Friday, May 26, 2006

Images as "Pictures"

In Wednesday's post, I wrote about the picture analogy for vision and the implicit (and possibly erroneous) assumptions about visual experience invites. Well, the picture analogy for visual imagery, visual imaginings, is even more pervasive. It almost doesn't feel like a metaphor to call an image a "picture" in the mind or to say "picture to yourself".

This struck me with particular force as I was reading through English translations of ancient Greek texts, looking for analogies between visual experience and paintings or pictures (I found exactly one in the whole corpus, though I can't pretend to have given it exhaustive coverage), pursuing thoughts relevant to the issues in Wednesday's post. I kept finding reference to "picturing" in the English translations, in discussions of imagery -- but, oddly, different translations would use that word in different places. Going back to the original Greek, it was evident in nearly every case I examined that it was the translator bringing in the metaphor (possibly not even aware of its metaphorical status or its potential interest in understanding the history of conceptions of imagery).

The analogy between images and pictures has worked its way so deeply into our thoughts about imagery that it's almost invisible as an analogy.

Now in some sense, surely, images are like pictures and the analogy between them is a good one. But let me suggest some ways in which images might not be like pictures (might not -- I don't think these issues have been adequately explored yet).

(1.) Images might be three-dimensional in some robust sense, while pictures are flat. Perhaps even (as some people have reported and as Borges describes in his story "The Zahir") images can be experienced simultaneously from multiple angles?

(2.) Images might be indeterminate in a way it's difficult for pictures to be. For example, it might be indeterminate whether the man you're imagining is wearing a hat, or what color his jacket is, or whether he has a beard.

(3.) Images might have their interpretations built into them in a way that pictures do not. For example, when we imagine an "ambiguous figure" like Wittgenstein's duck-rabbit or the Necker cube, the image might incorporate or involve one interpretation then another, while nothing in the "picture" before our mind changes. (This thought arose recently in conversation with Charles Siewert, as we were reading a paper by William Robinson that seemed to assume that there's nothing different in the visual imagery experience of a Necker cube interpreted one way and that same Necker cube interpreted another.)

As I mentioned in Wednesday's post, I'm inclined to think our metaphors for the mind often distort our understanding of it. So I wonder: Would people be more likely to accept features 1-3 as aspects of our imagery experience if the picture analogy weren't so deeply ingrained in our thinking?


Jonathan Jenkins Ichikawa said...

This strikes me as a very solid discussion. I think you're right that the underlying assumption is pretty questionable -- (2) and (3) seem clearly right, and (1) is an interesting idea too.

Colin McGinn goes to some (some) legnth to argue for the aptness of picturing language in his recent book on imagination, Mindsight. I don't find it very compelling, and I'm sure you won't either, but at least he treats it as an assumption needing defense.

I'll be at the SPP, by the way -- I hope to make it to your session.

Brad C said...

Very interesting. Two thoughts:

(1) I find the use of the noun 'image' to refer to the "object" of an act of imagination, a bit odd and worry it might invite confusion.

Consider, for example, your claim that images might be three dimensional while pictures are not. I can certainly imagine a three-dimensional object, but that does not mean that, by exercising my imagination, I create a new three dimensional object that was not there before; Three-dimensional objects exist in space and time, but the objects of our imagination do not (well not unless there is a sense in which they are identical to bodily events - neural firings etc.)

(2)More generally, I feel, but on reflection resist, the temptation to think about imagination like this: when exercising my imagination I conjure up images before my mind's eye.

I have trouble imagining (haha) how to cash out the metaphor here.

More importantly: on reflection I doubt that imagining centrally involves conjuring up "objects" with some sort of intentional/ representational content.

I might put my view like this: a picture is one thing and what it pictures is another, but when I imagine something there just one thing to which I attend - that which I imagine - there is not an "image" which I create with my imagination and then some further thing that it depicts or represents.

I worry that you accept something like the later view and argue that the problem with the "picture" understanding of imagination is that it assimilates "images" and the way and what they depict to "pictures" and the way and what they depict.

I think that the picture view is misleading in the first place because imagination need not involve conjuring up "images" that "stand for" something else when we exercise our imagination. I think your second point reinforces this - when the color of the hat is left indeterminate, it is not as if you have an "image" in mind that is colorless.

Eric Schwitzgebel said...

Thanks for the kind words, Jonathan. I haven't read the McGinn yet, but I hope to do so very soon. See you at the SPP!

