2

What's an intuitive way of thinking about type versus token identity?

I have read and understand the difference between those two terms: tokens are things in space and time; types are abstract and what tokens are occurrences of. So to enumerate tokens we look at things in space and time, and to enumerate types we group tokens together.

e.g., the sentence

a rose is a rose is a rose

as it appears on the screen has three tokens occurring of the type the word rose.

Please correct me if any of that is mistaken. The issue, however, is that I can't quite get inside, intuitively, the distinction.


e.g., why token identity is weaker than type identity. Given that there are more tokens then types, I don't understand why it's not the case that positing token identity groups more things together, so that it is the stronger claim. If the answer to that is that type identity trivially includes token identity, then what is type identity in addition to token identity?

  • See Types and Tokens: the issue of what Types are is not simple. – Mauro ALLEGRANZA Nov 5 '17 at 19:27
  • "Type physicalism... asserts that mental events can be grouped into types, and can then be correlated with types of physical events in the brain... token identity physicalism... argues that mental events are unlikely to have "steady" or categorical biological correlates" Wikipedia. So mental tokens of a type are identical to some physical tokens, which are of of course of some type, but they are not identical to them by virtue of the types involved, the type correspondence is accidental. – Conifold Nov 5 '17 at 21:09
  • Let us continue this discussion in chat. – user29299 Nov 5 '17 at 23:21
  • It’s not obvious that there are more tokens than types: on some views, English contains infinitely many sentences (sentence types), but only finitely many have been tokened. Also, what do you mean when you say that “token identity groups more things together”? As I understand it, token identity groups each thing together with itself only: if t1 and t2 are the same token (of type T), then they are the same object, no? Note that tokens belong to types (a bit like to a set) and aren’t identical to them. (Of course, some types might have other types as their tokens.) – MarkOxford Nov 6 '17 at 13:48
  • @MarkOxford There are a couple better ways of pointing out this ambiguity: 1) say your type is 'a dime', then your tokens are all the individual dimes. Those dimes are also tokens of the types of round things, things with ridges, things with faces on them, things that are silver, things with a fixed monetary value... Also 2) each token is at some microscopic level absolutely unique, so that microscopically exact description is itself a type with exactly one token. – jobermark Nov 6 '17 at 20:25
1

An example I have seen is to compare dogs with your dog (or any other pet). Your dog is a dog because it meets the characteristic requirements of the type dog. Your particular dog is a token of that type.
To illustrate further let us consider a piece of art. Say you have some oil paint on a piece of wood. This is an oil painting (Type) and a particular piece of art (Token). If the artist then adds other media to the wood (say cloth, metal, string, etc.) it stops being an oil painting (Type) and becomes mixed media (Type). It is still the same piece of art (Token) though.

  • interesting illustration. not sure how it links to 'token identity', but viable, thanks – user29299 Nov 7 '17 at 19:32
0
  1. Take a coin from your pocket. Suppose it is a 10-cent euro. In terms of type/ token analysis the euro is a type of monetary unit, different from the US dollar or the English pound. You can't spend the euro but you can spend, save or throw away the coin you are holding, which is a euro token, a physical object (unlike the euro itself, which is just a category of money).

  2. Tables can be switched. What is a type from one perspective can be a token from another. Back to the euro : if we take money (a medium of exchange, a measure of value, &.) as a type, then the euro, the currency, is a token, an instance of money.

  3. If the language is familiar, type/ token analysis can be seen as a version of the universal/ particular distinction. That's if one allows that distinction, which is another matter.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy