The Science of Autistic Communication

For a long time, autism research framed communication difficulties as a one-sided deficit: something autistic people “lack.” Over the last decade, a growing body of research has challenged that assumption.

The central shift is this: communication breakdowns between autistic and non-autistic people are often mutual, not unilateral. This article summarizes the core research findings behind that shift and what the evidence actually shows.

🧾 The key research this summary is based on

🧠 Milton D. (2012)
The Double Emppathy Problem
Introduced the theoretical framework that communication difficulties arise from reciprocal differences in perspective, not from a single group’s deficits.

🧠 Crompton CJ, et al. (2020)
Autistic peer-to-peer information transfer is as effective as non-autistic peer communication
Empirical testing of information exchange accuracy across autistic–autistic, non-autistic–non-autistic, and mixed groups.

🧠 Crompton CJ, et al. (2020–2022)
Follow-up experimental studies on rapport, trust, and interaction quality across autistic and non-autistic pairings.

These papers collectively form the empirical backbone of the Double Empathy framework.

🧠 The traditional deficit model (what research used to assume)

Earlier autism research often assumed:

🧠 autistic people have impaired social understanding
🧠 non-autistic communication is the “neutral baseline”
🧠 misunderstandings originate primarily from autistic limitations

Under this model, communication difficulty was treated as a property of the autistic individual rather than an interaction.

🔄 The Double Empathy Problem: the core idea

The Double Empathy Problem proposes a different mechanism:

🧠 autistic and non-autistic people often have different communication norms
🧠 each group finds the other less intuitive
🧠 misunderstandings arise from mismatched expectations, not lack of empathy

Crucially, the theory predicts that:
🧠 autistic–autistic communication should be more effective than autistic–non-autistic communication
🧠 non-autistic people should also struggle to interpret autistic communication

This makes the hypothesis testable.

🧪 What the empirical studies tested

The key experimental question was:

🧭 Is communication accuracy lower because autistic people lack skill, or because mixed-neurotype communication introduces mismatch?

Researchers tested this by measuring:
🧠 accuracy of information transfer
🧠 rapport and trust ratings
🧠 interaction comfort
🧠 mutual understanding

Across three group types:
🧩 autistic–autistic pairs
🧩 non-autistic–non-autistic pairs
🧩 autistic–non-autistic mixed pairs

📌 Core empirical findings

The results consistently contradicted the deficit-only model.

Key findings included:

🧠 autistic–autistic pairs communicated as accurately as non-autistic–non-autistic pairs
🧠 mixed autistic–non-autistic pairs showed lower communication accuracy
🧠 autistic participants reported higher rapport and comfort with autistic partners
🧠 non-autistic participants also reported reduced rapport in mixed pairs

This pattern supports the Double Empathy prediction:
communication difficulty emerges between neurotypes, not within autistic communication itself.

🧠 Rapport and social comfort findings

Beyond accuracy, researchers examined subjective interaction quality.

Reported patterns included:

🧠 autistic participants felt more at ease with autistic partners
🧠 non-autistic participants rated autistic partners as less “intuitive”
🧠 both groups reported reduced comfort in mixed interactions

This indicates that discomfort is reciprocal, even if it is not always recognized as such.

🧩 What this means for the concept of “social skills”

These findings challenge a core assumption:

🧠 social difficulty is not necessarily a lack of skill
🧠 it can be a context-dependent mismatch

In research terms:
🧠 social competence is interaction-relative, not absolute
🧠 effectiveness depends on shared norms, timing, and expectations

This reframes many autistic communication behaviors as differences rather than deficits.

🧠 Why autistic communication is often misread

Research discussions identify several contributing factors:

🧠 differences in prosody, timing, and non-verbal cues
🧠 different assumptions about indirectness and inference
🧠 reduced reliance on implicit social rules
🧠 more literal or explicit communication styles

When these styles meet non-autistic norms, misunderstanding increases—even if the autistic communication itself is internally coherent.

🔁 Broader implications discussed in the literature

Researchers note that the Double Empathy framework has implications beyond conversation accuracy:

🧠 clinical assessments may over-pathologize autistic interaction styles
🧠 workplace communication norms may disadvantage autistic people
🧠 “social training” focused only on autistic adaptation ignores reciprocity
🧠 mutual adjustment may reduce breakdowns more effectively than one-sided change

These implications are discussed conceptually rather than as intervention outcomes.

⚠️ Limitations of the current evidence base

The literature also acknowledges important constraints:

🧩 many studies use relatively small samples
🧠 experimental tasks simplify real-world communication
🧑‍🤝‍🧑 participants are often verbally fluent adults
🧩 cultural and linguistic diversity is underrepresented

So while the evidence strongly challenges deficit-only models, it does not claim universality across all contexts.

🧠 Research takeaway

Empirical studies testing the Double Empathy Problem show that autistic people communicate with each other as effectively as non-autistic people do with each other, while mixed autistic–non-autistic interactions show higher rates of misunderstanding and lower rapport. These findings support a relational model of communication difficulty, where breakdowns arise from mismatched communication norms rather than a unilateral lack of social ability in autistic individuals.

References

Milton, D. (2012).
On the ontological status of autism: The double empathy problem. Disability & Society, 27(6), 883–887.
https://doi.org/10.1080/09687599.2012.710008

Crompton, C. J., et al. (2020).
Autistic peer-to-peer information transfer is highly effective. Autism, 24(7), 1704–1712.
https://doi.org/10.1177/1362361320919286

Crompton, C. J., et al. (2022).
Neurotype matching and social interaction quality. Frontiers in Psychology, 13, 839270.
https://doi.org/10.3389/fpsyg.2022.839270