A Deliberative Primitive is a **gestalt-level** unit of deliberation that people and AI can grasp as a whole, transmit across contexts, and recombine into larger practices. It is “primitive” not because it is indivisible, but because it functions as a **unit of selection** in cultural space: it can be copied, varied, and retained while keeping its identity.
# Purpose To provide reusable building blocks for better governance by making deliberation **recognizable**, **teachable**, and **operational** within Living Documents and multi-agent settings.
# Core Properties - **Gestalt**: apprehended as a whole pattern, not only as parts. See Gestalt. - **Family resemblance**: recognized by overlapping similarities-in-use rather than a single essence, across a Language Game. See Family Resemblance. - **Memetic identity**: stable enough to pass from mind to mind while allowing variation. See Cultural Evolution. - **Path dependence**: includes a teachable assembly history (how it is learned and performed). - **Composability**: chains with other primitives to form complete deliberative workflows. - **Normativity**: suitable for reason-giving under tests of truth, rightness, and sincerity.
# Human and AI Recognition - **Human**: recognized through story → theory transfer, supported by empathy / Theory of Mind. - **AI**: recognizes a primitive by the pattern of moves and commitments it creates in dialogue (who asserts, who challenges, what is justified, what is revised), not by similar wording alone. See Deliberative Move Recognition.
# Contrast With Related Notions - Not just a “step” in a process; it is a **named pattern** that carries meaning across contexts. - Not merely a “prompt” or “macro”; it is **normed** for public justification in Deliberative Democracy. - Not only logical rules; it is **practice-first**, defined by use and transmission.
# Minimal Test (Checklist) - Can a newcomer learn it from one short exemplar and reuse it elsewhere. - Do others recognize “the same thing” when varied by medium (text, audio, video, interactive). - Is there a brief assembly path to teach and audit it. - Does it measurably improve outcomes when combined with other primitives. - Can both humans and AI agents detect and suggest it at the right time.
# Examples - **Mirror role-swap**: argue briefly from the most-affected standpoint; record concrete plan changes. - **Care ledger attach**: bind a proposal to named harms, obligations, triggers, and stewards. - **Boundary proof**: show the action fits within quantified Planetary Boundaries or local limits. - **Reciprocity check**: counterpart writes their understanding; resolve mismatches before proceeding. - **Repair plan**: specify rollback and compensation paths before launch. - **Interactive steelman**: generate the strongest opposing case, then adapt the plan. - **Premortem**: imagine failure, list causes, add mitigations now.
# Implementation in Living Documents - Each primitive is delivered as a small Living Documents guide: story exemplar, steps, media, metrics, and an evolution log. - Tools in the document suggest missing primitives, check obligations, and gate merges (“policy as code”).
# Metrics - Harm ratio, care adherence, boundary integrity, reciprocity score, intergenerational legibility, ecosystem impact.
# References and Inspirations
- Wittgenstein – family resemblance, language games 
- AI can help humans find common ground
- DeepMind – Habermas Machine project
- Science 2022 – CICERO Diplomacy agent
- Generative Agents – arXiv
- Living Reviews in Relativity
- BMJ – living systematic reviews
- OECD – Rules as Code primer
- Open Policy Agent – policy-as-code
- Planetary boundaries update
- Doughnut Economics – design tool
# Concept Map Deliberative Democracy · Argumentation · Policy as Code · Design Justice · Planetary Boundaries · Language Game · Family Resemblance · Living Documents · Theory of Mind · Gestalt
# See - The Sudden a Hitchhiker Story