Tuesday, January 23, 2007

Green is the Color

A while ago I read this post on Weatherson's blog. When I first saw Weatherson's powerpoint slides I thought "wow, that's kind of wierd." But, recently, I looked over more "color illusions" here and here. When I saw the spiral green and blue on the first of the sites linked to in the last sentence, I though "No way is that an illusion". I decided that any theory of color that entails that the green and blue spiral is an illusion must be a false view about color. I then came up with a rough thesis that would not have that unwanted consequence. Unfortunately, I discovered that Daniel Nolan presented roughly the same thesis in his comments on Weatherson's original post. In spite of that fact, I am going to talk about the thesis a bit any way and consider the best objection that I have come up with to that thesis.

The thesis is that, roughly, the color of an object sometimes depends on the things going on around the object. It might be put as a supervenience thesis. For example, perhaps the color of an object supervenes on the topological and geometric properties of that object and of the things nearby that object.

Thus, although the geometric and topological features that result of one of the spirals in the picture that has inspired me are roughly the same as the features in the other spiral, it seems that those featrues are slightly different in the surrounding objects. This is what makes one of the spirals green and the other one blue. Since the colors depend on the surrounding objects, we cannot demonstrate that they are the same color by covering up the surrounding area (as is often done). When we cover up things around the object, we change the color of the object.

Here is an objection, inspire by a conversation with Andrew a few days ago. Suppose we have one of our spiral pictures. We also have a wall situated several feet away from the spiral picture. The wall has holes in it so that an oserver on the other side of the wall can see the spirals but cannot see the stuff surrounding the spirals. Such an observer, the objection goes, would see the colors as the same. But an observer on the other side of the wall, that is, an observer who can see both the spirals and their surroundings, would see them as two different colors. One might think that we should say that the spirals are the same color (to the first observer) but different colors (to the second observer). But no two spirals can both be the same color and different colors at the same time. So, the view in question is incorrect.

Here are the responses that I am considering:

1. The first observer is just mistaken. Since the wall is partially blocking his view, he doesn't have the relavent information for determining the true colors of the sprials.
2. The first observer will in fact see the sprials as two different colors (this can be confirmed by expirment). This might be the case if the cause of the different colors observed is some kind of interference between the light waves bouncing off the spirals and the light waves bouncing off the surroundings. The interference will have already taken place before the light waves from the spirals travel through the holes in the wall. So, the first observer will in fact see the colors as different. This response will, of course, be subject to empirical confirmation or refutation.
3. Both observers are correct. One object (such as the spiral) can in fact be two different colors at the same time.

I like the first response best and the last response least. So, that is what I have been thinking about. What do you guys think?


Blogger Chris Tillman said...

Here's step 2 from Adelson's site for how to convince yourself that the squares are the same shade:

2) Cut out a cardboard mask.

By viewing patches of the squares without the surrounding context, you can remove the effect of the illusion. A piece of cardboard with two circles removed will work as a mask for a computerscreen or for a printed piece of paper.

8:41 PM  
Blogger Chris Tillman said...

Btw, here's step 1:

1) Try putting the image in Photoshop.

Using the eyedropper tool you can determine
that the RGB values of the grays in both square A and square B are 120-120-120.

I find this rather compelling. I'm inclined to direct realism, however, and I don't think it's at all implausible that something like modes of presentation are at work in perception cases. So maybe that's some autobiographical info on why I'm not compelled by the Nolan/Spencer thesis. But once we appeal to something like spectral analysis (or of course the vastly more scientifically sensitive tool called 'Photoshop') and we learn that the factors we think make for real differences in color are absent, we should accept that there is no difference in color here. (Yes I am equi-surprised by the spirals. I'm especially freaked out by the two circles where the inner one contracts and the outer one expands. I'm tempted to describe the phenomenology as follows: they seem to be moving, but at the same time they do not seem to be changing position. This is highly upsetting to me.)

8:47 PM  
Blogger Joshua said...

If the cardboard mask expiriment really results in the experiences purported, then this merely undermines my second response to the problem presented in the original post. It is still the case that my view might be defended by appealing to either the first or the second responses in the original post.

Any technique that involves covering the surrounding area (such as the photoshop technique) is insufficient for demonstrating the falsity of the view. Moreover, it doesn't matter if both squares have the same RGB values. The RGB values (according to the view being considered) may not correspond perfectly to particular colors.

