Source
- Huebner (2010), LIT-702, read in full in NOTE-553, from the author’s manuscript. Page numbers are the manuscript’s, not the journal’s 89–118.
What was actually shown
All of it is argument. There are no data of the author’s own, and the paper says it is conditional on collective representation being possible, which it defends elsewhere (n. 4).
The test. Huebner, LIT-702, separates holism, on which individuals’ emotions depend on their social relations, from collectivism, on which groups themselves are in emotional states (pp. 3–4, after Pettit). Collective guilt in Jaspers’s, May’s and Tollefsen’s sense establishes only the first. What would establish the second is an emotion “that could not be exhaustively characterized in terms of intentional states of individuals in aggregation” (p. 4).
What an emotion requires. Emotions are representations with a formal object and a particular one, indicative content and imperative force: “action-oriented representations that reflexively reorient behavior”. For fear, at least: a capacity to perceive things as dangerous; attention aimed at the threat and at planning; systems producing and monitoring system-wide preparatory change; and coordination of these into one state that guides coping (p. 10). Nothing in that description mentions neurons, so a distributed system that integrates routines this way represents as fear does.
It need not be felt. Huebner concedes that a ship’s crew cannot be conscious: it is too spatially distributed and too loosely integrated (p. 14). He argues instead that some individual emotions are non-conscious in a strong sense, with no felt emotion, only scattered felt bodily effects (§5, after Sizer on moods), so a collective emotion need not be felt either. The argument rests on describing a case, and the neuroscience it cites shows unconscious processing of fear stimuli, which is weaker (NOTE-553, Limitations).
Emotions are functional kinds. Against the objection that groups lack mechanisms homologous to ours, he argues that demanding homology rules out artificial and alien emotion by fiat and forces Griffiths’s eliminativism about emotion even within humans. Kinds are fixed at the computational level: “Emotions, like vodkas, require only the right sort of functional organization” (§6, p. 22).
The case fails, and the replacement is not evidence. The USS Palau crew, tracing Hutchins’s report of a loss of steam pressure, detects danger and copes, but training had removed the agitation. Huebner withdraws it: it is “a collective detection of danger in the absence of a fear response” (p. 23). He then offers the closing weeks of the 2008 McCain–Palin campaign as “an intuitively plausible case”, noting that “until the relevant anthropological data is collected, this appeal will be insufficient to conclusively establish the existence of collective emotions” (p. 23). There, he argues, members were agitated but their emotions varied, hopeful, angry, saddened, so the campaign’s fear was not an aggregate of members’ fear (p. 25).
So the paper concludes that “there are no principled philosophical reasons to deny that emotional states can be extended to some sorts of collectivities”, and that where a group lacks one it is because it is “not organized in the right way” (p. 25). That is a possibility claim with a research programme attached. No group has been shown to be in an emotional state.
What this does not say
- It does not say any actual group has an emotion. The account’s last clause says so, and the Palau failure is the only case worked through.
- It does not say a group’s emotion is felt. The account sides with the view that collective emotions need no phenomenology. So does Gilbert’s plural-subject account (LIT-701, Deferred and unread), which Huebner reports as holding that feelings are “inessential, though common concomitants” of emotion, so that collective emotions “require no specific phenomenology” (Gilbert 2002, 119, as reported on p. 2). The two locate the emotion in different bearers: for Gilbert a plural subject constituted by members’ joint commitment to feel “as a body”; for Huebner an integrated computational system, possibly including its technology. Huebner argues the plural subject’s emotion reduces to individual commitments plus aggregation rules (p. 4), and rejects Gilbert’s route to unfelt emotion because her psychoanalytic examples show inaccessible emotions, not unfelt ones (p. 15). Gilbert’s reply is not considered, and what is said of her here rests on Huebner’s report. Schmid (LIT-697, Deferred) is, by his abstract, the view that groups do feel, as plural pre-reflective self-awareness.
- It does not say convergence of members’ emotions is collective emotion, or that it is not. The convergence account, THEORY-126, answers a different question: whether and how members’ affect converges on one object. This account asks whether the group bears a state of its own. The two are compatible. Huebner’s own surviving case is one in which members’ emotions diverged, so a genuinely collective emotion neither requires convergence nor follows from it, and convergence can occur in a group that bears no emotion of its own. No relation is declared between them. In List’s taxonomy of collective attitudes, LIT-700, the difference has a name: a converged emotion is an aggregate attitude, or a common one when members are mutually aware, and a genuinely collective emotion would be a corporate attitude, the only kind that needs a group agent. List says only that there can be “aggregate fears and hopes, common fears and hopes, and corporate fears and hopes” (§1); the placement of these accounts in his kinds is the record’s.
- It does not say a group’s emotion would have the rationality profile of its judgments. List’s corporate attitudes must be rational, and the reason a group’s judgments come apart from its members’ is a theorem about deductively closed judgments (THEORY-127). Huebner’s test, a state not characterisable as members’ states in aggregation, is the informal analogue of that theorem’s failure of proposition-wise supervenience, and the Palau crew, routines each sensitive to one constraint and integrated on a chart, has the shape of List and Pettit’s distributed premise-based procedure. But an emotion’s coherence conditions are not deductive closure, and neither paper shows that the theorem carries over to emotions. The parallel is the record’s.
Connections
- THEORY-043 (minds of minded agents) agrees that functional states
survive composition and narrows the dispute to experience. This account
adds emotion, as a functional state, to what survives, and takes the
same line on experience by conceding that the crew is not conscious
because it is too distributed and too loosely integrated. That is an
architectural verdict asserted, not computed, like List’s low-Φ verdict
in LIT-401, so it does not meet THEORY-043’s
promote_when. Huebner’s footnote 22 states the conditional Schwitzgebel (LIT-159) argues for: on representational or higher-order theories, collectivities “can be conscious where they have the right functional organization”. This paper is the record’s first read statement of the macrocognition programme; the book (LIT-426) is unread, and THEORY-043 names it as the work most likely to move that account. - Emotion theory. The account needs emotion kinds fixed by function, with formal objects taken from Lazarus’s core-relational themes, and action-orientation in Griffiths and Scarantino’s sense. That is close to the appraisal account, THEORY-055, on which emotions are states of action readiness that differ by mode. Barrett’s constructionist account, THEORY-057, would deny that fear is a kind with a shared mechanism at any level, and on it whether a group’s state counts as fear would turn on how perceivers categorise it. Huebner does not cite Barrett. THEORY-045 is about felt affect, the experience of a body’s regulatory state, and this account claims nothing about it: a group’s emotion, on this account, is exactly an emotion without that.
Note of 2026-10-29: what holds it back
Read in full (NOTE-553). The paper argues a possibility and withdraws its
one worked case, the USS Palau, as detection without fear. Its replacement,
the McCain–Palin campaign, waits on anthropological data the paper says it
lacks. The record holds no group-level study of the kind promote_when
names, and no computational definition of agitation that can be applied to
a group without reading it off members’ feelings.