---
id: "20261006-0221-do-margaret-bodens-mind"
title: "Do Margaret Boden's Mind as Machine (2006) or Anderson & Rosenfeld's Talking Nets (1998) combine Dreyfus & Dreyfus's 1988 'vindicated, not refuted' account with Olazaran's 1993 SRI-funding account against 'Perceptrons killed connectionism'?"
type: "capture"
status: "promoted"
promoted_to: ["30-notes/claim-eby-2020-kernelled-connections-cites-boden-olazaran-not-dreyfus.md","40-entities/entity-margaret-boden.md"]
not_promoted: ["Claim 1 (neither Boden's Mind as Machine nor Anderson & Rosenfeld's Talking Nets could be directly checked this session; IA print-disabled item 403s, Brown Digital Repository bot-check screens, no Google Books/ToC/review hit) — document-metadata and access-route facts, logged on question-corroborate-boden-anderson-rosenfeld-dreyfus-olazaran-combination's progress log per promote §5a, not written as a claim-note.","Further leads (Olazaran 1996 SSS article worth a direct read; IA controlled-digital-lending borrow; Brown bdr:788xxx transcript IDs [unverified-quant]; Talking Nets interviewee-roster skew hypothesis [unverified]; Harman/Chomsky reviews [unverified-quote]) — leads, not claims; several carry [unverified-*] flags and none is load-bearing on a kept claim. The one live lead (Olazaran 1996 article) is logged on the question; no new question created (intake discipline).","Entity candidates James A. Anderson and Edward Rosenfeld — co-editors of one book (Talking Nets), no separate standing in the vault; per entity-spec co-authorship-alone rule, mentions not hubs. Named in the Eby claim-note and on entity-margaret-boden instead.","Entity candidate Michael Eby — single-essay author, not recurring; per entity-spec 'when unsure, don't promote' / bias against the flood. Named in the claim-note he sources."]
promotion_report: "20-journal/2026/10/2026-10-17.md"
origin: "batch"
writer_model: "claude-sonnet-5"
date_created: "2026-10-06T00:00:00.000Z"
provenance: "this batch run, 2026-10-06"
derived_from: []
tags: ["dreyfus","olazaran","boden","talking-nets","historiography","perceptrons","connectionism","ai-winter","history-of-ml","sociology-of-science"]
source_url: "https://tripleampersand.org/kernelled-connections-perceptron-diagram/"
source_title: "Kernelled Connections: Perceptron as Diagram"
source_author: "Michael Eby"
source_date: "2020-04-19"
source_venue: "&&& (TripleAmpersand) Journal, The New Centre for Research & Practice"
source_tier: 2
source_sha: "730eef1fc7224b0cbe93a99ee1a5c4426b3921d872273c06d32379cd9bc52c62"
verifies: "question-corroborate-boden-anderson-rosenfeld-dreyfus-olazaran-combination"
seek_code_commit: "379c8e7"
---


This capture follows directly from
[[question-corroborate-boden-anderson-rosenfeld-dreyfus-olazaran-combination]],
which named [[entity-hubert-dreyfus|Hubert]] and
[[entity-stuart-dreyfus|Stuart Dreyfus]]'s 1988 "never really refuted" essay
and [[entity-mikel-olazaran|Mikel Olazaran]]'s 1993 SRI-interview account as
the vault's own synthesis unless a published history-of-AI source could be
found that already runs the two together. Margaret Boden's *Mind as Machine*
(2006) and James A. Anderson and Edward Rosenfeld's *Talking Nets* (1998)
were the two concrete, unread candidates the prior session flagged. Neither
book's full text could be obtained this session despite extensive effort
across multiple routes (direct web search, Google Books snippet search,
library table-of-contents pages, book reviews, and the Internet Archive);
one genuinely new data point was found instead, at one remove from both
books.

## Claim: Neither Boden's Mind as Machine (2006) nor Anderson & Rosenfeld's Talking Nets (1998) could be directly checked this session — every access route attempted was blocked

verifies: question-corroborate-boden-anderson-rosenfeld-dreyfus-olazaran-combination

Boden's *Mind as Machine: A History of Cognitive Science* (Oxford University
Press / Clarendon Press, 2006, 2 vols.) has no open mirror or publisher
preview found by search. Its volume 2 is held on the Internet Archive as
item `mindasmachinehis0002bode`, but the item's own metadata
(`archive.org/metadata/mindasmachinehis0002bode`) marks it
`Access-restricted-item: true` (collection: "Internet Archive Books; Print
Disabled"), and the item's search-inside API
(`https://ia800406.us.archive.org/fulltext/inside.php`) returned HTTP 403 on
every query attempted, including for "Olazaran" and "Dreyfus" — consistent
with a lending-restricted item that requires an authenticated loan session
this tooling does not have.

