Source
- Block (1986), Advertisement for a Semantics for Psychology, LIT-648, read in full in NOTE-503: pp. 619–630 (narrow meaning, conceptual role, the social dimension), 635–639 (one factor or two) and 643–644 (meaning and reference). Page numbers are the journal’s.
- Block (1987), Functional Role and Truth Conditions, LIT-625, read in full in NOTE-492: pp. 167–178 (Twin Earth, private and public use, the convergence), n. 31 and the Appendix (pp. 179–181).
What was actually shown
Two factors. Block (1986), LIT-648, states the theory: meaning has “a conceptual role component that is entirely ‘in the head’ (this is narrow meaning) and an external component that has to do with the relations between the representations in the head … and the referents and/or truth conditions of these representations in the world” (pp. 627–628). Conceptual role is “total causal role, abstractly described”, limited to the relations that mediate inference, decision “and the like”. Narrow meaning is motivated by explanation: Teen and her Twin Earth duplicate share a thought about “Michael Jackson”, and two people who think “I am in danger” leap aside for the same reason; what they share explains what they do (pp. 619–625).
Role fixes the kind of reference, not the referent. Which theory of reference is true of names, causal rather than descriptive, is a fact about how names function in our thinking, so conceptual role “determines the nature of the referential factor” without determining reference; the same role maps to H₂O here and XYZ on Twin Earth (LIT-648, pp. 643–644). LIT-625 repeats it: “The functional role factor determines the nature of the referential-truth-conditional factor (without determining reference or truth conditions themselves)” (p. 170).
Long-arm roles reduce to two factors. Harman’s one-factor theory extends roles into the world. To handle a visitor to Twin Earth, the elm–beech case and borrowed reference to Aristotle, it must build in normal context, experts’ roles and causal chains running into the past. “If you take Harman’s long-arm conceptual roles and ‘chop off’ the portion of these roles outside the skin, you are left with my short-arm conceptual roles. If the outside-the-body part that is chopped off amounts to some familiar sort of theory of reference, then the difference between Harman’s one-factor theory and two-factor theory is merely verbal” (LIT-648, p. 637). Block (1987), LIT-625, makes this a general convergence. A private-use theory must add a referential factor to separate aluminum from twin aluminum, since the defining terms share extensions on both planets (pp. 168–169). A public-use theory, use “in the sense in which it could be recorded by a movie camera”, is identical on both planets, and becomes adequate only by importing a dominant causal source, borrowed reference and experts (pp. 174–178). The result is “two distinct (though not independent) sub-theories”: one “at least partly inferential” and inside the head, the other “very like a traditional theory of reference”. “The whole thing is equivalent to the two-factor theory we arrived at starting with private use” (p. 178).
Social meaning. For Burge’s arthritis, “two-factor theory can try to capture such phenomena in the referential factor”, the causal chain mediated by people who know more; “alternatively, two-factor theory may have to expand to three-factor theory, allowing a distinct social factor to meaning” (LIT-648, p. 630). The 1987 appendix gives the first treatment: what is needed is “functional role in a language”, cashed out through “the structure of deference”, not an average of speakers, and individual roles still matter (LIT-625, pp. 179–181).
What could have come out otherwise: a long-arm theory could have fixed reference with resources that are not a theory of reference in disguise, or the internal remainder could have differed between the twins. In Block’s cases it does neither.
What this does not say
- It does not say narrow content is explanatory. “Is narrow content an explanatory notion, either in scientific psychology or everyday life? I haven’t tried to answer these questions” (LIT-625, p. 178). The 1986 paper argues it is, from shared behaviour across Twin Earth; the 1987 paper only makes room for it, and lists the objections without answering them.
- It does not say one bookkeeping is wrong. Which to call the theory, one factor or two, is a matter of purpose: “The short-arm roles are better for psychological explanation, long-arm roles are better for thinking about translation and communication” (LIT-625, n. 31). The claim is that they are equivalent, not that the long-arm form is false.
- It gives no identity conditions for roles. The “crucial question”, what counts as same and different role, is posed and left; the 1986 paper calls itself an advertisement, and Block says he is “far from a true believer” (NOTE-503, Limitations).
- The social factor is not built. Whether it is part of the referential factor or a third factor is left open in both papers.
- It is not functionalism about experience. Block holds this account for intentional content while denying functionalism for qualitative states (LIT-648, p. 635), which is the split THEORY-095, filed alongside, relies on.
How it bears on the record
The reply NOTE-117 needs to give Toon. Drayson (LIT-112, read in NOTE-117) answers Toon’s metasemantic premise, that intentionality belongs first to public, socially normed representations and that inner representations are “locked away” from convention, by saying that a functional role semantics can let mental representations take content from social and communicative use, and she cites “Block 1987”, which is LIT-625. NOTE-117 records that this is “named, not developed”. This account is the developed form, and it gives the reply a sharper shape than Drayson’s sentence. On Block’s theory the content of an inner representation is two-factored. Its internal factor, conceptual role, is in the head and is shared by twins in different communities. Its referential factor, or a third, social factor, is where use in a community, borrowed reference and deference to experts enter. So the inner representation is not locked away from convention: one of its two factors is constituted by it. That is the reply. It is also a qualification of Drayson’s wording. On Block’s own conclusion the social dependence sits in the referential or social factor, not in the representation’s internal role, so a defender of RTM who wants the inner role itself socially fixed gets no support from Block. Block’s 1995 chapter (LIT-630, n. 4) adds the case in point: the reference of ‘Aristotle’ in the language of thought derives from the public word.
The record’s other holdings on intentionality. No theory in the record states an account of mental content or metasemantics, so no relation is declared. Among the works:
- Teleosemantics (LIT-149, Martínez) is in the family whose teleological “wheel” Block argues cannot be naturalised, since selection need not aim mechanisms at truth (LIT-648, pp. 659–660). That bears on it as an argument, not as a rival theory filed here.
- Interpretationism about language models (LIT-206) maximises truth or rationality by charity; Block rejects truth-maximising charity and derives a coherence-based charity from conceptual roles (p. 645).
- Drayson’s realism about beliefs as whole-person functional states (NOTE-117, C2) does not need representations at all; this account is about the content of representations if there are any. They are compatible, and Block’s own commitment to representations is the minimal, empirical one (pp. 641–642).
Inverted Earth. Block’s note 26 to LIT-634 uses this account’s equivalence: long-arm roles are short-arm roles plus a referential component. THEORY-095 asks the functionalist to supply, for qualitative content, the short-arm factor that this account supplies for intentional content.