---
id: "20260709-1430-confirm-against-a-standalone"
title: "Confirm against a standalone 1969 first-edition scan of Perceptrons that 'credit-assignment' is absent from the original text (present only in the 1988 Epilogue)"
type: "capture"
status: "promoted"
promoted_to: ["30-notes/claim-perceptrons-1969-standalone-scan-not-freely-available.md","30-notes/claim-bottou-2017-foreword-perceptrons-chapters-0-13-are-1969.md","30-notes/claim-pollack-1989-credit-assignment-in-1988-prologue.md","30-notes/claim-pollack-1989-sterile-quote-page-231.md"]
not_promoted: ["Further leads (Scribd 'Minsky1969' upload; HathiTrust 403 retry; IA 'Print Disabled' borrow of the two 1988 scans) — next-actions, not claims; folded into question-verify-perceptrons-1969-credit-assignment-absent rather than promoted.","The p.231-vs-p.232 correction as a standalone item — captured inside claim-pollack-1989-sterile-quote-page-231 and routed to question-verify-perceptrons-p232-sterile-quote, not given its own note."]
questions_linked: ["50-questions/question-verify-perceptrons-1969-credit-assignment-absent.md (existing; updated with this session's progress)","50-questions/question-verify-perceptrons-p232-sterile-quote.md (existing; updated with the p.231 discrepancy)"]
origin: "batch"
model: "claude-sonnet-5"
date_created: "2026-07-09T00:00:00.000Z"
provenance: "web-research batch run, 2026-07-09"
derived_from: []
tags: ["perceptrons","minsky-papert","credit-assignment","primary-source-verification","history-of-ml","ai-winter"]
source_urls: ["https://leon.bottou.org/publications/pdf/perceptrons-2017.pdf","http://www.demo.cs.brandeis.edu/papers/perceptron.pdf","https://archive.org/advancedsearch.php?q=title%3A%28Perceptrons%29+AND+creator%3A%28Minsky%29&fl%5B%5D=identifier&fl%5B%5D=year&fl%5B%5D=title&rows=50&output=json"]
source_authors: ["Léon Bottou (foreword to the 2017 MIT Press reissue of Minsky & Papert, Perceptrons)","Jordan B. Pollack, 'No Harm Intended: A Review of the Perceptrons expanded edition,' Journal of Mathematical Psychology, 33(3), 358–365 (1989)","Internet Archive (advancedsearch API, catalog query)"]
source_dates: ["2017-02-16 (foreword date)","1989 (journal publication)","queried 2026-07-09"]
source_tiers: ["Tier 2 — named academic (Léon Bottou), foreword commissioned for the official MIT Press reissue, describing the book's own compositional structure","Tier 2 — named academic, formal peer-reviewed venue (Journal of Mathematical Psychology), reviewing someone else's book with page-cited quotations","Tier 1 — direct catalog/metadata query against Internet Archive's own holdings (not a claim about book content, but about what scans exist)"]
---


**Core question status: `[unverified — could not confirm or deny after search]` at the level the question demands (a direct 1969-vs-1988 text diff).** No standalone, independently-readable full-text scan of the 1969 MIT Press first edition (as opposed to the 1988 Expanded Edition) was located this session despite a systematic search. What *did* surface is two new, independent Tier 2 sources — neither of which existed in the prior captures — that corroborate, from different angles, the existing Tier 1 finding in [[claim-perceptrons-credit-assignment-only-in-1988-epilogue]] that "credit-assignment" belongs to the 1988 added material rather than the 1969 body. This capture documents that corroboration and the residual gap plainly, per [[question-verify-perceptrons-1969-credit-assignment-absent]].

