A Systems Theory for Today

A living, forkable attempt to build a systems theory adequate to the present — by designed disagreement.

Contributing

Version: 0.2 · Status: Draft (author to ratify) · Last updated: Session 2

This is a living commons, not a finished book. It exists to be forked, strengthened, broken, and argued with. Theory C — the claim that a systems theory for today must be held plurally rather than authored alone — is only true if people actually contribute. So: welcome. Here is how.

As of Session 4k the repository is public — this is live, not aspirational. The author closed his build (a research program with strong hygiene and no data yet — the site’s self-audit page has the honest scoreboard) and opened the repo to hand the attempt onward, knowing a community may not come. If you fork it, you are the genuinely-foreign vantage the closed build conceded it could not supply itself. Begin at logs/OPEN_QUESTIONS.md.

First principles

  1. Disagreement is the product, not the problem. The most valuable contribution is often a sharp, well-argued objection — not a polite extension. If you think the whole project is misconceived, the best possible thing you can do is say so, precisely, in a way we have to answer.
  2. Dissent is never erased. No contribution is ever merged by silently deleting a live objection. If your argument survives scrutiny, it is preserved even if it is not adopted. If you dissent from a decision, your dissent is recorded at the foot of the relevant document, with your reasoning.
  3. Provisional, always. Nothing here is final — not the theories, not the definitions, not the roster, not these rules. Everything carries a version and a status; everything can be revised.
  4. Recruit for the fault line, not the fame. New voices (panelists, theories, objections) are admitted for the novelty of the disagreement they add, not for prestige.

The three kinds of contribution

1. Strengthen a theory. Take Theory A, B, or C (see outputs/CANDIDATE_THEORIES.md) and make it more precise, more falsifiable, better grounded, or better argued. Especially wanted: turning an evocative claim into a checkable one. (The single highest-value task right now is operationalizing Theory A’s “gap” as two measurable rates — see outputs/COMPREHENSIVE_PLAN.md, Phase 1.)

2. Break a theory. Try to falsify it. Find the counterexample, the smuggled value-judgment, the unfalsifiable move, the case where its prediction fails. A theory that survives a real attack is stronger; one that doesn’t should be revised or retired. Red-teaming is a first-class contribution, not an attack on the project.

3. Propose a rival. Seed your own candidate theory (Theory D, E, …). Use the anatomy every candidate follows (claim / master variable / exposition / the pressure-tests — the expanded, layered set in docs/GOALS.md and panel/SESSION_2_PRESSURE_TESTS.md, drawing the cross-layer arrows / falsifiable commitments / lineage / known weaknesses stated in your own text / how to fork it). Rivals are not renumbered once public; they stand alongside the others, and the commons decides — over time, by argument and evidence — how they fare.

How to contribute (mechanics)

Until the public repository’s issue/PR templates are live (Phase 2), the shape is:

  • Fork the repository.
  • Make your change in the relevant file, preserving its version/status header and bumping the version.
  • If your change touches a decision (adopting, revising, or rejecting a claim), add a short note in the relevant document’s change log and, if you are dissenting from an existing decision, add your dissent — with reasoning — at the foot of that document.
  • Open a pull request describing: which of the three kinds of contribution this is, what fault line or evidence it adds, and what (if anything) it asks the project to change its mind about.

Contributions are reviewed by the panel process (designed disagreement, steelman-first), synthesized by the chair, and ratified by the author. The review preserves your argument whether or not it is adopted.

Proposing a new panelist

The roster (panel/PANEL_ROSTER.md) is explicitly extensible. To propose a new standing or bench member, submit a one-paragraph case for the fault line they add — which of the five axes they sharpen, or which new axis they introduce, and specifically who on the current roster they would productively fight with. Novelty of disagreement is the admission criterion. “They are famous and smart” is not a case; “they would split the room on axis 4 in a way no current member does” is.

A note on how panelists work: members are reconstructions of real thinkers’ published positions, framed as intellectual positions, not the actual persons. Where a living figure’s current, real-world view is relevant, cite it — do not put invented words in a living person’s mouth. Contributions must respect this line.

Quality and conduct

  • Steelman before you strike. State the position you are attacking at its strongest before you attack it.
  • Cut before you add. A shorter, truer claim beats a longer, hedged one. Subtraction is welcome.
  • Ground empirical claims. Cite sources for facts, especially about the contemporary landscape (which goes stale fast). Don’t invent attributions; if unsure, omit.
  • Argue the idea, not the person. The panel fights hard over positions and treats contributors with respect. Bad-faith participation, harassment, or misrepresentation of others’ views are the only things that get a contribution rejected on non-intellectual grounds.
  • Do no harm (see docs/GROUND_RULES.md, §4). This project studies dangerous dynamics — propaganda, addiction, collapse — in order to understand them. Contributions that operationalize that understanding into tools for harm are out of scope.

License

By contributing, you agree your contribution is licensed under the same terms as the project: documents under CC BY-SA 4.0, code under MIT (see LICENSE.md). Share-alike means your improvements stay part of the commons, and the commons stays open. You are credited for your contributions.

The test of whether this project is what it claims to be is whether the commons actually forms — whether real rivals and real objections arrive, and whether the process catches what a lone author would miss. That test only runs if you take part. Fork it, break it, better it.