Thursday, May 31, 2007

Can You Have a Demonstrative Thought About a Color?

Last night the workshop met for the final time this academic year to discuss Rachel Goodman's paper "Demonstrative Thoughts as a Response to Lewis". Both the paper and discussion were complicated and interesting. I'll just summarize a few central topics here.

Rachel's target was anyone who wanted to respond to Jackson's knowledge argument by saying that what Mary acquires when she leaves her black and white room is the ability to have demonstrative thoughts about colors. Jason and David tentatively suggested that they were interested in that way of describing what happens to Mary when she leaves the room during the last meeting of the workshop.

Rachel's strategy was to try to show that there are disanalogies between a paradigmatic kind of demonstrative thought that concerns objects individuated according to their location in space and time and putatively demonstrative thoughts that concern colors. If the disanalogies are great enough then it would be a mistake to say that what happens to Mary when she leaves the room is that she acquires the ability to have demonstrative thoughts about colors.

First Disanalogy

The central disanalogy that Rachel wanted to argue for involved the possibility of a certain kind of failure that is present in the case of demonstrative thoughts about spatio-temporal objects that isn't present (she claimed) in the case of (putative) demonstrative thoughts about colors. That failure is the following:

It is possible to have the thought That cup is blue, while thinking about a BOTTLE, and still successfully have an object-dependent thought about the bottle. That is, you can apply the wrong sortal and still succeed in having a thought that is about an object (as long as it is in roughly the right place in space and time). Rachel wanted to say that in such a case you still succeed in having an object-dependent demonstrative thought.

In contrast, Rachel claimed, you can't have the same kind of failure in the case of a putatively demonstrative thought about a color. So, for example, it wouldn't be possible to think That color is beautiful, while getting the sortal wrong and still having an object-dependent demonstrative thought. It wouldn't make sense to say that you managed to have a thought about a TEXTURE or a SHAPE, for example, if you took yourself to be referring to a color. It was on the basis of this disanalogy that Rachel claimed it wasn't possible to have demonstrative thoughts about colors.

Members of the workshop objected to this line of reasoning in different ways.

Jason didn't think you could have an object-dependent demonstrative thought in the case where you apply the wrong sortal to the cup.

Justin suggested that there was a corresponding kind of failure in the case of a color, if the sortal was chosen correctly. So, for example, you might think That pastel is beautiful, and be mistaken about the fact that the color you demonstrated was a pastel (maybe it was flourescent or neutral).

Second Disanalogy

At another point in the discussion, Rachel said that unlike demonstrative thoughts about spatio-temporal objects, thoughts about colors didn't involve a "mapping" of egocentric features onto objective features. When you have a demonstrative thought about spatio-temporal objects, you think about That cup both as located in space relative to you and as located in objective space. But in the case of putative demonstrative thoughts about colors, Rachel claimed that there wasn't an analogous mapping of subjective features (in this case, something like color phenomenology) onto anything objective. I objected to this suggestion because insofar as someone can recognize a difference between how things seem to him (say I'm wearing 3-D glasses and everything appears either red or green) and how those things really are colored, then there is the possibility of a "mapping" of subjective features of experience onto (more or less) objective features.

There was also discussion of McDowell's notion that having a demonstrative thought about a color involved the presence of a sample. Jason and David discussed the possibility of a thought that depended not on the presence of the object that it is about, but on the presence of some other object (the sample). We didn't make much headway on this topic, however.

After the workshop, we watched a discussion between Gareth Evans and P.F. Strawson on the nature of truth, filmed for the Open University in 1973.

This was the last meeting of the mind workshop for this year. The workshop will resume in the fall, with a new grad student organizer: Will Small.


Zed said...

Maybe I'm missing something obvious, but aren't Rachel's considerations really reason to think that Evans's account of demonstrative thoughts is designed specifically for thoughts about objects and, as such, cannot be automatically applied to thoughts about properties? Color's not really the issue at stake. What's at stake is whether Evan's account of demonstrative thoughts, and its particular forms of failure (and, hence, the externalist conclusions Evans draws about it), applies equally well to properties. I, for one, think that Rachel's worries show that someone who wants to apply Evans's account of demonstrative thoughts to the case of properties must identify the external-world-dependent features of such thoughts that make them similar to demonstrative thoughts about objects. The real point is that Evans's account of object dependent thoughts doesn't automatically generalize to thoughts about properties.

Nat Hansen said...

I could be wrong, but it seemed that Rachel did want to say that there was something particularly problematic about color, as opposed to other kinds of properties.

Would her worry about not being able to map "egocentric" or "subjective" modes of presentation of colors onto objective modes of presentation apply to other properties, like mass, or shape, for example? It doesn't seem like it would.

She was interested in the idea that colors lacked the kind of objectivity that characterized spatio-temporal objects. Some properties (the primary qualities?) seem about as objective as you're going to get.

Zed said...

There's obviously a lot going on here, so let me first try and separate some different issues.

First: mapping subjective modes of presentation onto objective modes. Well, nothing could be more complicated than this, but one way to think about what this involves in the case of color is in terms of an ability to map the colors one sees onto the color solid (i.e., the contemporary descendant of the color wheel).

Second: drawing as is/looks distinction for color. It's tempting, but I think ultimately misguided, to want to conflate this question with the first. They seem similar, until one reminds oneself that there are facts about how things look from particular points of view, for both colors and shapes. And we can get these facts wrong. I can think your face looks circular from the side, but be wrong about that. Similarly, I can think your car looks black in the sun, but be wrong about that. The point is that we can't explain shape or color illusions just in terms of how things look, more must be involved.

