Thursday, August 8, 2013


I used to think that the problem of consciousness can be plausible reduced to the problem of representation: that to have a quale of red might just be to represent something as being a certain way. But this can't be right. For we have both conscious and unconscious representations. My mind represents Beijing as the capital of China even when I am not consciously thinking about Beijing, capitals or China. And the point seems to generalize: the representative content that a quale of red is supposed to correspond to can surely be had unconsciously, e.g., as when I have a non-occurrent belief that paradigmatic tomatoes are red.

But perhaps there is something in the vicinity that might work. Let me try this. There are two kinds of representative mental states: fundamental and derivative ones. Non-fundamental representative states get their representative content from other representative states (ultimately reaching back to the fundamental ones) in the way that a book gets its representative content from the representative states of human language users.

Then to have a quale is nothing but to have a fundamental representation. Thus, when I see the tomato as red, I am having a fundamental representation of redness (or maybe of the reflectivity/emmissiveness with respect to a certain range of wavelengths of light). But when I have the non-occurrent belief that a tomato is red, my non-occurrent belief is a derivative representation, which derives its content from conscious and hence fundamental representative states. Perhaps it derives its content from my having perceived tomatoes and red things. Or perhaps it derives its content from a propensity to produce certain kinds of conscious images, including red ones, in my imagination. (The imagination seems to me to have played a rather bigger role in philosophy of mind in past centuries. Perhaps that role should return.) Or perhaps, if I have never seen or imagined anything red, it derives its content from someone else's fundamental representative states—say, from someone else's having seen red things.

On this picture, we have two axes of differentiating qualia. There is a representative content axis, which differentiates the quale of red from that of green or of squeaky. And there is a mode of representation axis, which differentiates non-occurrent beliefs involving redness from occurrent perceptions of red things. On the above story, mode of representation difference reduces solely to the difference between fundamental and derivative representations.

But maybe a further differentiation is needed: to imagine (or remember) a red tomato is phenomenologically different from seeing a red tomato, yet both are conscious. What accounts for this difference? If there is a difference of mode without a difference of content here, my above account won't account for this, since it only licenses a binary distinction, not a ternary one between non-conscious representation, imaginary representation and perceptual representation. But maybe the difference between the imaginary and perceptual is simply one of detail of content? Given that we so easily drift between imagination and dream, and that dream is phenomenologically like the perceptual but with less detail, this is not implausible.

So this suggests the following way to make the distinctions:

  1. Thought of red without consciousness of the quale of red = A representative state whose red-content component is only derivatively representative.
  2. Perception of red = A fundamental representative state with detailed red content.
  3. Imagination/memory of red = A fundamental representative state with non-detailed red content (and maybe some truth-canceling "this isn't real/present" content? Spinoza has something like that).

Deep question: Are fundamental representative states metaphysically fundamental or only representatively fundamental?

No comments: