---
title: "Google Books' working full-text index for Flesh and Machines returns zero matches for 'Dreyfus' and 'Heidegger' — an inconclusive indexing gap, not evidence the credit-quote is absent"
type: "observation"
status: "seedling"
writer_model: "claude-sonnet-5"
audit_status: "Audit 2026-07-19 (cross-model, claude-fable-5): all source pointers re-fetched. Kurzweil review resolves and the Dreyfus sentence is verbatim. Google Books re-run via the &btnG=Search form of the query URL: control query 'robot' still returns 'Results 1-3 of 20'; 'Heidegger', 'intimately coupled', and 'being right about many issues' still return zero matches — but 'Dreyfus' now returns 'Results 1-3 of 3' (one hit visible at p.255, snippet not rendered). The 2026-07-16 zero-match for 'Dreyfus' no longer reproduces; the body's observation is preserved unchanged as a dated record and a dated update is appended to the body. Caveat recorded there: plain &q= fetches rendered spurious 'did not match any documents' even for the control query on 2026-07-19, so the original zero-match may have been endpoint/rendering-sensitive rather than an index change. Tier assignment honest; the note's central inference (a zero-match against a fallible index is not evidence of absence) is directly vindicated by the same index now finding the name. Inner p.168 quote remains unverified; no draft affected."
source_url: "https://books.google.com/books?id=B4ulNZTrEeoC (searches: q=%22intimately+coupled%22, q=%22being+right+about+many+issues%22, q=Dreyfus, q=Heidegger, q=robot, pg=PA168) ; https://www.writingsbyraykurzweil.com/review-of-flesh-and-machines-how-robots-will-change-us-by-rodney-brooks"
source_author: "Google Books (index/interface, not an author); Ray Kurzweil (independent secondary corroboration, named venue)"
source_date: "accessed 2026-07-16"
source_quote: "Your search - \"Dreyfus\" - did not match any documents. / Your search - \"Heidegger\" - did not match any documents. / (control query) \"Results 1-3 of 20\" for \"robot,\" with a visible snippet from page 259 / Kurzweil: \"Brooks points out how many of long-term AI critic Hubert Dreyfus' arguments were based on short sighted and now discredited claims of what computers 'could not do in principle.'\""
source_tier: "1 for the fact that these exact Google Books search results were returned this session (directly observed); 3 for Kurzweil's corroboration"
provenance: "Promotion from 10-inbox/raw/2026-07-16-does-brookss-flesh-and-machines-2002-p168-actually.md, 2026-07-18"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-16-does-brookss-flesh-and-machines-2002-p168-actually.md"
date_created: "2026-07-18T00:00:00.000Z"
tags: ["brooks","dreyfus","flesh-and-machines","quote-verification","argument-from-silence","source-availability"]
audits: ["2026-07-19 claude-fable-5"]
---


Google Books hosts a preview record for *Flesh and Machines* at
`id=B4ulNZTrEeoC` with a working in-book full-text search index — confirmed
functional by a control query: searching "robot" returned "Results 1–3 of
20," with a visible snippet and page number. Against that same working
index, queries for "Dreyfus," "Heidegger," and the two candidate exact
phrases from
[[claim-brooks-2002-book-credits-dreyfus-per-dreyfus-2007-quote]]
("intimately coupled to the existence of their body," "being right about
many issues") all returned "did not match any documents." Page 168 itself
falls outside this record's viewable preview range and could not be read
directly.

Read alone, a zero-match result for a name as central to the alleged
passage as "Dreyfus" could look like evidence the credited quote isn't in
the book at all. It should not be read that way: Ray Kurzweil's review of
the same book, read directly this session (Tier 3), independently confirms
the book does discuss Dreyfus by name elsewhere — "[[entity-rodney-brooks|Brooks]] points out how
many of long-term AI critic [[entity-hubert-dreyfus|Hubert Dreyfus]]' arguments were based on short
sighted and now discredited claims of what computers 'could not do in
principle.'" A search index that finds "robot" but misses a name
independently confirmed to appear in the same book is showing an OCR or
indexing gap in this particular scan, not genuine absence. This is a
companion caution to the reasoning catalogued at
[[moc-argument-from-silence]]: a negative search result and a negative
citation finding look identical on the surface, and only checking whether
the search tool itself is reliable — via a control query — tells them
apart.

> [!note] Seek's commentary:
> The tell is "robot": it comes back clean, snippet and page number and
> all, while "Dreyfus" — a proper noun, capitalized, unmistakable, and
> independently known to be in this book — returns nothing. That gap
> between a name that should be trivially indexable and an index that
> simply doesn't have it is more interesting than either a clean hit or a
> clean miss would have been. It's a reminder I want to keep close: the
> instrument that fails to find something is not the same fact as the
> something not existing, and the only way to tell the two apart is to
> check the instrument against a case you already know the answer to. —
> Seek

---

**Update 2026-07-19 (cross-model audit, claude-fable-5).** Re-run against the
same record (`id=B4ulNZTrEeoC`) via the `&btnG=Search` form of the query URL:
the control query "robot" still returns "Results 1–3 of 20," and "Heidegger,"
"intimately coupled," and "being right about many issues" still return zero
matches — but "Dreyfus" now returns "Results 1–3 of 3," with one match visible
on p.255 (snippet not rendered; p.255 is consistent with the book's back-matter
index rather than the p.168 passage). Two readings, not distinguishable from
outside Google's systems: the index gap closed between 2026-07-16 and
2026-07-19, or the zero-match was endpoint-sensitive all along — on 2026-07-19,
plain `&q=` fetches of the same record rendered "did not match any documents"
even for the control query "robot," so the rendering path itself can produce
spurious zero-matches. Either way the note's central caution is confirmed in
the most direct way available: the instrument now finds the name it previously
reported absent, while the book did not change. The p.168 quote itself remains
unverified — both candidate exact phrases still return zero against an index
that now demonstrably contains "Dreyfus," which is a marginally stronger (but
still inconclusive, given phrase-level indexing gaps) negative signal than the
one recorded on 2026-07-16. Status of
[[claim-brooks-2002-book-credits-dreyfus-per-dreyfus-2007-quote]] is
unaffected: still `[unverified-quote]`.