Become a Fan

« Expressing repugnance in reviewing | Main | The "Kuhn Loss" in the demise of natural philosophy... »

04 June 2012


TrackBack URL for this entry:

Listed below are links to weblogs that reference Minds on Monday: Perceptual Models Without a Viewpoint:


Feed You can follow this conversation by subscribing to the comment feed for this post.

Carolyn Dicey Jennings

so perception "produces," "is generated from," and "employs" models, but why think that models are perceptual? Could we say that perception produces, is generated from, and employs memory, knowledge, or, gosh, the self, without any of these things being properly perceptual (whatever that means)?

Chris Rawls

A great reference for this discussion, at least something that has struck me profoundly currently, is the new work by Alva Noe 'Out of Our Heads: Why You Are Not Your Brain and the Biology of Consciousness' released this year. It is extremely well done, and as Dan Dennett and Oliver Sacks write on the back cover, it should be required reading for anyone who thinks about thinking.

Mohan Matthen

You're right. And yes, maybe, for all of those questions about memory, knowledge, etc. But I am less confident, at this very moment—maybe not tomorrow—about what "properly perceptual" is.

Mohan Matthen

yeah, right.

Taylor Murphy

It is interesting that you use olfaction in your examples, for many people cannot form olfactory images. Maybe you're an exception. I've seen a number of studies that suggest that olfactory imagery does not improve olfaction (e.g., in detecting threshold olfactory stimuli) and nor does priming with words (like 'cinnamon'). Except, that is, in people who are quite good at naming olfactory stimuli. Incidentally, this does not generalize to olfactory working memory (given an olfactory stimulus, holding it in one's mind does help subsequent threshold detection of smells), and I believe fMRI results still (always?) show increased blood flow to primary or secondary olfactory areas (I do not remember if these were good imagers, or if this just reflects concentrating on olfactory stimuli without any relevant olfactory images). It has been about a year since I looked at the scientific literature on olfaction though, and this is just my recollection.

In any case, I agree that there are non-perspectival ('allocentric') images or models, and that these are located in space by adding an observer in vivid imagination. But I wonder whether there are perhaps several kinds of perspectivalness. One kind has to do with how the object is oriented with respect to oneself; its heading (e.g., seeing the front or back, etc). But then there is a second sort, which is where in space the object is located with respect to oneself (i.e., distance & magnitude). Third, there is its egocentric heading, which is whether it is to the left or right of oneself or one's visual field. These are standardly distinguished in the spatial cognition literature I am familiar with, and I wonder how claims of perspectivalness (simpliciter) relates to these perspectival aspects of perception and imagination.

If I form an image of a banana, it's clear that it has a heading and it's clear that it has an orientation, but it's not clear whether this image is a meter away, or if it is a larger banana 2 meters away, and so on. I wonder if it could be at a distance of 0 or be undefined, so that I am not imagining any particular size/distance for the banana. If that is so, it is quite like the auditory case, where the sound is not necessarily located anywhere in space with respect to oneself (i.e., no particular loudness and distance away from oneself).

Thoughts on this?

John O'Dea

In support of Taylor Muphy's point, perhaps it is useful to add that auditory perception seems fairly tolerant towards indeterminacy, at least more so than vision. We often hear a sound perfectly well without having much idea of where it came from, but this is surely not a case of non-perspectival perception since we may hear that sound to be coming from some particular distance away. The best thing to say might be that the sound is heard to come from somewhere, but nowhere in particular. This is obviously not the same as saying that the sound is not heard to come from anywhere.

Mohan Matthen

Very interesting point about olfaction. Do you think that most, or even all, people have non-imagistic add-ons (captioning or the like) for smells? I certainly did think that I can recall the smell of fresh baking get stronger as I approach the kitchen, but now you have introduced a strong doubt: perhaps the memory is semantic (or something similar).

A couple of questions:

1. What about the Proust phenomenon of smells evoking memorial images? This strongly suggests that olfactory memory is episodic at least insofar as it carries source information. This is compatible with there being no olfactory images. But since episodic memory is usually imagistic, it does suggest that a smell can at least form a part of a larger image.

2. What happens when we identify smells? Admittedly, we are not very good at this. But we can make broad identifications, e.g., we can identify cinnamon or lavender, etc. Again, this is compatible with there being no stored image. But unless the smell associates with a name in these cases, it seems to suggest some kind of Hebbian stored activation patterns

(Can you point me to some literature on this?)

