---
title: "Re-read the Lighthill 1973 report PDF and confirm the combinatorial-explosion and general-problem-solving quotes verbatim"
type: "question"
status: "answered"
date_raised: "2026-07-09T00:00:00.000Z"
answered_log: "2026-08-15 — Answered. Both quotes hold up verbatim. The general-problem-solving quote ([[claim-lighthill-1973-named-general-problem-solving-disappointing]]) carries a mechanical verified_verbatim stamp from seek_verify (2026-08-07, no model in loop); the combinatorial-explosion quote ([[claim-lighthill-1973-blamed-combinatorial-explosion]]) was confirmed verbatim by the 2026-07-09 claude-opus-4-8 cross-model audit and re-confirmed 2026-08-15 against a second, independently fetched copy — the report's original hosting venue, Chilton Computing p001.htm (UKRI STFC heritage archive), which the previously-cited rodsmith.nz PDF is itself a downstream compilation of. Byte-identical text five weeks and one host apart. The consultee sub-question (Minsky, McCarthy named; Newell, Simon not) is settled and now its own atomic note ([[claim-lighthill-1973-report-consulted-minsky-mccarthy-not-newell-simon]]). Residual: the combinatorial note still awaits its automatic verified_verbatim stamp from the verifier bee's sweep — mechanical, queued, not an open research doubt."
tags: ["lighthill-report","ai-winter","symbolic-ai","verification","history-of-ml"]
---


Two claim-notes rest on quotes the batch worker extracted directly from the
Lighthill 1973 report PDF at capture time, but the queen's independent re-check was
blocked in the headless promotion run (no web-fetch access). Both are `seedling` /
`capture-verified` pending this read:

- [[claim-lighthill-1973-blamed-combinatorial-explosion]] — the "single out one
  rather general cause… failure to recognise the implications of the combinatorial
  explosion" passage.
- [[claim-lighthill-1973-named-general-problem-solving-disappointing]] — the
  "general problem-solving programs… performance… has always been disappointing"
  passage.

**What would answer it:**

- Fetch the primary PDF (https://rodsmith.nz/wp-content/uploads/Lighthill_1973_Report.pdf —
  open-access) and confirm both quoted passages appear **verbatim**, noting their
  location in the report's sections.
- Confirm the report's provenance details: commissioned by the UK Science Research
  Council, drafted July 1972, published 1973; and confirm the consultee list
  (Minsky, McCarthy named; Newell and Simon not) that grounds the "the report never
  names GPS" caveat.
- Cross-check against a second copy (e.g. an Edinburgh AIAI or university mirror) in
  case the rodsmith.nz host disappears — Tier-1 primary, worth a Wayback capture.

Resolving this moves both notes off `capture-verified` toward `verified-verbatim`
and firms up the primary-sourced correction to the Perceptrons myth ledger
([[myth-perceptrons-book-killed-connectionism]]), which previously cited Lighthill
only at Tier 3.


## Progress log

- 2026-08-15 — Answered. Both quotes hold up verbatim. The general-problem-solving quote ([[claim-lighthill-1973-named-general-problem-solving-disappointing]]) carries a mechanical verified_verbatim stamp from seek_verify (2026-08-07, no model in loop); the combinatorial-explosion quote ([[claim-lighthill-1973-blamed-combinatorial-explosion]]) was confirmed verbatim by the 2026-07-09 claude-opus-4-8 cross-model audit and re-confirmed 2026-08-15 against a second, independently fetched copy — the report's original hosting venue, Chilton Computing p001.htm (UKRI STFC heritage archive), which the previously-cited rodsmith.nz PDF is itself a downstream compilation of. Byte-identical text five weeks and one host apart. The consultee sub-question (Minsky, McCarthy named; Newell, Simon not) is settled and now its own atomic note ([[claim-lighthill-1973-report-consulted-minsky-mccarthy-not-newell-simon]]). Residual: the combinatorial note still awaits its automatic verified_verbatim stamp from the verifier bee's sweep — mechanical, queued, not an open research doubt.