On Philosophy

October 30, 2006

Scraps: Homophonic T Sentences

Filed under: Language — Peter @ 12:00 am

Currently I am engaged in writing a critical response to the paper “Knowledge of Meaning and Theories of Truth” by Richard K. Larson and Gabriel Segal. It is quite an interesting paper, so interesting that I had more to say about it than I had space. However, I have no such constraints here. So what follows is an omitted section, responding to their claim that even a homophonic T sentence is informative. (A homophonic T sentence is one in which the object language and the metalanguage are the same language).

No matter what role they are to play in understanding meaning it is clear that the T sentences yielded by a T theory should provide us with some new information; they shouldn’t be trivial. Triviality would imply that the T sentences are things that could be known a priori, much like an assertion to the effect that a = a. Although such an assertion is true it isn’t informative about the objects themselves, and is something that we can know to be true without any knowledge of those objects. So if T sentences are trivial it isn’t possible that generating them via a T theory could be informative. Larson and Segal claim that the triviality of T sentences is just an illusion generated by homophonic T sentences, which appear trivial only because if understand the homophonic T sentence we already understand both the object language and the metalanguage, and hence there is nothing for them to inform us about. Clearly T sentences in which the metalanguage isn’t English don’t appear trivial, and so the triviality of the homophonic T sentences must be an illusion, and that they really do say something substantive.[1]

Or must they? Just because some statements in a form are non-trivial doesn’t mean all statements in that form are non-trivial. Consider equality. a = b isn’t trivial, but a = a is, it is not the case that because a = b isn’t trivial that a = a really says something substantive. Given this it seems perfectly reasonable to suppose that homophonic T sentences are trivial, even if their multi-linguistic counterparts aren’t. There are two conclusions we can draw from this. One is that T sentences only are informative about meaning when they act in a translational capacity, i.e. when the person possessing the T sentence yielding theory speaks only the metalanguage then they convey the meaning of the previously unknown object language. The other is to assume that the metalanguage of apparently homophonic T sentences is some kind of pure propositional content, which we express in English because we are unable to communicate it directly.[2] Thus possessing such T sentences would allow the person who knows them to go from sentences in a language to facts about the state of the world, and this might be entertained as a description of how meaning in a speaker’s primary language works. Neither interpretation would invalidate the remainder of Larson and Segal’s paper, so long as we remember that homophonic sentences are simply stand-ins for the real T theories.


1. Specifically they claim “But it certainly does not follow from this that what is said by such a T sentence [a homophonic one] is not highly substantive.”

2. It sounds easy, but there may be a hidden problem with deciding that the metalanguage is pure propositional content; a representation of some state of affairs in the world. If it was the case that the metalanguage was pure propositional content then “is true” would have to be something that can be part of a state of affairs. However there is some reason to believe that the notion of truth is a linguistic construction. Which is not to say that truth doesn’t exist, only that it is something that can be predicated of language, and not part of the pure propositional content proper. We could adapt our theory to say that in the sentence “‘Q’ is true if and only if P” P alone is pure propositional content, and that the metalanguage is some hybrid of pure propositional content and something else. A bit ad hoc though.


Blog at WordPress.com.

%d bloggers like this: