TVSS Proof of Concept!

Those of you who have read my paper Neurobiological Dispositionalism: A Naturalistic Account of Color Perception will recall a lengthy discussion about Bach-y-Rita's hypothetical Tactile-Visual Substitution System (TVSS) device. In short, a TVSS is a device that uses a camera (or something like that) to take in information about the environment, then transmits that information to a human user by way of coded tactile information (timed vibrations, small electrical stimulations, etc.). When Bach-y-Rita wrote about it, the TVSS was (as far as I know) still very much hypothetical--well no longer! Thanks to researchers at the University of Tokyo, we now have a working TVSS, called "haptic radar" by its inventors. Watch the video of it in action, then some discussion:



In the video, the presenters tout the TVSS as a "6th sense;" it is this intuition that I argue against in the section of Nerubiological Dispositionalism, and it is this intuition that I think leads to some very wrong conclusions. Snip from the paper:

What this story [about sensory modalities] perhaps shows most strongly, then, is that the nature of the subjective experience of a color (or indeed any perception at all) enjoys no necessary connection to that quality or object in the environment that causes or produces the experience[1]. Some[2] have postulated that there must be some necessary link between the object represented by the color experience and the representational content itself—that is, that when I perceive an apple as being red, I perceive something that is literally true of the apple: namely, its redness. This example, I think, shows that this notion is fundamentally mistaken; there is nothing in the phenomenal experience that we call our “perception of redness” that has some special metaphysical link to the nature of the red object; to suppose otherwise seems to make one guilty of anthropocentrism in the extreme.

Alva Noë has pointed out in conversation with me that this example might be taken as question-begging—we cannot, he claims, understand color (or any perception) just in terms of what the experience of it is like. To make this point, he invokes Bach-y-Rita’s example of the tactile-visual-substitution-system (TVSS)[3]. The TVSS is, in short, a device that changes data from the “ambient optical array” into tactilely experienced data; the subject wears a video camera on his head, which is connected to electrodes on (say) his tongue. The camera and the electrodes are connected in such a way as to produce an orderly and predictable pattern of tactile stimulation on the tongue when the camera is presented with visual data. Noë notes in his book that “For subjects who are active […] it becomes possible, in a matter of hours, to make quasi-visual perceptual judgments. For example, subjects can report the number, size, spatial arrangement, and so forth of objects at a distance from themselves across the room.”[4] This, Noë claims, should prompt us in the direction of objectivism about visual perception in general (and thus color perception in specific); the TVSS shows (he claims) that what defines visual perception cannot be the phenomenological experience—it is the source of the information being perceived (e.g. light waves), or the mode of interaction between the subject and the environment that makes this definition coherent. My Ups, he objects, come naturally equipped with a kind of TVSS—information about what Gibson might call the “ambient optic array”[5] is translated into certain bodily sensations. I think, however, that there is a certain element of disanalogy here—namely that the Ups are, to reiterate, equipped with such a device naturally, while users of the TVSS rely on technology that is anything but natural.

We might draw an analogy between the TVSS as it has been described and the cane that has been used by the blind for centuries, for what is the TVSS but a more sophisticated way of obtaining information about the environment tactilely? Let us take a moment to consider the similarities. In both cases, information about the environment—information that would be obtained visually by a person with normal visual capabilities—is being obtained tactilely; in the case of the TVSS, this is happening by way of minute electrical signals on the tongue, whereas in the case of the cane it is happening by way of the user sensing contact between the cane and some object. Again, in both of these cases, it seems that the actual mode of perception is tactile; we certainly would not confuse the blind person’s stick as operating as a kind of “new sense” simply because information is being obtained without the stick’s user directly touching any object (aside from the stick itself). Why, then, do we seem to intuit that the TVSS is actually a case of visual perception, rather than an analogous case of (enhanced) tactile perception?

To discern the origin of this intuition, let us now examine the differences between the TVSS as described and the blind person’s stick. Perhaps foremost among these distinctions is the fact that the TVSS seems to employ light as a vital part of the perceptual causal structure, whereas the stick does not. In other words, perhaps the true modality of a given perception depends—at least partially—on the portion of the environment from which the information is obtained; since the TVSS depends on light to function (just as vision does) perhaps we can say that it counts as a case of visual perception. After all, Noë notes that users of the TVSS seem to be able to perceive facts about their environments that otherwise could only be known through a veridical case of visual perception—namely, the size, arrangement, and number of objects at range. However, this definition of “visual perception” as “any mode of perception which depends on light to function” seems to open new difficulties. Consider a scenario in which a person’s vision is rendered unusable (say, as by a blindfold), but instead of being equipped with a TVSS, his environment is simply changed such that there is a very bright light mounted on top of each object, with a brightness corresponding to the object’s size. If these lights were of the proper kind (e.g. incandescent bulbs), it seems that our individual would be equipped to navigate through his environment just as well as an individual equipped with the TVSS simply by sensing the heat emanating from the bulbs, and using that information to form judgments about the size, number, and location of objects. It might be objected at this point that our test subject in the situation just described would not be navigating by any sort of visual sense at all, but only by his tactile sense of heat; I would tend to agree. However, this intuition seems to present a problem for the defenders of the TVSS qua visual perception, as it is not clear why exactly this new case is not analogous; light is playing a causal role here, just as it is in the use of the TVSS, and our new subject can navigate his environment nearly as well (if not just as well) as his TVSS-equipped counter-part.

---

[1] This statement is somewhat misleading. I do want to say that there is a necessary link between a certain kind of stimulus and the subjective experience produced in an organism with a specific set of biological features. I do not, however, want to admit the point that whatever quality it is in red objects that causes our subjective experience of red has any logically necessary connection to our (i.e. the normal human perceiver’s) experience of red.

[2] John Campbell, 1994; P.M.S. Hacker, 1987

[3] Bach-y-Rita, 1996

[4] Noë, 2004, page 111

[5] Gibson, 1973

Let me reiterate my point a little bit here (ignore the discussion of the Ups and Downs; it refers to another part of my paper, and isn't really relevant for this point), as I realize that the method of presentation in the paper might not be the most appealing for less philosophically inclined readers.

I think it's a mistake to call the TVSS a kind of visual perception (as Alva Noe seems to want to) or a new sense that is neither visual nor tactile (as the Tokyo researchers seem to want to). Instead, it is a device that interacts with the environment in such a way as to provide the wearer with tactile information, which he can then use to infer certain facts about his environment (e.g. that someone's hand is about to touch the back of his head). I compare the TVSS to a blind man's stick--in both cases, information is being obtained tactilely, but without the traditional tactile interaction (i.e. touching the object you're trying to learn about). Just as we wouldn't call the stick a "new sense," we shouldn't call the TVSS--as amazing as it is--anything but a clever melding of IR technology and tactile information transmitters.