Wednesday, May 4, 2011

There ain't nothin' about Mary

The knowledge argument purports to show that physicalism is false. The arguments exploits our intuition that Mary learns something new when she sees red for the first time for its rhetorical force. But if we can show that that intuition may not shared for a group with different native linguistic-conceptual color scheme the argument may lose much of its persuasiveness. The linguistic color scheme one employs is a historical contingency. It may have been that one was born into a society that employs a different scheme. If one's intuitions on philosophical issues turn out to be different based on that contingency, that would cast some doubt as to the substantiveness of findings based on that intuition. The argument goes something like this:

Mary is a brilliant scientist who is, for whatever reason, forced to investigate the world from a black and white room via a black and white television monitor. She specializes in the neurophysiology of vision and acquires, let us suppose, all the physical information there is to obtain about what goes on when we see ripe tomatoes, or the sky, and use terms like ‘red’, ‘blue’, and so on. She discovers, for example, just which wavelength combinations from the sky stimulate the retina, and exactly how this produces via the central nervous system the contraction of the vocal chords and expulsion of air from the lungs that results in the uttering of the sentence ‘The sky is blue’.… What will happen when Mary is released from her black and white room or is given a color television monitor? Will she learn anything or not? It seems just obvious that she will learn something about the world and our visual experience of it. But then is it inescapable that her previous knowledge was incomplete. But she had allthe physical information. Ergo there is more to have than that, and Physicalism is false. (From the Stanford Encyclopedia of Philosophy article “The Knowledge Argument)

Before I turn to my basic argument, I need to give a little primer on linguistic relativism. It has been known to linguists that many language groups have different color terms than ours. For example, native English and modern Chinese speakers both have the same number of basic color terms and both languages partition the color spectrum in essentially the same way but other languages do not partition the color spectrum the same way.

Similarly, languages are selective when deciding which hues are split into different colors on the basis of how light or dark they are. English splits some hues into several distinct colors according to lightness: such as red and pink or orange and brown. To English speakers, these pairs of colors, which are objectively no more different from one another than light green and dark green, are conceived of as belonging to different categories.[1] A Russian will make the same red-pink and orange-brown distinctions, but will also make a further distinction between sinii and goluboi, which English speakers would simply call dark and light blue. To Russian speakers, sinii and goluboi are as separate as red and pink or orange and brown. (wiki article on color terms. See also wiki article “linguistic relativity and the color naming debate” and “linguistic relativity”)

Some languages such as certain Trans-New Guinea languages only have two basic color terms (dark and light!) and all colors are considered a shade of those two basic colors. Likewise, some languages have more than we do and native speakers tend to speak of, and presumably think of, more basic colors concepts. Keep in mind that it has been shown by anthropologists and cognitive psychologists that people all over the world can see and make color distinctions the same way we do (unless they are colorblind of course). It's the way that they partition and group the colors on the spectrum that is culturally/linguistically distinctive. It should be noted that "color relativism," that is, this notion that color term differences reflect differences in cognitive differences in color concept classification and that this classification is in turn a function of cultural/linguistic norms is still a controversial thesis among linguists, anthropologists and cognitive scientists. But if the relativists are correct, even partially, my argument follows.

My skeptical question as relating to the issue of color qualia is this: let's say that there is someone, call him Larry, who has lived an otherwise normal life except that for whatever reason, Larry has never seen a particular shade of yellow before. This shade only spans one nanometer on the color spectrum and may be a rare shade of yellow so it may not be a big deal that Larry has yet seen such a shade. One day he comes across this shade which we will call 'mellow.' Has Larry learned anything new by perceiving mellow for the first time? Has he learned anything new much as Mary presumably learns upon seeing red?