In visual images, I agree that distance might be optional, as direction is in auditory images.

Mohan Matthen

I agree that audition is more indeterminate with respect to sound location than vision. And this is why vision dominates audition in spatial tasks. Active sound perception, in which I am allowed to move and look, as well as listen, is a good bit better. (It could even defeat the ventriloquist illusion.) And it's implicit in my story that the formation of allocentric sensory images involves active perception. (I meant, but forgot, to include a link in this story to my piece last week on perceptual exploration and active perception.

John O'Dea

Your point about about active perception improving determinacy seems very intuitive to me. I guess my point was just that less determinate does not mean less perspectival, and perhaps even that complete indeterminacy need not be perspective-free; especially if the indeterminacy is part of the experience (we experience that there is indeterminacy). Audition seems to me this sort of case.

The link to your earlier post is interesting, since there ought to be link a between perspectival experience and perceptual constancy. After all, the point of constancy is to filter out the perspective - or at least parse it, to borrow a phrase. And there is no evidence for spatial constancy in the case of smell (there is odour constancy, but it is not spatial; if you breath in more sharply, the smells do not seem stronger). I take that as evidence in your favour smell is non-perspectival (but not therefore, of course, amodal).

Incidentally, the house I grew up in had a sort of complicated shape. Although I have no problem imagining walking around it, and seeing it from any position on the ground, it is incredibly difficult to imagine looking at it from above; I feel like I have to construct it in my head piecemeal from the various ground-based views. Doesn't that suggest that the shape of the house is not stored amodally, only lots of individual perspectives?

Chris Rawls

The Noe book is worth one's time and investment. I don't think in images. I have only met a few others with this phenomena, one is in robotics at CMU in the states. For that reason, diagrams and models of systems don't work well with aiding in my understanding much typically. I dream in images, but in waking hours I don't think in images that I am aware of hardly at all. If I try to imagine something, I don't see any details in the picture I am seeing much at all. I see amorphous shapes and colors at best. This may be a poor analogy, but I cannot focus the lens to see any image more clearly in waking hours. None-the-less, I think about philosophical systems and many other overlaps on an interdisciplinary level daily. I do well thinking conceptually and I tend to describe the ideas I have as a feeling more than an image or representation. Just food for thought...

Mohan Matthen

A number of interesting point in your comment 9.

I agree with you that indeterminacy of location doesn't imply freedom from perspective. But it might mean that perspective is indeterminate. Take the sound of an ambulance blaring its siren on city streets. Drivers don't know where it's coming from, and this is part of the experience. (Very frustrating!) But this does not mean that there is no perspective. What it means is that you don't know what your perspective is. By the way, I think this is compatible with my overall picture: what I was suggesting is that the elimination of perspective comes with the increased accuracy that comes from perceptual exploration.

Constancy is, as I understand it, some kind of sameness of experience despite different conditions of viewing. When you locate smells or sounds through active exploration, you eliminate the different conditions of viewing, i.e., you have information that doesn't depend on these conditions, and thus it's not entirely appropriate to talk of constancy.

Finally, your childhood house. I take it you never saw it from above. So when you try to image it that way you are performing a complicated act of construction. By the way, I don't think that a perspective-free model is amodal.

Mohan Matthen

It is very interesting that you don't have mental images. I thought that only happens with frontal lobe damage, and you don't have that, do you? Do you have auditory images? Can you listen to a piece of music in your head?


Mohan, I don't see what's wrong with what I assume you take to be the standard picture, namely, that we don't generate and store non-perspectival representations, it's all perspectival and when we imagine moving around the visualized house we use the existing (and stored) perspectival representations to create new perspectival representations on the basis of them. So what's wrong with this view?

Mohan Matthen

I can't refute that view, but I think it gives an implausible account of how you generate dynamic images. For instance, suppose you image walking around the house clockwise. How does the system call up the appropriate images in the right order. They aren't stored such that the view from position P is naturally followed by the view from P + ∂, because such a storage system would defeat counter-clockwise perambulation. My theory is that the views are being generated from a single underlying substrate, since this is computationally equally complex regardless of order.

Since the dynamic imaging process is isotropic, the data-structure must be isotropic too. Is that very naïve?

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Comments are moderated, and will not appear until the author has approved them.