Positive Heuristic

Definition

The positive heuristic is a research programme's partially articulated plan for how to develop its models over time — "a partially articulated set of suggestions or hints on how to change, develop the 'refutable variants' of the research programme." It tells researchers which idealization to lift next, largely independent of which anomalies happen to be loudest at the moment, because the plan already anticipates that each stage will be refuted before it calculates the next one.

In the Book

The showpiece is Newton's planned sequence of solar-system models: first a point-sun and a single point-planet; then multiple planets with mutual perturbations; then planets as extended balls; then spinning, bulging spheroids; then wobbling axes and interplanetary tides. Newton knew each stage was false the moment he calculated it — the failures of the crude model were foreseen and already queued as the next stage's problem. Lakatos argues this explains something Kuhn's "puzzle-solving" leaves unexplained: not why scientists in a paradigm keep working, but why they choose these problems in this order — "the relative autonomy of theoretical science," where the programme's own plan, not nature's complaints, drives the sequence of papers.

Why It Matters

This is the difference between a roadmap and a backlog. A backlog reacts to whatever surfaces; a positive heuristic embodies a planned relaxation of assumptions — v1 assumes one region, one currency, benign users; v2 lifts one assumption; v3 the next — so that most incoming complaints can be ignored rationally, because the plan already schedules them for a later stage. It gives a diagnostic for organizational drift: if every anomaly reorders your priorities, you don't have a programme, you have a complaint queue: the heuristic has either never existed or has run out, and Lakatos concedes that programmes do eventually exhaust theirs and drift into reactive mode.