Brad: I find your comments very interesting. Perhaps an even more radical rejection of the picture metaphor is called for than I envision here -- a rejection of the very idea of an "image" that is experienced in the course of imagery (as a picture might be experienced in the course of seeing).

Indeed, I don't want to say that there is a image that exists separately from the imagining of it -- an image that might, for example, have properties the imaginer has trouble "making out". But does the very use of the noun "image" suggest this? Maybe it does tempt that reading.

But consider a "dance". There is no dance separate from the dancing of it; it is created in the process of dancing and is nothing more than that process itself -- at least in one way of talking and thinking about dances. There's also a more generic sense of dance in which two people can dance the same dance or one can get a dance wrong; and perhaps there's a parallel to imagery here also (two people might imagine a Necker cube, one might get the image wrong).

This doesn't address all your interesting points -- but I hope I'm not committed to disagreeing with you just because I want to say things like "visual images might be three-dimensional or two, might be precisely detailed in a particular respect or may leave those details unspecified".

Brad C said...


I like the dance example and agree you were not obviously committed to the "mental image" view of which I am skeptical.

You wrote: "I hope I'm not committed to disagreeing with you just because I want to say things like "visual images might be three-dimensional or two, might be precisely detailed in a particular respect or may leave those details unspecified"

After thinking about it a bit more, I think the first claim, but not the second, might get you into trouble here.

You wanted to draw a contrast between pictures and images by appeal to the fact that picture are (at least centrally) two dimensional. Clearly, you do not mean to draw the contrast at the level of the intentional objects: a picture can depict (be about) a 3d object just as I can imagine a 3d object.

So the guess that the contrast has to be between the 2D nature of the picture - the thing that does the depicting, so to speak - and something that can be 3D in the imagination case. And I have trouble seeing what that could be except a mental image of the sort I was casting doubt on.

But maybe I am not getting what you mean when you say that then image can be 3D --- thanks for providing me with an occasion to read some Borges tonight!

Eric Schwitzgebel said...

Thanks for pressing me on this, Brad, with your usual insightful comments!

On whether an image might be three-dimensional or two: There can be no debate here about whether images (like pictures) can *represent* things as three dimensional (they can). Nor can there be any debate about whether they exist like material objects in three-dimensional space (they don't, setting aside complications about their possible equivalence to 3D brain states).

So what could I have in mind when I say an image might be three-dimensional or two? I didn't clarify this very well, but I was thinking along lines analogous to the debate about whether perceptual visual experience has a two-dimensionality about it -- whether it involves projection onto a plane, per Hume, Tye, etc. (See the Mon & Wed posts from last week.)

So, if you're imagining a coin viewed from an angle, is there something elliptical in your imagery experience? If you imagine a line of streetlights running off into the distance, is the "imagined size" of the nearer ones (in some sense, in *any* sense of "imagined size") larger than that of the more distant ones?

Gary Isaac Wolf said...

In the 7 November, 2008 issue of Science, in an interesting review article, "Consciousness and Anesthesia," the authors assert that anesthesia causes loss of consciousness in part through reducing the brain's ability to integrate information. Then they make an interesting assertion: "every experience is an integrated whole that cannot be subdivided into independent components. For example, with an intact brain you cannot experience the left half of the visual field independently of the right half, or visual shapes independently of their color."

I know that, as stated, the authors' example is almost a commonplace, but Eric's interrogation of "Melanie" in his recent book leads me to think it is not true, as she confidently defends her experience of particular components of imagined visual scenes, and disclaims experience of other components that would seem to be inseparable. (Seeing, in her imagination, a person standing, without any experience of the position of his legs, for instance.) It would seem that color, too, could be "unspecified" in ones experience of a shape.

I don't notice anything important in the authors' paper that rests on the validity of this passing assertion about the unity of experience. In fact, by proposing that anesthesia reduces consciousness by interfering with integration, and proposing a simple, fairly abstract model for measuring integration that would lead to the ability to grade/measure degrees of consciousness, the theory (theories) reviewed seems to support the idea that conscious experience can be partial.

Eric Schwitzgebel said...

Thanks for the comment, Gary! I'm not opposed to some sort of integration view of consciousness. It's an intriguing idea that these sort of half dis-integrated experiences might be "partially conscious". It does seem to me that consciousness must come in degrees and admit of in-between cases, like virtually everything else, though on the other hand I don't feel I have a very good intuitive handle on what it would mean to be "partially" conscious. I also wonder whether imagining shape without color and the like is the kind of dis-integration that would interfere with full consciousness. An interesting tangle of issues here!