Token, Type, Proposition: Write-Up of Some of Yesterday's Dialog

I am enjoying the pleasure of a three-day visit  from Dr. Elliot Crozat who drove out yesterday from San Diego.  The following expands upon one of the topics we discussed yesterday.

How many sentences immediately below, two or one?

Snow is white

Snow is white.

Both answers are plausible, and indeed equally plausible; but they can't both be right. There can't be both two sentences and one sentence. The obvious way to solve the problem is by distinguishing between token and type. We say: there are two tokens of the same type. One type, two tokens. That's a good proximate solution but not a good ultimate one. It is a stop-gap solution.

For the solution gives rise to problems of its own. And these may be expected to be as bad as, or worse than, the original one. We made a distinction between sentence-token and sentence-type to avoid contradiction.   But what is a sentence-type and how is it related to its tokens?  You see the tokens above but you do not see the type. The tokens are in space and bear spatial relations to each other and to other things.  The type is not in space. It is obvious that the tokens came into being and will pass out of being. But that is not obviously the case with respect to sentence-types. Such types are arguably a species of platonica. The tokens exist contingently, but this is presumably not the case with respect to  their type. The tokens are temporal items, but it doesn't follow that the type is. If the concrete is the causally active/passive, then the tokens  are concrete whereas the type is not and is therefore abstract in the Quinean though not the traditional sense.   If there are both abstracta and concreta, are both sorts of entity in time? Or only the abstracta?

Now consider:

    Snow is white

    Schnee ist weiss.

Here we have two different sentence-tokens of two different linguistic types.  It is reasonable to maintain that such types are necessarily tied to their respective languages, English and German, in the sense that, were the languages not to exist, then neither would the types exist.  But 'surely' human languages are contingent in their existence. If so, then the linguistic types are contingent in their existence, in contradiction to the strong tendency to view them as platonica, and thus as necessary beings.

Puzzles are erupting like weeds in Spring. I can't hope to catalog them all in one entry. 

But let's throw a couple more into the mix.  The two sentences lately displayed, the one in English, the other in German, express the same proposition or thought (Gedanke in Frege's lingo). Or at least they express the same proposition when assertively uttered or otherwise tokened by a speaker/writer competent in the language in question, a speaker/writer with the appropriate expressive intentions.

We now have token, type, and proposition to understand in their interrelatedness. It is obviously not enough to make distinctions; one must go on to inquire as to how the items distinguished gear into one another, fit together, are 'related.' To avoid this task would be unphilosophical.

One more set of questions. How do we become aware of types and propositions? We see with our eyes the sentence tokens on the page or we hear them with our ears when spoken. But we have no sense-acquaintance with abstracta. Do we 'see' them with the 'eye of the mind'? And how does that 'eye' hook up with the 'eye of the head'?

This ties in with another topic Elliot and I discussed yesterday evening: the difference between the literal and the metaphorical.   Is talk of the 'eye of the mind' and of visio intellectualis metaphorical or is it literal? What does it mean to be literal? Is the literal the same as the physical? And what is the difference between metaphorical talk and analogical talk? Can food literally be healthy? Or is food healthy only metaphorically or figuratively? Some dead meat is good food. But no dead animal or its flesh is healthy. For an animal, being alive is a necessary condition of its being healthy.

Are analogical statements about God literally true?  

 

Comments

[Elliott]

Hi Bill. I’m back in San Diego. Thanks again for your hospitality. We had a good time engaging in our own kind of peripatetic philosophy!

This was an interesting discussion. On the drive home, I thought about what you wrote here:

“It is reasonable to maintain that such types are necessarily tied to their respective languages, English and German, in the sense that, were the languages not to exist, then neither would the types exist. But 'surely' human languages are contingent in their existence. If so, then the linguistic types are contingent in their existence, in contradiction to the strong tendency to view them as platonica, and thus as necessary beings.”

The argument is something like this:

1. The existence of a linguistic-type is modally tied to the existence of its corresponding language.
2. All languages exist contingently.
3. Thus, all linguistic-types exist contingently.
4. Thus, no linguistic-type is an abstract/platonic object.

I agree that there is a strong tendency to view linguistic types as platonica. To support this tendency, what if we distinguish between a language-type and a language-token? English and German are historically developed systems of communication and therefore contingent language-tokens. But suppose there are language-types, such as English-type and German-type, and that these exist as platonica. With this distinction, it seems we can argue:

A. The existence of a linguistic-type is modally tied to the existence of its corresponding language-type.
B. All language-types exist necessarily.
C. Thus, all linguistic-types exist necessarily.

One problem with the idea of a language-type is that the relation between type and token is usually a one-to-many relation. But the relation between a language-type and a language-token seems to be a one-to-one relation. What's the deal with that? Perhaps one could address this problem by positing just one language type but many language tokens, such as English, German, etc.

Another problem with the idea of a language-type is that it raises a host of difficult questions similar to the ones you noted about sentence-types and how they relate to sentence-tokens.

[BV]

Hi Elliot,

I am happy to hear that you made it home safely. It was great meeting you and the three days were memorable. I hope we do it again.

Setting aside dialects, let us suppose we are dealing with one well-defined language system such as Hochdeutsch, high German, with its vocabulary and formation rules. And I suppose that we would have to relativize that system to a time since languages evolve. I am having trouble understanding how that language system viewed as a type could have even one token. Who could possibly token it? I 'll think about this some more. No time now.