Douglas R. Hofstadter and the Fluid Analogies Research Group (1995), Fluid
Concepts and Creative Analogies: Computer Models of the Fundamental
Mechanisms of Thought, New York: Basic Books (Library of Congress 93044294,
call number BF311 .H617 1995; ix, 518 pp., illustrated, references pp.
[493]–501, index). The Library of Congress record spells the title Fluid
concepts & creative analogies; the publisher spells it with “and”. The
hardback ISBN is 0-465-05154-5, as Bruce Burns’s review in AI Magazine
16(3), Fall 1995, gives it with “Basic Books, New York, 1995, 518 pp.”
The paperback (ISBN 978-0-465-02475-9) carries the same 1995 copyright in
its catalogue record, and the publisher gives its on-sale date as 22 March
1996, at 528 pages. No source I reached gives a day or month for the 1995
hardback, so published: carries the year only, as the first of January.
Key takeaways
Registered from the publisher’s description, the catalogue record and one review, not a reading. What follows is what the book, its cataloguers and its reviewer say about it.
- The publisher’s description (Basic Books), verbatim: “From the Pulitzer Prize winning author of Gödel, Escher, Bach, a stunning exploration of the mechanisms of intelligence through computer modeling. … Readers of earlier works by Douglas Hofstadter will find this book a natural extension of his style and his ideas about creativity and analogy; in addition, psychologists, philosophers, and artificial-intelligence researchers will find in this elaborate web of ingenious ideas a deep and challenging new view of mind.”
- The Library of Congress subject headings: analogy; artificial intelligence; cognitive science; computer simulation.
- Contents, from the contents note in the paperback’s catalogue record
(cataloguing source DLC), with the authors it gives. Chapters 2 to 10 are
each preceded by a preface the note does not attribute:
- Prologue: “The When, the Where, and the Who of This Book” (the note garbles it as “Why, the when, the where, and the who of this book”).
- 1, “To Seek Whence Cometh a Sequence” (Hofstadter).
- Preface “The Unconscious Juggling of Mental Objects”; 2, “The Architecture of Jumbo” (Hofstadter).
- Preface “Arithmetical Play and Nondeterminism”; 3, “Numbo: A Study in Cognition and Recognition” (Daniel Defays).
- Preface “The Ineradicable Eliza Effect and Its Dangers”; 4, “High-Level Perception, Representation, and Analogy: A Critique of Artificial-Intelligence Methodology” (Chalmers, French and Hofstadter; the note misspells Chalmers as “Chambers”).
- Preface “Conceptual Halos and Slippability”; 5, “The Copycat Project: A Model of Mental Fluidity and Analogy-Making” (Hofstadter and Mitchell).
- Preface “Two Early AI Approaches to Analogy”; 6, “Perspectives on Copycat: Comparisons with Recent Work” (Mitchell and Hofstadter).
- Preface “Retrieval of Old and Invention of New Analogies”; 7, “Prolegomena to Any Future Metacat” (Hofstadter).
- Preface “Analogy-Making in a Coffeehouse”; 8, “Tabletop, BattleOp, Ob-Platte, Potelbat, Belpatto, Platobet” (Hofstadter and French).
- Preface “The Knotty Problem of Evaluating Research in AI and Cognitive Science”; 9, “The Emergent Personality of Tabletop, a Perception-Based Model of Analogy-Making” (Hofstadter and French).
- Preface “The Intoxicating World of Alphabets and Their Styles”; 10, “Letter Spirit: Esthetic Perception and Creative Play in the Rich Microcosm of the Roman Alphabet” (Hofstadter and McGraw; the note has “MGraw”).
- The note stops there. A third-party transcription of the contents (egodeath.com, with section headings) agrees with this order and adds an Epilogue, “On Computers, Creativity, Credit, Brain Mechanisms, and the Turing Test”. That transcription is not the publisher’s, and the epilogue is the only thing taken from it here.
