Wednesday, September 20, 2006

Can People Imagine Things from Multiple Angles at Once?

Here's another question about imagery experience -- related to Monday's post about whether images have subjective location. Can people (at least some people, in some circumstances) imagine things from multiple angles at once? Francis Galton, in his seminal study of imagery experience (1880, 1907), says that some of the best imagers report being about to do this. Jorge Luis Borges describes a similar phenomenon in a fictional character obsessed with a coin he calls a "Zahir":

There was a time when I could visualize the obverse, and then the reverse. Now I see them simultaneously. This is not as though the Zahir were crystal, because it is not a matter of one face being superimposed upon another; rather, it is as though my eyesight were spherical, with the Zahir in the center.

Now is this really possible? I can't claim ever to have had such an imagery experience myself; but that doesn't mean others can't do it. On the other hand, I don't think we should simply take people at their word when they make unusual claims about their experiential lives.

Here are two reasons one might think multiple-angle imagery is impossible:

(1.) If images are located in subjective space, as some people report -- say, near your forehead -- then it seems natural to suppose (if not strictly implied) that we have a single visual angle on those images, presumably the angle from the center of one's subjective self to the image in question. (Now that I see these words in print, though, I must say there seems to me something a little fishy in them!)

(2.) If images are instantiated in the brain (or caused by the brain) in accord with some topography of either subjective or objective space (e.g., the right side of the image is created by this location in the brain, the left side by this other location), then that topography may well require a single visual angle or point of view (e.g., in "circular vision" right and left might not be well defined).

I don't mean to say that either of these points is decisive -- not by a long shot. I wonder: Have any readers of this blog have had experiences they would describe as imagery from more than one angle simultaneously?


Tad said...

Eric -

This is a good question. I have very poor visual imagery, so I rarely know what people are talking about when they make such claims.

A related question I've always wondered about is the phenomenon of synesthesia (spelling?). Some report that they can see sounds or hear colors, and that this is an immediate experience, not some kind of metaphorical leap. Nabokov famously claimed to be able to do this, and some have claimed it's common among writers. I think Ramachandran may actually have done some empirical research on this recently.

What do people know about this? Does anyone who reads this blog experience synesthesia?

I find the idea of hearing colors or seeing sounds as confusing as the idea of seeing things from multiple angles at once. Do you first hear middle c, and then imagine seeing red? Or does your whole visual field turn red when you hear middle c? When you're shown a red apple, is it harder for you to hear because of the noise in your 'mind's ear'? Are there imaging studies that confirm unusual activation in inappropriate cortices when synesthetes (?) are shown stimuli?

Anonymous said...

Hi Eric,
I just wanted to mention experiment by Perky C. where the subject were asked to imagine a banana on a screen. Unknown to them a projector was put behind the screen, and slowly the intensity of the projector was intensified. The subjects didn't realize that they are looking at the real picture, while the projected banana image was clearly visible to other people.
BTW, tried to find some link about this on web, but you will never guess what shows up for "Perky C" search :)

Tad, you can watch Ramachandran talk on synaesthesia here. Very interesting stuff.

Eric Schwitzgebel said...

Thanks for the comments, folks!

Tad: I had been inclined to be skeptical -- and not to know what to make of -- reports of synaesthesia. Then Ramachandran and Hubbard came out with a very cool experiment suggesting that synaesthetes found it easy to see shapes arranged by number, where non-synaesthetes don't. For example, if you have an array of letters scattered across a field, with the H's in a triangle, most people will not easily detect that the H's are in a triangle, whereas their synaesthetes (they reported) could detect this very quickly -- as non-synaesthetes also would if the H's were all colored red and the other letters were non-red. This suggested there might be some validity to their claims that they literally see different colors with different numbers!

About a year ago, though, someone told me that others have had trouble replicating Ramachandran's and Hubbard's work on this; so I don't know.

