Narrative asymmetry

“But you never said anything.”

The sentence seems simple, almost irrefutable.

A situation has repeated itself for several months, sometimes several years; a person has accepted it, adapted to it, not really protested, then one day explains that this situation did not suit them, that they experienced it differently, or that they now understand something they were not yet able to name while they were in the middle of that relationship.

For the person opposite, there is then an apparent inconsistency.

If it really was a problem, why not have said so earlier? Why keep going along with it? Why accept something several times before explaining that it cost us something?

And how can two people who took part in the same conversations and lived through the same events end up telling two such different stories?

Our first reflex is often to look for which of the two interpretations is the right one.

I wonder whether the question should not start somewhere else.

Because before asking why someone did not defend a position earlier, we might need to ask whether we all have the same resources to object, to negotiate a limit, or simply to make our experience understandable enough to be heard.

Among some autistic people, this distinction seems to carry particular weight.

Not objecting does not necessarily mean agreeing

We all make a fairly natural inference between what we see and what we imagine to be the other person's intention.

Someone agrees to a request, so we assume they agree with it. They do not protest, so the situation probably does not pose a major problem for them. They continue taking part in a relationship, so we infer that its workings suit them at least well enough that there is no need to question it.

This reading probably works often enough that we have no reason to be wary of it.

Yet it is not always reliable.

In 2022, Amy Pearson, Jon Rees and Samantha Forster published a qualitative study devoted to the experiences of interpersonal victimisation among autistic adults. Its title is already particularly interesting: “This Was Just How This Friendship Worked”.

Among other things, the authors describe in some participants a form of heightened compliance, that is, an increased tendency to comply with requests that were nevertheless considered unreasonable, associated among other things with fear or avoidance of confrontation.

The testimonies also show that some people had had difficulty identifying other people's negative intentions, or had gradually come to regard certain relational patterns as normal.

This finding obviously does not mean that autistic people would be incapable of knowing what they want, that they would be naturally naive, or that they would systematically accept whatever others ask of them.

Such a reading would be both a caricature and scientifically unjustified.

What interests me is much more subtle:

the cost of objecting may not be equivalent for the two people.

Two individuals can know perfectly well what they want, while not at all having the same ease in defending it within an interaction.

For one, refusing may consist of quickly expressing disagreement, arguing for a few minutes, then bearing a relatively transient social tension.

For the other, the same interaction may require precisely understanding what has just caused discomfort, turning that experience into words, organising an argument while also following the other person's, decoding their reactions, and simultaneously bearing the emotional load produced by the conflict.

The question is therefore not simply:

“did this person know how to say no?”

It could be:

“what was the price of that no, at that precise moment and in that precise relationship?”

Because when the cost of disagreement becomes high enough, accepting can sometimes be less costly in the short term than defending one's position, even when acceptance does not actually correspond to what one would have wished.

And when this situation does not happen once, but regularly, it can gradually start to organise the relationship itself.

A repeated concession sometimes ends up no longer looking like a concession

Let us imagine something very simple.

A asks B for something.

B would prefer to refuse, but opening a confrontation requires enough energy that accepting seems, in the moment, more accessible.

A few weeks later, a similar request comes back and B accepts again, then another time.

Nothing spectacular may have happened.

No threat, no explicit constraint, perhaps not even a conscious wish to obtain more than the other wanted to give.

Only a succession of small interactions in which one of the two people had a relatively banal advantage: defending their interest cost them less.

What was initially a concession can then gradually become a habit.

Habit creates expectation.

Then that expectation sometimes ends up resembling an implicit rule.

The initial question — “Would you agree to do this?” — turns almost imperceptibly into “This is how we work.”

From then on, the two people can tell something true.

B can say:

“I adapted for a very long time because I really couldn't manage to defend my position.”

While A replies:

“I thought it suited you, since you had always agreed.”

These two sentences do not necessarily describe two different sets of events.

They give two different meanings to the same events.

For one of the two people, the successive acceptances are the traces of everything they absorbed to keep the relationship going.

For the other, they have become the proof that this arrangement was consensual.

