Re: What are the neural correlates of consciousness?

Jon-

     Thanks for asking the easy questions! It's a really interesting one, though. I did a bit of reading to get the ideas flowing, mostly the papers of David Chalmers at the University of Arizona. He is a philosopher as opposed to an empiricist, but he does a good job of citing and presenting the work of some seminal experiments. Christof Koch and Francis Crick have also been thinking a good deal about this problem.

     I'll first describe some of what Chalmers and others seem to think (it's a reasonable starting point), what I think about it, and then some random, completely untested ideas of my own. I hope that's what you're looking for!

     Chalmers writes quite a bit about the search for the Neural Correlate of Consciousness (NCC), and how to possibly go about testing it (see this paper). He debates the pros and cons of using illusion, direct stimulation, and lesion studies to determine what and where a NCC is. Each has its limitations concerning the question of where it could be.

     He offers the very intriguing case presented by Milner and Goodale.

For example, a subject with visual form agnosia (e.g. Milner and Goodale's patient D.F.) cannot consciously identify a vertical slot, but can "post" an envelope through it without problem; while subjects with optic ataxia (e.g. those with Balint's (1909) syndrome) can identify an object but cannot act appropriately toward it. The dissociations here appear to go along with damage to the ventral and dorsal pathways respectively.

     This idea left me wanting more, however. It seems to me that a true NCC isn't going to be situated in the major processing area of one modality. Re-reading Chalmers text, you see that he uses the phrase "neural correlates of the contents of visual consciousness." "Contents" is the key word. He (and others) divides up NCC to certain modalities to say that we are 'conscious of' a certain visual or auditory or tactile percept, but not an integrated feeling of 'consciousness.' So he states that there may be many NCCs.

     He does qualify his terms and says that to define the NCC truly "depends on whether we are talking about neural correlates of arbitrary conscious states, or about the more constrained case of neural correlates of conscious contents." However, I'm of the opinion that breaking up the percepts of individual modalities isn't so helpful in terms of describing the true NCC. While our perception is critical to our consciousness, it does not define our consciousness. We can be conscious of our memories and thoughts - states which are, at the time of entering one's consciousness, independent of a sensory percept. Furthermore, while Milner's patient D.F. is not 'conscious of' the envelope slit, in this example she is undoubtedly 'conscious.' She is still able to engage in self-reflection, awareness of most other surroundings, etc. The elements that he describes are clearly critical to a complete conscious experience (given the sensory parameters of the standard human), but seem more like inputs into rather than the 'seat' of consciousness.

     I believe the issue comes down to finding good operational definitions. The terms/phrases 'perception,' 'consciousness,' and 'being conscious of' are somewhat loaded and ill-defined in many situations. I think when Chalmers describes someone as 'being conscious of' something, he is really describing whether they are 'perceiving' something. The consciousness or 'conscious experience' of the individual is still not being described. I'll describe what I think of consciousness below. But, in the end, I can't say that I have the end-all, be-all definitions at hand (I spend most of my time looking at neurons in a dish!), but it does strike me that this point is a major source of confusion.

     When I think of consciousness, I do think of what Chalmers might define as "the neural correlates of arbitrary conscious states." That is the sort of self-awareness or self-reflection that (at least in my opinion) is the hallmark of true consciousness. If I could describe a conscious state, I would say I am 'attentive' to something in working (or short term) memory, and it is available for 'cognitive analysis.' It is not instant reflection on a perception - there is a time delay. For instance, when I am speaking (particularly in a presentation or lecture), I find myself conscious of the words I say after I say them - and often it'll be a strange contraction of two ideas at once. I am conscious of a percept (or internal thought, IT) a very short, but significant time after it has occurred. Thus, it seems to me, that I have shifted my attention to an internal representation of a percept or idea that is in my working memory. This could involve many modalities at once. Chalmers description of a being 'conscious of' something involves bringing the fully parsed percept (from IT) into working memory.
     Perhaps one could look for this by having subjects think about what they had for dinner, to think about some visual percept being presented, about a auditory percept (etc), and look for areas which are co-active in all cases. One would properly have to control for 'general attention' or when one's 'mind wanders', I suppose. It is quite likely that this 'attention' center will be interacting with many of the functional subregions of the brain (visual, auditory, etc), and will be hard to pin down. Perhaps attention itself is the seat of consciousness. However, I don't have a good operational definition of what truly constitutes attention (in terms of neural activity), so that might not be right.

     I've thought a reasonable amount about other ways people have been thinking of testing for the presence and location of consciousness - particularly those studying sleep for answers. This initially seems like a fruitful line of investigation, but I think it has a major experimental pitfall in that it too closely ties 'non- responsiveness' with 'unconsciousness.' Perhaps REM sleep, as it is closest to the awake state, can lead to some insights, but it still has the confound of non-responsiveness. While I can't say I know what it would be like, it seems plausible that one could observe a situation in which a person is not conscious, but is still responsive. (think of the lecturing example I gave above - you can clearly communicate ideas without being completely aware of what's pouring out of your mouth). And one could be non-responsive, but still conscious - particularly if someone has major brainstem/spinal cord injury (check out this fascinating book - it's a quick read and a thought-provoking one, at that).
     You can probably see that one of the hardest problems about consciousness is not only defining it, or knowing where to look for it, but also how to assay it. We tend to equate responsiveness with consciousness, but I do believe that they could be distinct. But how we could distinguish former in the absence of the latter would take some 'looking under the hood,' so to speak. But we'd have to know what to look for! It would become somewhat tautological. Someone will have to provide a quantum leap in our way of thinking about this (which I'm clearly incapable of!) for us to really move forward. Perhaps there are states which could induce this dissociation, like having people be extremely sleep deprived (I'm not saying this concept is proven at all - this is just a wacky idea based on how I've felt after not sleeping for a long time).

     I must say, for the record, that I most certainly respect the ideas of those who have given this a lot more thought that I have (i.e. Chalmers and Koch). Perhaps I ought to drop them a line to discuss this sort of stuff, and see what they think. I'm just trying to offer my analysis of what they've presented. That's easier to do than coming up with the ideas in the first place!

     I hope this spiel is what along the lines of what you sought. I think we as a scientific community still have a lot of work to do in describing and defining the basic phenomenon. I'd be happy to continue conversation - my e-mail is below!

-Alex G
cgoddard@fas.harvard.edu