In any case, neural activation can spread between regions of the cortex associated with different sensory modalities; there's no reason to suppose that some people might have greater interconnection. I don't know if this has been studied with neural imaging in (claimed) synaesthetes; but I think there's some evidence for this related to the synaesthetic experiences associated with LSD.

Thanks for the links and references, Tanasije and Pete! I think I know the Perky experiment you're talking about, Tanasije; but I don't recall it being connected to multiple visual angles.

The possibility of representing four-dimensional images is cool, Pete. And why not, if the brain is plastic enough? Maybe I'm not getting something, but I don't see how the Hinton case does it, though.

Anonymous said...

Oops, sorry Eric, that Perky reference was supposed to be about your previous post.

Eric Schwitzgebel said...

Ah, right! Yes, that's pretty interesting. Since visual objects are subjectively located and since -- if Perky [and Titchener her teacher and Hume] are right -- imagery experiences are very much like faint perceptual experiences, so much so that one could mistake one for the other, then it seems to follow that imagined objects are also in subjective space.

Nice thought!

Anonymous said...

Hi Eric, this isn't really related but your post reminded me of an old paper by Ulric Neisser (and someone else) on the phenomenology of memory (perhaps you've seen it). The results coincide with my own experience: the more recent the memory the more likely we visualize it from the first-person perspective (i.e., from behind our eyes); the more distant the memory, the more likely we see it from a third-person perspective (i.e., as a "movie" in which we appear). There are, of course, questions about the reliability of the reports, but those are the reports (and that's what I would report). My guess (and I don't remember if Neisser suggests this too) is that it may have something to do with the difficulty of "taking on" a different body image (e.g., our childhood body)? It'd be interesting to know whether this phenomenon is influenced by the prevelance of photographs and home movies.

(Now that I've written this, I can't remember if you were there when I was talking about it at SPP--my memory of that discussion is already shifting to the third-person perspective. Oh, that's why your post reminded me of this--it also seems to me that I sometimes remember events simultaneously from both perspectives, or maybe it just shifts back and forth quickly.)

Anonymous said...


The crucial idea behind Hinton's cubes is something like this:

Voxels in a 3-D array are to 4-D objects what pixels in a 2-D array are to 3-D objects.

Eric Schwitzgebel said...

Hi Eddy -- thanks for the comments! The Neisser work you mention sounds vaguely familiar, but I don't know if I've read the original. (One great thing about this blog are all the cool articles people point out to me.) And, indeed, if your last point is correct, it might be relevant to the issue of this post.

Pete: Sure, but why does that make it a 4-D imagining (if that's is what's being proposed), instead of just a 3-D imagining? Aren't you just restoring the dimension lost in a flat projection?

Maybe the thought is that we can imagine 3D objects by entertaining 2D visual images, so likewise we can imagine 4D objects by entertaining 3D visual images? I'm not sure if that's true; but even if it is, the image itself remains three-dimensional.

Anonymous said...


I think you get the main points. It's not that Hinton has 4-D objects in his head, it's that his cubes emulate the 3-D hypersurface of a 4-D creature's retina. I'm assuming, I guess, to flesh out the analogy, that normal visual imagery involves the emulation of the 2-D surface of a human retina.

I know that you've blogged on the important lack of flatness of human retina, but hopefully the above conveys the drift of what Hinton thought he was up to.

Genius said...

If I visualize a three D image I end up doing it with a rotating see-through object.
I htink that is because I have seen them before on TV etc.
I dont know if I can imagine seeing somthing fundimentally different from what I have seen. probably what I would need to do is not actually try to see it and just "think it". I guess

Anonymous said...

Regarding synesthesia, as a child I saw numbers as though they were colored. This ability has disappeared with age. I have a friend that still see numbers as colored. He spots a shape or a single number in an array of numbers very quickly. So, synesthesia is for real.

Eric Schwitzgebel said...

A belated thanks for your comments, genius and brian! (I was on hiatus and only now got to checking this post.)

I think your comments speak for themselves, and I would encourage other readers to send in relevant descriptions of their experiences (in this post or in any other), which I find an interesting source of basic data.