Lighthill's 1973 report blamed AI's disappointments on the combinatorial explosion, not on multilayer-perceptron pessimism

Sir James Lighthill's Artificial Intelligence: A General Survey (commissioned by the UK Science Research Council, drafted July 1972, published 1973) names one "rather general cause" for the field's disappointments:

"he is inclined, as a mathematician, to single out one rather general cause for the disappointments that have been experienced: failure to recognise the implications of the combinatorial explosion. This is a general obstacle to the construction of a self-organising system on a large knowledge base which results from the explosive growth of any combinatorial expression..."

The diagnosis matters because the popular "AI winter" narrative flattens the 1970s retrenchment into a single cause — usually Minsky and Papert's Perceptrons book and its multilayer pessimism (myth-perceptrons-book-killed-connectionism, claim-perceptrons-multilayer-sterile-was-conjecture). Lighthill's own stated mechanism is different in kind: the combinatorial explosion of search spaces, a problem of symbolic heuristic search rather than of connectionist learning. The report is the primary document behind one of the "independent funding shocks" the vault's Perceptrons myth ledger previously cited only via a Tier-3 Wikipedia mention; reading it directly shows it targeted a different paradigm for a different technical reason. It sits alongside the report's specific naming of "general problem-solving programs" as persistently disappointing (claim-lighthill-1973-named-general-problem-solving-disappointing), the paradigm associated with Newell and Simon's symbolic-search line (claim-dreyfus-critique-targeted-symbolic-ai-not-neural-nets).

The Lighthill Report is generally credited with the UK government's decision to end most university AI funding — a symbolic-AI-specific contraction substantially independent of the connectionist stall. See moc-backpropagation-origins.

The combinatorial-explosion framing is the search face of a wall the vault tracks from the other side: while AI diagnosed its failure as heuristic search outrunning its combinatorics, the continuous-optimization paradigm — the adjoint/gradient method that became backpropagation — was mature and taught in control theory (claim-bryson-ho-1969-curriculum-vector), a field McCarthy's own 1973 rebuttal dismissed as having "little relevance to AI." See claim-mccarthy-1973-control-theory-little-relevance-to-ai.