CONCEPT
Metacommunication
Bateson's term for communication about communication — the signals that frame, calibrate, and contextualize the content of messages, and whose absence in AI output produces characteristic circuit pathologies.
Metacommunication is Bateson's term for the signals that tell a receiver how to interpret a message — whether it is serious or playful, confident or tentative, sincere or ironic. In human conversation, metacommunication is constant and largely unconscious: tone of voice, facial expression, body posture, the thousand small cues that carry the meta-level signal. These are not decoration; they are constitutive of meaning. The same words — 'I love you' — mean different things depending on the metacommunicative frame. For the AI age, the framework exposes a structural feature of human-AI circuits that produces characteristic pathologies. The AI produces outputs syntactically indistinguishable from human collaboration but without the metacommunicative shading that would normally accompany such outputs. Uniform confidence regardless of uncertainty. No tonal variation between domains where the system is reliable and domains where it is extrapolating. The signals that would calibrate reliability are simply not produced.
In The You On AI Field Guide
The metacommunication problem maps onto Bateson's logical typing framework. The circuit produces