Show the full module text
Module 9 — Finding Your People
What the community offers, how to walk into it, and what to expect once you are there.
Self-Identification · Part Four — Other people
The big idea
Short on capacity today? The big idea: Most courses put community at the end, in the tone of an afterthought — and of course, connecting with others can be helpful. That framing is wrong, and the evidence says so. Connection to other autistic people is one of the few things that repeatedly shows up as associated with better wellbeing in this population.
This module is about walking in with your eyes open: what is actually there, what the arguments are, what the community cannot do for you, and how to enter it if the idea of a group makes you want to lie down.
Step 1 — The lesson
A. Why this is not a soft extra
The dominant explanation for poor mental health in autistic adults used to be that autism and distress were simply linked — that the misery came with the wiring. A 2020 study tested a different explanation, borrowed from research on other stigmatised minorities: that some of the distress is produced by the social position rather than the neurology.
In a sample of 111 autistic adults, three minority stressors — everyday discrimination, internalised stigma and concealment — each significantly predicted poorer mental health, and did so after controlling for general stress exposure. That last clause is the point. It is not that autistic people have harder lives and therefore feel worse; it is that being discriminated against, believing the negative account of yourself, and hiding each carry their own cost on top. Two of the three are what other people in the same position are best placed to shift.
Diagram — A · Put that down. These three stay. The dominant explanation for poor mental health in autistic adults used to be that autism and distress were simply linked — that the misery came with the wiring. A 2020 study tested a different explanation, borrowed from research on other stigmatised minorities: that some of the distress is produced by the social position rather than the neurology. In a sample of 111 autistic adults, three minority stressors — everyday discrimination, internalised stigma and concealment — each significantly predicted poorer mental health, and did so after controlling for general stress exposure. That last clause is the point, and it is why the sack in the drawing is on the ground rather than on the back. It is not that autistic people have harder lives and therefore feel worse; it is that being discriminated against, believing the negative account of yourself, and hiding each carry their own cost on top. Two of the three are what other people in the same position are best placed to shift. And there is evidence that they do: a 2017 study compared 272 autistic participants against 267 non-autistic comparison participants, and found that stronger identification with autism as a social identity was positively associated with personal self-esteem, through the perceived positivity of that identity, with significant negative indirect effects on both anxiety and depression by the same route.
And there is evidence that they do. A 2017 study compared 272 autistic participants against 267 non-autistic comparison participants. The autistic group had lower self-esteem and higher depression and anxiety, which everyone expects. The second finding is the interesting one: stronger identification with autism as a social identity was positively associated with personal self-esteem, through the perceived positivity of that identity, with significant negative indirect effects on both anxiety and depression by the same route.
A 2022 study of 196 autistic adults put a sharper edge on it. Higher masking related to poorer wellbeing; higher autistic community connectedness related to more positive wellbeing. But the study was testing whether connectedness buffers the harm of masking, and there was no interaction effect — connectedness was good for wellbeing on its own account, and did not neutralise the cost of hiding. It also found that higher connectedness correlated with more masking, not less. A 2024 study of 409 autistic adults pointed the same way: concealment related to worse depressive symptoms, and that ran through greater experienced stigma, not through community connectedness.
Diagram — B · Good ground, no roof. A 2022 study of 196 autistic adults found that higher masking related to poorer wellbeing and higher autistic community connectedness related to more positive wellbeing. But the study was testing whether connectedness buffers the harm of masking, and there was no interaction effect: connectedness was good for wellbeing on its own account, and did not neutralise the cost of hiding — which is why the ground in the drawing is warm and the umbrella is an outline the rain goes through. It also found that higher connectedness correlated with more masking, not less. A 2024 study of 409 autistic adults pointed the same way: concealment related to worse depressive symptoms, and that ran through greater experienced stigma rather than through community connectedness. So the honest version of the claim is narrower than the one on the poster, and still worth acting on: connectedness to other autistic people is consistently associated with better wellbeing; it is not a shield that makes the rest of it stop mattering; and the studies are cross-sectional, so nobody has established which way the arrow runs. It is not an extra. It is also not a cure.
So the honest version of the claim is narrower than the one on the poster, and still worth acting on:
Connectedness to other autistic people is consistently associated with better wellbeing. It is not a shield that makes the rest of it stop mattering, and the studies are cross-sectional, so nobody has established which way the arrow runs. It is not an extra. It is also not a cure.
