THEORY-625: Institutional facts are status functions, imposed by collective acceptance on things that cannot perform the function by their physical structure alone; the status carries deontic powers that give desire-independent reasons for action, and it exists only as represented, so institutions require language; 'X counts as Y in C' is the common case of the more general 'We accept (S has power (S does A))'

Source

John R. Searle (2006), “Social ontology: Some basic principles”, Anthropological Theory 6(1):12–29, LIT-482, read in full in NOTE-388 in the parallel English printing, Papers: Revista de Sociologia 80:51–71. Page numbers are Papers’ pages. The Construction of Social Reality (LIT-485), which this paper restates, is not lawfully readable and was not read.

What was actually shown

The paper has no theorems. Its results are distinctions and definitions, argued informally or by example.

• Objectivity (pp. 52–55). Observer-relative facts are ontologically subjective but can be epistemically objective, and the attitudes that create them are themselves observer-independent. That, Searle says, is why social science is possible.
• Three primitives (p. 56). Collective intentionality, the assignment of function, and constitutive rules. A social fact is any fact involving the collective intentionality of two or more agents, so a wolf pack hunting is one. Institutional facts are the subclass involving status functions (p. 57).
• Deontic powers (p. 59). Accepting a status function is accepting rights, duties and permissions. These are “desire-independent reasons for action”, “the glue that holds society together”. Desire-based reasons such as wanting to be rich presuppose them (p. 66).
• Language (pp. 60–62). The fact exists only as represented, its content is deontic, the deontology persists after creation, and the institution must be recognised as well as its instances. The dog can see a man carry a ball across a line but cannot see a touchdown (p. 60).
• The revision (pp. 63–66). Status indicators matter in their own right (De Soto’s squatters, whose powers “stop at the point where the larger society requires some official proof”). Some Y terms are free-standing, with no X: a corporation is created by declaration, and money needs only recorded values. There “the real bearer of the deontology is the participant”. “We accept (S has power (S does A))” is now ranked as “a much more general account” (p. 65).
• Scope (pp. 69–71). Friendship is institutional “because friendship carries obligations”. A date, a colour word or a unit of weight is not.

What could have come out otherwise. The paper names its own falsifier: an institution of another structure would refute it (C7). It offers none and finds none. The revision shows the original form already failed for corporations and ledger money.

What this does not say

• That collective intentionality is irreducible. It is a primitive here, “a genuine biological phenomenon”, and the analysis is deferred to LIT-490 (p. 57, fn. 3). Gilbert, Bratman and Tuomela & Miller are not mentioned.
• That the three primitives suffice. “I am going to be very brief because I am repeating things I have said elsewhere” (p. 56). Asserted and illustrated.
• That animals lack a deontology. Held “as far as we can tell”, and if some primates have one, “so much the better for them” (p. 62).
• That “X counts as Y in C” is now derivative. The paper ranks the general form higher but does not say whether the old form is derivative or merely the common case, and a status function is still defined as imposed “on a person or an object” (p. 57).
• That a corporation is an agent. It is a deontic construction whose powers go to individuals. Whether it can act is a different question, and the paper does not ask it.
• That the taxonomy of status functions is settled. “I still do not have a taxonomy … that I find satisfactory” (p. 66).

Connections

• Rival: rules in equilibrium (THEORY-414, LIT-481). Hindriks & Guala hold that a constitutive rule is a regulative rule with a name added, that institutions such as etiquette need none, and that “desire-independent reasons” is a platitude or an unargued anti-Humean thesis. Searle’s own free-standing Y terms, with powers sent “directly to the individuals” and bills as “markers” of power, treat the Y term much as they do. He does not draw their conclusion (NOTE-388).
• Epstein’s survey (LIT-487, THEORY-487) reports collective acceptance as one theory of anchoring and “X counts as Y in C” as a frame principle of one narrow form.
• Tomasello (LIT-125, NOTE-185; THEORY-495) locates the first felt obligation in children’s joint commitments from about age three. That is the empirical side of the human/animal line Searle draws by analysis, but Searle’s collective intentionality includes hunting wolves and Tomasello’s shared intentionality is species-unique (NOTE-388).
• QUESTION-007 asks what grounds normativity across agents at one level. This account says institutional obligations are grounded in collectively accepted status. No answers edge is filed: the desire-independence that would make that a grounding of normativity, rather than a description of what people accept, is asserted here and disputed by the rival.
• THEORY-057 uses this paper to say Barrett’s comparison of emotion categories with money needs collective acceptance and a status.
• The group-mind strand (THEORY-043). A society exists as that society only because its members represent it; a brain’s parts do not represent the brain. NOTE-388 notes this could be the kind of asymmetry THEORY-043 asks for, as the reader’s inference, not Searle’s.

Note of 2026-10-30: what holds it back

Its promote_when is still unmet. Searle’s 1990 paper (LIT-490) is still unread, but Zaibert (LIT-1875, read in NOTE-1431) quotes it page by page, and on his quotations it gives no irreducibility argument: Searle offers irreducibility as an intuition his arguments “suggest” (CIA 401), says “I have not demonstrated that no such analysis could ever succeed” (CIA 406), and rejects Tuomela and Miller by one counterexample (CIA 404). So the first deferred argument is unlikely to come from LIT-490 either; the second, that a status’s deontology exceeds what a regulative rule in equilibrium gives (THEORY-414), is still missing, and the record holds no institution the analysis cannot state.