David K. Lewis (1969), Harvard University Press (xii + 213 pp.; LCCN 69012727). Reissued by Blackwell, 2002; Crossref’s chapter pages for the reissue also end at 213 — DOI-10.1002/9780470693711
Standing in the record
Filed on 2026-10-02 at the owner’s request, as part of filling out the
record’s coverage of social-ontology (ADR-017), which names Lewis’s
Convention among the works the record lacked.
Unavailable because no lawful full text was reachable, so it is not read
and no NOTE is filed.
Contents (chapter titles and pages from Crossref’s records for the 2002 reissue):
- Introduction (1–4)
- I Coordination and Convention (5–51)
- II Convention Refined (52–82)
- III Convention Contrasted (83–121)
- IV Convention and Communication (122–159)
- V Conventions of Language (160–202)
- Conclusion (203–208); Index (209–213)
What it is said to argue, attributed:
- Per the SEP entry (LIT-487, §4.2.2): conventions are solutions to coordination problems in situations with more than one equilibrium, such as driving on the left or the right. They need no explicit or tacit agreement, only beliefs, knowledge, expectations and preferences among the members. Three critiques are listed: many social phenomena are not solutions to coordination problems; the analysis is too mentalistic; and it requires behavioural regularity (Burge 1975, Millikan 2005).
- Per Hindriks & Guala (LIT-481, §3): Lewis modelled conventions as solutions to coordination games with multiple, mostly symmetric equilibria, the “first major breakthrough in the equilibria approach”. They generalise it to correlated equilibria. Guala’s book, in its sample chapter 1 (LIT-472), adds that Lewis made precedent the source of salience, building on Schelling’s focal points, and quotes Lewis’s account of why conventions are “a species of norms” (1969: 98). That quotation was read in Guala, not in Lewis.
Topic. social-ontology first: convention is named in the topic’s
blurb and is one of the anchoring theories the SEP entry surveys.
philosophy-of-language holds chapters IV–V. social-science holds the
game-theoretic analysis, which economists took up (Schotter, Sugden).
Neighbours. Hindriks & Guala (LIT-481) build their theory on it. Bicchieri (LIT-493) works in the same equilibrium tradition, on social norms. Tomasello (LIT-125) and the record’s joint-commitment material are a rival tradition that the SEP entry places beside Lewis as another account of how social categories are set up. Lewis’s later work on the content of theoretical terms, which Hindriks & Guala use for reference to institutions, is a different work and not in the record.
Anthology. The book carries no instruction for machine-learning practice. Lewis signalling games are used in research on emergent communication between learned agents, but nothing here or in the anthology’s literature.d (grepped for “Lewis” and “signaling game”) makes that link, so it is not flagged.
What a reading should check. How Lewis defines common knowledge (Binmore’s objection, reported in Hindriks & Guala’s reply, is that it is not needed for coordination). Whether chapter III’s contrasts (convention versus agreement, norm, rule) anticipate Searle’s constitutive rules. Whether the “species of norms” passage is a theory of normativity or a description.