Topics Language

Question For any given term or concept, is it possible to formulate a correct definition? Some people claim all definitions are equally valid and subjective. I can't believe this though because if we can't agree on a definition, then you can't transmit your exact meaning to me through words, and the whole idea of communication is shot. How can definitions be rooted in reality and truth?

Accepted:
March 25, 2006

Comments

Take the concept of apples. There is no cosmic connection between that concept and the term 'apple'. It's just a convention, and indeed it's only a convention for those who speak English. Among English speakers, however, it does seem important for the sake of communication that we are following the same convention. Otherwise when you ask me for an apple I might give you a hamster.

There is however a deep philosophical worry about how it is possible for you to entertain a concept. The issue here is not how you know that other people's concepts are the same as yours, but how you can have any general concept at all. Take the concept of apple again. How do you manage to think, not just of this or that apple you may have seen, but of the set of all apples, the great majority of members of which you will never encounter? The trouble is that the apples you have seen are not just a subset of the set of apples, but also a subset of innumerably many other sets, such as the set consisting of apples you have seen plus hamsters you haven't seen. What have you got when you have the concept of apples that makes it different from what you would have if your concept was of apple-hamsters instead? This question is remarkably difficult to answer. To get a better idea why, have a look at Saul Kiripke's wonderful Wittgenstein on Rules and Private Language.

I will just talk ab0ut words, but the ideas apply to concepts as well.

It is very reasonbale to suppose that a typical word, such as 'apple', has a definite meaning. But then it can't be that all definitions are equally valid, since many will be inconsistent with the truth about what the word means, thus e.g. "'apple': a small glass or vessel for applying lotion to the eye."

I guess the question 'how can definitions be rooted in reality and truth' then boils to the quesiton of what determines the meaning of a word. And that's one of the really big issues in the philosophies of language and mind.

There are real issues about the extent to which words can be defined. People sometimes start out thinking that each word or concept has a definition in terms of conceptually necessary and sufficient conditions, on the model of "vixen': female fox". But in fact the vast majority of words seem not to have definitions like that: we just can't find ways of saying exactly what they mean.

It is not clear that we need agree on a definition if we are to be able communicate. Maybe we can communicate perfectly well even if one or both of us don't have any good ideas about how to define the words we are using. E.g. we both use the word 'apple' to talk about apples, but we don't know how to define the term. Notice also that there is a risk of regress. A definition for a word is just more words. So if we are to agree on a definition, then we have to understand the same things by the words in the definition. But then it looks as though we need to agree on definitions for those. etc. etc.

My own view is that if we disagree 0n a definition of a word we are using, then this does raise problems for communication: what I mean by it will differ from what you mean by it. But actuallythat view somewhat controversial. I discuss the matter a bit in my paper 'Cognitive Content and Propositional Attitude Attributions', which is on my website.

I will just talk ab0ut words, but the ideas apply to concepts as well.