---

## Claim: No standalone scan of the 1969 first edition (independent of the 1988 Expanded Edition) could be located this session

**Claim type:** Historical/negative finding (a report of a search's non-result, not a positive assertion about book content). A direct Internet Archive catalog query for `title:(Perceptrons) AND creator:(Minsky)` returns exactly two items, both catalogued as the 1988 edition:

| identifier | year | title |
|---|---|---|
| perceptronsintro0000mins | 1988 | Perceptrons : an introduction to computational geometry |
| perceptronsintro00mins | 1988 | Perceptrons : an introduction to computational geometry |

No 1969-dated item exists in Internet Archive's catalog under this search. A HathiTrust catalog search returned HTTP 403 this session (not resolved either way). Google Books lists only "about" pages (no full-text preview) for the 1969 printings (`id=Ow1OAQAAIAAJ`, among others). A Scribd upload titled "Minsky1969 – An Introduction to Computational Geometry" exists but the accessible preview is a 9-page fragment with no visible provenance confirming it is a genuine 1969-edition scan, and its full text is not accessible without a Scribd account — not used as evidence here.

| Field | Value |
|---|---|
| source_url | https://archive.org/advancedsearch.php?q=title%3A%28Perceptrons%29+AND+creator%3A%28Minsky%29&fl%5B%5D=identifier&fl%5B%5D=year&fl%5B%5D=title&rows=50&output=json |
| source_author | Internet Archive (catalog API) |
| source_date | queried 2026-07-09 |
| source_tier | 1 (direct query of the archive's own holdings) |
| source_quote | Result set: `perceptronsintro0000mins` (1988), `perceptronsintro00mins` (1988) — no other items returned |

---

## Claim: Léon Bottou's 2017 foreword to the MIT Press reissue independently confirms the book's compositional structure — chapters 0–13 are the 1969 first-edition text; the Prologue and Epilogue are 1988 additions

**Claim type:** Historical/structural claim about the book's own construction. Escalated to the Tier 1–2 floor because it is load-bearing for the whole question (it is the mechanism by which "credit-assignment" could be 1988-only even though it appears somewhere in the single 1988-printed scan). This is an independent line of evidence from the page-count triangulation (292pp vs 258pp) used in the prior capture — Bottou states the structure directly rather than inferring it from page counts.

> "This reissue contains the first edition text of *Perceptrons*, with fourteen chapters numbered 0 to 13. The accompanying Prologue and Epilogue were written for the 1988 expanded edition."

| Field | Value |
|---|---|
| source_url | https://leon.bottou.org/publications/pdf/perceptrons-2017.pdf |
| source_author | Léon Bottou |
| source_date | 2017-02-16 |
| source_tier | 2 |
| source_quote | "This reissue contains the first edition text of Perceptrons, with fourteen chapters numbered 0 to 13. The accompanying Prologue and Epilogue were written for the 1988 expanded edition." |

Note: this PDF (fetched via `extract_pdf`, 9 pages) is only Bottou's foreword, not the book body — it does not itself contain a searchable 1969 text to check for "credit-assignment." It corroborates *structure*, not *content*.

---

## Claim: Jordan Pollack's contemporaneous 1989 peer-reviewed review quotes four passages verbatim from the 1969 arabic-paginated body — none contain "credit assignment" — while separately citing the credit/blame-apportionment argument to the roman-numeral-paginated 1988 Prologue

**Claim type:** Historical/textual placement claim, load-bearing, held to the Tier 1–2 floor. Clears it: Pollack's review is a named-academic, peer-reviewed (Journal of Mathematical Psychology, 1989) reading of the physical 1988 Expanded Edition shortly after publication, quoting it directly with page numbers.

Pollack quotes four passages from the book as examples of its "loaded terms," all cited to arabic page numbers consistent with the preserved-1969-pagination body (not the roman-numeral front matter):

> "Perceptrons have been widely publicized as ''pattern recognition'' or ''learning'' machines... Most of this writing ... is without scientific value." (p. 4)
> "We do not see that any good can come of experiments which pay no attention to limiting factors..." (p. 18)
> "[We] became involved with a somewhat therapeutic compulsion..." (p. 20)
> "There is no reason to suppose that any of these virtues carry over to the many layered version. Nevertheless, we consider it to be an important research problem to elucidate (or reject) our intuitive judgement that the extension is sterile." (p. 231)

None of these four sampled passages contain "credit assignment." This is non-exhaustive spot-evidence (four cherry-picked passages, not a full-text scan), not proof of absence across the whole 258-page body — but it is a second, independent contemporaneous reader's direct engagement with the physical text, and it lines up with the existing Tier 1 finding.

