Philosophy, Film, Politics, Etc.

Sunday, July 11, 2010

Dreams and Consciousness

Richard Brown, of Philosophy Sucks!, has posted some interesting ideas about dreams and consciousness, and whether or not we have (or can have) evidence of consciousness in dreams. I posted the following in the comments section:

The question of consciousness in dreams is very interesting, and I like the idea of testing it by determining whether or not people can report dream experiences while dreaming. But I think the only way this could be done is if the dreaming subject were reporting the experiences to somebody else--that is, there must be interaction between the dreaming subject and another subject within the dream. The experimenter would have to interact with a person as a part of their dream, like in Dreamscape (1984). We have no evidence that this sort of interaction is even possible, so it follows that we have no evidence that consciousness of this sort is present in dreams.

But, as I suggested in my last sentence, perhaps we can talk about dream consciousness of another sort--perhaps what David Chalmers would call "phenomenal consciousness"--that is, a dreaming subject can have experiential states with qualitative character, but lack the psychological abilities we associate with self-awareness and action. This would have to be an unreportable sort of experience, but which left reportable traces in our memories. But I don't think we could test for that. We'd have to establish a causal connection between neural activity and phenomenal experience first and then look for it in dreaming subjects. But even if we found the right brain activity occurring in dreaming subjects, we couldn't be sure that the activity was sufficient for phenomenal experience. And if we didn't fine the right activity, we still couldn't be sure that some activity qualified as sufficient for a phenomenal experience. So it seems hopeless.

What this suggests, I think, is that the very notion of a phenomenal consciousness independent of psychological consciousness is inherently untestable.