Thinking ahead to the planned discussion on the how of our interactions, and the complexities of DEI as a practice rather than a tick box exercise
I’ve been reflecting on a recent study about Persian taarof, and its relevance to how we communicate around diversity and systems change. I’d really like to explore how we might design communication protocols that make our forums more inclusive — across cultures, genders, and neurotypes.
Hidden Rules, Neurodiversity, and Communication in Systems Change
This study looks at how AI is failing to address the Persian practice of taarof—a ritual dance of politeness where what is said often diverges from what is meant. For example, when an Iranian taxi driver waves away your payment with “be my guest,” the expected response is to insist several times before paying. What looks like generosity is in fact a coded exchange of respect.
Researchers (Sadr et al., 2025) built TAAROFBENCH to test how well AI models navigate these scenarios. Native Persian speakers scored 82% accuracy, but large language models managed only 34–42%. The AI systems defaulted to Western directness: completely missing the underlying cultural compression.
This resonates deeply with my own experience. Communication is not just about words—it’s about hidden codes. As Gregory Bateson (1972) reminds us, meaning is always contextual, carried in patterns of relation. Speakers compress meaning, and listeners must decompress it correctly. When we don’t share the same frame, misalignment easily arises.
Neurodiversity as a Parallel Case
For neurodiverse people, this isn’t an occasional cross-cultural challenge—it’s daily life. Research on the “double empathy problem” (Milton, 2012) shows that misunderstandings between autistic and non-autistic people are mutual. Both sides misinterpret each other’s codes.
I’ve seen this play out in many forums. Someone will speak with what seems to be bluntness, intensity, or excessive detail. Others read it as domineering or aggressive. To me, it often looks like a neurodiverse communication style: deeply ethical, but inflexible. And more than once, I’ve watched such people get excluded—not for the substance of what they said, but for how they said it.
In practice, these ND traits are often misread:
-
Bluntness taken as rudeness
-
Insistence taken as aggression
-
Excessive detail taken as domination
-
Ethical clarity taken as rigidity
This isn’t malice—it’s a different compression scheme. Just as AI misses taarof, groups miss the ethical and conceptual value of ND modes of expression.
A Case in Point: Greta Thunberg
The global response to Greta Thunberg illustrates these dynamics vividly. Thunberg communicates with directness, intensity, and uncompromising ethical clarity. Her blunt insistence that leaders “listen to the science” and her refusal to dilute urgency to fit into established diplomatic norms resonate with many as moral leadership. Yet critics often dismiss her as rude, naïve, or inflexible.
Scholars link this style to autistic communication traits: low tolerance for hypocrisy, strong adherence to principle, and difficulty playing along with the rituals of political performance (Rosqvist, Chown, & Stenning, 2020; Yergeau, 2018). To many, this comes across as moral leadership. To others, it provokes ridicule and hostility.
The reaction to Thunberg mirrors what the taarof study and the “double empathy problem” describe: it’s not a failure of skill, but a clash of compression codes. Her moral urgency doesn’t fit conventional politeness rituals, and so she is both celebrated and dismissed.
This example underscores why international forums addressing systemic crises cannot rely on hidden Western, or academic rules of civility as the only valid mode of discourse. Doing so risks excluding precisely those who carry the strongest ethical compass in the room.
Why This Matters for Systems Change
Systems-change and metacrisis research emphasises plurality, reflexivity, and what Bateson (1972) called “double descriptions.” Nora Bateson (2016) likewise insists that we need multiple patterns of framing to see into complexity. If our forums reproduce hidden defaults of communication—whether cultural or neurotypical—we risk suppressing exactly the diversity of perspective that’s most needed.
For me, the exclusion of ND voices is not just a question of fairness. It narrows the ethical bandwidth of our collective. In a time when our decisions carry planetary consequences, that seems like an unacceptable loss.
