Blog
Fineness of Grain and McDowell's response
Sunday, May 31, 2009
Evans argued that the content of experience is non-conceptual because things can look to have fine-grained shades that we don't have colour concepts for.
McDowell responded that we have colour concepts like 'that shade' which we can use to pick out the fine-grained shades.
However, what if the phenomenal character of visual experience is so fine-grained that you can't even pick out the differences with demonstrative concepts? Imagine you have an experience that has a range of colour phenomenal character from one shade of green to another nearby shade of green, say green23498 to green23499. The range might be tiny, and your cognitive abilities to focus on the differences may be so limited, that you think, incorrectly, that this is a uniform range of colour phenomenal character. Given that you can't attend to the differences between the shades in this range, it seems hard to see how you could form demonstrative thoughts picking out each shade.
Another way of thinking about the limits of your own cognitive discriminatory abilities, as applied to your own experiences, is like this. If you have an experience which has a certain range of colour phenomenal character, that range will be associated with a certain region of your visual experience. There may well be a smallest region of your visual experience that you can attend to cognitively - call a particular region of that size R. By hypothesis, you can't attend to any sub-regions of R. If one sub-region of R, say R1, is associated with phenomenal character green21, and another sub-region of R, say R2, is associated with phenomenal character green22, then you won't be able to attend to the instance of the green21 visual phenomenal character, because it's associated with R1, and R1 is a region of your visual experience that is too small for you to attend to.
In the above case, your cognitive abilities may be fine-grained enough to discriminate between green21 and green22 in principle, but they can't discriminate between the instances of those two kinds of phenomenal character in the particular case in question, because each one of those instances is associated with a region of your visual experience that is too small for you to cognitively attend to. Given that the R1 and R2 regions are too small for you to attend to, it seems that you can't demonstratively refer to either of them, or the shades that are associated with them.
In the first case of the range of colour phenomenal character, the problem is that your cognitive capacities are too coarse-grained to pick out the differences between the fine-grained kind of colour phenomenal character. In the second case, your cognitive abilities are fine-grained enough to discriminate in principle between the kinds of colour phenomenal character in question, but in the particular the case in question, you can't discriminate between those kinds of phenomenal character, because the kinds of phenomenal character are associated with regions that are too small for you to cognitively attend to.
If it is hard to imagine that our cognitive discriminatory abilities, as applied to our own experiences, might be this coarse-grained, then I found it helpful to imagine this process. Imagine a visual experience with a range of phenomenal character from green20 to green30. At the moment you can discriminate the different kinds of colour phenomenal character in this range. Now imagine keeping your visual experience and its visual phenomenal character fixed, and gradually degrading your cognitive abilities (perhaps to imagine this process, it's helpful to imagine a mental disease that gradually degrades cognitive abilities). It seems feasible that one can degrade the cognitive abilities enough so that one can no longer discriminate the differences in the kinds of colour phenomenal character of one's experience, even though, by hypothesis, those differences are still there.
The Relational View of Experience
Sunday, May 31, 2009
The relational view of experience holds that the phenomenal character of an experience is constituted by the objects, properties and relations in a subject's environment.
I've never been exactly sure what the term 'constitutes' means, but I think it must mean at least that the phenomenal character of an experience supervenes on the objects, properties and relations in a subject's environment.
What objects, properties and relations belong in this supervenience base? If it is all objects and properties, including properties of the subject, then the relational view becomes trivial: the phenomenal character of a subject's experience supervenes on the subject's properties and the properties of the environment around the subject. (Or if the supervenience base is just all 'physical properties of the environment around S, and of S herself', then the relational view just reduces to the claim that the mental supervenes on the physical).
If the supervenience base is supposed to be a subset of the objects and properties in the subject's environment, which subset is it?
Here are some examples that I have been considering. Relationalists often appeal to relations to the light source to explain illusions. E.g. in the case of a white wall that looks yellow because it is illuminated by yellow light, a relationalist may say that the experience is of a white wall that is being illuminated by yellow light. Suppose a subject is wearing yellow contact lenses: can those contact lenses be appealed to in the description of the experience, according to the relationalist? I.e. will a relationalist describe the experience of subject S as being of a white wall looked at by S whilst S is wearing yellow contact lenses?
Or suppose S is looking at the Muller Lyer illusion. Will a relationalist explain this experience by saying that it is an experience of two straight lines with arrows at each end, viewed by a subject with an visual system that interprets that the retinal data in such-and-such a way?
Generally, the question here is to what extent properties of the subject's brain, and their visual system, can be invoked in specifications of the experience, according to the relationalist. If any property of the subject and her environment can be invoked - i.e. if the supervenience base is unrestricted - then the relational view becomes trivial. If only a subset of properties can be invoked, e.g. if there was some way of excluding properties relating to S's brain, then it becomes harder to account for illusion, as often the source of the illusion is in the brain, as in the case of the Muller Lyer illusion.