The best thing that could happen for me professionally is if a bunch of smart philosophers make arguments for interesting philosophical conclusions that use dubious sporting analogies which can be exposed by those who spend large chunks of time watching the sports in question. I don’t really expect that this will become a commonplace in philosophy, but I’m glad to say I found one instance in Michael Glanzberg’s “Against Truth Value Gaps”:http://philosophy.ucdavis.edu/glanzberg/againstgaps.pdf.

Michael wants to argue that assertions that express propositions are either true or false, not in some way gappy. He wants to endorse what he calls a _one-set_ view, i.e. that you can characterise the truth conditions of a proposition by a single set of worlds, the worlds in which it is true. This is opposed to a view in which you need to say which worlds the proposition is true and then make a separate characterisation of the worlds in which it is false.

The core argument for this is that when we look at what the point of making assertions is, we see that they just have correctness conditions, not correctness conditions and separate incorrectness conditions.

bq. I shall argue below that once we take seriously the idea that a proposition is the content of an assertion a speaker might make, the pro-gaps intuition loses its force. Indeed, once we think about matters in these terms, there is a strong countervailing intuition against truth-value gaps. There seems to be no more sense to be made of a proposition that is neither true nor false than there is sense to be made of an assertion that neither succeeds nor does not succeed in saying something correct. There is none. Let us call this the no-gaps intuition.

Correctness conditions for an action are determined by the purpose of an action. Purposes are most easily understood by analogy to the purpose of sporting endeavours.

bq. Having an intrinsic purpose is a feature of a wide range of actions, but not all. Natural actions, such as moving one’s hand, seem not to have intrinsic purposes. To have purposes, they need to acquire them from the agents who perform them. But types of actions which are created by practices (in the sense of Rawls (1955)) typically do have intrinsic purposes. Clear examples are to be found in games, as both Rawls and Dummett have stressed. Hitting a ball may have no intrinsic purpose, but batting does. Its purpose is to hit the ball in whatever way produces home runs. Shoeless Joe Jackson may have an ulterior purpose of his own when he is at bat, but it is still the case that the purpose of batting is to produce home runs. In the case of games, the purpose of a move in a game is fixed by the game’s more general purpose: to win. Other practices may have more complex purposive structures.

The argument is then that the purpose of assertion is expressing truths, and from that the no-gaps conclusion follows.

bq. Once we see that truth is the intrinsic purpose of assertion, we may extract the consequence for the nature of truth values given in the [no-gaps] conclusion of the argument above. Assessing a proposition for truth is assessing whether the intrinsic purpose of assertion has been achieved. Though we can certainly make sense of truth outside of actual acts of assertion (as with all the better things we might have said), the truth of a proposition still comes down to the matter of whether an assertion of it—a genuine move in the practice—would have been successful in achieving its intrinsic purpose. Assessing for truth in a given world comes down to assessing whether an assertion would have been successful had things been some given way. Assessing for truth is assigning a truth value. Hence, what truth values there are and how they are assigned is fixed by the intrinsic purpose of assertion, and what ways there are of achieving or failing to achieve it.

It seems to me that these kinds of considerations do not really tell against the gaps view. Or at least they don’t tell against a substantive degrees of truth view, of the kind that I hold. And the baseball analogy helps us see this.

It simply isn’t true that the purpose of batting is to hit home runs. To be sure home runs are a nice outcome to hitting. But it’s not true that whether you hit a home run is the only consideration relevant to determine how successful a hitter you are. There’s a big difference between a 3-run triple and a strikeout. If we said that the point of hitting is _just_ to hit home-runs, we would say that both the triple and the strikeout were unsuccessful examples of hitting, which is not what we’d normally say. (Unless the team needs 4 runs and the next batter is, say, me, in which case the batter better hit a home run. But that’s not the _normal_ situation.)

The same thing I think is the case with assertion and truth. The most successful kind of assertion one can make is when one asserts something that is plainly true. But nothing follows from that about truth value gaps unless we say that all the ways of not making a completely successful assertion are ways of being equally unsuccessful. That’s not what happens in baseball, and the independent reasons we have for believing in degrees of truth are reasons for thinking it isn’t true in assertion either.

Having said all that, I tend to agree with Michael that many of the alleged examples of truth-value gaps are really examples of not making assertions. And I agree that 3-valued approaches to vagueness are not particularly plausible. But the no-gaps argument is meant to tell against more views than those, and I don’t think it really does.