Friday, February 4, 2011

Is this Walk-on-able? 2



Does the glass floor structure light in such a way as to enable a human to visually perceive stand-on-ability?

Charles wondered, at one point, whether my exploding box example was just some philosophical exotica, like imagining a wizard, but as I said earlier, the point is completely pedestrian.  The basic point can be transferred to plenty of other examples.

I guess it's unfortunate that the EPists have been too busy to test one of the fundamental assumptions of Gibson's theory, namely, whether the composition and layout of surfaces constitute affordances, or the TSRM assumption that anchoring properties structure light.

14 comments:

  1. I guess it's unfortunate that the EPists have been too busy to test one of the fundamental assumptions of Gibson's theory, namely, whether the composition and layout of surfaces constitute affordances, or the TSRM assumption that anchoring properties structure light.
    You should probably read Bill Warren's work on stair climbing and aperture passing before getting too worried about how busy we are. Oh, and Karen Adolph's work on affordances for locomotion down slopes. Oh, and Bingham's recent work on throwing, and maybe even the stuff he's done with Mark Mon-Williams on prehension.

    Oh, and if you're worried about the structuring of light issue: you should read David Lee's work on tau and the large body of work that followed in on the visual information for time-to-contact. Oh, and all the studies on the outfielder problem and the analysis of information available to the fielder (Bill Warren again, plus Mike McBeath). I'd also read the biological motion literature (Troje's work is great, but you should go back the the original Johansson work) as well as the event perception literature (Bingham's in on this again, plus Mats Lind).

    Also, if you're into non-visual perception, you might want to look at Farley Norman's extensive work on the haptic perception of shape, etc. He's an excellent psychophysicist.

    I guess it's unfortunate you were too busy to do a literature review to test your fundamental claim that ecological psychology hasn't been studying what Gibson suggested might be the things to study.

    ReplyDelete
  2. Sorry, Andrew, I was just taking your word for it.

    ReplyDelete
  3. if you're worried about the structuring of light issue

    That's exactly what I have been on about for days.

    you should read David Lee's work on tau and the large body of work that followed in on the visual information for time-to-contact.

    Now you are talking sense. Maybe there is something here that is like "see-ability" or "useability as an occluder", rather than edible, stand-on-able, climb-up-able, etc. Some affordances might be constituted by the composition and layout of surfaces, even though many the EPists have looked at are aparently not.

    ReplyDelete
  4. I really have no idea how you drew that conclusion from a comment about time-to-contact. What do you mean?

    But seriously, you should go read this stuff. You keep asking this question, the answers are here. I thought you liked your psychology experimental.

    ReplyDelete
  5. Ok. Forget time-to-contact.

    Seriously. I don't need to read more to figure out that the composition and layout of surfaces do not typically constitute an affordance. This is pedestrian physics, chemistry, and biology.

    If there answer were in these papers, you might have told me by now.

    ReplyDelete
  6. "exploding box example was just some philosophical exotica, like imagining a wizard"

    This mixes two ideas that shouldn't be mixed, and - I humbly submit - shouldn't be dismissed.

    The "philosophical exotica" had to do with my impression (at the time) of the "affordance" concept as being slippery - like the "truth" concept. Andrew and Sabrina have been patiently clarifying the concept for me, and it appears that it does in fact need some tweaking.

    But affordances are associated with a continuous dynamic process involving a perceiver acting in a time-varying environment. From that perspective, to present a static picture of some distant surface and ask if it offers to the viewer (acting as a proxy perceiver) an "affordance" isn't particularly meaningful. There is a continuum of intermediately actualizable affordances, and actualizations of some of those, necessary before a real perceiver's relationship to the surface becomes such that s/he could even address that question, never mind answer it. Andrew and Sabrina have introduced the idea of a perceiver having a "goal", which apparently plays the role of that "something perceived at a distance which motivates subsequent actions" (what my earlier terms "apparent/perceived affordance" were intended to capture).

    The point of introducing the "wizard" was that intervening events caused by an imperceptible wizard with magical powers obviously shouldn't affect one's opinion about a theory of perception, so neither should intervening events caused by any other imperceptible causal actor, even one less fantastic.

