Add 64 references on thought, language, and representation; correct 6 entries

[jeremymanning]

Adds 64 references and corrects 6 existing entries, in support of a manuscript on the
geometry of thought and language. python bibcheck.py verify --verbose reports
looks good!; the commit message was generated by python bibcheck.py commit.

What was added

64 entries across: symbolic and rule-based representation (Newell & Simon, Fodor &
Pylyshyn, Smolensky, Harnad, Minsky, SOAR); early chatbot systems (ELIZA, PARRY,
SHRDLU, A.L.I.C.E.); anti-representationalist critiques (Brooks, van Gelder, Ramsey,
Chemero, Brette, Dennett); representational geometry (Shepard 1962 I & II, Gärdenfors,
Kriegeskorte & Kievit, Kriegeskorte & Wei, Chung & Abbott); deep learning and LLM
geometry (Rumelhart/Hinton/Williams, LeCun/Bengio/Hinton, Silver, Jumper, Ethayarajh,
Hewitt & Manning, Park et al.); and the nature of thought (Whorf, Carruthers,
Boroditsky, Fedorenko & Varley, Fedorenko/Piantadosi/Gibson, Kosslyn, Churchland,
Vyas, Newell 1990, Jonas & Kording, Guest & Martin).

Every entry was verified against CrossRef — title, authors, journal, volume, issue,
pages, year — or against the publisher's own record where CrossRef has none (Fodor
1975, Whorf 1956, Newell 1990, Russell & Norvig 1995, Minsky 1975).

Corrections to existing entries

Entry	Problem	Fix
ViswEtal17	First author's surname misspelled "Viswani"; it is "Vaswani". The key was derived from the misspelling.	→ VaswEtal17, surname corrected, Volume = {30} and Pages = {5998--6008} added
Krus98	ALCOVE is Psychological Review 99(1), 22–44, 1992. The entry said volume 1, year 1998, and the author field was a bare "Kruschke" with no initials.	→ Krus92, volume/year corrected, Number added, author J K Kruschke
McClEtal86	Typed @article with Journal = {Explorations in the Microstructure of Cognition}, Pages = {216--271}, Volume = {2}. It is a book.	→ @book, MIT Press, Cambridge MA, with the real volume 2 subtitle
RumeEtal86	Volume 1 carried a bare title while volume 2 gained a subtitle. Identical titles plus identical author surname sets is what the duplicate rule targets.	→ RumeEtal86a, given its real volume 1 subtitle and an address
AndeEtal04b	Author field read JR Anderson.	→ J R Anderson (initials must be separated)
MontEtal12	Monti, Parsons & Osherson (2012) also keys to MontEtal12, and the suffix rule requires every key sharing a base to carry a suffix.	→ MontEtal12a; the new entry is MontEtal12b

Breaking changes

Four keys no longer exist. A document citing one loses the citation silently:

Gone	Now
ViswEtal17	VaswEtal17
Krus98	Krus92
RumeEtal86	RumeEtal86a
MontEtal12	MontEtal12a

MontEtal12a is the only rename of an entry unrelated to the manuscript that prompted
this work.

Duplicate checking

The additions were checked against all 6,330 prior entries four ways: duplicate keys;
key collision across all a–z suffix variants (a bare-key check reports "no
collision" while the work is already present as, say, AndeEtal04b); the duplicate
rule itself (same author surname set and same title); and fuzzy title similarity.
Suffix series were also checked for gaps and for a base existing both suffixed and
bare. Two prepared entries were dropped as already present under suffixed keys
(AndeEtal04b, NormEtal06b), and one candidate was dropped in favour of the existing
Jame90.

One thing left for a maintainer

CrutEtal13 and HealEtal14a share a title and an author surname set:

• CrutEtal13 — Crutchley, Healey & Kahana (2013), Context and Episodic Memory
  Symposium Abstracts — a conference abstract
• HealEtal14a — Healey, Crutchley & Kahana (2014), JEP: General, 143(4), 1553–1569

bibcheck passes on the pair and they are arguably distinct works, but it is the exact
shape the duplicate rule was written to catch. Untouched here, since deleting either
breaks anyone citing it.

🤖 Generated with Claude Code