Finally, I am not sure why direct realism has any bearing on whether the view is true or not. I can believe direct realism and yet think that the color of an object supervenes on the light reflective properties of the object and of the surrounding objects. Consider an analogy. I can believe direct realism and claim that I can see when one object has the property of being five feet from something even though having that property supervenes on both the object in question and the surrounding area.

Perhaps there is some way to combine the evidence that results from the cardboard mask expiriment and the claim that direct realism is true to show that my view is false. But I don't immediately see how to do it. I'll keep thinking about it.

10:49 AM  
Blogger Joshua said...

One last thing. I am not sure I know what you mean by "Direct Realism". I took it that direct realsim was a denial of indirect realism. Moreover, I thought that indirect realism was best put as a sort of perceptual analogue to representational theory of mind with respect to propositional attitudes. That is, representational theory of mind says that we believe a proposition in virtue of standing in some special believing* relation to a certain kind of entity that represents the propositions. Similarly, indirect realism says that we percieve an object in virtue of staning in some special percieving* relation to a certain kind of entity that represents the object of perception.

But if this is, roughly, the best way to state the indirect realism thesis, then the claim that modes of presentation are involved in perceptual processes is a kind of indirect realism.

I understand that many indirect realists leave out the part of the view that says that there is some kind of special percieving* relation. But a view that leaves that part out is pretty stupit and charity pushes to me think that they mean something more like what I was indicating above. So, in short, I think that, under the best construal of indirect realism, you are an indirect realist rather than a realist.

10:57 AM  
Blogger Chris Tillman said...

A. I was not providing any argument that your claim was false. I realize now it is consistent with direct realism, as I was thinking about it, and I am also aware (as I was when I wrote the comment) that a proponent of the Nolan/Spencer view would deny that color properties depend solely on things like RGB values. My comment included the undefended crucial premise that things like RGB values are all that matters to color. I did not support that claim and I know that you are prepared to deny it. I find the claim intuitive, however, so I was merely registering a reaction that leads to an inclination contrary to the Nolan/Spencer thesis.

2. I was assuming that direct realism/indirect realism was a matter of the relata of the perceiving relation. This may have been a false assumption. I assumed that if the second relatum of the relation is something like a sense datum or a mode of presentation, then indirect realism is true, and if the second relatum is "worldly" stuff, then direct realism is true. These assumptions may have been incorrect. I concede that it doesn't matter much, since N/S is compatible with direct realism as I was thinking about it. I guess part of why I mentioned it is that one who is inclined to think that only RGB values matter to color could handily explain why things seem otherwise in terms of modes of presentation.

A long-standing concern that I've had is when appeals to modes are good moves and when they are bad moves. I don't know whether there are general principles about this. It might just be case-by-case.

12:34 PM  
Blogger Chris Tillman said...

Would you hold, similarly, that these actually are moving: http://www.psy.ritsumei.ac.jp/~akitaoka/rotsnakee.html

9:06 AM  
Blogger Joshua said...

I just got a chance to look at those crazy snakes. It was not quite was I was expecting. I would like to know the underlying cause of this apparant movement and also the underlying cause of the apparantly different colors in the original cases. If both are caused by the same sort of color interference, then I might be willing to accept that there is color movement in the snakes case.

I have to admit though that the snakes case is a stranger situation. Partly this is due to the fact that in addiction to apparant motion it also seems as if the colors are stationary! I noticed that if I stare right at a snake coil, it looks stationary and the ones on the perefery seem to move. this fact seems to suggest to me that if I am to uphold my position and if this illusion is the result of the same phenomenon as the earlier illusions, then I should probably admit taht these snake coils change colors simply when I move from looking directly at one to directly at another. That is wierd!!

but, It might be mitigated if I take the position that some objects can be multiply colors all over and that sometimes we are unable to percieve certain of those colors.

11:40 AM  
Blogger Chris Tillman said...

Yes I realize that one can maintain your position in light of moving snakes. It just becomes an increasingly implausible view at this point. It seems even more clear in the snakes case than in the swirls case that my interaction with the picture brings about the effect--not the interaction of parts of the picture with each other alone. But this makes it seem overwhelmingly plausible to me that it really is an illusion and that color movement is no mind-independent feature of the picture. I know that's sort of pound fist, stomp foot, but it really does seem like the better explanation to me.

9:41 AM  

Post a Comment

<< Home