Opiniatrety

Post with Ulterior Motive--Do Not Waste Time Reading

If I were going to try to make a methodological point by linking to this discussion, it would be this:

To some extent, we are free to stipulate that our terms mean whatever we want them to mean, so long as we are clear that we are so stipulating. I could stipulate that I am using the word 'know' so that "S knows that p" is true if and only if p is true--as long as I say warn you up front, you can't protest that the word 'know' isn't used that way. I'm using it that way--that's what I just stipulated. (Speaking of upfront warnings, reread the title of this post.)

Yet some of these stipulations are bad stipulations. The foregoing stipulation is bad because it takes a word that's in ordinary use ('know') and stipulates that it has a meaning that really can be taken care of by another word in ordinary use ('true', give or take a lambda-abstraction). Other stipulations will just be plain awkward to work with. If I stipulate that "S knows that p" means either that S truly believes that p or that S was told p by a one-armed parrot dealer, then the things I go on to say using 'know' are not going to be useful or interesting.

So patently illogical stipulations are bad. Nevertheless, there can be cases in which one stipulation may be slightly more or less logical than an alternative; but the best thing to do is simply to settle on a stipulation upfront. The gain from using the more logical stipulation is less than the loss from determining that it is more logical, arguing that it is more logical, etc.

That's essentially what I was trying to do here. Or so I'll claim if anyone asks me about it.

(Really, the purpose of this post is to try and open up the CT comments to Google, so that this search--AOTW yielding no hits--makes some progress.)