The qualitative work fills in what the numbers cannot. Twenty autistic people, interviewed in person, by video, by text and by email — a design built so that people who cannot do an interview could still be in the study. The researchers found three separate things hiding inside the word “community”: belongingness (the sense of similarity), social connectedness (actual individual friendships) and political connectedness (alignment with the community’s equality goals). Participants named the benefits as increased self-esteem, a sense of direction, and a sense of community they had not found anywhere else. Where connectedness was absent, what showed up instead was ambivalence about being autistic, and internalised stigma.
Those three elements come apart. You can have a great deal of one and none of another. Knowing which one you are short of is more useful than deciding you need “community”.
B. Where it actually is
Not a directory. A description of the kinds of places, because the specific ones change every eighteen months and the kinds do not.
Comment sections. For a large share of people this is genuinely the first entry point, and it is worth saying so rather than treating it as a lesser version of the real thing. You read something, you scroll down, and forty strangers are describing your childhood. Nothing is asked of you.
Large open forums and social platforms. High volume, wide range of quality, low barrier. Good for the specific practical question — the sensory problem, the workplace wording, the what-do-I-say-to-my-GP question. Less good for anything needing continuity.
Smaller moderated spaces. Invitation, application or simple obscurity keeps them small. This is where the second element — actual friendships — tends to form, because the same names recur and people remember what you said last month. Usually found through the large spaces.
Special-interest groups that happen to be full of autistic people. Not autism spaces at all: the model railway group, the tabletop game, the language exchange, the choir. Plenty of people find their people this way and never once discuss autism — social connectedness without the identity part.
Peer-led groups, conferences and events. Run by autistic people rather than for them. Some are superb; some are one person’s unexamined opinions with an audience, and the tell is whether the group can tolerate disagreement. A conference is a concentrated dose — often the first time someone experiences a room organised around their sensory needs rather than against them, and also expensive, exhausting and not a starting point for most.
And if you cannot face a group at all. Reading without posting counts. It is not a failed version of participation, it is a form of it, and it is how most of any online community behaves. Following a handful of autistic writers and never interacting is a real relationship in one direction, and one-directional relationships do real work — it is the ordinary way people acquire language for their own experience. If your entire engagement for the first year is reading, you have still made the change that matters: you now know the account of your life is not unique to you.
Diagram — C · Six rooms in one house. Twenty autistic people were interviewed in person, by video, by text and by email — a design built so that people who cannot do an interview could still be in the study — and the researchers found three separate things hiding inside the word community: belongingness, the sense of similarity; social connectedness, actual individual friendships; and political connectedness, alignment with the community's equality goals. Participants named the benefits as increased self-esteem, a sense of direction, and a sense of community they had not found anywhere else; where connectedness was absent, what showed up instead was ambivalence about being autistic and internalised stigma. Those three elements come apart, and knowing which one you are short of is more useful than deciding you need community. As for where it actually is: comment sections are genuinely the first entry point for a large share of people, and nothing is asked of you. Large open forums and social platforms are high volume and low barrier, good for the specific practical question and less good for anything needing continuity. Smaller moderated spaces are where actual friendships tend to form, because the same names recur and people remember what you said last month, and they are usually found through the large spaces. Special-interest groups that happen to be full of autistic people — the model railway group, the tabletop game, the language exchange, the choir — give social connectedness without the identity part. Peer-led groups, conferences and events are run by autistic people rather than for them: some are superb, some are one person's unexamined opinions with an audience, and the tell is whether the group can tolerate disagreement. And reading without posting counts. It is not a failed version of participation, it is a form of it, and it is how most of any online community behaves.
C. The language question
You will hit this in the first hour, and it is presented as settled far more often than it is.
Identity-first is “autistic person”. Person-first is “person with autism”. The community-side argument is that autism is not detachable from the person, and that the grammar of separation implies something removable. The clinical-side argument is a long-standing disability-practice convention meant to avoid reducing a person to a condition. The evidence, reported honestly:
A 2016 survey of 3,470 UK residents — autistic people, parents, family, friends and professionals — found the most endorsed terms overall were “autism” and “on the autism spectrum”, with consensus across groups. The disagreement was elsewhere. “Autistic” was endorsed by a large percentage of autistic adults, family members and parents, but by considerably fewer professionals. “Person with autism” was endorsed by almost half of professionals, but by fewer autistic adults and parents. The authors concluded that no single term is universally accepted and that some disagreements appear deeply entrenched.
