The intersection of Hoffman's rule of generic views and Noë’s sensorimotor contingencies

It’s been a while since I last wrote a blog post! Lots has happened since then: I’ve gotten to meet some top Canadian perception and action researchers in the course of interviewing at graduate schools, wrote my honour’s thesis, and finished my undergraduate degree. In the fall, I’m going to move to Germany where I will begin my graduate studies in Cognitive Neuroscience at the Max Planck School of Cognition!!

In the time between finishing my last final exam and starting work this summer, I have been making an effort to read as much as possible. After all, I have so many books that I haven’t yet read that I’m going to have to part ways with while I live abroad over the next five years. I recently finished reading Donald Hoffman’s Visual Intelligence (1998). I bought it at a discount bookstore in Toronto, and initially thought it may have been one of those self-help books that bizarrely co-opt perceptual language. It turned out to be a great, scientifically sound, book on visual perception, though. I haven’t seen it discussed much, and I wanted to talk a bit about it and the similarities between Hoffman and Alva Noë’s (2004) theories of vision.

 Hoffman’s central thesis is that vision is a constructive process. This is hardly a revolutionary claim for vision scientists. I can only barely imagine what non-constructed vision could even be: an optical cable acting as a periscope for a homunculus deep in our brain, perhaps? Where Hoffman makes a neat contribution, however, is in proposing concrete rules grounded in geometry and perspective, which we seem to use to construct visual experience. Many of these rules concern how we determine the boundaries and spatial extent of objects on the basis of edges. The fundamental underlying rule of vision, according to Hoffman, is the Rule of Generic Views: “Construct only those visual worlds for which the image is a stable (i.e. generic) view.” (p. 25). Below are three line drawings of a cube. Take a look at them and consider any differences between them.

(Hoffman, 1998, p. 22)

Each line drawing is a legitimate potential view of a cube, but most likely, only the cube in the middle really appears to be one. The idea behind the rule of generic views is that in everyday viewing, the likelihood of several edges or vertices of a cube overlapping like they do in the left and right objects is so tiny that it is only in cases where there are discrepancies between the edges and vertices of the object that we naturally perceive it as a cube, as we do with the central object.

What stood out the most to me about this is the way that the rule of generic views compliments Alva Noë’s idea of sensorimotor contingencies. Hoffman’s rule of generic views (and many of his other rules of perceptual organization) explain how we discriminate and represent objects on a snapshot-view, whereas Noë’s idea of sensorimotor contingencies extends this across time, and considers the role of action in perception. Underpinning the rule of generic views is the idea that we have an innate understanding of the appearance of objects as a function of our relationship to them. If the proximal stimulus of the cube does not match our hypothesis of what it should look like (mismatched edges and vertices), we reject the hypothesis that the object is a cube. But, where Noë comes in, if we move just ever so slightly, the cubes on the left and right will start to look more like the cube in the middle. Not only can we test the in-the-moment genericity of the cube, but we can also gauge the effect of our own movements on the proximal stimulus of the cube.

I don’t have anything extremely insightful to add, but I just wanted to establish that I feel that Hoffman’s rules for “visual intelligence” and Noë’s writing on the role of action in perception are compatible and complimentary. Where Hoffman establishes geometric rules for perceptual segmentation, Noë picks up effectively where Hoffman leaves off, bringing action into the mix and discussing some of the theoretical implications of such a view of perception. I definitely recommend them both! 

References

Hoffman, D. D. (1998). Visual intelligence: how we create what we see (1st ed). W.W. Norton.

Noë, A. (2004). Action in perception. MIT Press.