Wednesday, September 26, 2007

Millianism = Platonism?

I have fairly little background on this, and hopefully I'm way off (because being right would make me sad).

Here's the checklist of similarity:
(1) There are concrete things: check. (duh)
(2) There are also abstract things: check.
(3) These abstract things exist in a publicly accessible domain: check.
(4) These abstract things are not mind-dependent: check? P:check. M:tentative check?
(5) These abstract things are the objects of thought: check.
(6) When we speak we express a token of the type: check.

(7) Do we token when we think, or do we interact with the type directly? I don't know what either says about this.

Chris, or anyone for that matter, please tell us we haven't regressed a few thousand years. What's the difference?

Also, I'm guessing Fregianism denies (4), and maybe (3). Correct?


Adam said...

To deny (4) would be to hold that the contents of thoughts are existentially mind-dependent.

I believe Frege would deny this.

From the little Frege I have read, it seems clear that he thinks that (i) thoughts are public, (ii) thoughts are eternal, and (iii) we 'grasp' or apprehend (public) thoughts when we think stuff. (I think that) Frege thinks this explains how you and I can have the same thought. Most of this stuff comes from the 'The Thought' essay.

Given this, it also seems likely that Frege would not have assented to the negation of (3), if by 'abstract things' we here mean such abstract things as thoughts or concepts.

Why "regress"? Isn't philosophy just a big footnote to Plato anyways?

Just kidding. Why 'regress'?

Jay Christie said...

Hmmmm, that seems like a reductio to me. If x = platonism, then x stinks.

As for why 'regress'? Well, at least for me, it's sort of disheartening to think that the modern theory is just a revamped 2k year old theory. To think that meanings exist without language users is, I find, preposterous. Why would we want to posit that in our universe? In the universe without language users it plays, by definition, no causal role. Much like we wouldn't posit things like 'beauty' and 'good' without judgers, but the platonist, and it seems the Millian, finds this perfectly acceptable.

I was thinking Frege, or at least Fregianism, would deny (4) because the contents of thoughts are modes of presentation; straightforwardly so for individuals, and the same sense/references distinction can be made for predicates (which is an option which Chris mentioned some Fregians take).

This allows a division between concrete properties (the referents, when the sense actually has a referent) and existentially mind-dependent senses which are the objects of thought. If this is an accurate account, that's far more palatable to my tastes.

This is armchair armchairing since my Frege knowledge is limited, but given Chris' presentation last class, this account seems to accord with it.

Dan said...

Also armchair armchairing here, but I thought I'd pipe in.
Perhaps I missed something, but I don't think the linguistic theses of millianism alone commit you to much metaphysically. The starting point was sentences, and the distinction between sentence tokens and sentence types. Never was it outlined what these are meant to be in a metaphysical sense (at least not in class). If one thinks there are no un-instantiated properties (and that being of a type is a property of a token), then it fits in perfectly with millianism that there are no sentence types without tokens. If one believes that tokening is dependant on a speaker knowing what he's doing, then ALL of this is dependant on comptetent language users.
This is all speculation of course. I'm just noting that the metaphysics hasn't been laid out here. My guess is that millianism read from certain metaphysical viewpoints commits one to little more than actually instantiated properties and relations.

Jay Christie said...

Where's Chris? I bet he's read this and is sitting back quietly, letting the students fill in the gaps (thus encouraging us to read or something silly).

As Dan alludes to, if the initial tokening brings a type into existence (dependent upon language users), that wouldn't be full-on platonism and considerably less spooky. However, although the metaphysics hasn't been fully laid out, if we wish to avoid having types which have not yet been instantiated, I think we are going to have causality problems.

We have 2 events; a token (a concrete object) being instantiated, and a type (an abstract object) being instantiated. If concrete things can't causally interact with abstract ones, how is this causal chain going to be linked?

Note: if you don't think tokens are concrete, run the same argument and insert a person muttering sounds as a 3rd event which is concrete and needs to be linked.

If the type is the cause, then it exists pre-tokening. But we don't want that.

If the token is the cause, well, it isn't able to create an abstract object, since that would be a causal interaction.

Having them pop into existence simultaneously prompts the question 'what was the cause of them?' That cause was either concrete or abstract, and then we have the problem anew.

Dan said...

I think Chris likes to encourage students to puzzle these things out for themselves (at least on the blog).
I think if we're to consider abstract objects in the first place, it would be a mistake to think they're causaly related to concrete objects (or each other for that matter). These aren't ghosts after all. If a property is simply a way something is, and if properties don't exist until something is that way, it comes into existence when something becomes that way. That's rough I know, but it doesn't seem entirely implausible.
Take the classic example of blueness (substitute your favorite property if you think blueness is riddles with problems like mind-dependance, vagueness, subjectivity and whatnot). So, reasonably (along these lines) if nothing is blue then blue doesn't exist. Now something becomes blue. There was no causal interaction there between an abstract and a concrete. The relationship is completely different. I instantiated properties exist but uninstantiated properties don't, it seems perfectly natural to say properties come into being when they're first instantiated (ignore necessary existence for now).

Chris Tillman said...

This paper interestingly bears on the causal argument and may be illuminating to all sides here:

And yes, I think there's more educational value for all parties when you discuss issues as opposed to me just saying something and all discussion coming to a halt.

That said, some brief remarks:

1. Adam's right about what Frege's view is but I think Jay's concern is at least in part with what the view should be. Frege, for his part, was an unabashed Platonist and held that modes of presentation were publicly accessible, mind-independent entities.

2. Dan's right that Millianism per se is not committed to much metaphysically. Remember that if it's just a thesis about semantic content of proper names and the like, it's not even committed to the existence of propositions.

But Russellianism or the more full-blown view may seem to have more substantive metaphysical commitments. Here's where one contemporary Russellian argues that no such commitment is incurred:

The most relevant bit is in section 3.1, but the entire entry is excellent and reinforces some of the points briefly touched upon in class.

So I think the degree to which Jay should worry about Russellianism is probably exactly the degree to which Caplan and Matheson and King are wrong. A worthwhile endeavor would be to carefully sort out whether that is the case.

Finally (for now), Russellianism or Fregeanism would not merely be Platonism warmed over if the most plausible versions of the views turned out to involve a commitment to mind-independent abstract entities. It's wrong to suggest that Fregeanism and Russellianism would thereby constitute no advancement of our understanding of semantics over that of Plato's.

Okay, I'll stop and sincerely hope I haven't stopped this important discussion. Thanks to Jay, Adam, and Dan for their contributions so far. Others should feel welcome to add what they think. Please don't let apprehension prevent you--we're all friends here!

Jay Christie said...

I'm reading the Ben & Carl paper and will post back when complete.