Friday, September 13, 2019

Informative characterizations

It is hard to characterize an “informative characterization”. Here is an instructive illustration.

Ned Markosian in his famous brutal composition paper says that an informative, or non-trivial, characterization of when the xs compose something is one that is not synonymous with the statement that the xs compose something. But by that definition, here is a non-trivial characterization of when the xs compose something:

  • water is H2O and the xs compose something.

This statement is not synonymous with the statement that the xs compose something. Nor are the two statements provably equivalent. Nor are they a priori equivalent. But they are metaphysically necessarily equivalent.

Van Inwagen in Material Beings proceeds seemingly more restrictively. He wants a characterization of when the xs compose something that doesn’t use mereological vocabulary. But here is such a characterization:

  • the xs have the property expressed by the actual world’s English phrase “compose something”.

This characterization mentions mereological vocabulary, but doesn’t use it. And if we want, we can avoid mentioning mereological vocabulary as well:

  • the xs have the property referred to in the second bulleted item in this post in the actual world.

Obviously, none of these characterizations of “compose something” are informative.

1 comment:

Michael Gonzalez said...

It's interesting that you say the first statement (about H2O) is "metaphysically necessarily equivalent" to "the x's compose something". Do you take metaphysically necessary equivalence, in a case like this, to mean something like "if we fully understood every part of the sentence, and what is being conveyed, we'd see that nothing more or less is conveyed by it than by the other statement"?