- Burns’s review (AI Magazine 16(3), 1995, pp. 81–83, read in full from AAAI’s own PDF) is the one account of the content I read. It says the book is a collection of articles, many already published, given “a coherent vision” by Hofstadter’s prefaces. It names two underlying ideas: a computational one, that cognition can be modelled by a “parallel terraced scan”, in which codelets chosen probabilistically from a coderack build and break structures while a falling temperature makes coherent structure harder to tear down, “without the need for any top-level control”; and a psychological one, that “cognition is recognition”, high-level perception interacting with the processing of what is perceived. Burns recommends the book but reports two criticisms: every model works in a microdomain (letter strings, table settings, word puzzles, letter fonts), and the critiques of rival programs lean on analogy, rhetoric and intuition. On one point the review and the catalogue differ: Burns calls the evaluation preface “Preface 5”, while the contents note places it before chapter 9. Which is right needs the book.
Standing in the record
Filed on 2026-10-01 at the owner’s request for the Fluid Analogies Research Group’s programme that analogy-making is the core of cognition, Copycat and its successors. This is the programme’s main collection. No anthology topic holds it, and the anthology does not hold it (checked by grep). It carries no instruction for present machine-learning practice. Its nearest approach is the methodological argument of the Eliza-effect and evaluation prefaces, about when a program’s apparent success is credited to the program; an anthology reader on evaluation might want those, and that is a judgement for the anthology, not a tag here.
Deferred because it was not read. This record files a book from its
publisher’s description and contents when no lawful text can be read, and
does not rebuild a book from what others say about it. The one review above
is quoted as the reviewer’s account, not as the book’s.
Its parts that the parallel filing covers, named in prose because those works have no codes yet:
- Chapter 4 has the title and authors of Chalmers, French and Hofstadter’s 1992 critique of AI methodology, which is being filed at the same time. Whether the chapter reprints it unchanged is for a reading.
- Chapters 5 and 6 are on Copycat, the program of Mitchell and Hofstadter’s 1990 paper, also being filed now.
- The book’s own claims about analogy as the core of cognition are carried further in Surfaces and Essences, in this batch, and in Hofstadter’s 2001 essay of that title in the parallel filing.
The publisher calls the book “a natural extension” of Hofstadter’s earlier
ideas about creativity and analogy. That is the publisher speaking, and no
extends: is declared to Gödel, Escher, Bach.
Tags. complex-systems is there because, per the review and the contents
titles (“The Emergent Personality of Tabletop”, Copycat’s “statistically
emergent” behaviour in the transcription), the architecture is about how
coherent behaviour arises from many small, locally acting parts. Rizi’s
proposal that emergence is a coarse-graining that stays predictive
(LIT-150) is the record’s nearest reading on what “emergent” should mean;
the pairing is mine, and a reading should check whether the book’s
“emergent” means anything more than “not explicitly programmed”.
Access when seeded:
- Library of Congress catalogue record (lccn.loc.gov/93044294, through the loc.gov JSON API): title, publisher, date, extent, references pages, call number and subjects. The MARCXML and MODS links on lccn.loc.gov returned HTTP 404, and the LoC SRU server timed out.
- The public MARC record of the Internet Archive’s catalogue copy (fluidconceptscre0000hofs_w7o9, cataloguing source DLC): the paperback ISBN and the contents note above.
- Hachette Book Group’s page for the paperback (ISBN 9780465024759): the description, on-sale date and page count. Its pages for the hardback ISBN redirected and gave nothing.
- Open Library edition records for ISBNs 0465051545 (1995, ix, 518 p.) and 0465024750 (1996), both with LCCN 93044294 and OCLC 29429061.
- Burns, “Fluid Concepts and Creative Analogies: A Review”, AI Magazine 16(3), 1995, from AAAI’s own PDF, read in full.
- The Internet Archive scan is lending-only (access-restricted), so I did not open it. HathiTrust holds two copies, search-only.
- The ACM Digital Library’s record (10.5555/218753, not a registered DOI) returned a JavaScript challenge (HTTP 403). Crossref has no DOI for the book, only for reviews of it.
- A PDF offered by a book-summary service is a reconstruction, and I did not open it.
- WorldCat returned HTTP 429. Google Books’ API refused for an exhausted daily quota.