5.542 It is, though, clear that “A believes that p”, “A thinks p”, “A says p” are of the form “’p’ says p”: And here it is not a question of a coordination of a fact and an object, but rather of the coordination of facts by way of the coordination of their objects.

My relevant dictionary entry on this says: “Wittgenstein’s view seems to be that one cannot judge or believe or think a piece of nonsense. One can believe that Bush is a good President or that Bush is not a good President, but one cannot believe that cockadoodledoo. This is a matter of logic, not psychology, as Wittgenstein sees it. To think (that) a proposition p (is true) is somehow (it does not matter how, except to a psychologist) to represent, express, or picture p, just as saying “p” is. So “A thinks p” means something like “Something represents/expresses/pictures/says p.” This something could as well be a sentence as anything else, such as a person. So, as Wittgenstein puts it in proposition 5.542, “A thinks p” has the same form as “‘p’ says p.””

Mounce (pp. 85-86) says that Anscombe makes a mistake here. She fails to distinguish between the contingent fact that a person A happens to have uttered p, and the non-empirical fact that p means p. Of course, the sounds or marks that make up p might have meant something else, or nothing, but given their meaning, it is not contingent that they mean p. According to Mounce (p. 86): “The point is simply that B can convey to us what A says (or thinks) simply by telling us what sounds he utters. How is this possible? Well, first, because these words possess logical form; and second because, since we ourselves have a grasp of logical form, understand a language, we do not have to be told what these say; we can tell that for ourselves.” Of course, Mounce also points out that what is believed is not an object in the ordinary sense, because what is believed must make sense.

Anscombe (p. 88) says: “It is perhaps not quite right to say that ‘A judges p’ is of the form ‘”p” says that p’; what he should have said was that the business part of ‘A judges that p’, the part that relates to something’s having as its content a potential representation of the fact that p, was of the form ‘”p” says that p’: ‘A believes p’ or ‘conceives p’ or ‘says p’ must mean ‘There occurs in A or is produced by A something which is (capable of being) a picture of p’. We should here remember the letter to Russell in which he said he did not know what the constituents of thoughts were, but he was certain that a thought must have constituents corresponding to the words of language.” The letter in question is discussed on p. 28 of Anscombe.

Friedlander (p. 113): “Only an internal connection between the act of thinking or judging and the constitution of the judgment is capable of explaining why a subject cannot judge what is not sense, what is nonsense.”

Cf. PI §358. One cannot mean a senseless string of words, and so an act of meaning is not what gives sense to (otherwise meaningless) strings of words.

Black (p. 300) says that the fact in question is the fact that A spoke the words A spoke and that the object is A.