAutistic studiesHistory of Autism: Landmarks, Concepts and Transformations
B — Critical examination
Observation, interpretation, causation, power and normalisation
The documentary landmarks used in this part come in particular from sources S7, S11, S14–S16, S19–S23 and S28. The methodological propositions are argued analyses made in this publication.
B.1 — The first danger: reading the past with present-day categories
Anachronism arises when Bleuler’s Autismus, psychoanalytic uses of the word and contemporary diagnostic categories are treated as a single stable object. Each must first be placed back within the conceptual language and populations of its period before any comparison is made.
B.2 — Observation, interpretation and causation
A child does not answer immediately when spoken to: observation.
Saying that the child “refuses the relationship”: interpretation.
Saying that the child refuses the relationship because of an unconscious conflict caused by a parental relationship: causal hypothesis.
These three levels do not have the same status.
The history of autism shows how often they have sometimes been confused.

Accessible description of the illustration: Three cards connected by arrows illustrate three levels that should not be confused: observation, for example “the person leaves the room”; interpretation, for example “they are rejecting the relationship”; causality, for example “they act this way because of X”. The graphic reminds the reader that an observed behaviour does not automatically contain either its interpretation or its cause.
B.3 — Underdetermination: the same behaviour can have several causes
A person who leaves a room may:
- be experiencing sensory overload;
- be in pain;
- not understand the situation;
- be afraid;
- be tired;
- be trying to regulate their state;
- not want to participate;
- be experiencing a psychiatric difficulty;
- or combine several of these factors.
Observable behaviour does not, by itself, contain its explanation.
The more atypical or inaccessible the person’s communication is to the observer, the more interpretive caution is required.
B.4 — Can a theory be contradicted?
A classic difficulty with some psychoanalytic systems lies in the possibility of interpreting almost every reaction as compatible with the theory.
Agreement may confirm the interpretation; disagreement may be described as resistance.
The philosophical discussion of the falsifiability of psychoanalysis is more complex than the slogans surrounding it, but the methodological question remains essential:
what could force us to abandon our hypothesis?
A theory that allows no clear answer risks becoming a closed system.
B.5 — Interpretive circularity and authority
The danger increases when the theory assigns in advance a meaning to every behaviour: silence, refusal, attachment, repetition, protest.
If the person can never correct the interpretation because their correction is itself interpreted as a symptom, the professional effectively becomes the owner of meaning.
This structure is particularly problematic in autism, where communication differences may already reduce the person’s ability to challenge what is said about them.
B.6 — Individual history and general theory
An autistic person may have traumas, conflicts, defence mechanisms, relationship difficulties and a complex psychic history.
Recognising this psychic life does not demonstrate that these phenomena cause autism.
An explanation that is valid for an individual case does not automatically become a general theory of aetiology.
B.7 — Why could psychogenic models be convincing?
They emerged in a context in which the developmental mechanisms of autism were very poorly understood.
They offered a coherent narrative for behaviours that were not understood.
They belonged to prestigious schools of thought.
They gave the impression of transforming the mysterious into the intelligible.
But the coherence of a story is not a causal demonstration.
B.8 — Parental blame and the impossibility of defending oneself
“Refrigerator mother” theories illustrate the dangers of an asymmetrical interpretive system.
A parent’s protest can be read as denial; acceptance as confirmation.
An accusation that is almost impossible to refute then becomes an institutional mechanism of blame.
The historical role of parent movements should also be understood as a challenge to this power structure.
B.9 — Therapeutic usefulness and theoretical truth
Psychotherapy can be helpful through the relationship, listening, stability, verbalisation or narrative reconstruction without all the causal hypotheses of its school being demonstrated.
“This helps this person” and “this explains the origin of autism” are two independent propositions.
B.10 — The biological turn does not resolve problems of category and norm
The retreat of psychogenic explanations profoundly changes causal hypotheses, but leaves other questions open: the heterogeneity of the populations grouped together, the actual scope of biological associations, the choice of outcome criteria and the normative objectives of interventions.
Changing the level of explanation therefore does not remove the need to examine separately the construction of categories, causality and the practical effects of interventions.
B.11 — Visible behaviour and quality of life
The disappearance of a behaviour does not automatically demonstrate improvement.
A person can learn not to show overload without the overload decreasing.
They can produce the expected response while expending considerable effort.
They can stop a self-regulatory movement because it is punished.
The evaluation of an intervention must therefore include its effects on understanding, autonomy, communication, distress, safety and quality of life — not only visible conformity.
B.12 — The problem of normalisation
When an intervention defines success as becoming “indistinguishable from peers”, the reference norm explicitly becomes a non-autistic appearance.
We must then ask: why should resemblance be an objective in itself?
Objectives more directly connected with the person’s interests can be formulated: learning, communicating, understanding, reducing distress, exercising choice, being autonomous, accessing an activity or living safely.
B.13 — Psychiatry: resource or totalising framework?
The psychiatric indication should be linked to the identified problem: depression, psychosis, bipolar disorder, catatonia, suicidal crisis, disabling anxiety disorder or another specific clinical situation.
The useful formulation is therefore: psychiatry for an autistic person when they need it, rather than a “psychiatry of autism” that would make autistic existence itself the target of treatment.
Contemporary recommendations notably state that medication should not be used as a treatment for the “core features” of autism in adults. [S21–S22]
When the person communicates little or differently, additional safeguards are needed: look for pain, epilepsy, infection, adverse effects, sleep problems, constipation or another somatic cause; collect observations from several contexts; check for changes from the person’s usual functioning; and do not confuse inability to explain verbally with absence of a concrete cause.
