Kelly on Perceptual Constancy
Sean Kelly came to Essex last Thursday and spoke about perceptual constancy. An earlier version of his paper is here (Essex access only). He argued against so called 'dual content' accounts of perceptual constancies. These views, such as that of Alva Noe, hold that when viewing a coin from an oblique angle I see it as both circular and as eliptical at the same time. Kelly himself favours the view according to which our ordinary, everyday experience of the coin presents it as circular. According to him we only really see it as eliptical when we adopt the 'painterly attitude' towards it. This requires, he thinks, an aspect change analogous to duck-rabbit type examples.
What was interesting about Kelly's approach was that he thinks the debate open to empirical adjudication. He produced a number of empirical hypotheses that would help settle the matter. Most prominently is the shape priming paradigm. Essentially, the question is whether, having been shown a coin from an angle, subjects are primed for circles, elipses or both. It appears that subjects are primed for circles, not elipses, and this, argues Kelly, supports his view.
Discussion ranged accross a number of issues. Might it not be significant that the experiment used a 2D representation of a coin rather than a real 3D object? Wasn't Kelly underplaying the role that beliefs play in determining what properties something is perceived as having? Might not Noe have resources to hand that would allow him to maintain that it would be very likely that only one of the presented properties (the real one, not the apparent one) would have a priming effect - an example might be the selective advantage of having only those shapes prime that feed into the representations controlling one's bodily action (when one reaches for the obliquely presented glass one's hand forms a circular shape, not an eliptical one).
A thought that I had since was that there is a respect in which the analogy with aspect change breaks down. For, if I understand it correctly, the positive idea is that object's real properties are presented in the engaged attitude but not in the painterly attitude. But now it becomes less easy to 'swap' attitudes in this case than it is in the aspect change case. For one of the attitudes that one is being asked to put oneself in specifically precludes attending to the qualities of one's experience (on some understaning of that phrase). So, whilst the commands "See it as a duck" and "See it as a rabbit" are equally easy to follow, the commands "See it as a circle" and "See it as an elipse" are not.
What was interesting about Kelly's approach was that he thinks the debate open to empirical adjudication. He produced a number of empirical hypotheses that would help settle the matter. Most prominently is the shape priming paradigm. Essentially, the question is whether, having been shown a coin from an angle, subjects are primed for circles, elipses or both. It appears that subjects are primed for circles, not elipses, and this, argues Kelly, supports his view.
Discussion ranged accross a number of issues. Might it not be significant that the experiment used a 2D representation of a coin rather than a real 3D object? Wasn't Kelly underplaying the role that beliefs play in determining what properties something is perceived as having? Might not Noe have resources to hand that would allow him to maintain that it would be very likely that only one of the presented properties (the real one, not the apparent one) would have a priming effect - an example might be the selective advantage of having only those shapes prime that feed into the representations controlling one's bodily action (when one reaches for the obliquely presented glass one's hand forms a circular shape, not an eliptical one).
A thought that I had since was that there is a respect in which the analogy with aspect change breaks down. For, if I understand it correctly, the positive idea is that object's real properties are presented in the engaged attitude but not in the painterly attitude. But now it becomes less easy to 'swap' attitudes in this case than it is in the aspect change case. For one of the attitudes that one is being asked to put oneself in specifically precludes attending to the qualities of one's experience (on some understaning of that phrase). So, whilst the commands "See it as a duck" and "See it as a rabbit" are equally easy to follow, the commands "See it as a circle" and "See it as an elipse" are not.
posted by Joel at 9:43 am
2 Comments:
Hi Jonah,
The way I understood the dialectic was this: The dual content theorist maintains that the coin is presented as both circular and eliptical at the same time (of course, these properties are given in different ways - for the object does not look to have an impossible shape). Kelly then turns round and says, "No! Only one property is given at a time. An aspect switch is required to move from seeing the coin as circular to seeing it as eliptical." But Noe's response to this was just to deny that as accurate phenomenology.
My point is that there is a sense in which Noe's response is the right one (although this doesn't support his positive view). You cannot just switch from the painterly attitude to the engaged one in the same way as you can switch from the duck to the rabbit. The reason for this is that (as I understand Kelly's own view - and I may be misrepresenting him here), once you're in the project of attending to your experiences in order to decide about their phenomenological character you are thereby in the painterly attitude. To recognise that you ordinarily see the object as round (but not eliptical) you would have to - as Merleau-Ponty says at one point - catch the experience in the corner of your eye. This, of course, is related to what we were chatting about in the pub the other day.
Joel
I think I agree with much of what you've said - although I would need to go back to read both Noe & Kelly again (incidentally, you can dowload Noe's article "Real Presence" from his website). Interestingly, it has struck me that the most that could be shown by Kelly's shape priming experiment would be that ordinarily and pre-reflectively we are only aware of the coin as circular. It would not show that, when we have adopted the painterly attitude and see the coin as eliptical, we are not also marginally aware of it as circular. So something very similar to what you suggest is consistent with (at least that part of) the evidence that Kelly presented.
Post a Comment
<< Home