But I don't know if either of these are really what's at stake, because I'm still hung up on the distinction between objects and properties. What appears to be Rachel's central claim is that there's a disanalogy between demonstrative thoughts about objects and putatively demonstrative thoughts about colors: namely, that there's a certain kind of failure present in the object case that's lacking in the color case. But here's my point: substitute "shape" in for "color" in Rachel's own example:

'That shape is beautiful.'

Everything Rachel wants to say about her color example is true of this example as well. So she's not talking about anything particular about colors.

Finally, as for the idea that any of these considerations show that colors lack some kind of objectivity that spatio-temporal objects possess...well, I think that's a bunch of hooey. Maybe I should now get back to what I was doing before writing this comment, which was working on the chapter of my dissertation in which I argue that it's a bunch hooey.

chauncey said...

1. I have not read Rachel’s essay.

2. I’m not sure about this claim: “If the disanalogies are great enough then it would be a mistake to say that what happens to Mary when she leaves the room is that she acquires the ability to have demonstrative thoughts about colors.” A. Suppose that demonstrative thoughts about color are disanalogous to demonstrative thoughts about standard spatio-temporal objects. It’s not clear what conclusions one can draw from this. B. It’s especially not clear that one can conclude that Mary doesn’t acquire “the ability to have demonstrative thoughts about colors”.

Part of my problem here might be that Rachel has argued for the conditional that I am here calling into question. Maybe Nat omitted that for ease of exposition. Or maybe I’m just missing something (which is plausible since I don’t know well Evans’s stuff on demonstrative thoughts.)

3. It seems plausible to me that upon leaving the room, Mary acquires the capacity to apply color terms in predicative positions by “direct experience”. So, rather than acquiring “the ability to have demonstrative thoughts *about* colors”, Mary acquires the ability to apply color terms in predicative positions in demonstrative thoughts, an ability she plausibly lacked in the room.

If you think of the version of the story where Mary is initially color-blind, you might argue in the following way against me. Mary had the ability to use color terms in predicative positions in demonstrative thoughts while in the room. Specifically, she could use color terms as predicates in demonstrative thoughts reached inferentially via her masterful scientific knowledge. She could have thought ‘That is red’ while “ostending” some sample that she knows to have a certain spectral value.

Thus, I might modify my claim: Mary acquires the ability to noninferentially use color predicates in demonstrative thoughts. That seems more plausible.

From here, we might think that she acquires the ability to have demonstrative thoughts with color terms in the subject position or with colors in the subject position.

4. What is the form of a demonstrative thought *about* a color? 1. ‘That is red’ or 2. ‘That [color] is beautiful’. Clearly, the logical subject of 1 is typically a spatio-temporal object and thus 1 seems to be about that object and not about the color red. However, it’s not obvious that 1 need be about that object since there is a fairly standard use of 1 to identify redness for a novice. Thus, there is a sense in which 1 could be about red. And this is relevant to Mary’s case. When she departs the room, Mary develops the ability to think and say ‘That is red’ where she is not especially commenting on the object of which she is predicating redness, so much as she is exhibiting her ability to identify redness or red things. (I wonder whether this could be helpfully connected with Sellars distinction between reporting and fact-stating uses of color predicates in EPM.)

chauncey said...

woops. At the end of 3, I meant to say "...or with colors as the referents".

Jason said...

But here's my point: substitute "shape" in for "color" in Rachel's own example:

'That shape is beautiful.'

Everything Rachel wants to say about her color example is true of this example as well. So she's not talking about anything particular about colors.

No, because one of the things she wants to say about her example is that it concerns color. Ha! Gotcha!

Zed said...

Fine. I amend what I said before. I should've said "Everything that's relevant to a discussion of the limitations of generalizing Evans's account of object-dependent demonstrative thoughts can equally well be said about 'That shape is beautiful' as about 'That color is beautiful'."

You're right that Rachel wants to talk about colors. My point is that colors have nothing to do with her point. And your point is humorous, but beside the point.

Rachel said...

Thanks for these comments--sorry it's taken me so long to read them. Zed, you haven't missed something--this is an issue I've been thinking about a lot.
The paper I sent out was very much a first pass at these issues and there were two lines of argument, one of which was more developed than the other.
The more developed line was, I think, the one Zed suggests: The reason there are problems with extending Evans's account of demonstrative thought to putative demonstrative thoughts about colours is that colours are properties. It is hard to make sense of the idea that a thought that's about a colour (not a coloured object) is object dependent.
The line Nat has mentioned here is also one I'm interested in, though. Evans's acccount of demonstrative thought relies on the notion of egocentric concepts of space being mapped onto objective concepts of space. This enables a demonstrative thought about an object to fulfill the Generality Constraint because the fundamental ground of difference for physical objects is their objectively specified location in space and time. In giving an account of demonstrative thoughts about colour one would face the fact that the fundamental ground of difference for a colour might not be objective.
If Zed's right and it's a bunch of hooey (which, I take it, means nonsense) that colours lack objectivity, then this would be relevant. I'm going to do some more reading on colour over the summer and I'd love to talk more about it at some stage, Zed.
At the moment, I'm inclined to think the line of argument that concentrates on the difference between an object and a property is more promising for establishing that perception based thoughts about colours do not have the characteristics of demonstrative thoughts (that is, object-dependence and special possibilities for failure within the scope of success).
So, Zed, I agree with a lot of what you say, but I still think the objective/subjective issue is involved in some interesting way. A quick point about the examples you raise (thinking a face looks round or a car looks black but being wrong about how these things look). The problem with these examples seems to be that, in order to make space for the error, we need to be talking about the properties of objects: the face doesn't really look round, the car doesn't really look black. When we try to just talk about 'that shape' or 'that colour' things get difficult. Without an object, it doesn't seem like there can be the right sort of mistake. Not to sound facile and unhelpful, but objects seem to have something to do with objectivity. Anyway, I don't want to just give up on the objectivity question yet.