In Jackson’s knowledge argument, Mary is a talented color scientist, who was prevented from seeing color since birth. Despite that deprivation, she has studied all there is to know about color processing. When she finally sees something red for the first time, will she learn something new? The argument goes that she probably will, and it means that there is something to having a conscious experience that outstrips representational knowledge, and perhaps also functionalism as a result.
Empirically, we know that sensory deprivation of this kind can severely limit normal brain development. After years of such deprivation, chances are her ability to see color may be permanently gone, or drastically changed somehow. To assume that she can see red exactly like we do is to force our imagination through some plausibly incoherent assumptions.
So instead, why don’t we consider a more realistic example that would serve the same purpose: Maru is a young child who has never tasted natto before. Genetically, she is a supertaster. Like many young children in her culture, she can cook simple meals for herself. She is deeply interested in food and culinary art. But her own parents dislike natto, which is not so uncommon in her culture. So she has never tried it. She heard that it has such a distinctive flavor that either she will love it or hate it. She has been told that natto is basically a kind of fermented soybean. So it is in a way like miso, although the flavor driven by the fementedness is a lot more intense. It is not necessarily as salty, but it is even funkier than old cheese, or Chinese fermented bean curd. In fact, it is on the level of Taiwanese stinky tofu, although for natto the intensity is more on the palette than on the nose. It has a gooey texture, like it is mixed with raw egg white or something. It’s basically ineffable what it is like, she’s been told. She has to try it to find out herself.
The inquisitive and imaginative Maru-chan asks a lot of questions about this curious food. She regularly thinks about what it would taste like. One morning, after a slumber party at her best friend’s, they have natto and rice for breakfast. So she finally gets to try it for the first time. Does she learn anything new?
Obviously, that depends on how Maru-chan’s ‘research’ has gone. But one possibility is, she may say: this is exactly how I always imagined it to be! Based on the other experiences that are familiar to her, with some remarkable level of imagination, she might have actually figured out what it would taste like. On reflection, hasn’t that also happened to us sometimes, for other stimuli? Experiencing something for the first time doesn’t always feel so surprising. So it is possible that she will learn nothing fundamentally new. Tasting natto just confirms what she already knows without first-hand conscious experience. In that case, conscious experiences don’t necessarily outstrip representational knowledge.
Alternatively, it could be that Maru does learn something significantly new. Natto isn’t quite like what she’s imagined it to be. From here, perhaps she gains the ability to imagine it correctly. She also learns some new self-knowledge: this is what natto tastes like to her. At a subpersonal level, some mechanisms in her brain acquire the new information that this is what the relevant sensory vehicle is like, for natto - it is a bit like the vehicles for this other taste, that other taste, nothing like the vehicles for the tastes of ikura, karaage, etc. This is not the sort of stuff one can learn by reading books. Even if one learns the information as a person, it doesn't mean the relevant mechanisms in the brain will get it too.
So in either case there’s no challenge to functionalism.
P.S. This story is inspired by my experience of once dining with an ardent anti-functionalist in a Japanese restaurant in downtown Manhattan. I tried to convince the adventurous philosopher not to order the natto dish, but I failed.
Maru would obviously still learns something new when she tastes natto for the first time, even if it tastes just as she imagined. She still didn't know how it would taste until she tasted it. The fact that she guessed right is not significant.
ReplyDeleteHer ability to imagine a taste she has never tasted should not be confused with the actual experience of tasting. We all can imagine (or dream) things that we've never experienced and, most of the time, we can tell that what we are imagining (or dreaming) is not real. The fact that we can know an imagined "experience" is not real tells us that imagination and experience are different.
I think what’s key here is the weight we put on the term ‘learn’ and whether we can associate new experience with learning. In other words, does a new form of consciousness result in intellectual development? And if that’s not what we mean by ‘learning’ then please identify what else is meant by it. If it is a question of intellectual development, to what end if any end at all? In other words, is the conscious registering of experience something that is useful to human life? His question, of course, is central to utilitarian philosophy that does ground a system of justice and social governance on the principle of the conscious registering (‘learning’ in your sense here, I believe) of pleasure.
ReplyDeleteOn a different but related note, this older article by Malcolm Gladwell on Ketchup might be interesting to you and could be a good teaching resource for undergrads to get them thinking about these issues: https://www.newyorker.com/magazine/2004/09/06/the-ketchup-conundrum
I agree with you entirely Hakwan, about the empirically implausiblity of the Jackson example. E.g. if you were to compare vision in the mononchromat with vision of the trichromat, it's not just a matter of 'nno colour' versus 'lots of colours', as if monochromatic vision is the same except for missing colour. It's a different visual system in many crucial ways.
ReplyDeleteThat said, I think that a central difference between the Mary and Maru case is that the 'data' Maru uses to make her prediction is a set of tastes and their known causes. She's had lots of taste experiences with other fermented foods, and she draws her inference about the taste of this new fermented food based upon something like a similarity space of tastes. So the Maru case seems to be parallel to Locke's missing shade of blue' problem. There Locke asks whether a person who was familiar with every other colour save for a certain 'missing shade of blue', could truly know what it was like to see this blue prior to seeing a surface/light of that particular shade of blue. I think most philosophers would say that if colour were represented by a physical similarity space of some kind, then, in principle, inferring the missing shade of blue could well occur. But whether it did would depend upon things like the resolution of that colour space, our ability to remember precise colours over long periods of time (we can't), our access to that space and so on. (Actually, I take back my comment about most philosophers!) I can't remember Jackson's exact phrasing of his example, but suppose Mary does have access to the (completed) psychophysics of colour. Here, Mary's visual experience is limited to her experience of, well,a grayscale or in terms of surfaces, albedo. The question about Mary then becomes, can she 'interpolate' the colour purple based upon her knowledge of "all of colour science", including the psychophysics of colour? Put this way, the problem is much harder. Presumably Mary can see albedo, although probably not very accurately without chromatic information. In any event, the parallel case for taste in Maru would be this: we have Maru, who has tasted one flavour, marshmellow, and we've asked her to imagine how this new fermented food will taste. Can she do so? It seems impossible, even for the neuroscientifically inclined, because she lacks the requisite flavour space. BTW, I also with you on the Mary question, but that's a long story.
thank you for the feedback! sorry i only saw this now. let's hope we meet and chat about all this some time
Delete