    ReplyDelete
  7. Charles, the picture is only an aid to the point. I'm surprised at the number of comments that seem to think I am asking about a picture (be it of dirt, salt, sand, sugar, candy, or plastic), rather than the thing pictured. Of course, the picture on your monitor is not edible. The screen on your monitor is not walk-onable because it's vertical.

    But, really, I am asking about a real glass floor. Does the surface of the glass structure light in such a way as to specify whether the physical glass floor is walk-on-able?

    Go ahead, let the viewer move all around this glass floor. Free viewing welcome. Let's have no aperture vision. Does that help to cultivate information in the light that specifies whether the floor is walk-on-able? No.

    In fact, let a perceiver lick the floor, smell the floor, put her ear to the floor, touch the floor lightly with her hand. Just don't let her walk on it (or put, say, 15o lbs on the floor). It ain't gonna help.

    I get the point about the wizard; that's why I've provided several examples that I hope allay that concern.

    ReplyDelete
  8. I of course understand that the question is about a "real" surface, but without explicit mention of a perceiver's mobility it isn't clear that it's being assumed.

    So, assume mobility. Before a walk-on-able affordance can be perceived, one must perceive and actualize several affordances leading to perception of an immediately actualizable step-on-able affordance. At the time of actualization of that affordance, any of mumerous consequences can result. One is that the surface doesn't support the perceiver, who then falls. But that doesn't negate the step-on-able affordance, which was in fact actualized. It merely precludes some subsequent affordances, in particular walk-on-able. But whereas you insist on saying that walk-on-able therefore was never an affordance for that perceiver, I'm content with saying only that no walk-on-able affordance ever became immediately actualizable by the perceiver.

    OTOH, if the surface holds, a walk-on-able affordance becomes immediately actualizable (assuming we define "walk" as two or more steps, which allows for the possibility that the surface collapses/breaks subsequent to the first step).

    My only quarrel is the leap from "the concept of affordance needs some tweaking" to "the whole enterprise fails". To the limited extent that I understand the Gibson approach, it seems conceptually reasonable. Whether preferable to other approaches - or even implementable - are separate questions.

    ReplyDelete
  9. I don't mention the free movement, since it's evident that it does not matter. No need to clutter the example with irrelevant details.

    Before a walk-on-able affordance can be perceived, one must perceive and actualize several affordances ...

    This is not part of the TSRM or Gibsonian accounts I have been challenging. I don't know where you are getting this.

    But, stop thinking of this as a problem in perception. It's not. It's fundamentally physics. The surface of the glass does not reflect light in such a way as to specify the weight it will bear, right? Maybe this is even an engineering fact.

    Now apply this physical fact to EP. If the surface of the glass won't structure light in such a way as to specify what weight the glass will bear, then it won't specify whether the glass affords walk-on-ability.

    ReplyDelete
  10. Well, I thought I had made this clear, but apparently not:

    I AGREE that walk-on-able is not a visually determinable feature, whether it's called an affordance, a disposition, or a Fred. At least not at a distance and/or not without doing a lot of inferring.

    Same with edible. So with respect to walk-on-able, I don't see a need to consider the detailed structure of light reflected from surfaces. I'm ready to move on to some other candidate affordances. (But recall that I reject imperceptible inhibitors.)

    ReplyDelete
  11. Ken, do you have a take on how Chapter 6 of Tony's RECS might relate to all this? I'm working up a blog post on it now, and while I'm not actually convinced there's a need to replace TSRM information-via-laws with Barwise & Perry information-via-constraints, it seems that the problem Tony's tackling is precisely analogous to the issues here.

    ReplyDelete
  12. Actually, I've not read that far into the book, but this gives me a reason to press on.

    But, a priori, I would guess that it will not. The basic problem is light is structured by the outsides of objects and affordances are anchored by the insides of objects.

    ReplyDelete
  13. As I read it, the situation semantics constraint approach allows situations that don't lawfully specify to still be informative. So the outside could be sufficiently informative about the inside.

    As I said, I'm not convinced this is a good move, and I'm not advocating this as a solution. I think there are other ways to solve this problem and I'm working on a post now, but this chapter clearly overlaps with this discussion, Tony's basically trying to tackle the problem you're posing. The chapter stands fairly well on it's own, so it's readable without the earlier sections.

    ReplyDelete
  14. Andrew, you are right that it makes room for informational connections that are less than lawful, but I still don't see that it addresses the problem. Check the post I've just added.

    ReplyDelete