Part II · The Evolution Interface Factor 4 of 12
Criticism Hardens Conjecture into Knowledge
5 min read
Principle

A conjecture becomes knowledge only by surviving serious attempts to kill it.

In brief
  • Conjectures are guesses. They earn the status of knowledge only by withstanding severe attempts at refutation.
  • The quality mark of an explanation is that it is hard to vary — every part does load-bearing work, so it cannot be quietly patched to fit any outcome.
  • Criticism is the system’s quality regulator: without it, unfalsifiable and endlessly patched explanations silently fill the knowledge base with noise.

#The claim

A conjecture is not yet knowledge. However elegant the guess, it earns that status only by surviving criticism. This is Popper’s engine: knowledge advances through bold conjectures subjected to severe attempts at refutation — we attack our best guesses as hard as we can, and what withstands is what we provisionally keep.Popper (1963), Conjectures and Refutations. Theories can “die in our stead” — criticism as selection applied to ideas, not people. The word provisionally is doing real work. Nothing that survives today is exempt tomorrow; survival is a status, not a certificate.

The criticism funnel Many conjectures enter a funnel of criticism; most are refuted, marked with crosses, and only a few survive at the bottom as knowledge. conjectures criticism knowledge
Figure 1. The funnel earns its narrowness: most conjectures die under criticism, and the few that pass through the bottom are what a system may provisionally call knowledge.

#The mechanism

What separates the survivors from the casualties? Deutsch’s criterion: a good explanation is hard to vary.Deutsch (2011), The Beginning of Infinity. Good explanations are hard to vary while still accounting for what they account for. Every part of it does load-bearing work, so you cannot quietly adjust a component to accommodate an unwelcome observation without the whole structure failing. Bad explanations are the opposite — easy to vary. Rescue one with an ad-hoc amendment and it survives anything, which means it explains everything, which means it explains nothing.

Hard to vary versus easy to vary Left panel: a rigid truss where every strut is load-bearing. Right panel: an amorphous dashed blob with arrows showing it reshaping to fit anything. hard to vary easy to vary every part load-bearing reshapes to fit anything
Figure 2. Deutsch's criterion. The truss cannot absorb an amendment without collapsing — so surviving criticism means something. The blob absorbs every objection — so its survival means nothing.

Verifiability is the gate on the funnel. A claim that forbids no observation cannot be corrected by any observation, and a system that admits unfalsifiable, endlessly patched explanations silently fills with noise — its knowledge base grows while its knowledge does not.

#Why it matters

Criticism is the regulator of quality for the entire framework. Exploration proposes; criticism disposes; and the loop of the two is the evolution interface — the process by which a civilization’s stock of explanations improves without anyone decreeing which are true. It keeps the body of knowledge honest, discarding what cannot be tested and hardening what can. And occasionally criticism does its most dramatic work: forcing a paradigm shift that does not merely replace an answer but reframes the questions themselves.Kuhn (1962), The Structure of Scientific Revolutions. Accumulated anomalies force a reframing of the questions, not just the answers.

The stakes recur at every scale of this framework. The institutions of Factor 11 are, at bottom, machinery for keeping this loop running at civilization scale — and the noise that unfalsifiable claims inject is exactly what the market of Factor 10 must pay verification costs to filter.

The factor, in full

A conjecture is not yet knowledge. It earns that status only by surviving criticism. Knowledge advances through bold conjectures subjected to severe refutation.1 We attack our best guesses as hard as we can, and what withstands is what we provisionally keep. The mark of a good explanation is that it is hard to vary.2 Every part does load-bearing work, so it cannot be quietly patched to fit any outcome. Bad explanations are easy to vary. Rescue them with an ad-hoc amendment and they survive anything, explaining everything and so nothing. Verifiability is the gate. A claim that forbids no observation cannot be corrected, and a system that admits unfalsifiable, endlessly-patched explanations silently fills with noise. Criticism is the regulator of quality. It keeps the body of knowledge honest, discarding what cannot be tested and hardening what can, and occasionally it forces a paradigm shift3 that reframes the questions themselves.

References

  1. Popper, K. R. 1963. Conjectures and Refutations: The Growth of Scientific Knowledge. London: Routledge.
  2. Deutsch, D. 2011. The Beginning of Infinity: Explanations That Transform the World. London: Allen Lane.
  3. Kuhn, T. S. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press.