A genuinely adapted approach is therefore transdisciplinary: general medicine, neurology, psychiatry when necessary, psychology, speech and language therapy, occupational therapy, education, augmentative and alternative communication, social work, sensory and environmental analysis, peers, rights, and above all detailed knowledge of the person.
B.14 — Internal symptom or response to the environment?
A difficult behaviour can arise from an internal condition, an unsuitable environment, or an interaction between the two.
Treating only the individual risks making invisible noise, pain, an incomprehensible rule, an abusive relationship, overload or lack of accessibility.
A shutdown, sudden silence, agitation or apparent refusal may, for example, follow intense noise, an uncomfortable temperature, pain, an unpredictable wait, an accumulation of demands, unexpected physical contact or a misunderstanding. The final behaviour alone does not distinguish an internal, environmental or interactive cause.
The appropriate sequence is therefore: describe precisely what happened, reconstruct the chronology, look for somatic and environmental causes, try the least intrusive adaptations when possible, and only then attribute a more general mechanism.
B.14bis — A psychoanalysis “adapted to autism”?
As a general causal theory of autism, no: history does not provide a sufficient basis for making psychic conflicts, a parental relationship or a failure of symbolisation into a general explanation of autistic existence.
An autistic adult may nevertheless choose psychoanalytic, Jungian or psychodynamic psychotherapy to work on their history, relationships, conflicts or emotions. Evaluation then concerns what the person requests, understands and chooses, as well as the effects observed.
Minimum safety conditions are particularly important:
- genuine consent and the possibility of stopping;
- no blaming of the person or their family for autism;
- sufficient knowledge of autism and communication differences;
- adaptation of pace, language and implicit assumptions;
- an explicit right to challenge the therapist’s interpretation;
- investigation of somatic, sensory and environmental explanations when plausible;
- evaluation of benefits and harms in the person’s real life.
The central point is interpretive authority: a therapeutic interpretation must remain a contestable hypothesis, never an explanation that becomes true precisely because the person rejects it.
B.15 — Classifications are historical tools
Official classifications are operational and historical tools: they serve communication, research organisation and access to rights, while their boundaries and categories evolve. The transition from childhood schizophrenia to the spectrum provides a direct example.
B.16 — Who produces knowledge?
The entry of autistic people into knowledge production corrects a historical asymmetry: for a long time, the people being studied had very little power over how their behaviours were interpreted.
Their participation provides a direct source of information for formulating questions, choosing variables and interpreting results.
B.17 — Methodological rules derived from this history
- Define the meaning of words in their historical period.
- Separate observation, interpretation and causality.
- Ask what evidence could contradict the hypothesis.
- Avoid systems in which every objection becomes confirmation.
- Do not confuse the usefulness of a therapy with proof of its causal theory.
- Do not confuse autism, psychosis and schizophrenia.
- Do not confuse reduction of a behaviour with improvement in life.
- Look for concrete environmental and physical causes.
- Give the person concerned a genuine right of correction.
- Keep every interpretation revisable.
- Do not treat Freud, Jung, psychoanalysis or psychiatry as homogeneous blocs: specify the author, proposition, period and practice concerned.
- Do not infer from the failure of a psychogenic theory that every biological, genetic or behavioural hypothesis is automatically correct.
- Explicitly preserve the possibility of psychiatric or medical support for associated problems when indicated.
- When the objective permits, prefer learning, accessibility, autonomy and adaptation to assimilation or mere normalisation of appearance.
- Distinguish criticism of institutional power from denial of real suffering or disorders.
B.17bis — Frequent misunderstandings and brief answers
Did Jung discover autism? No. He is historically close to Bleuler and Freud, but he is not the founder of the modern concept of autism.
Did Freud invent autism? No. His vocabulary and influence belong to Bleuler’s intellectual context, but the term Autismus was introduced by Bleuler within his theory of schizophrenia.
Are autism and schizophrenia the same thing? No. Their former historical proximity explains some confusions, but contemporary classifications distinguish them.
Are autistic people “psychotic” by definition? No. An autistic person may or may not have an associated psychotic disorder, just like other people.
Did Popper “prove Freud wrong”? The question is more precise than that: Popper criticised the falsifiability of certain theories; Grünbaum and others discussed that criticism. Each proposition and its evidence must be examined.
Did Kanner invent the “refrigerator mother”? The history is more complex. Some of Kanner’s remarks contributed to a context in which parents were viewed with suspicion, but the most famous spread of the blaming model is notably associated with Bettelheim. Kanner also opposed simplistic parental blame.
Does the word “spectrum” correspond to a single axis? No. The profiles described under this category are multidimensional and needs vary across several axes and depending on context.
Is the disappearance of a behaviour sufficient to evaluate an intervention? No. Evaluation must also address understanding, distress, autonomy, consent, communication, safety and quality of life.
B.18 — Conclusion of Part B
The history of autism is also the history of explanatory frameworks that have sometimes transformed real observations into insufficiently demonstrated causal claims.
The main correction is not to choose a definitive school that would finally possess the whole truth.
It is to continually improve the quality of distinctions: fact / interpretation; characteristic / distress; individual / environment; help / normalisation; difference / associated disorder; diagnosis / lived reality.
Part C examines what this history becomes when the conceptual framework of Autismology is explicitly applied.