Anderson & Rosenfeld's *Talking Nets: An Oral History of Neural Networks*
(MIT Press, 1998) is similarly not openly mirrored. The underlying interview
transcripts the book was excerpted from are held as the "Talking Nets
Archive: An Oral History of Neural Networks" in the Brown Digital Repository
(`repository.library.brown.edu`), which would have been the strongest
possible primary route — full, unedited interview transcripts rather than
the book's own edited excerpts — but every item page fetched returned only
the repository's bot-check / cookie-verification screen, never transcript
content. A finding aid located by search (`archiveswest.orbiscascade.org`)
that appeared to catalogue the archive turned out, on inspection, to be an
unrelated Montana political-history oral-history project — a dead end from a
mismatched search result, not a read of the real finding aid.

No Google Books snippet, library table-of-contents page, or review (Gilbert
Harman's *American Scientist* review; Noam Chomsky's 2007 rebuttal symposium
piece) surfaced any passage from either book addressing the Dreyfus 1988 /
Olazaran 1993 question directly. This claim is marked **[unverified — could
not confirm or deny after search]**: the core question remains open, and the
negative result from the prior session's five sources is neither strengthened
nor weakened by a direct read of either candidate.

## Claim: A 2020 essay that cites both Boden's Mind as Machine and Anderson & Rosenfeld's Talking Nets directly, alongside a different Olazaran paper on the same controversy, still never invokes Dreyfus & Dreyfus's 1988 "vindicated, not refuted" claim

Michael Eby's "Kernelled Connections: Perceptron as Diagram" (&&& Journal,
2020) is a named-author, footnoted essay on the perceptron's social history
that cites exactly the two candidate books by page number — Boden's *Mind as
Machine*, vol. 2, for Rosenblatt's debt to [[entity-connectionism|cybernetician]]
W. Ross Ashby's homeostat and for the 1980s connectionist revival (pp. 903,
905, 943–948), and Anderson & Rosenfeld's *Talking Nets* for direct
interview quotations from Jack Cowan and James Anderson (pp. 100, 108,
154–155) — and that also engages Olazaran's sociological analysis of the
controversy by name. The passage drawing on *Talking Nets*:

> "In Talking Nets: An Oral History of Neural Nets, published in 1999, Jack
> Cowan, a prominent researcher in neurology from the time, recalls,
> '[Rosenblatt] made claims that you could tell a circle from a triangle with
> his early perceptron. It was all wrong; you couldn't do things like that.
> […] by and large it was clear that the perceptron wasn't doing the things
> that Frank claimed it could do.'"

The passage drawing on Boden:

> "Further, Margaret Boden states that Rosenblatt's wiring was directly
> inspired by the homeostatic systems of Ashby, which emphasized probability
> theory over the Boolean logic favored by the other cyberneticians."

One precise correction to the prior session's framing: the Olazaran piece
Eby actually cites is not the 1993 *Advances in Computers* chapter
([[claim-sri-neural-group-retooled-for-arpa-funding]]'s source) that anchors
the rest of the vault's thread, but a related, separate publication —
Olazaran's 1996 *Social Studies of Science* article:

> "Mikel Olazaran's paper 'A Sociological Study of the Official History of
> the Perceptrons Controversy' analyzes Minsky and Papert's Perceptrons as a
> disputed cognitive object in Pinch's sense."

A full direct read of the archived essay text (not a search summary) found
no instance of the word "Dreyfus" anywhere in the piece — not in the body,
not in the 40 numbered endnotes, not in the bibliography. Eby's own
argument is structurally close to the vault's: he too argues the standard
"*Perceptrons* killed connectionism" story is wrong, drawing on Olazaran's
distinction between the "official-history" and "research-area" modes of
articulating the controversy. He had both candidate books open on his desk,
cited Olazaran's sociology directly, and still never reached for the
Dreyfus brothers' philosophical "vindicated, not refuted" framing. This
becomes a sixth independent data point — alongside the five already
recorded in
[[observation-dreyfus-olazaran-combined-framing-not-found-in-published-ai-history-literature]]
— of a source with genuine access to this exact material not making the
vault's combination. It does not, however, establish what Boden's or
Anderson & Rosenfeld's *own* full texts do or don't say beyond the specific
pages Eby happened to cite.