And this difference can remain invisible for a very long time.

Until the day B starts saying no.

When starting to set limits suddenly looks like a change of personality

From the moment B starts more clearly defending their needs, something that seemed stable stops working.

From B's point of view, the change can be relatively simple:

“I am finally starting to express what I couldn't manage to defend until now.”

For A, the experience can be very different:

“You weren't like this before.”

Or:

“Until now, this had never been a problem.”

And A does not necessarily need to be lying to think this.

What seemed to them like an agreement genuinely disappears.

Something they regarded as part of the normal functioning of the relationship is now being questioned, so that the newly expressed limit can be experienced not as the belated correction of an asymmetry, but as the sudden removal of something that seemed a given.

There is a paradox here that seems important to me:

when a person has adapted for a long time, their assertiveness can be experienced as a change precisely because their adaptation had ended up becoming invisible.

The concession was no longer perceived as such.

It had become the normal way things worked.

And this is where the relationship can start to produce two very different stories from the same facts.

Two people can live through the same interaction without giving it the same meaning

This question connects directly with the evolution of our understanding of the communication difficulties linked to autism.

For a long time, these were mainly described as a problem located in the autistic person: they would supposedly misunderstand intentions, implicit meanings, emotions, social codes, or other people in general.

In 2012, Damian Milton proposed an important shift with the double empathy problem.

This shift is explored in more detail in Autism and double empathy: when two people do not experience the same interaction.

His idea is not to deny that autistic people may encounter certain difficulties with social understanding, but to point out that when an interaction brings together two people with sufficiently different ways of perceiving, interpreting and communicating their experience, misunderstanding can become mutual.

The autistic person may misunderstand the non-autistic person, but the reverse is also possible.

More than that, the communication failure may lie precisely in the gap between their two ways of functioning.

Several experimental studies have since explored this hypothesis.

In 2020, Catherine Crompton and colleagues studied the transmission of information within chains made up solely of autistic people, solely of non-autistic people, or alternating between the two neurotypes.

In this study, the mixed chains showed a greater loss of information, whereas chains made up solely of autistic people did not differ significantly from chains made up solely of non-autistic people. Participants in the mixed chains also reported weaker interpersonal rapport.

These results are interesting, but they should not be turned into a general law: later studies have not always reproduced exactly the same differences depending on chain composition.

They nonetheless help to shift an assumption long considered almost self-evident: when an autistic person and a non-autistic person fail to understand each other, we cannot automatically conclude that the misinterpretation lies on the autistic side.

Misunderstanding also belongs to the relationship.

Let us return to our example.

A said:

“You accepted.”

B replies:

“I gave in.”

If we look only at what was visible from the outside, the two sentences do in fact describe the same behaviour.

But they do not tell the same experience.

And no camera placed in the room would, on its own, have been able to resolve this difference.

The problem begins when the two interpretations do not carry the same weight

Two people can understand a relationship differently without this being in itself pathological or particularly worrying.

We all reconstruct events from our own experience, from what we perceived, from what we did not know, and from what we have learned since.

The question becomes more delicate when the two interpretations do not have the same resources available to be recognised.

One person may be quicker verbally, more comfortable with conflict, more able to summarise, more persuasive, or simply closer to the social codes considered usual in their environment.

The other may need time, distance, to reconstruct a chronology, to understand certain events after the fact, or to provide much more context before arriving at what they are actually trying to explain.

The two stories are then not merely different.

They do not start out with the same chances of being heard.

And this is where another notion becomes useful: that of epistemic injustice.

Who holds authority over what was lived?

The philosopher Miranda Fricker uses the term epistemic injustice to describe situations in which a person suffers an injustice precisely in their capacity to be recognised as someone capable of knowing, interpreting or bearing witness to their own experience.

One of the forms she describes, testimonial injustice, occurs when a person's testimony receives less credibility than it should.

This question takes on particular importance when we look at autism.

Various theoretical works have shown how autistic people can sometimes be seen as less reliable narrators of their own experience, or have their interpretation reclassified by someone else as essentially the product of their autistic functioning.