It seems that most folks would say "no.” If mellow is similar enough to a shade of yellow Larry is familiar with, Larry may not even realize he is seeing a shade of yellow he has never encountered before. It would be a wholly uneventful event for Larry. The “no” answer seems plausible for assume that the answer is yes. Larry gains a new color concept corresponding to the qualia associated with mellow. He has a new color experience analogous to the first experience of seeing red in the original knowledge argument. The color segment of the color spectrum may be divided into shades and these shades may be further divided into sub-shades of those sub-shades further divided and so on perhaps ad infinitum (I'm assuming that the color spectrum is continuous but even if it is not but divisible into very fine sub-shades, my reductio argument follows). So ex hypothesi knowing and experiencing a segment of the color spectrum for the first time, even if a tiny segment corresponding to a shade of some familiar basic color as Larry presumably has in my scenario is to know and grasp a new color concept, then we can ask what happens when instead of seeing mellow, Larry sees a shade of yellow that is also a shade of mellow perhaps occupying half the length on the color spectrum as mellow. If he also learns a new color in this scenario (we will call this sub-shade 'sub-mellow') then we may reiterate the question mutatis mutandis of whether he learns a new color concept and so on. There seems to be no clear line to draw when Larry does not learn of a new color concept from ever so fine divisions of shades of yellow. But if there is no line to draw and the color spectrum is continuous then seeing any shade for the first time of arbitrary spectrum span implies learning an (uncountable) infinite number of new color concepts. That I think is absurd.

Each of us normally sighted individuals may see many new shades of familiar colors everyday. And each of those shades may be divided into sub-shades. But it is preposterous to say that we learn an infinite number of new color concepts on a daily basis. So Larry does not learn a new color concept simply by seeing that shade of yellow for the first time. He simply sees yellow, not gain a new “mellow” color concept.

Now imagine that a tribe of people that makes only the two dark-light basic color distinction and their possible reaction to the Mary though experiment. Will they say that Mary learned anything new by seeing red which presumably they think is just a shade of their super-color "dark"? It may be that they would say "no" much as we do to the hypothetical case of mellow I gave above and for the same reasons we give. For them, red is just some shade in the basic super-color dark. Our intuitions about the knowledge argument seem to be sensitive to the linguistic color scheme we actually employ. If accidents of our birth influences the direction our intuitions point regarding some thought experiment, there may be good reason to be suspicious of insights those intuitions are supposed to offer us.

This so far of course, a limited application of the Sapir-Whorf hypothesis for color terms to explain qualia. Much as many think that the Sapir-Whorf hypothesis is a skeptical hypothesis regarding realism of natural properties or kinds, my attempt has been to use the hypothesis as a skeptical attack on qualia. Thus, if my skeptical attack is correct we, under another color schemes may not agree that Mary would learn anything new upon seeing red. She may simply dismiss it at most simply as a perhaps odd shade of "dark" or "black" much we under our color conceptual scheme presumably view the person first seeing the uneventful mellow for the first time.

One possible response to my argument thus far is that it only shows that red qualia (or any color qualia not light or dark since all known linguistic communities have at least these two basic color concepts/terms) does not have the epistemic properties ascribed to it by the knowledge argument and thus it is not effective as an argument against physicalism in whole. Though there are no red qualia, there are still white and black qualia for no known linguistic communities lack these basic color concepts and thus all communities presumably concede that there are qualia corresponding to white and black qua basic colors.

The argument I made is meant to show that the quintessential example of color qualia red can be doubted that it has the private epistemic property it is alleged to have by culturally relativising red. If red qualia goes by way of this kind of relativism, what is to stop all color qualia including light or dark qualia going with it? What makes these more special as qualia that red is not? Now the burden is at least shifted. This is the main rhetorical point of my argument.