A 2023 US survey of 728 autism stakeholders found 87 per cent of autistic adults preferred to self-identify using identity-first language, and named the other figure in the same sentence: a sizeable minority, 13 per cent, preferred person-first for themselves. Professionals were more likely to use, like and choose person-first terms.
A 2025 systematic review pooled 19 studies, N = 6,350. Of the fourteen that assessed preferred terminology, ten found more participants preferred identity-first — with person-first endorsement running between 4 and 39 per cent. Where “no preference / either” was offered, between 4 and 37 per cent chose it. Both studies finding a person-first preference were conducted in Dutch. Every study was online, and few participants had intellectual disability.
So: identity-first is the majority preference among autistic adults, by a wide margin, in English-speaking online samples. It is not unanimous, the minority is not trivial, the professional convention is a convention rather than an insult, and it may not travel across languages.
The practical rule is one line. Use what a person uses about themselves. If you do not know, ask, in the plainest possible way — what do you prefer? — and then use that, including when it is not what you would have picked. About yourself, use whatever is true; nobody gets a vote.
Diagram — D · Ask the person, not the room. Identity-first is autistic person; person-first is person with autism. The community-side argument is that autism is not detachable from the person, and that the grammar of separation implies something removable; the clinical-side argument is a long-standing disability-practice convention meant to avoid reducing a person to a condition. The evidence, reported honestly. A 2016 survey of 3,470 UK residents — autistic people, parents, family, friends and professionals — found the most endorsed terms overall were autism and on the autism spectrum, with consensus across groups; the disagreement was elsewhere. Autistic was endorsed by a large percentage of autistic adults, family members and parents, but by considerably fewer professionals, while person with autism was endorsed by almost half of professionals but by fewer autistic adults and parents. The authors concluded that no single term is universally accepted and that some disagreements appear deeply entrenched. A 2023 US survey of 728 autism stakeholders found 87 per cent of autistic adults preferred to self-identify using identity-first language, and named the other figure in the same sentence: a sizeable minority, 13 per cent, preferred person-first for themselves. Professionals were more likely to use, like and choose person-first terms. A 2025 systematic review pooled 19 studies, N = 6,350; of the fourteen that assessed preferred terminology, ten found more participants preferred identity-first, with person-first endorsement running between 4 and 39 per cent, and where no preference or either was offered, between 4 and 37 per cent chose it. Both studies finding a person-first preference were conducted in Dutch, every study was online, and few participants had intellectual disability. So identity-first is the majority preference among autistic adults, by a wide margin, in English-speaking online samples — it is not unanimous, the minority is not trivial, the professional convention is a convention rather than an insult, and it may not travel across languages. The practical rule is one line: use what a person uses about themselves. If you do not know, ask, in the plainest possible way, and then use that, including when it is not what you would have picked. About yourself, use whatever is true; nobody gets a vote.
D. Gatekeeping, and the argument you will meet in week one
Some spaces will ask whether you have a diagnosis. Some will state that they are open to self-identifying people. Some will not say, and you will find out by watching.
Neither policy is a scandal, and it helps to decide that in advance rather than in the moment. A space that requires a diagnosis is usually not making a claim about your legitimacy; it is solving a moderation problem, or protecting a function — an assessment-support group where everyone is post-diagnosis has a coherent reason to be built that way. A space that is open is not being careless; it is acting on the view that the queue is why most people are undiagnosed.
Hold on to Module 1 when you meet it. The comparison of 147 self-identifying against 115 diagnosed adults found more than 93 per cent of both groups scored at or above the same threshold on the same trait instrument, and that 68.7 per cent of the self-identifying group wanted a diagnosis and could not get one. That is not a rhetorical weapon. It is the reason you need not take a closed door personally: you already know what the population looks like.
The argument itself will find you within a week.
The good-faith case against open self-identification: the categories mean something, the community’s political claims rest on describing a real population, a trait threshold is not a diagnosis, and spaces that cannot tell the difference get flooded and lose their usefulness to the people who need them most — particularly people with higher support needs, already under-represented in the very online spaces where this is being decided.
The good-faith case for: the queue and the cost mean that requiring a diagnosis selects for money, geography, gender and how well you presented at seven years old rather than for anything about autism. A gate that filters on access to services excludes exactly the people it claims to protect.