Empirical research is also beginning to document these experiences.

In an article published in the 2026 volume of Sociology of Health & Illness, Aino-Mari Uisma, Tuija Virkki and Minna Ylilahti study the experiences of autistic young adults in their interactions with the healthcare system.

The title of their article sums up part of the phenomenon:

Misunderstood, Minimised, Misrepresented.

Misunderstood.

Minimised.

Misrepresented.

The participants describe in particular situations in which their understanding of their own experience was diminished, reformulated or interpreted on their behalf.

Caution is obviously needed here: this study concerns the healthcare context and does not demonstrate that the same phenomenon automatically recurs in family, romantic, friendship or professional relationships.

It does, however, show something fundamental:

being the one who lives through an experience does not guarantee that one will be regarded as the one who has the best reading of that experience.

And when an autistic person tries to revisit a relationship several months or several years later, this question of credibility becomes central.

A simple story easily wins out over a story that requires context

Let us now suppose that B is trying to explain why they view the relationship differently today.

They go back over past events, clarify the chronology, explain that they had not understood certain things at the time they were happening, or that they felt them without managing to put them into words, and acknowledge having sometimes accepted situations that, in hindsight, seem deeply unbalanced to them.

For them, all this context is necessary.

For the person listening, it can have the opposite effect.

“If it was really a problem, you would have said something at the time.”

“You're changing your story.”

“You're analysing everything differently now.”

“Why bring this up after all this time?”

At that point, the difficulty is no longer merely telling what one has lived through.

One must also explain why not having been able to tell it properly at the time does not make this experience any less real today.

The person then finds themselves having to defend both their interpretation and their legitimacy to interpret.

One of the stories can fit into a single sentence:

“It has always worked like that.”

The other needs to explain everything that was invisible behind that arrangement:

“Yes, it has always worked like that, but here is what that arrangement required of me in order to keep existing.”

The first story is immediately legible.

The second needs context.

And in social life, simplicity often holds a considerable advantage.

When an explanation also threatens the image we have of ourselves

There remains another dimension, less specific to autism but essential for understanding what may follow.

When a person says:

“This relationship was deeply unbalanced for me,”

their intention may simply be descriptive.

They may be trying to put words to something they understand better today.

But the person opposite may hear something else.

“You took advantage of me.”

Then, almost immediately:

“You are someone who takes advantage of others.”

The discussion then leaves the terrain of experience to touch on moral identity.

We have known for a long time in social psychology that human beings seek to preserve a certain consistency between their behaviours, their beliefs and the image they have of themselves.

Leon Festinger's work on cognitive dissonance in particular showed how much the coexistence of incompatible representations could produce psychological tension.

I can sincerely think:

“I am a fair person.”

And hear at the same time:

“Our relationship worked in a way that was unjust for me.”

These two statements are not impossible to reconcile.

But they require psychological work.

I can acknowledge that my intentions differed from the effect produced, admit that I had not understood what the other person was going through, or accept that I benefited from a situation I had not consciously sought to create.

But there is also another way to reduce this tension:

to consider that the other's interpretation is incorrect.

Albert Bandura's work on moral disengagement describes various mechanisms through which we can preserve an acceptable image of ourselves when our behaviours come into tension with our moral standards: minimising the consequences, shifting responsibility, justifying our conduct, or attributing part of the problem to the person who suffered its effects.

Here again, precision is needed.

To my knowledge, no study has demonstrated a specific sequence in which a person would first take advantage of an autistic person's difficulty in defending their limits, before consciously or unconsciously constructing a narrative designed to preserve their own image.

That would claim more than the data allow.

On the other hand, when we bring together work on compliance with certain requests, victimisation, mutual difficulties of understanding, epistemic injustice and the general mechanisms for protecting our moral identity, one question becomes hard to avoid:

what happens when two people have not merely two different versions of a relationship, but different power to have their version recognised?

Narrative asymmetry

I propose here an expression for thinking through this phenomenon:

narrative asymmetry.

I am not using it as an established clinical concept, but as a reading lens.

