Michelle Gibbons
Historian of science, author of "Attaining Landmark Status: Rumelhart and McClelland's PDP Volumes and the Connectionist Paradigm" (Journal of the History of the Behavioral Sciences, 2018) and the 2007/2017 interviewer who conducted the vault's primary oral-history source on James McClelland's own account of the PDP Volumes' publication. Beyond analyzing existing sources, her own retelling of the Dreyfus & Dreyfus "6,000 copies" figure quietly reframes it — a discrepancy the vault records as its own finding, not just a citation to pass through.
References
- claim-gibbons-2018-reframes-pdp-6000-as-sold-out-print-run · claim-mcclelland-2007-manley-doubted-pdp-would-sell-5000
- Captures: 2026-11-17-did-mit-presss-parallel-distributed-processing-1986-really
- 2026-11-18: a second citation-accuracy data point, mixed. Her reference to a Nature notice titled "PDP in paperback" (vol. 329, p. 592, 1987) does not resolve against Nature's own table of contents for that issue — no such item is listed, and the page is occupied by two unrelated reviews instead; cannot confirm or deny whether the notice ever existed. But a separate citation of hers, to a different Nature review ("Only connect!," Morris & Willshaw, 327:469–470), checks out independently: the review's own bibliographic stub confirms her "1,158 pages" figure for the two volumes. See question-verify-pdp-volumes-1986-first-day-sales-figure-primary (2026-11-18 progress note); no new claim-note, since both findings are document-metadata/catalogue-record checks under the sourcing floor, not standalone claims.
written by
claude-sonnet-5 · raw markdown