Postscript on Diachronic Discrimination Failure

This postscript to the Transcending Zombies series is primarily a follow-up to the remarks on Raffman-style nonconceptualism. Would my objection to the Raffman-style case against conceptualism be defeated by an experimental design that tried to better control for possible context effects of the presentations of the colors? The sort of redesign I here have in mind might go as follows. The stimuli presented in each distinct presentation in the diachronic discrimination case would be one of figures 1 and 2.

colors1

figure 1.

colors2
figure 2.

The task put to the subject is to make a “same as before, yes or no?” judgment about colors appearing on the right side of each display. Synchronic discrimination tasks could use just one of figures 1 and 2 and ask, say of figure 1, if the left and right regions contain the same color.

Such an experimental design is aimed at avoiding the accusation that the colors presented in the synchronic and diachronic contexts are colors presented in different contexts and it thus may not be assumed that there is a color-appearance that is constant across contexts. In this new experiment, the color context of the right-hand color in figure 1 is arguably the same as the color context of the left-hand color in figure 2 since figures 1 and 2 are just spatial rotations of each other.

Does such an experimental design help to defeat the conceptualist? One point in favor of the conceptualist is that in the experiments using figures 1 and 2, there may no longer be a failure of diachronic discrimination. The subject, in being presented with figure 1, is in a position to conceptualize the color on the right as the lighter of the two. Further, the subject may re-conceptualized the diachronic task as, in seeing fig 2 after fig 1, judging whether the lighter of the two has changed its relative spatial location.

3 Responses to “Postscript on Diachronic Discrimination Failure”

  1. Robert Briscoe says:

    Hi Pete,

    Sorry again for the slow turn around, but I was away from my computer much of the weekend.

    Some of this goes back to earlier posts.

    Re: It seems that the hard-core conceptualist could easily respond that such data is consistent the following multi-part claim: (1) B25 chips seem the same as one another in simultaneous presentations of B25 chips, (2) B27 chips seem the same as one another in simultaneous presentations of B27 chips, and (3) the way B25 chips seem in context (1) is *not* different from the way B27 chips seem in context (2).

    In the set up I had in mind, the observer, e.g., looks at a screen containing a B25-colored target and then matches the target to a B25-colored sample on another screen, rejecting B24, B26, and B27 (again on a screen separate from the screen on which B25 appears) as matches. The samples are presented to the observer individually, not side by side, so as to avoid context effects. I was assuming that the memory delay between saccades wouldn’t interfere with discrimination, while longer delays do.

    If, however, by hypothesis, ANY memory delay results in the observer’s inability to discriminate B25 and B27 (which I think is actually the hypothesis you had in mind), so that they are only discriminable side by side, then my response is irrelevant and I am enough of a verificationist to concede your point.

    For what it’s worth, the color patches in your stimuli in the last post are discriminable in the set up I described. But this of course doesn’t mean that there aren’t patches that are not so discriminable.

    Re: The Raffman stuff seems to offer a much more powerful line of thought against conceptualism than does the Evans stuff.

    You’re right. I should have been more careful in my response. As Heck often points out, the Evans stuff only supports the state view and not the content view.

    Re: [I]f you can serially discriminate B17 and B25, then you can also serially match two presentations of B17, and thus satisfy the criterion. Doesn’t this make sense? If you are remembering B17 enough to reject a subsequent presentation of B25 as a match, then you should be able to remember B17 well enough to match tokens of it across an equivalent memory delay.

    I don’t think I agree. Suppose I meet Judy at a party. I may remember Judy well enough the next day to recognize that Sally was not the person I met the night before, but not well enough to recognize that Jill, Judy’s sister, was not the person I met the night before. My assumption is that the same is possible with colors. I may know that the shade of red I am seeing is not Coke can red (it may seem too orangey or whatever), but may nonetheless not be able reliably to recognize Coke can red when I see it. (I seem to recall evidence that people are actually quite bad at identifying the color of a Coke can.)

