Source
Katherine Ritchie (2013, online 10 October 2012), “What are groups?”, LIT-479, read in full in NOTE-390. Page numbers are those of the Online First PDF (1–16). Carried forward, with an amended membership definition, in Ritchie (2020, online 2018), LIT-486, read in full in NOTE-389.
What was actually shown
The criteria (p. 3). Any theory of groups must allow that (1) a group can have different members at different times and (2) in different worlds; (3) groups exist temporarily and (4) contingently; (5) they are located; and (6) Coincidence: two groups of the same basic kind can be extensionally coincident and non-identical. Each is supported by one ordinary example (pp. 2–3). Only Coincidence is said to separate groups from tables and trees (p. 4).
The refutations, strong given the criteria.
- Non-singular pluralities fail (1) and (2), since some people are not identical to more or other people, and meet (6) only if arrangements are added (pp. 5–6).
- Fusions (Sider) fail (6), since fusion identity is extensional. The Supreme Court and a committee of exactly the Justices, with the same members at every time, would be one group (p. 7). Counterparts “qua” a sortal do not rescue them (p. 8).
- Burge aggregates fail (1) for member-components and fail (6) when the member-components are the same (pp. 9–10).
- Effingham’s sets fail (5) unless some sets are concrete, meet (3)–(4) only by stipulation, capture (6) only “in spirit”, and face a Benacerraf problem (pp. 10–12).
The positive view (§6). A structure is nodes joined by functional relations (pitch-ball-to, has-more-power-than). Some things realize it iff each occupies a node and every node is occupied (p. 14). Two groups are identical at a time only with the same members in the same intra- and inter-structural relations (p. 15). It meets all six criteria (p. 15). It also explains why groups usually have many members: most group structures contain an irreflexive relation, which needs two occupants (p. 14).
The 2018 amendment. LIT-486 keeps the view for “organized social groups” and makes membership collective, so that two groups can share a structure (its n. 22). It adds that occupation can be normative: members are functionally related “or at least normatively bound” as the structure requires (p. 411).
What could have come out otherwise. One of the four rivals could have met Coincidence without giving up its account of identity. None did. The fusion case is the cleanest: its identity condition is extensional by design.
What this does not say
- That the criteria are established. Each is defended by one case. Coincidence rests on the claim that a student “is in two clubs” (p. 3). A reader who holds that the chess club and the nature club with the same members are one group under two descriptions loses nothing to the refutations. Uzquiano’s alternative is set aside, not answered (n. 1).
- That groups are structures. Groups are realizations of structures. The grain of structures is left open among three options (p. 13), and persistence through a change of structure is not given beyond “continuity of the realization” (p. 14).
- That the view survives vacancies. Occupation is defined holistically: one thing occupies a node only if every node is occupied (p. 14). A team one player short has, as written, no members (NOTE-390, Limitations).
- That these are social groups. The intention condition that would separate social groups from structured material objects “might be added to any of the views” (p. 4) and is not part of the view.
- Anything about agency or minds. The paper never mentions them. That a group is distinct from its members holds of every structured group, agent or not. So “over and above the members”, in the sense of numerical distinctness, is settled before agency comes up, and List & Pettit’s claim (LIT-391) says something further (LIT-479, Standing). That reading is the record’s.
Connections
- Extended by the two-kinds account from Ritchie 2018 (THEORY-575), which keeps this view for organized groups and adds feature groups beside it.
- Epstein’s survey (LIT-487, NOTE-393) reports the same problem for the fusion view and Epstein’s 2019 argument that distinct groups can necessarily coincide (§5.1).
- THEORY-043 uses this paper in its body: it is one reason “the group”
a criterion of consciousness is applied to is not fixed by its members.
NOTE-390 says it does not move THEORY-043’s
promote_when, since its structures are social roles, not causal architecture. - QUESTION-118 asks what must stay causally continuous for a collective to persist through the replacement of all its members. This account allows a group to have different members at different times, but its positive view has no persistence conditions, so it does not answer the question.
- Schwitzgebel’s United States (LIT-159, NOTE-381). If one population realizes several structures, each a distinct group, there are several coincident candidates for what citizens represent. The application is the record’s (NOTE-390).