Jonathan Edwards
Senior Member (Voting Rights)
I'm not excited about philosophical theories about the metaphysics of consciousness and experience. I think we will learn more about it by studying simple forms of awareness like how bees know how to communicate and locate food sources or the AI of computers. By building up that knowledge, I suspect scientists will come to a better understanding of what consciousness is. Don't think that philosophical speculations about the substance of consciousness will be of much use to that.
I think I might have held this view in my forties. I cannot really remember. You are expressing the view of people like Christof Koch, at least until recently, and the Patricia and Paul Churchland, I think. The problem is that neuroscientists are completely stuck precisely because they have not engaged with these fundamental issues. Michael Hausser, who is perhaps the most eminent neuroscientist in London, recently co-authored a review in Nature saying how stuck everyone was and that people from a wider range of disciplines needed to work together.
The central problem is that you cannot work out what awareness is going to be until you have a hypothesis of what physical dynamic unit is actually aware. It is not a person or a bee that is aware - that is just lay parlance. Awareness has to be a property of whatever inside a central nervous system gets the information it is aware of. That is just a brute causal imperative. Yet neuroscientists dance around refusing to say what it is they think gets the information. Giulio Tononi and colleagues have recently tried to get around this by proposing 'systems' that have 'integrated information' (Integrated Information Theory, IIT). However, the Journal of Consciousness Studies has just put out a volume of essays on IIT, several of which point out that the theory's causal structural basis is invalid. I am preparing my own paper with a student where we try to pin down precisely where the error in causal dynamics is made.
As William James pointed out 'Metaphysics means nothing but an unusually obstinate effort to think clearly.' The metaphysical issues around consciousness are issues around how the physics questions can even make any sense. As indicated, the key issue is what 'bit' is going to be aware? How are you going to address that without having some idea of what candidates would even qualify?
Good or bad is just something monkey brains developed to further their social interactions in evolutionary terms.
I don't think that is right. Monkeys' good and monkeys' bad are just something monkey DNA encodes in the developing brain to further social interactions. Tigers' good and bad may be very different. But the sense of good or bad is something quite different - it is the sense that we allocate to whatever our DNA tells our brains to allocate it to. Assuming that we believe that physical laws cover all the causal dynamic interaction in the universe we then have to discover what interactions give rise to this sense, irrespective of the fact that it is triggered by different things in monkeys' brains and tigers' brains.
So we have the question: what dynamic unit X in a brain will undergo a physical interaction which to X feels 'good' or 'bad'. For a crossbill that event may be triggered by seeing a spruce cone, for a parrot crossbill it will be a pine cone and for a two barred crossbill it will be a larch cone, but that is another issue.
Put differently, the metaphysics is not extraneous to the physical arguments, it is the base they stand on. The problem with neurobiology at present is not that it is too 'physical' but that it is not physical enough. It does not respect the basic causal requirements that you need for a physical theory to be true or false.