There is narrative asymmetry when two people have taken part in the same interaction or the same relationship, but do not have the same resources available to make their interpretation understandable, credible and socially acceptable.

One may be quicker verbally, more comfortable with conflict, more concise, more persuasive, or simply closer to the communicational expectations of the environment.

The other may need distance, to precisely reconstruct a chronology, to understand certain events after the fact, to provide a great deal of context, or have a way of expressing their experience that corresponds less to dominant social expectations.

The important point is not to decide immediately that one is telling the truth and the other is lying.

A dominant story does not need to be untruthful to be incomplete.

“You never said no” can be factually accurate without telling us anything about the cost that saying no would have entailed.

“We have always worked like that” can likewise be true without explaining how that arrangement came to be built, nor whether both people truly had the same capacity to negotiate it.

“I never forced you” can be accurate without proving that both parties had an equivalent capacity to defend their needs.

We sometimes look for the lie when what should interest us is instead:

the frame from which a truth is told.

Up to this point, everything I have just described could occur in a friendship, a couple, a family, or any relationship in which two people do not have the same resources to have their experience recognised.

But there is one environment in which this asymmetry takes on an additional dimension.

Work.

Because in a company, the two people do not merely have different power to defend their interpretation.

They may also not have the same power to give it consequences.

When narrative asymmetry enters the workplace

An employee and their manager can have two different readings of the same situation, exactly like two friends or two partners.

The difference is that, in a company, these narratives do not exist in a neutral space.

A manager can evaluate, formalise a problem, assign tasks, write a report, escalate a situation to human resources or take part in decisions that will have a direct effect on someone's professional trajectory.

The two interpretations therefore do not necessarily carry the same institutional weight.

This organisational framework is developed in my Guide — Neurodiversity at work: understand, manage, transform

Let us take a relatively ordinary example.

An autistic employee regularly accepts schedule changes announced at the last minute.

He carries them out, produces no visible conflict and continues to meet the expected objectives.

His manager can reasonably conclude that this way of functioning, even if not ideal, does not really constitute a problem.

What he may not see is what happens around it: the time needed to mentally rebuild the day, the energy spent absorbing the change, the recovery needed afterwards, or everything that had to be shifted elsewhere to make this adaptation possible.

Then one day, the employee says he needs these changes to be anticipated more.

He refuses a last-minute change.

Perhaps several.

Something that was previously invisible suddenly enters the professional relationship.

And it becomes very easy to tell this shift as follows:

“He has become much less flexible.”

Whereas another description could be:

“He has stopped silently absorbing the cost of our lack of predictability.”

The two sentences may be describing exactly the same shift.

But they do not place the problem in the same place at all.

“And yet, up to that point, everything was fine”

There is a particularly interesting sentence in this type of situation:

“And yet, up to that point, everything was fine.”

It sometimes conflates two very different things.

The absence of an observable problem.

And the absence of an invisible cost.

A person can meet their objectives while struggling, attend every meeting while spending enormous energy recovering afterwards, take on a workload for several months without it being truly sustainable, or never ask for an adjustment because they do not yet know that another way of working is possible, because they do not wish to disclose their diagnosis, or because they anticipate too high a social cost for making that request.

Research on masking in the professional context reminds us precisely that part of the work carried out by some autistic people consists of adjusting or masking their behaviour in order to match the social norms of their professional environment.

This adaptive work has an almost paradoxical characteristic:

the better it works, the less the environment sees that it exists.

An organisation can then end up using a person's past adaptation as proof that they never needed any adjustment.

The world of work adds a power asymmetry

The professional difficulties faced by autistic adults do not seem to be reducible to their individual characteristics alone.

The literature on employment highlights the importance of organisational culture, the working environment, the quality of relationships with colleagues and managers, access to adjustments and the fit between the role and the person.

This evidence matters because it challenges a still very present conception of inclusion, according to which the main issue would be teaching the autistic employee to better express their needs.

Of course it is useful to be able to express them.

But we must still look at the environment in which this expression must take place, and what it can cost the person who carries it.

