Methodology & Epistemology
Postulates, not prophecies
This page is the test lab. The founding document explains why the discipline exists. Here we show how to stress-test a law without inventing dated event forecasts.
For the full argument, read the founding document. This page stays on method. Founding document →
I do not claim to predict events. I claim to identify the constraints that make some outcomes far more probable than others.
Postulates vs universal truths
We describe how systems tend to behave under stated conditions. When a postulate fails in the field, we document the falsifier and tighten the scope. We do not defend dogma.
Constraints vs event predictions
Law 15 is the spine: understand the system well enough to name probable outcomes — not a calendar of press releases from the future. Investors can use that. Prophets get embarrassed.
Descriptive vs normative layers
The 16 postulates describe how economic systems behave. The 7 Laws on /about prescribe how to build a healthy enterprise. Law 16 links both. Mixing them without naming the difference collapses back into opinion.
Falsifiability protocol
Each law page carries signals and kill-switches. A postulate without a falsifier is a story. Stories belong to Level 1. Laws need a way to die.
- Define scope (civilization → individual).
- List measurable signals an operator can track.
- State what observation would kill the postulate.
- Link Level 1 observations as empirical anchors.
- Use the result for a decision — or revise the postulate.
Worked example — Law of Survival on a PME cohort
1. Hypothesis
Companies do not die because a technology appears. They die when their system stops functioning. AI can be the trigger. It is not the cause.
2. Cohort
Take 200 French SMEs (€2–20M revenue) that bought an “AI transformation” package in the last 24 months. Split by founder-absence test pass/fail and Sellability Index band.
3. Metric
Primary: 12-month survival of core operating metrics without founder heroics. Secondary: transferability score delta. Not “AI seats deployed.”
4. Falsifier
If high-dependency firms systematically outperform low-dependency firms after the same AI stack, the survival law is wrong for this scope — or the stack is not what we think.
5. Decision
Capital: fund system capital before automation spend. Operators: refuse AI projects that skip process and delegation. If falsifier hits, rewrite the law — do not defend the deck.
Present — Past — Future
Editorial method used across insights. It keeps noise (Level 1 chatter) from pretending to be law.
THE PRESENT — what is actually happening on the board right now. Numbers, constraints, games being played.
THE PAST — the structure that produced this present. Cycles, selection, persistence. Not nostalgia.
THE FUTURE — probable outcomes under constraints. Not dated prophecies. Direction and types of winners.
Three corpora compared
| Aspect | 16 Postulates | 7 Normative Laws | Observations |
|---|---|---|---|
| Nature | Descriptive — how systems behave | Normative — how to build a healthy system | Empirical — field evidence |
| Level | Macro / meso / systemic dynamics | Micro / enterprise / autonomy | Concrete cases |
| Status | Testable postulates | Construction principles | Facts and analysis |
| Test | Falsifiable on cohorts, markets, cycles | Testable via BE Fit / autonomy metrics | — |
Normative laws live on /about. Law 16 links both layers. Measure health on the operational framework. Country proxies: System Index (illustrative while the pipeline expands).
Test protocol
- Identify which postulate applies to your decision context.
- Check signals — do they line up with the mechanism?
- Hunt falsifier conditions before committing capital.
- Cross-check Level 1 observations for empirical grounding.
- Decide — or revise the postulate. No theatre.