TTW Nest · College, work, and adult lifeNovember 2026
Home / Research / Questions / Reading
Research and evidence

Is autistic communication actually deficient?

A communication mismatch is not necessarily one person's failure.

It is remarkably easy to label a conversation “difficult” when the other person communicates differently. It's also easy to assume the other person should adapt. That's not always a useful place to start.

The idea

Damian Milton argued that autistic and non-autistic people can misread one another because they are working from different assumptions. The phrase double empathy problem asks us to examine the interaction, not diagnose a single person's shortcomings.

A theory helps us ask a question. It does not answer every empirical question for us.

What was tested

Crompton and colleagues used a structured information-sharing task with 72 adults divided into autistic, non-autistic, and mixed groups. Information moved effectively between autistic participants in this task. That result complicates the story that autistic people are simply poor at communication.

But a laboratory diffusion-chain exercise is not a staff meeting, a breakup, or a classroom discussion. It cannot tell us that every autistic person understands every other autistic person, or that neurotype matching guarantees a good relationship.

What to do with it

When a group project falls apart, have each member explain what they thought the assignment required and what they understood the others to have promised. Write down the agreed plan. When somebody misses a cue, clarify it rather than interpreting a personality trait.

This also matters for the language we use to describe people. Botha and colleagues argue that labels are not merely cosmetic and that autistic people should have a say in how they're represented.

What would change our minds? Replications in more natural settings, with different ages, communication modalities, and relationships, plus research on the practical tools that help people repair misunderstandings.