A Systems Theory for Today

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

A Systems Theory for Today

A living, open, collaboratively-built attempt to construct a systems theory adequate to the world of the 2020s — by new methods, because the old ones no longer reach.

Founder / author: Hulki Okan Tabak — with Claude Status: a research program — three untested seed hypotheses, one borrowed falsifiable chain, and no data yet. (Honest stage, named at the author’s S4k ratifying contact — decision 1. The evocative title stays; this line sits above it so nothing over-claims. Build history: Foundation S1; field surveyed + pressure-tests expanded + shuttle designed S2; living-document + Theory A operationalized S3; Theories B & C operationalized S4; the empirical turn, its instruments, and three delegated sessions S4d–4j; the author closed the build at S4k — panel/SESSION_S4K_RATIFICATION.md.) License: docs under CC BY-SA 4.0, any code under MIT (see LICENSE.md)

Read it online → hulkiokantabak.github.io/systems-theory-today · the reading site is the project’s artifact (S4k decision 6); this README is the maintainer’s front page behind it. (The site publishes when the repo is pushed and Pages is enabled; the contribution repository was opened to the public at Session 4k — the founding wager enacted, the attempt handed onward — see docs/REPO_OPEN_CHECKLIST.md.)

The question

Since Heidegger, no single thinker has produced a total explanatory system — a philosophy that tries to account for everything: nature, history, society, mind, meaning. For most of human history we had such systems (theological cosmologies, then the great rational systems of the Enlightenment and German Idealism). Then, roughly a century ago, the ambition fractured.

This project starts from a hypothesis:

The pace of change is now so fast, the relevant fields so numerous, and the interdependencies so dense that a total systems theory can no longer be built by the old method — a single mind writing a single book. But perhaps it can be built by a new method: living, distributed, plural, computationally-augmented, and continuously revised.

We test that hypothesis by trying to do it.

We use today’s symptoms as the pressure-test. If a systems theory is worth anything, it should illuminate:

  1. Why are people not giving birth?
  2. Why has the dopamine rush become so central?
  3. Why is “post-truth” insufficient to describe the level of propaganda?
  4. Why is what is happening today happening at all?
  5. Why are we destroying the climate?
  6. Why can we no longer coordinate and communicate?
  7. Why are populism and authoritarianism rising?

And underneath all of them: what explains the apparent loss of shared ideology, belief, and coherent understanding?

(These seven are the project’s origin and are preserved as “the founding pressure-tests.” In Session 2 the panel re-debated them, admitted five more — machine intelligence, the wealth pump/inequality, institutional decay, generalized ecological overshoot, and anomie — and restructured the flat list into four coupled layers plus a master condition, presented as a contested causal hypothesis. See docs/GOALS.md and panel/SESSION_2_PRESSURE_TESTS.md; the map is docs/DIAGRAMS.md §4.)

What this is (and is not)

  • It is a research program run as designed disagreement: a standing panel of ten thinkers, backed by a twenty-two-person advisory bench, argue, rate, vote things down, and dissent — and a chair (Claude) synthesizes without voting, while the author ratifies.
  • It is a living document. After publication it is meant to be forked, cloned, contradicted, and extended. The website is a front door, not a fence.
  • It is not a claim to have found the answer. Our own candidate theories (see outputs/CANDIDATE_THEORIES.md) are seeds, explicitly provisional, offered for others to break.
  • It is not neutral-to-the-point-of-emptiness. Panelists take sides. The project has opinions. It also has a discipline for catching and correcting its own mistakes (logs/).

