Wednesday, January 26, 2011

The perception of stand-on-ability

So, take to boxes 3' x 3' x 3' in size. They are physically identical on the outside, but one is rigged with a touch sensitive device so that it will explode if you stand on it.

I say they are visually perceived to be the same, but that they afford different things (one affords stand-on-ability; the other does not). So, one does not perceive stand-on-ability.

14 comments:

  1. Is it supposed to be that to visually perceive a thing is to perceive all of its affordances?

    ReplyDelete
  2. Andrew has suggested that EPists do not think that all affordances are perceived. He concedes that. But I am going after more. I am going after the idea that many of the things that EPists think are affordances that we perceive are in fact perceived.

    ReplyDelete
  3. I object to the word 'concedes'; the idea you might need to do some learning is built right into Gibson so it's not like this idea is a point in your favour.

    This is just the shark, redux: to the shark, both the electrical field and the fish produce information that specifies 'edible'. To the perceiver here, both boxes produce information that specifies 'stand-on-ability' or whatever is your affordance idea de jour. The fact that there is no information about the other properties that enable the shark or person to tell the difference just isn't the problem you think it is. Perception is good, but it isn't magic, and just because (from a third person perspective) you can sensibly claim that the exploding box can't be stood on, doesn't mean this description has any place in a theory of perception.

    This objection isn't anything new; these are all points that have come up before.

    I'm willing to bet that people would quite happily attempt to sit or stand or pick up either box and there would be no detectable difference in their behaviour.

    ReplyDelete
  4. You are exactly right, Andrew, that the argument here is just like the one for the exploding boxes and pick-up-ability. You agree that in this case one does not perceive pick-up-ability. So, you agree that all the song and dance about scope, information, etc. does not save the Gibsonian or TSRM view in that case.

    Now, I'm just showing that the same considerations work for stand-on-ability, sit-on-ability, leap-overability, etc.

    At this point, maybe you should unconcede the case of pick-up-ability.

    This is indeed the shark revisited and TSRM have the wrong analysis of that case too. Patterns in the electric fields don't specify edibility; just as the structure of light does not specify pick-up-ability, sit-on-ability, stand-on-ability, leap-over-ability, edibility, etc.

    ReplyDelete
  5. You agree that in this case one does not perceive pick-up-ability
    No, I was happy to say it may not be only visually specified; or that maybe pick-up-ability wasn't the right affordance. These things all need empirical investigation, is all I ever said.

    I just don't agree with your analysis, is all. I think all you've found is that you can't perceive something for which there is no information (which is true) and you think you've found that there can't be information for pick-up-ability because the box rigging makes it up-pickupable (which is false). The issues of scope and information make all of these perfectly comprehensible.

    Gibson's ontology isn't 'what's objectively the case', if you recall; it's 'what is real for a perceiving organism?'. Your third person perspective is the wrong view point.

    So right; nothing new to add, I guess.

    ReplyDelete
  6. Ok. You agree that in this case one does not visually perceive pick-up-ability.

    I think all you've found is that you can't perceive something for which there is no information

    Ok. Fine.

    So, in the exploding box case there is no information for the visual perception of pick-up-ability.

    In the big exploding box case, there is no information for the visual perception of stand-on-ability.

    In the big exploding box case, there is no information for the visual perception of sit-on-ability.

    Reformulate the problem, but it remains. Most of the things that TSRM thought were affordances that one can visually perceive are not affordances that one can visually perceive.

    ReplyDelete
  7. No, there's no visual information about the exploding nature of the box (is what I'm saying. I know you're saying what you said).

    ReplyDelete
  8. I'll renege on my promise and make one last comment on exploding, collapsing, etc, whatevers. (A promise I'll definitely keep.)

    I see no difference between those intervening events and Hume's "end of the world" event. But since no one likes that equivalent or my argument against it, I'll try another one.

    In the environment is an invisible (ie, imperceptible) wizard. Just as one is about to pick-up, sit-on, jump-over something the wizard waves his magic wand and the something disappears.

    Ken's challenges have raised some technical questions in my mind about the concept of affordance as it's been described, and for that I'm grateful since one way or another I'm going to end up with a better grasp of the concept. But to claim that nothing can be a visually detectable affordance because of invisible wizards who make things disappear before the affordance can be actualized seems a bridge too far. I can't imagine anyone taking seriously an argument based on that premise.

    ReplyDelete
  9. But, Charles, I'm using the exploding stuff just for philosophical drama.

    The exact same point can be made with completely pedestrian examples. Look at two different clear liquids. Just by looking at them from a lot of different angles --just from the structure of light--, you cannot tell which one (if either) is potable. Take two white powders. One cannot tell just by looking at them which one (if either) is edible. Take two cups of green mushy stuff. One cannot tell just by looking at them which one (if either) is edible.

    I have a strategy for generating these counterexamples. I just pick cases in which the surfaces of objects do not specify what they afford. Maybe rather than having me provide you with the example, you provide yourself with the example. (Remember, Gibson proposes that surfaces constitute affordances. Just think of your own counterexamples.)

    Andrew did this. His example was ice. The surface of ice (which structures light) does not specify walk-on-ability. That's just common sense. You can't tell whether you can walk on ice without falling in to the underlying water just by looking at the surface.

    ReplyDelete
  10. And I'm with you on many of the pedestrian ones. Potable is a perfectly real example of something that is in general imperceptible. So, I agree that it isn't - or at least from my perspective, logically shouldn't be - an affordance - although one can imagine a circumstance where there is information in the light - color, rising fumes, presence of foreign particles, animal skeletons lying nearby - that combined with items in the perceiver's knowledge base allow inference of nonpotability. Same with edible - maybe approachable, growl-at-able, bite-able, but edible isn't a perceptible feature of an object unless the object is recognized as something known from experience to be so. And even then I'd probably prefer consumable - it could be rotten, poisoned, etc, as you suggest.

    But I'm not so sure about others, eg, sit-on-able. Depends on how "sit" is defined how much additional info you aquire as you get closer, and whether you incorporate some concept like my "conditional affordance" in order to accommodate changed circumstances. One can sit on a stool notwithstanding that it may collapse when you do if you interpret "sit" as an event that occurs at a point in time rather than over a time interval. However, I might make this distinction. If upon close inspection (ie, perception) it were impossible to discern any reason that the stool should collapse, then I'd put the fact that it does into the exploding box category - imperceptible features of the environment aren't relevant to a theory of perception.

    Same with thin ice. You can step on it despite it's breaking when you do. But I might agree that walk-on-able shouldn't be an affordance unless supported by detailed perceptual information and possibly a process of inference.

    ReplyDelete
  11. The point can be made even simpler. Affordances don't typically structure light; surfaces do.

    ReplyDelete
  12. that combined with items in the perceiver's knowledge base allow inference of nonpotability.

    Ah, but this is precisely the kind of move that EPists loathe. Perception is not a matter of inference, computation, combination with memories, etc. etc. Admitting a shortage of information regarding affordances starts us on the road to cognitivism. That's the right path, but EPists will only go that way kicking and screaming.

    ReplyDelete
  13. The surface of ice (which structures light) does not specify walk-on-ability. That's just common sense.
    Actually the walk-on-ability is specified; friction isn't. These are obviously related but aren't the same thing, and these sorts of careful details matter in a discussion.

    And yes, we don't do inference. If you think information isn't good enough, just wait till you have to make educated guesses! I'll be watching you fall over from this box I found to sit on :)

    ReplyDelete
  14. I'm thinking of pond ice. You can't tell just by the way the surface structures light whether it is thick enough to walk on without falling through.

    ReplyDelete