    Re: the Heck article. I think that the preceding wall example in his article is better as regards the point he is making about what is implicitly and explicitly represented in perceptual experience, in particular, the claim that “x is the same color as y” may be explicitly representable only under certain circumstances. (This point has to do with whether perceptual contents satisfy the generality constraint.)

    The 10×10 array example admits of two interpretations, I think. Here is what Heck says:

    Imagine a 10 × 10 array of color patches of various shades of blue. Imagine looking at the patch in the upper left corner and then at the one at the lower right. Suppose that the two patches are in fact the same color and that they are actually represented in one’s experience as being of the same color. We may even suppose that the lighting has been carefully controlled, so that the two patches reflect the same spectrum. Even under these ideal conditions, one need not be able to say with any confidence whether the two patches are the same color nor even whether they look to be. The content of perceptual experience is thus not ‘transparent to introspection’ in the way the content of belief is. At least to the subject, then, it does not seem as if a single ‘concept’ of a shade is being deployed in the characterization of both patches.

    Now Heck could intend either:

    1) SameColor[P1, P2] (or B25[P1] & B25[P2]) is explicitly represented in a single, momentary experience of the array, but this is not introspectively accessible.

    or

    2) B25[P1] is explicitly represented in the experience E1 had when I look at/attend to P1 and B25[P2] is explicitly represented in the experience E2 had when subsequently I look at/attend to P2, but SameColor[P1, P2] (or B25[P1] & B25[P2]) is not introspectively accessible.

    Interpretation 1), I agree, seems quite weird. I don’t accept that conscious visual experience is “rich” in the sense of having contents outrunning what is introspectively accessible.

    2), on the other hand, seems plausible to me and is consistent, it seems, with the broader point Heck is trying to make about whether perceptual contents satisfy the generality constraint. It is also consistent with the supposition in the preceding wall example that large angular distances prevent explicit representation of two patches as having the same color. When angular distances are large, visual experience may be “silent” about whether they are the same or different. Indeed, if the supposition is that large angular distances prevent explicit representation of two patches as having the same color, then 1) is disqualified as an interpretation, and Heck is not countenancing determinate aspects of phenomenology that are inaccessible to introspection. (At least, he is not countenancing determinate aspects of the phenomenology of a single, momentary visual experience that are inaccessible to introspection.) This speaks in favor of 2).

  2. Pete Mandik says:

    Thanks again, Robert. This continues to be quite helpful to me.

    Re: Heck stuff. I think what you say about 1) and 2) and the superiority of 2) as an interpretation of Heck makes a lot of sense. Note, though, how much Heck’s 10×10 grid case, under interpretation 2, becomes similar to a Raffman-style argument about what can make it across a memory delay. I would think it would be similarly vulnerable, then, to the kinds of defenses of conceptualism that I rehearse against Raffman.

    Re: Can two shades be diachronically discriminable without thereby satisfying a reidentifiability criterion on concepts?

    I don’t quite follow the mini-story about Sally, Jill, and Judy. But maybe I can just focus here on the case of the Coke can. I take it that the way in which people are poor at recognizing Coke-red is that there is a set of colors all relatively close to Coke-red and including Coke-red, such that people are poor at picking out which one is Coke-red. Each of these would, I presume, be synchronically but *not* diachronically distinguishable from Coke-red. But now let’s focus on a color that that *is* diachronically distinguishable from Coke-red (CR): your too-orangey shade (TO), for instance. If a subject *is* able to diachronically distinguish TO from CR, and, if given both TO and CR and asked to pick out which of the two is CR, would you expect that they’d be able to do so? And would this count as a re-identification of CR? I would say ‘yes’ to both questions.

    Is this getting convincing yet?

  3. Robert Briscoe says:

    Count me as convinced. I’m still not a conceptualist, but I think that you’ve got a good reply to Raffman-style arguments.