Thursday, July 12, 2007

An elementary question

Why is the use/mention distinction important? As far as I can tell, it goes back to Quine, who was very big on it. Some people since him have picked up on it and used it to argue for various things, such as Harman's sharp distinction between inference and implication. I have a vague sense that there is a connection between it and Quine's extensionalism. Both are used as criticisms of modal logic. Modal logic was, according to him, conceived in the sin of use/mention confusion. But, modal logic has been vindicated. What has the use/mention distinction given us? Is it just an artifact of curmudgeonly Quinean philosophy? I currently have no idea.

On a related note, I had an odd experience recently when I realized that my entire undergrad logic education, including a couple modal logic courses in the Amsterdam style, never once touched upon use/mention.

4 comments:

Kenny said...

I think it's actually not that important in logic per se, but more in language and semantics. Obviously, a word and what it refers to are (generally) different things, and if you're going to talk about any sort of correspondence between words and the world, you have to keep straight which you're talking about at any point.

Shawn said...

That might be the right approach. What makes me a little hesitant is a recent review of Goldfarb's new logic book. The book includes a discussion of use/mention and the review praises it as being a big deal that a logic book finally covers it in depth.

I agree that one needs to keep straight what one is talking about when doing semantics or philosophy of language, but, I worry that the people who make much of use/mention take that point farther than it needs to go. For example, Harman takes the simple point and runs with it, drawing significant conclusions about logic. (I'm not comfortable saying much more on him since I need to go through his papers/books in greater detail). It is quite possible that I'm confused on how the use/mention distinction fits into Quine's philosophy of logic...

Matti Järvinen said...

I was under the impression that use/mention-distinction reflects the object/metalanguage distinction. It's important to be aware when you are operating in the metalanguage and referring to object-language expressions, and when you are operating in the object-language.

Shawn said...

The object/meta-langage idea might track use/mention to a certain degree. However, on the Quinean picture of modal logic, first or second grade (I forget which), we have object language use and mention. In philosophy of language generally I don't think use/mention tracks object/meta-language though.