Read in this order

# File What it gives you
1 ARCHITECTURE.md The map of the whole project and how the documents fit together
2 docs/HISTORY_OF_SYSTEMS_THEORIES.md The intellectual spine: every previous total system, who built it, when, and how it fell
3 docs/LANDSCAPE_OF_CONTEMPORARY_SYSTEMS_THEORIES.md What’s out there now — cliodynamics, complexity science, Luhmann, polycrisis/metacrisis, Big History, metamodernism, integral theory — what each reaches, can’t, and the gap we aim at
4 docs/GOALS.md What we are trying to do — and the expanded, layered pressure-tests
5 panel/PANEL_ROSTER.md The ten-person core panel + twenty-two advisors, and the fights they will have
6 panel/SEVEN_ROUND_DISCUSSION.md The panel’s founding seven-round deliberation, enacted
7 panel/SESSION_2_PRESSURE_TESTS.md The panel re-debating and expanding the pressure-tests (7→13, layered), with dissent
8 outputs/INITIAL_EVALUATION.md The panel’s preliminary hypotheses
9 outputs/COMPREHENSIVE_PLAN.md How we intend to build the theory, step by step, toward code
10 outputs/CANDIDATE_THEORIES.md Our own seeded systems theories (A, B, C)
11 outputs/THEORY_{A,B,C}_OPERATIONALIZED.md Each theory turned into a measurable, falsifiable claim with a pre-registered first study — A (the adaptation gap), B (optimization intensity; the cheapest to test), C (the reflexive ablation, turned on the project itself)
12 logs/OPEN_QUESTIONS.md Every live disagreement and open question — the most revealing part of the record; where to start if you want to contribute
13 docs/DIAGRAMS.md The project’s logic and flows, drawn (renders on GitHub): governance, learning loop, the pressure-test map, the field positioned, the path-to-Code decision tree, the Chat/Code shuttle, and Theory A as a signed causal loop
14 docs/CHAT_CODE_WORKFLOW.md How the project runs across two tools once it outgrows a single context: Chat deliberates, Code executes, the repo remembers — registers + handoff templates
15 docs/THE_LIVING_DOCUMENT.md What makes this a living document rather than a static one — its metabolism, honest test, and failure modes
16 logs/REFLECTIONS.md The project’s honest retrospective on its own work — what’s working, what’s fragile, what’s quietly wrong
17 logs/handoffs/COOPERATION_LOG.md The append-only ledger of the Chat↔Code collaboration — who did what, what crossed the boundary
18 viz/systems-theory-map.html · viz/diagrams.html Interactive maps of the landscape/panel and of the logic diagrams

Supporting: docs/GROUND_RULES.md, docs/METHOD.md, docs/METRICS.md, docs/SKILLS.md, docs/VISUALIZATION.md, docs/GLOSSARY.md, logs/CATCHES.md, logs/LEARNINGS.md, SKILL.md, CONTRIBUTING.md.

The reflexive wager

The form of this project is an argument for its content. If no single mind can hold the whole, then the way to think the whole is a commons of sense-making — many perspectives, in structured disagreement, integrated and continuously revised. This repository is a small prototype of exactly the method our third candidate theory proposes. The medium is the thesis. But the wager is, so far, asserted — not demonstrated. The “many perspectives” are, for now, one model in many roles (a single reasoner prompted into a panel), so the plurality is downgraded to one standpoint until a genuinely foreign, project-blind grader — or an opened repository — earns the word back; the medium can only become the thesis when a real outside vantage sits down (logs/OPEN_QUESTIONS.md Q-016; logs/REFLECTIONS.md, Session 4f).

Building & running this repository (the Code layer)

Added in Session 4, when the project stood up “Light Code” — a durable repository and a public reading site — while keeping deliberation in Chat. How that works is docs/CHAT_CODE_WORKFLOW.md: Chat is the mind, the repo is the memory, Code is the hands.

  • Start at the cockpit. dashboard/index.html is the repository control room — canonical sources, generated outputs, the key commands, live validation status, and the guardrails, each linked to its evidence. It is generated from live repo measurement (npm run dashboard; don’t hand-edit it). Open it before changing code. It is a maintainer surface — not published on the public site.
  • Two surfaces (docs/THE_LIVING_DOCUMENT.md §7). Both are now public: the repository is the contribution surface — opened at Session 4k, free to fork, contradict, and extend — and the website is the reading surface. This directory tree is the source of both.
  • The reading site is an Eleventy build over the repo’s own Markdown (site/README.md): npm install then npm run build (or npm run serve). It renders every doc, embeds the nine figures (viz/diagrams.html), and deploys to GitHub Pages via .github/workflows/pages.yml when pushed. The repository is the site’s source.
  • Analytics & install. The public site carries privacy-respecting, cookieless GoatCounter page-view counting (the author’s shared hulkiokantabak.goatcounter.com dashboard — no cookies, no IP logging, no personal data), and an Apple / Android home-screen icon + web-app manifest (site/assets/) so it installs cleanly to a phone home screen.
  • The standing checks (scripts/README.md) run before every commit — npm run check (or node scripts/run-checks.mjs): reference-integrity, count-reconciliation against docs/METRICS.md, a Mermaid-parse + diagrams-render check, and a structural-change → ripple helper. They have zero dependencies.
  • The shuttle is versioned under logs/handoffs/: the work-order (WORK_ORDER_S4.md), the append-only cooperation log (COOPERATION_LOG.md), the session digest, and R3 (R3_GIST.md) — the standing one-paragraph compression of the whole, loaded by every Chat session.
  • The studies (studies/) are gated scaffolds — pre-registrations + data/analysis skeletons for the three pre-registered studies (B, then A, then C). They do not run; running any of them crosses into heavy Code and needs a fresh, author-ratified work-order.