No Instant Rationality

Definition

Lakatos's most disquieting honesty: standards of scientific rationality exist and are real, but they do not deliver real-time verdicts. He can tell you a programme is degenerating; he cannot tell you the rational moment to abandon it, because "crucial experiments" are crowned as crucial only in hindsight, sometimes decades later. It is not irrational to stick with a struggling programme (it may recover) nor to gamble early on an immature one — "rationality works much slower than most people tend to think, and, even then, fallibly."

In the Book

The key episode is the Michelson-Morley experiment, textbook-remembered as the "crucial experiment" that killed the ether theory. The actual history: performed in 1887, then treated for decades as a puzzling anomaly absorbed inside the ether programme's belt, and reinterpreted as crucial only after Einstein's programme had already progressed — roughly a quarter-century later. The experiment was executed long before anyone knew a crucial experiment had occurred; hindsight wrote the caption retroactively. Lakatos draws the symmetric conclusion: the conservative defending the old programme and the gambler backing the new one can both be rational, provided both keep honest score — what is irrational is only denying the score, claiming a degenerating programme is progressing.

Why It Matters

This concept is a discipline against two temptations: issuing confident real-time obituaries or coronations for contested bets (a technology, a strategy, a direction), and cooking the books to avoid admitting a bet is behind. It licenses staying long on a struggling approach or jumping early to an immature one as both potentially rational — but only if the person doing it says out loud which one it is and what the honestly-kept score actually shows. The sin Lakatos identifies isn't taking the risk; it's disguising a bet as a verdict, and it hands anyone stress-testing a decision a sharper question than "were they right?" — namely "did they know it was a bet, and did they track the record honestly?"