Lastly I'd like to see an X-PHI experiment comparing survey responses of two linguistic groups: a group composed of native language speakers of a language that does not recognize the color red as a unique primary color but a shade of some supercolor such as "dark." The other group does recognize red as a primary color such as native English speakers. The speakers of the language that doesn't recognize red should be from a group that also has not had much influence from western culture or any cultures that partitions the color spectrum in a different way. Of course that will be hard and in practice no culture is so discrete from the linguistic and cultural influence of others but there still are some cultures that are relatively isolated. Many speakers of the Trans-New Guinea languages are so isolated even today. They would be a good group to sample because as I've mentioned, they only recognize two basic primary colors (dark and light). They would serve as a good comparison group with Native English speakers, e.g. To test the intuitions on the knowledge argument compared to the control group of native English speakers would be a good way, in my view, to see if there is anything substantive and sound about it. Let's say that the non control group doesn't think that Mary learns anything new once she sees the apple for the first time, that would go in some ways in showing that it is conceivable Mary learns nothing new when she sees red for the first time contrary to Jackson's original claim in his argument.


  1. For a while I've taken Mary to be most salient as an argument about know-how vs. know-that than an argument about "qualia." Even if Mary knows perfectly well that red is equivalent to R-fibers firing in the brain (or whatever), she lacks the know-how to tell when her R-fibers are firing. In the same way, Barry the Bicycle Scientist might know that you ride a bike by pumping the peddles, etc. but he will still tip over the first time he tries to ride a bike because riding a bike requires the know-how of working your legs and balance systems together in harmony.

  2. In other words, what Mary learns is less "This is what it's like for R-fibers to fire" than "Oh this is how you can tell if your R-fibers are firing."

  3. I believe that would be a dispositional account of color vision. I think there is a debate between Jackson and the dispositionalists that has since blossomed. I have never read much of that literature so I can't tell you anything about it.

    But I did read this one interesting paper relating to this debate and which explicitly defended Jackson's argument. It's a really interesting paper and I recommend it when you get the time. It called "Knowing How" by Jason Stanley and Timothy Williamson.

    They basically argue that all knowledge is just species of knowing that like knowing something akin to the Fregean sense. However, it seems to me that Stanley and Williamson could have argued as easily that all knowledge is just a species of knowing how or dispositional instead of cognitive knowledge. 

  4. Thanks for the link.

    My instinct is to agree with you about dispositional vs. cognitive issue. The trouble with monisms is that if it's all water, why can't you just as well say it's all ice or it's all fire? If all sentences are declarative, what's the difference between that and saying they're all interrogative instead? I'm with the later Wittgenstein in thinking that we don't always need to unify everything. Sometimes unifying too much causes you to lose the point of the category.

  5. I think that physicalists are very much in the minority now in philosophy of the mind. They have been put on the defensive by arguments such as Jackson's but also many others of a different type. The bottom line is that they haven't found even the slightest beginnings of a way to reduce the phenomenal properties of consciousness into a material framework. There is an explanatory gap.

    Now as for your objections to the framework of Jackson's thought experiment. It doesn't take much to adjust for your objections while leaving the point that Jackson was making intact.

    This is a thought experiment so we can do whatever we want.

    Suppose Mary was brought up in a completely white room with no other color or pigment of any kind. Mary herself was painted completely white every night while she slept. She then ventured out into the world one day and saw a color, any color, for the first time. Did she learn something new? For that matter, say they suddenly introduced black into her room one day. Did she learn something new?

  6. @Casey: you're probably wrong about the popularity (or lack thereof) of physicalism among philosophers of mind. In the PhilPapers survey, 60% of all respondents whose AOS is philosophy of mind accept or lean towards physicalism, while only 26% accept or lean towards non-physicalism (among target faculty, the split is slightly more uneven: 61% to 22%).

  7. Casey, it would be more vivid as an example to use a Mary case that didn't involve vision at all such as a blind Mary that is suddenly granted vision (by surgery perhaps) and had learned all she knew prior by auditory means. I was arguing against color qualia specifically but there may or may not be something like a generic "visual" qualia. In your example, it doesn't seem possible for her to make any kind of visual distinctions if she is only allowed to see one shade of white (how they get rid of shadows etc will be an intractable pragmatic problem here) and thus she will not be able to learn anything visually prior to her release from her room except that one white shade's qaulia if such thing exists.

    I also agree with Jura that most philosophers seem still to be convinced by physicalism.