Monday, March 02, 2009

Do Things Look Flat?

I've posted a draft of Chapter Two of my book in progress (tentatively titled Perplexities of Consciousness). The chapter is titled "Do Things Look Flat?" and is a revision of my 2006 essay of the same title, with a little more nuance and historical depth and a longer discussion of the view (dating back to Ptolemy but peaking, evidently, circa 1900) that most things appear doubled in the visual field. Comments/suggestions/criticisms welcome, of course, either as email or as comments on this post.

Abstract:
Does a penny viewed at an angle in some sense look elliptical, as though projected on a two-dimensional surface? Many philosophers have said such things, from Malebranche (1674/1997) and Hume (1739/1978), through early sense-data theorists, to Tye (2000) and NoĆ« (2004). I confess that it doesn’t seem this way to me, though I’m somewhat baffled by the phenomenology and pessimistic about our ability to resolve the dispute. I raise a geometrical objection to the view and conjecture that, maybe, the view draws some of its appeal from the over-analogizing of visual experience to painting or photography. Theorists writing in contexts where vision is analogized to less projective media – signet ring impressions in wax in ancient Greece, stereoscopy in introspective psychology circa 1900 – seem substantially less likely to attribute such projective distortions to visual appearances. Stereoscope enthusiasts do, however, seem readier than scholars in other eras to attribute a pervasive doubling to visual experience – like the doubling, perhaps, of an unfused image in a stereoscope.

17 comments:

dejan said...

I think that acknowledging the perceptual reality of apparent shapes and sizes, as Tye and Noe seem to do, does not commit one to the notion that everything looks flat. Instead, one can talk about visual angles, as Tye does, or extents subtended in the visual field. I can judge (or see, or be aware) that a nearby car subtends a larger part of my visual field than a duplicate car further away, but still, at the same time, see them as 3D bodies (Also, contrary to Kelly, I can at the same time judge that they have equal objective sizes; whether my judgments are correct or not, is a different question). Similarly, I can judge that the shape that the penny subtends is elliptical (in the sense, say, that I could hold an ellipse perpendicular to the line of sight, whose border would coincide with the contour of the penny).
The reason that you are more pulled to the dual aspect view in the case of lampposts may not have to do with the difference between size and shape perception, but with different features of the two situations. With lampposts, you have a set of objects that you know are equal, so that the aspect in which they appear different becomes salient and easily attended to. With the penny, you have only one object, with nothing to compare it to. If you would have more pennies, but with different tilts with respect to you, the differences in their apparent shapes may become more salient.

Eric Schwitzgebel said...

Thanks for the comment, Dejan. That's an interesting suggestion about my take on the lampposts and the penny. I'll see if I can set up an array of coins and try it out!

I'm not sure that I fully understand your first point. Of course dual aspect people like Noe and Tye say that one sees depth at the same time one sees the ellipticality -- so things don't look as fully flat on their view as on the views of Locke, Hume, and Broad. But isn't there going to still have to be a sense in which things look flat for them -- the same sense in which the penny looks elliptical (while granting that it also in a different sense looks circular in 3D space)?

dejan said...

I did not express myself clearly. Of course Noe and others say that one sees depth. What I meant was not the standard case, in which depth is associated with seeing the penny as round, but rather the situation when one sees the penny as elliptical (by the way, the penny is a poor example for the world-looking-flat discussion anyway, because the penny is a flat object in the first place, as you noted (that is why I used the example with the car, which is a bulgy object)). What I should have said is that when one claims to perceive ellipticity when looking at the penny, one is not necessarily committed to seeing the penny itself as a flat elliptical object (in which case my next question would be, what is the orientation of that elliptical penny? Is it lying at the table, tilted with respect to your line of sight, just as the round penny? Or is its orientation perpendicular to the line of sight?). Rather, one can think of the perceived ellipticity in terms of visual angle and say, as I wrote, that the penny subtends an elliptical shape in the visual field, the same as the shape subtended by an upright ellipse. But such visual angles are not objects, such as pennies or cars, which can be flat or bulgy, or tilted at different angles. Rather, visual (spherical) angles can be conceived as so-called visual cones (a standard notion in perspective), whose apex is in the eye of the observer and whose mantle is formed by rays grazing the contour of the observed object. Of course, I don't mean that one sees such an imaginary geometrical construction when one looks at the tilted penny, but rather that it helps elucidate the notion of visual angle.

The projection of the penny at a spherical surface, such as the back of the eye, is not a conventional ellipse, as you noted, because the projection surface is not flat. However, it does have a rather ellipse-like shape, as it is an oval with two unequal axes, and in this sense subtends an elliptical shape on the retina. The information about this shape is preserved in the visual system upstream from the retina, and presumably forms the basis of our seeing ellipticity when looking at the penny. On the other hand, I don't think that the information about the concave shape of retina is preserved in this way. Thus seeing ellipticity does not need to entail seeing concavity, that is, the world need not look concave just because the retina is concave.

Eric Schwitzgebel said...

Ah, I see, Dejan! That makes more sense to me. I like your concluding point, especially, about what is and is not likely to be preserved in information flowing from the retina. That's a nice way of getting the visual angle idea and the ellipse too, without any gross geometric boo-boos. Although you're not yourself defending any form of flattism, I see how a flattist might be able to make use of that idea to escape my geometric objection....

dejan said...

I am not a flattist, I am a visual anglist...