This course will not adjudicate that for you, and be suspicious of anywhere that does it in one sentence. Both positions are held by autistic people arguing in good faith. What you can notice is that it is an argument about spaces and their functions, not about whether you exist.
E. What the community is not
Three things, said plainly, because omitting them is how people get hurt.
It is not a clinical service. It cannot assess you and it cannot treat you. It can tell you what happened to them, which is valuable and is a different kind of thing.
Taking medical or legal steering from a forum is a bad idea. Not because the people are foolish — they are frequently better informed about the process than the professionals they describe. But medication decisions, diagnostic conclusions, employment law and benefit claims are all jurisdiction-specific, situation-specific and consequence-heavy, and confident strangers have no access to your situation. What a forum gives you is the question to ask and the phrase to use. Take that; leave the verdict.
It is not uniformly kind. It is a large group of human beings, many injured, many tired, some running on a grievance that was justified when it started. There are hierarchies, there are people who are certain, there are pile-ons. Arriving expecting sanctuary and finding an argument is a common early disappointment, and it is worth knowing in advance so it reads as normal rather than as a verdict on you.
Diagram — E · A chair pulled out. First, work out which of the three things you are short of: belongingness, individual friendship, or political alignment. If what you need is people like me exist, a comment section does it in an evening; if what you need is a friend, a large forum will not produce one and a small recurring space might; if what you need is for something to change, that is a third activity. People burn months in the wrong kind of room because they never separated these. Second, read for a month before you post: every space has a local dialect, a set of exhausted arguments, and things that are unwelcome for reasons invisible from outside, and a month of reading tells you whether this is a place you want to be at zero social cost. Third, decide your sentence about the diagnosis question before anyone asks it — something plain and unembarrassed, such as I have not been assessed, I am on a list, or I have worked it out for myself and I am not claiming a clinical opinion. Fourth, take the phrasing from the community and the decision to a professional: a forum is an excellent source of the words for what is happening to you and a poor source of what to do about your medication, your appeal or your contract. Fifth, leave rooms that make it worse, without making it a rupture — not every autistic space is good for every autistic person, and a space that is excellent for someone else can be actively bad for you, which is why the chair in the drawing is pulled out rather than bolted down. Three things the community is not, said plainly, because omitting them is how people get hurt: it is not a clinical service and cannot assess or treat you; taking medical or legal steering from a forum is a bad idea, because medication decisions, diagnostic conclusions, employment law and benefit claims are jurisdiction-specific, situation-specific and consequence-heavy; and it is not uniformly kind, because it is a large group of human beings, many injured, many tired, with hierarchies and pile-ons. Arriving expecting sanctuary and finding an argument is a common early disappointment, and it is worth knowing in advance so it reads as normal rather than as a verdict on you.
F. What helps
1. Work out which of the three things you are short of.
Belongingness, individual friendship, or political alignment. If what you need is people like me exist, a comment section does it in an evening. If what you need is a friend, a large forum will not produce one and a small recurring space might. If what you need is for something to change, that is a third activity. People burn months in the wrong kind of room because they never separated these.
2. Read for a month before you post.
Every space has a local dialect, a set of exhausted arguments, and things that are unwelcome for reasons invisible from outside. A month of reading tells you whether this is a place you want to be, at zero social cost — and it is a complete form of participation, not a warm-up.
3. Decide your sentence about the diagnosis question before anyone asks it.
Something plain and unembarrassed — I have not been assessed; I am on a list, or I have worked it out for myself and I am not claiming a clinical opinion. Prepared, this is a two-second exchange. Unprepared, it produces a defensive paragraph or a silence, both of which feel far worse than the question deserved.
4. Take the phrasing from the community and the decision to a professional.
The rule that prevents the most damage. A forum is an excellent source of the words for what is happening to you and a poor source of what to do about your medication, your appeal or your contract. Both sentences are true at once.
5. Leave rooms that make it worse, without making it a rupture.
Not every autistic space is good for every autistic person, and a space that is excellent for someone else can be actively bad for you — too angry, too certain, too much of one kind of story. You are allowed to stop reading. No announcement is required.
Module 10 is about the other relationship this raises, which is finding a professional who does not need this explained to them from the beginning.
Up next
Module 10 — A Therapist Who Gets It