Mechanisms That Might Help
Drawing on both intercultural communication and neurodiversity studies, I think we could experiment with mechanisms like:
-
Surfacing hidden rules — making group norms explicit, not assumed.
-
Multiple channels of participation — balancing live dialogue with asynchronous text, transcripts, and reflective writing.
-
Normalization of clarification — treating “Did you mean…?” as a sign of respect, not a stumble.
-
Gentle structuring — time-boxing or rotating turns so intensity is balanced without shaming.
-
Reframing inflexibility as principle — seeing rigidity as principled commitment, which can strengthen ethical dialogue.
-
Practicing double empathy — recognising that misreadings go both ways, NT and ND alike.
Towards Multicultural Humility
Practicing awareness of these hidden rules—whether in Persian taarof or ND communication—does more than prevent exclusion. It surfaces the assumptions we normally take for granted. It shows us how easily meaning slips between words and intentions, and reminds us that our own interpretive frames are limited.
This is what I would call cultural humility: recognising that we don’t know what we don’t know. For systems-change work, that humility may be one of the most important protocols we can develop.
Extending to Gender and Other Variation
It’s also worth remembering that compression codes differ not only across cultures and neurotypes, but also along gendered lines. Communication research consistently shows that what is read as “assertive” in one gender may be judged “aggressive” in another, or what is praised as “detailed” from one speaker may be dismissed as “long-winded” from another (Tannen, 1990).
The taarof study itself (Sadr et al., 2025) found models reproducing gender stereotypes even when prompts were not gendered—suggesting that bias is baked into training data as well as human interpretation.
If we are to design protocols for genuinely inclusive forums, then we must attend not only to neurotypical/ND differences, but also to how gender, culture, and power intersect to shape which communication styles are heard as legitimate.
Some Things We Might Discuss:
-
Hidden Rules
-
What hidden assumptions or “compression codes” do we notice operating in our own group?
-
Where might these be excluding certain ways of speaking or knowing?
-
-
Plurality of Styles
-
How can we make space for different tempos, levels of detail, or forms of insistence without losing flow?
-
What might balance look like in practice?
-
-
Protocol Design
-
What protocols could we try that actively welcome and surface differences, rather than smoothing them over?
-
How might we structure time, turn-taking, or asynchronous follow-up to support inclusion?
-
-
Ethics and Urgency
-
How do we make sure voices of strong ethical clarity (like Greta Thunberg’s) are not dismissed as “too rigid” or “too much”?
-
Can we treat these voices as a compass, even if the style feels uncomfortable?
-
-
Humility and Learning
-
What practices might help us remember that we don’t know what we don’t know?
-
How can curiosity and humility become norms in our dialogue?
-
Further Reading
-
Bateson, G. (1972). Steps to an Ecology of Mind. Chandler.
-
Bateson, N. (2016). Small Arcs of Larger Circles: Framing through Other Patterns. Triarchy Press.
-
Clark, H. H. (1996). Using Language. Cambridge University Press.
-
Laszlo, E. (2001). Macroshift: Navigating the Transformation to a Sustainable World. Berrett-Koehler.
-
Milton, D. (2012). On the ontological status of autism: The ‘double empathy problem’. Disability & Society, 27(6), 883–887.
-
Rafiee, H. (1991). Polite verbal wrestling: The pragmatics of taarof in Persian. Iranian Studies, 24(1–4), 19–30.
-
Rosqvist, H. B., Chown, N., & Stenning, A. (2020). Neurodiverse communication and the politics of directness: Insights from the Greta Thunberg phenomenon. Autism in Adulthood, 2(1), 6–14.
-
Sadr, N. G., et al. (2025). We Politely Insist: Your LLM Must Learn the Persian Art of Taarof. Brock University / Emory University.
-
Tannen, D. (1990). You Just Don’t Understand: Women and Men in Conversation. William Morrow.
-
Yergeau, M. (2018). Authoring Autism: On Rhetoric and Neurological Queerness. Duke University Press.