LIT-702: Genuinely collective emotions

Bryce Huebner (2010), European Journal for Philosophy of Science 1(1):89–118 (online 30 November 2010; January 2011 issue) — DOI-10.1007/s13194-010-0006-2

Key takeaways

• The question is collectivism, not holism. Many arguments for collective emotion show only that individuals’ emotions depend on their groups (Jaspers, May, Tollefsen’s collective guilt). A genuinely collective emotion must be a state that is not individual states in aggregation, and Huebner makes that the test (pp. 3–4).
• Emotion is an integration achievement. Fear needs danger detection, redeployment of attention, system-wide preparatory changes, and the coordination of all of these into one state with indicative content and imperative force (p. 10). If a distributed system integrates routines that way, nothing in the computational story stops it being afraid.
• Two objections are answered and one is conceded. Consciousness: some individual emotions are non-conscious in a strong sense, so a group’s need not be felt (§5). Homology: emotion kinds are fixed at the computational level, not by brain homologues (§6). But the paradigm case, the USS Palau crew, shows “collective detection of danger in the absence of a fear response”, because training removed the agitation (§7). The case that replaces it, the McCain–Palin campaign of 2008, rests on no data.

Standing in the record

Filed on 2026-10-04 at the owner’s request, with Gilbert (LIT-701), Schmid (LIT-697), List & Pettit (LIT-705, LIT-700), Wimsatt (LIT-707) and Isaacs (LIT-698), to connect the record’s emotion holdings to its group-agency and social-ontology holdings. No anthology topic holds it, and it carries no instruction for machine-learning practice.

It is the paper; Huebner’s book Macrocognition (LIT-426) is held Deferred and unread. The paper’s footnote 4 defers its account of collective representation to work “in prep”, and its integration criterion (specialised routines integrated into skilful coping) is the one LIT-426’s publisher description states. So this paper is the record’s first read statement of the macrocognition programme, applied to emotion.

The three accounts of collective emotion. The owner asked for the debate between Gilbert’s plural-subject account, this distributed account and Schmid’s phenomenological one.

• Against Gilbert (LIT-701), declared rivals. Huebner reports Gilbert’s definition, “for us collectively to feel guilt over our action A is for us to be jointly committed to feeling guilt as a body over our action A” (Gilbert 2001, 139, quoted p. 4), and her claim that collective emotions need no phenomenology (Gilbert 2002, 119, p. 2). He grants that the definition comes closer to a collective emotion than membership guilt does, then argues that a plural subject’s emotion reduces to individual commitments plus rules for their aggregation, while an individual’s emotion does not, since subpersonal routines do not “traffic in commitments” (p. 4). He also rejects her route to unfelt collective emotion, “pace Gilbert (2002)”: the psychoanalytic examples show only inaccessible emotions, not unfelt ones (p. 15). The two accounts locate a group’s emotion in different things, joint commitments against integrated computation, and each excludes the other’s bearer. Neither tests the other, so the relation is rivals, not corrects. Gilbert’s own text is unread here, and what is said of it rests on Huebner’s report.
• Against Schmid (LIT-697), no relation declared. Schmid’s chapter came four years later and is unread. By its abstract it holds that a group with intentional attitudes has consciousness, as plural pre-reflective self-awareness, and that groups can have “slow” emotions but not “fast” ones. Huebner concedes that his crew cannot be conscious (p. 14) and builds unfelt collective emotion instead. His fast case (fear at sea) fails and his surviving case is fear sustained over weeks of a campaign. That is consistent with Schmid’s fast/slow line, but the two disagree on whether the group feels anything. A relation waits on a reading of Schmid.

Group agency. Huebner’s test, a group state that “could not be exhaustively characterized in terms of intentional states of individuals in aggregation” (p. 4), is the informal version of what List and Pettit prove for judgments: a rational group’s attitude on a proposition cannot in general be a function of its members’ attitudes on that proposition (LIT-705). Their distributed premise-based procedure, where subgroups each settle one premise and nobody need judge the conclusion, has the shape of Hutchins’s navigation crew, where each sub-routine is “sensitive to some one-dimensional constraint” (p. 12). List’s three-way taxonomy (LIT-700) gives Huebner’s holism/collectivism split a sharper form: holistic emotions are aggregate or common emotions, and a genuinely collective emotion would be a corporate one. Neither List paper mentions emotion beyond listing “fears and hopes” as attitudes (LIT-700, §1), so no relation is declared.

Emotion theory. The paper takes emotions to represent Lazarus’s core-relational themes, by way of Prinz, and to be action-oriented in Griffiths and Scarantino’s sense. It lists Frijda (1986) among accounts that require many component capacities (n. 12). The record holds Frijda’s 1986 book unread (LIT-546) and his 2014 paper read (LIT-540). An appraisal-style functional kind is what lets the argument run, and Barrett’s constructionism (LIT-542, THEORY-057) would deny it a natural kind to extend. Huebner’s own §6 already turns that against the homology objection: Griffiths’s eliminativism about emotion as a kind is offered as the cost of insisting on homology (pp. 20–22).

The book-length neighbours are Tollefsen’s Groups as Agents (LIT-483) and Schwitzgebel’s US-consciousness argument (LIT-159). Huebner’s footnote 22 says that on representational or higher-order theories collectivities “can be conscious where they have the right functional organization”, which is Schwitzgebel’s conditional. List’s reply to Schwitzgebel (LIT-401) quotes Huebner’s book on the same intuition.