Just to finish with my thoughts on your paper, which, as you see, I found quite stimulating, although I do not agree with some points. You mentioned Price's idea of the nearby objects appearing convincingly 3D but faraway objects rather flat, and criticized it by claiming, in effect, that in that case the intermediate distance objects should appear as having some in-between depth (for example with respect to perceived angle sizes), which did not seem plausible to you. However, what you call 'gradual progression towards flatness' is not the only possibility compatible with Price's view, because there is another possibility, which may escape your critique. Note that for nearby space, all of the many perceptual depth indicators (binocular disparity, convergence, accommodation, parallax etc) agree with each other, providing excellent depth information for observed objects. For faraway space, most of the depth indicators are likely to be knocked out, providing no depth cues for observed objects, thus again agreeing, but this time on flatness. For the in-between space, what could happen is that some of the depth indicators provide some depth cues, while the others don't, in which case they would not be in agreement as to the depths of objects at those distances. This might leave you not with an impression of orderly, gradually flatter depth with distance, but rather with somewhat murky, ambiguous depth impressions.

Finally, your example with the graph of the sphere projecting as larger from a sideways position is related to similar examples which are well known in the perspective literature, and which already troubled Leonardo da Vinci. For example, see Figure 8.11 in http://www.webexhibits.org/arrowintheeye/marginal3.html

Eric Schwitzgebel said...

Thanks for the very thoughtful and helpful comments, Dejan! I was not aware of Da Vinci's discussion of this issue. Maybe I'll work that in. Of course, to assume that discussion of the projective geometry of painting is tantamount to discussion of the projective geometry of visual experience is to assume the very thing that I'd like to bring into question; so except when flat-media visual artists claim to that the geometric laws governing their art are the same that govern visual experience itself, I take them not to be discussing the topic at hand. Of course, many painters do claim that visual experience has the same projective geometry that one sees in painting.

On Price: It is Price himself who says that the middle distance is partly flattened. Perhaps I wasn't very clear about this in the chapter draft. Your alternative view is a nice one -- somewhat more appealing to me.

Mariana Soffer said...

An analogy for flat surfaces, involving flatness itselff.

Monks live in the gaps between words and meditators look for the gaps between thoughts. Every inner "technology" for awareness starts with slowing down thoughts, stimuli and cravings looking for empty spaces.

They don't

Eric Schwitzgebel said...

Mariana, it looks like your comment was cut off in the middle....

Mariana Soffer said...

Yes, sorry I do not remember what I was going to say, anyway:

As far as the laws of mathematics refer to reality (either round shapes or square ones), they are not certain, and as far as they are certain, they do not refer to reality

Thoughts said...

You may be interested in the role of our sense of time in the perception of space. See Time and conscious experience

Ben Bronner said...

Hi Eric. You might be interested in the passage below, from Visual Experience: Sensation, Cognition, and Constancy (eds. Hatfield & Allred).

(The reference to the "attitude" of a painter, towards the end of the passage, reminded me of your ideas.)

'Although the doctrine of the two-dimensional sensory core was widely held in the eighteenth and early nineteenth centuries, the doctrine had largely been rejected by the second half of the twentieth century (Hatfield and Epstein 1979). But the phenomenology of a two-dimensional projection did not totally fade away. Rather, it was relegated to a special place. James Gibson contended that observers can achieve an experience that is like a two-dimensional perspective image by adopting a special attitude of perception. Gibson (1950, ch. 3) contrasted this experience of a "visual field" with the normal experience of visual perception, which he held to be of a "visual world" that exhibits full phenomenal constancy. He defined the visual field as a "picturelike phenomenal experience at a presumed phenomenal distance from the eyes, consisting of perspective-size impressions" (1952), 151). The two-dimensional visual-field structure arises when the perceiver takes a special attitude, sometimes called the "painter's attitude," toward the visual world, resulting, in the ideal case, in the experience of a perspective projection of the objects in the field of view.' (36)

Eric Schwitzgebel said...

Thanks for the passage! Reminds me of Kelly on Noe, too.

Ben Bronner said...

Sure, though after reading the rest of the chapter, it looks like I gave you the least relevant passage! Hatfield describes several studies allegedly showing that objects do in fact look smaller in the distance, and that distance constancy is a cognitive rather than perceptual feat. Sounds implausible to me, but I thought you might want to know about the studies and Hatfield's interpretation. They come up starting with the section "Phenomenal size and cognition of objective size," which if you are interested you can read here:

http://books.google.com/books?id=vsTNwfaEhKEC&lpg=PA42&ots=9uU7E7EKXf&dq=phenomenal%20space%20contraction&pg=PA47#v=onepage&q&f=false

Arnold Trehub said...

For a biological explanation of our phenomenal perspective on the visual world see "Where Am I? Redux", here:

http://theassc.org/documents/where_am_i_redux

Eric Schwitzgebel said...

Thanks Ben & Arnold!

Ben: I know some of Hatfield's earlier stuff, which is a kind of intermediate view between flattism and full-blown veridical 3D. Unfortunately, the punchline of this most recent article isn't showing in my view. I'll have to get a copy of the book!

Ben Bronner said...

It's kind of a hassle to read this way, but if you're up for it you can get the rest of the chapter (except for the second-to-last page) by switching to Amazon, here:

http://www.amazon.com/Visual-Experience-Sensation-Cognition-Constancy/dp/0199597278/ref=sr_1_1?s=books&ie=UTF8&qid=1342499374&sr=1-1&keywords=visual+experience+hatfield

(To pick things up at the right page, use Amazon's "search inside this book" feature, search for "Ebenholtz", and click on the second result listed.)

Eric Schwitzgebel said...

Thanks, Ben, that's helpful!