A person may have to decide whether or not to disclose their diagnosis, explain what it involves without being reduced to a list of symptoms, identify the necessary adjustments themselves, convince others that they are reasonable, renegotiate them whenever a manager changes, and then sometimes keep demonstrating that they constitute legitimate working conditions rather than a favour.

At this point, the question:

“Why didn't he ask for help sooner?”

may become less interesting than:

“How costly does our system make it to explain that one needs help?”

Within a company, a story can become a record

This is probably where narrative asymmetry takes its most concrete form.

In a private relationship, two people can remain for a long time with two incompatible stories of what they experienced.

In an organisation, an interpretation can become a record.

“Difficulty handling change.”

“Problematic communication.”

“Lack of flexibility.”

“Relational difficulties.”

“Resistance to management.”

“Attitude that has become confrontational.”

None of these observations is necessarily false.

The problem is not to assume that every negative remark made about an autistic employee is automatically the product of a failing environment.

That would simply reverse the bias.

The important question is rather:

who had the opportunity to contextualise the behaviour before it became a characteristic of the person?

Because once an interpretation enters a report, an appraisal interview or an HR discussion, it changes in nature.

It no longer merely describes a past event.

It becomes a piece of information from which other people, who were sometimes not even present when the events occurred, will come to understand the employee and make future decisions.

At that point, narrative asymmetry becomes institutional.

The person who finds it easiest to tell the story no longer merely holds more power over the interpretation of the past.

They can take part in building the very framework from which the future will be interpreted.

And when the person finally tells a different story?

Let us now suppose that the employee decides to explain precisely what he experienced.

He goes back over several months of functioning, provides a great deal of context, explains that a situation he used to accept was already difficult but that he did not yet know how to put it into words, or that he had long thought the problem simply came from his own inability to adapt.

For him, these details are sometimes necessary to make the mechanism visible.

For the person across from him, they can produce the opposite impression.

“Why bring up events from several months ago?”

“Why didn't you say so at the time it happened?”

“That's not how you presented things back then.”

The person then no longer has to explain only what they experienced.

They have to explain why their difficulty in putting it into words sooner does not make their experience any less legitimate today.

And they may find themselves having to become an excellent advocate just to obtain the basic right to be considered a reliable witness to their own experience.

An inclusive organisation should perhaps learn to ask other questions

We teach autistic people a great many things.

Identifying their needs, communicating more clearly, requesting adjustments, asserting themselves, developing their self-advocacy, learning to say no and making their functioning intelligible to the people around them.

All of this can be important.

But if our thinking stops there, we risk placing most of the responsibility on the person who already faces the obstacle.

A genuinely inclusive organisation should perhaps learn to ask other questions.

Not only:

“Why didn't you ask for this adjustment sooner?”

But:

“What, in our organisation, made this request difficult to formulate?”

Not only:

“Why did you accept this way of working if it did not suit you?”

But:

“What was this adaptation costing you, and why did we not see that cost?”

Not only:

“Why has your behaviour changed?”

But:

“Is it possible that what changed was not your need, but your capacity to defend it?”

And when a conflict arises, perhaps we should ask ourselves one last question before turning a behaviour into a characteristic of the person:

“What story are we in the process of writing from what we see, and what part of the experience remains invisible from where we are looking?”

Training managers and HR professionals in autism should therefore probably not consist solely of teaching them a list of characteristics, symptoms or good practices.

It also involves learning to shift one's gaze.

Not to confuse observable behaviour with inner experience.

Not to take past adaptation as proof that the environment was adapted.

To distinguish an intention from a consequence.

And above all, to look at what occurs between a person and their environment before deciding that the problem necessarily belongs to one or the other.

Because when someone has spent years adapting, the day they start defending their needs can give the impression that they have become more difficult, less flexible or simply different.

But sometimes, it is not the person who has just changed.

It is simply the asymmetry that, for the first time, becomes visible.

To go further

The question of reciprocal interpretation in autistic interactions is developed in more detail in Autism and double empathy: when two people do not experience the same interaction.

To place these mechanisms back within the context of management and work organisation, see also my Guide — Neurodiversity at work: understand, manage, transform