Lakatos

Philosophy

Criticism should change testable predictions, and programmes should be compared through new predictions and settled outcomes.

Architecture & control flow

1. 01

   Generate provenance-bearing candidates and apply PG discrimination checks

2. 02

   Create paired settleable bets for competing explanations

3. 03

   Turn anomalies into repair obligations and new questions

4. 04

   Use L1/L2/L3 repairs on lemmas, definitions or programmes

5. 05

   Compare programme progress and reallocate budgets

Architecture outline derived from this version’s control flow.

What this version changes

Connects refutation, anomalies, agendas and programmes while limiting template self-promotion.

Inputs & outputs

Inputs

Candidate topics, provenance records, discriminating experiments and anomalies.

Outputs

Provenance gates, paired bets, anomaly ledgers, question lists, programme states and budget allocations.

Implementation & evidence scope

Default candidates and some anomalies/settlements remain synthetic smoke cases; provenance fields or two copies do not establish independent external generation. Its strength is protocol structure and conservative downgrading.

Code & bundled material

The introduction draws on bundled notes, changelogs and central code. Software tests, synthetic diagnostics and scientific effectiveness use different evidence standards.