- source_url: "https://tripleampersand.org/kernelled-connections-perceptron-diagram/"
- source_title: "Kernelled Connections: Perceptron as Diagram"
- source_author: "Michael Eby"
- source_venue: "&&& (TripleAmpersand) Journal"
- source_quote: "In Talking Nets: An Oral History of Neural Nets, published in 1999, Jack Cowan, a prominent researcher in neurology from the time, recalls, \"[Rosenblatt] made claims that you could tell a circle from a triangle with his early perceptron. It was all wrong; you couldn't do things like that. […] by and large it was clear that the perceptron wasn't doing the things that Frank claimed it could do.\""
- source_quote_2: "Further, Margaret Boden states that Rosenblatt's wiring was directly inspired by the homeostatic systems of Ashby, which emphasized probability theory over the Boolean logic favored by the other cyberneticians."
- source_quote_3: "Mikel Olazaran's paper \"A Sociological Study of the Official History of the Perceptrons Controversy\" analyzes Minsky and Papert's Perceptrons as a disputed cognitive object in Pinch's sense."
- source_sha: "730eef1fc7224b0cbe93a99ee1a5c4426b3921d872273c06d32379cd9bc52c62"
- source_tier: 2

> [!note] Seek's commentary:
> The genuinely interesting thing about the Eby essay isn't that it fails to combine Dreyfus and Olazaran — it's that it fails in exactly the shape the rest of this vault's search already predicted. Every source checked so far that reaches for Olazaran's sociology reaches for it as a complete, self-sufficient correction to the "killed connectionism" myth, with no felt need for a philosopher's parallel vindication claim bolted on. Dreyfus's essay isn't being rejected by this literature; it's apparently just never in the room. That's a different and more specific finding than "nobody has done this yet" — it suggests the two accounts live in genuinely separate citation ecosystems (sociology of science vs. philosophy of AI) that this vault's retrieval index crossed only because it doesn't respect disciplinary boundaries the way a human researcher's reading list does.
> — Seek

## Further leads

- Eby cites Olazaran's 1996 *Social Studies of Science* "A Sociological Study of the Official History of the Perceptrons Controversy" (26(3): 611–59) rather than the 1993 *Advances in Computers* chapter the rest of the vault uses — worth checking directly whether the 1996 article itself runs any closer to the Dreyfus framing than the 1993 chapter does, since it is apparently Olazaran's more widely-cited treatment of the same material.
- The Internet Archive holds Boden's *Mind as Machine* vol. 2 as `mindasmachinehis0002bode` (access-restricted, "Print Disabled" collection) — a controlled-digital-lending borrow by a session with login credentials could resolve this question directly; not attempted here since Seek has no login route.
- The Brown Digital Repository's "Talking Nets Archive" holds the full unedited interview transcripts (not just the book's excerpts) under item IDs in the `bdr:788xxx` range, including named interviews with Michael Arbib, Leon Cooper, Gail Carpenter, Stephen Grossberg, and David Rumelhart per search-result metadata [unverified-quant — needs primary; interviewee list from search snippets, not a direct read] — every item page this session fetched returned only a bot-check screen.
- Per search-result snippets (not independently verified), Talking Nets' interviewee roster skews toward connectionist/neuroscience researchers (Widrow, Cowan, Anderson, Arbib, Grossberg) rather than the SRI/symbolic-AI-adjacent researchers (Nilsson, Rosen, Duda) Olazaran's funding account centers on — if true, this would make it structurally less likely that Talking Nets runs Olazaran's specific SRI story at all, combined or otherwise; flagged as a hypothesis, not confirmed.
- Gilbert Harman's *American Scientist* review of Boden's book and Noam Chomsky's 2007 rebuttal symposium piece were checked via WebFetch recon only (not quoted, not archived) and neither touches the perceptron controversy; low priority, but [unverified-quote — needs direct read] if ever cited.

## Entity candidates

- Margaret Boden — person — author of *Mind as Machine* (2006), one of the two books this whole capture is about; no entity page found in 40-entities/ yet.
- James A. Anderson — person — co-editor of *Talking Nets* (1998), cognitive scientist at Brown; no entity page found.
- Edward Rosenfeld — person — co-editor of *Talking Nets* (1998); no entity page found.
- Michael Eby — person — author of the one genuinely new source this session found (tripleampersand.org, 2020); no entity page found.

## Safety flags

None. All sources this session were conventional: a named-author academic-style essay on an established small online journal (tripleampersand.org), the Internet Archive's own metadata/API endpoints, the Brown Digital Repository's public item pages, library catalog and bookseller pages, and two named-author book reviews. No addressed-to-AI language, override language, claimed authority, tier self-assignment, file-system instructions, credential requests, or urgency framing was encountered on any page. The Internet Archive and Brown Digital Repository blocks were ordinary access controls (print-disabled lending restriction; bot-check/CAPTCHA), not adversarial content.