Fregean Sense and the Opacity of Intentionality Re-evaluated — Epoche C2
What opacity is, and what Frege's semantics says about it The phenomenon at issue is the failure of substitution in belief reports: from the truth of "Sam believes Hesperus is visible at dusk" and the truth of "Hesperus is Phosphorus", one may not infer "Sam believes Phosphorus is visible at dusk", although the two names pick out the same planet. Contexts in which such inferences fail are called opaque, and the compressed version of this essay described them incorrectly in its opening paragraph. It said that a Fregean semantics presumes that two expressions with the same sense are interchangeable salva veritate — preserving truth value — and that the failure of that interchange constitutes opacity. That is not the phenomenon. Opacity is the failure of substitution for expressions with the same reference and different senses; co-sensical expressions do substitute in belief contexts on Frege's account, and that is a consequence of the theory rather than a difficulty for it. The compressed version's own numbered list gets this right, so the mistake is confined to the framing, but repairing it reverses the dialectic. Frege's semantics is not caught out by opacity. It entails it. The apparatus is worth restating precisely, because the argument later depends on details of it. In "Über Sinn und Bedeutung" (1892) Frege begins from the difference between an identity statement of the form $a = a$, which is trivial and known without inquiry, and one of the form $a = b$, which may be a discovery. If the semantic contribution of a name were exhausted by the object it stands for, the two statements would say the same thing, and one of them would not be a discovery. Frege's response is to distinguish the Bedeutung of an expression, the object it stands for, from its Sinn , the mode of presentation of that object — the way in which the object is given. His own illustration is geometrical rather than astronomical: let $a$, $b$ and $c$ be the lines joining each vertex of a triangle to the midpoint of the opposite side. The point of intersection of $a$ and $b$ is the point of intersection of $b$ and $c$; one object, two ways of being given. The step that does the work here is the second one, and it is the step the compressed version never stated. Frege holds that in oratio obliqua — the indirect discourse of "Kepler believed that ..." — words do not have their customary reference. They refer instead to what is customarily their sense. So in "Sam believes Hesperus is visible at dusk", the term "Hesperus" does not stand for Venus at all; it stands for the sense of "Hesperus". "Phosphorus" stands for a different sense. In that position the two terms are therefore not co-referential, and the principle licensing substitution — that co-referring terms may replace one another without change of truth value — does not apply to them. Nothing has been given up. The very principle whose apparent failure defines opacity is preserved, and the appearance of failure is explained by a shift of reference that the theory independently postulates. The two examples, and which of them is a case of opacity The compressed version offered two illustrations, and they are not of the same kind; separating them shows where the theory is untested and where it is under strain. The first is the easy case. Someone in Jakarta may know Dr Citra as the physician who runs the local clinic, and know Ms Dewi's daughter as the young woman who organises activities at the orphanage, without knowing that these are one person. "The community trusts Dr Citra" may then be believed while "the community trusts Ms Dewi's daughter" is not. This is a genuine failure of substitution, but the Fregean's account of it costs nothing, because one of the two expressions is a definite description. A description is built out of its parts, and its sense is composed of theirs — "the daughter of Ms Dewi" presents its bearer by way of a relation to another person. That descriptions have senses distinct from their bearers has never been the disputed claim. The hard case is two proper names, "Hesperus" and "Phosphorus", or Frege's own pairing of the morning and evening star. Here the Fregean owes an account of what the sense of a name is, and the compressed version's appeal to "the very richness of its conception of sense" is where the argument needs support rather than assertion. That is taken up in the next section. The second illustration needs correcting rather than expanding. A child in rural Vietnam may believe that the spirit of the forest protects the village, and separately believe that Ông Tư, the conservationist who prevents illegal logging, is a good man. As written, this is not a case of opacity at all, because opacity requires that the two terms share a referent, and whether they do was left open — the compressed version said only that Ông Tư "is regarded by some" as the spirit's embodiment. Two quite different cases lie on either side of that ambiguity, and the essay must choose. If the identity holds, the case is exactly parallel to Hesperus and Phosphorus, with the additional feature that the child has no route by which the identity could be discovered, so the two beliefs can never be integrated. If it does not hold, the case illustrates something else that Frege's apparatus handles, and handles rather elegantly: a sense may exist without a reference. Frege's example is "Odysseus was set ashore at Ithaca while sound asleep", which he takes to have a sense but no truth value, because the name has no bearer. Yet the belief report about the child remains true or false, and this is a point in the theory's favour rather than an embarrassment: within the report, "the spirit of the forest" refers to its customary sense, and that sense exists whether or not any spirit does. What the account costs Two objections have been pressed against this machinery, and an essay that claims the machinery is adequate has to meet them rather than record that it is rich. The fir