Although Michael Dummett recognises the problem of ambiguity when it comes to our linguistic expressions and utterances, he still believes that thoughts or propositions aren't ambiguous. He accepts contextuality, for example; though thoughts or propositions include context or get the context right:
"It depends on the context how flat a thing has to be to be rightly called flat. Driving along you could say the country is very flat, but that doesn’t mean that there’s no variation in it. On the other hand, if it’s a billiard table, there have got to be pretty exacting standards. With all these things, we judge what someone meant from all sorts of indications, including what he is likely to have meant, what he is likely to have been saying, and I think there’s no way in which it is possible to circumscribe the context in such a way that it will determine a unique interpretation. The interpretation is a matter of selecting what thought, in Frege’s sense, a person was expressing or was asserting to be true. And so it determines the conditions for truth." (220)
Perhaps the speaker didn’t mean anything more to what he actually said. Why assume he meant more even if his utterance isn't precise or explicit? Why assume that there must be a proposition or thought behind it, as it were? To ‘judge what someone meant’ may not even make sense. If a person meant something else by what he said, then wouldn’t he have said something else? Isn’t there a certain philosophical arrogance to the idea of thinking that someone actually meant X when he said Y? There may be logical or philosophical reasons for why he should have said X rather than Y. This doesn't mean, however, that he meant X by Y. Indeed if you tell the speaker that he really meant X by Y, perhaps he will be surprised or dismissive of the claim that someone else really knows what he meant; though he himself didn’t really know what he meant or that he meant X when he actually said Y.
If there is any ‘interpretation’ of another person’s utterance at all it is only an interpretation which we would require in order to understand him. Either that, or the interpretation is simply a matter of what we think, perhaps as philosophers, he should have said. Thus, in a sense, if he doesn't make sense at all or if he is not precise and exact enough for us, then if we can interpret him at all, we can only do so if we apply Donald Davidson’s ‘principle of charity’ to him in the sense that we must believe that he must have been trying to communicate with us successfully and that he could have done so had he only tried harder, as it were.
Dummett actually accepts that there will be no ‘unique interpretation’ of the speaker. However, the interpretation will still depend on the true Thought expressed by the utterance. This is a strange conclusion. If there is no unique interpretation, then why is the thought or proposition expressed seen as explicit, precise and determinate at all? If we need to ‘select the thought’ surely this hints at the possibility that there is no determinate thought being expressed in the first. If the thought is determinate, then we wouldn't need to select it; we would simply know it or discover it. If, on Dummett’s own admission, we must select the thought that's expressed, then we may select the wrong thought or proposition.
The word ‘selection’ implies that there are various thoughts or propositions to select from. If that’s the case, why do we see thoughts as determinate in the first place? If there is a multiple choice, then at one time we may select thought X, and at another time we may select thought Y. Presumably there will be a difference between thought X and thought Y. Thus, whence the determinacy, precision and explicitness? Perhaps X and Y are radically different from one another. Alternatively, if there really is a selection process going on in our interpretations of other people’s utterances, perhaps this should lead us to suspect the very existence of propositions or thoughts in the first place. In that case, the utterer meant what he meant. He didn’t mean what we think he meant. And he didn’t mean what we think he should have meant. We can't impose an ought-of-meaning on an is-of-meaning. However, this doesn't stop us from saying that he should have said X rather than Y. What we can't say is that because he should have said X, that must be what he actually did mean by his Y.
Dummett, in a sense, goes on to stress the problems I've been stressing about interpretation and the alleged existence of determinate thoughts or propositions behind, as it were, our expressions. He says:
"But I think the fact that there is no way of laying down exact rules for what interpretation is to be adopted, that is a strong motive for saying that truth attaches to thoughts or propositions and not to sentences." (220)
From the conclusion that there are no "exact rules for what interpretation to be adopted" when it comes to people’s utterances, Dummett concludes that there must be thoughts or propositions behind our utterances. However, if there are no exact rules for interpretation, perhaps this should lead us to the contrary conclusion: that there are no thoughts or propositions to get right in the first place.
Not only that: Dummett also concludes that because of this lack of exact rules for the interpretation of people’s utterances that it must be the case that "truth attaches to thoughts or propositions and not to sentences" (220). That is, because of our sentential interpretations aren't bound by exact rules and therefore there is nothing to stop us offering various or many sentential interpretations of what someone said, then this must mean, according to Dummett, that truth can't be ‘attached’ to sentences at all but only to propositions or thoughts. The very fact that we can come up with various or many sentential interpretations of a person’s utterance means that we must not, or should not, attach truth to these sentential interpretations at all. After all, they can’t all be true. Instead, it's only the genuine thought or proposition which hides behind these various or many sentential interpretations that is the legitimate owner of truth.
Again, these various or many sentential interpretations should lead us to conclude that there is no proposition or thought at all behind them, not that they lead us to conclude that there must be such things behind them. My conclusion about the multiplicity and variability of sentential interpretations of an utterer’s thought should really lead us to a conclusion that is the exact inverse of Dummett’s own conclusion. Rather than multiplicity and variability leading us to the propositions or thoughts that are hiding behind our expressions or interpretations, they should lead us away from a belief in the existence of such abstract entities.