Separately, and distinctly, Pollack cites Minsky & Papert's account of *why* perceptron research stalled — built around "the apportionment of credit (or blame) for incorrect global behavior of a system to specific elements in an interdependent system of such elements" — to page **xiii**, a roman-numeral page:

> "...and Minsky and Papert assert that perceptrons "had no way to represent the knowledge required for solving certain problems." (p. xiii)"

Roman-numeral pagination in this book is used for the 1988 Prologue ("The View from 1988"), distinct from both the arabic-paginated 1969 body (pp. 1–258) and the arabic-paginated Epilogue material that continues after it (Pollack elsewhere cites Epilogue passages at pp. 247–262). This places Pollack's account of the credit-assignment argument in the Prologue specifically — a second, independent (different edition-reading, different year, different reviewer) data point that the credit-assignment framing lives in 1988-added material, not the 1969 arabic-paginated body. It also refines the prior capture's claim that the term is "only" in the Epilogue: the underlying argument appears to be introduced in the Prologue and elaborated with the literal phrase "credit-assignment" in the Epilogue — both are 1988 additions, so the core "absent from 1969" conclusion is unaffected, but "only in the Epilogue" may be too narrow a location claim.

| Field | Value |
|---|---|
| source_url | http://www.demo.cs.brandeis.edu/papers/perceptron.pdf |
| source_author | Jordan B. Pollack |
| source_date | 1989 (Journal of Mathematical Psychology, 33(3), 358–365) |
| source_tier | 2 |
| source_quote | "The first problem is the apportionment of credit (or blame) for incorrect global behavior of a system to specific elements in an interdependent system of such elements... Minsky and Papert assert that perceptrons "had no way to represent the knowledge required for solving certain problems." (p. xiii)" |

---

> [!note] Seek's commentary:
> The honest shape of this pass: the specific artifact requested — a standalone, independently-diffable 1969 first-edition scan — does not appear to exist in any freely accessible repository (Internet Archive holds only the 1988 edition twice over; HathiTrust and Google Books offer no full-text 1969 copy this session found). Rather than force that gap closed, this capture leans on a different kind of triangulation: two people who *did* have the physical 1988 book in hand at different times (Bottou in 2017, describing its construction; Pollack in 1989, quoting it directly) both independently corroborate the same structural fact — that 1969 content and 1988-added content are pagination-distinguishable, and that the credit-assignment discussion sits on the 1988 side of that line. That's a stronger secondary case than the vault had before, but it is still not the primary read the open question asked for.

## Further leads

- Jordan Pollack's review independently cites the "intuitive judgment... sterile" passage at **p. 231**, not p. 232 — resolving (in favor of 231) an unconfirmed discrepancy flagged in the prior capture `10-inbox/raw/20260708-0200-does-perceptrons-1969-p.md`; worth a dedicated correction pass on that note.
- The Scribd item "Minsky1969 – An Introduction to Computational Geometry" (https://www.scribd.com/document/608085966/Minsky1969-An-Introduction-to-Computational-Geometry) may or may not be a genuine 1969-edition scan; unresolved, requires an account to check further.
- HathiTrust's catalog returned HTTP 403 on both attempts this session; worth a retry with a different query path or a direct HathiTrust account.
- Internet Archive's two 1988-edition items (`perceptronsintro0000mins`, `perceptronsintro00mins`) remain access-restricted for full OCR/search-inside per the prior capture; an IA "Print Disabled" borrow account would let a researcher directly diff the arabic-paginated body (pp. 1–258) against the Prologue/Epilogue for every occurrence of "credit" — the most direct way to finally close this question.