Part of any good theory of consciousness, says David Chalmers, is a satisfactory explanation of experience. It is insufficient to let the theory stop with an explanation of how we convert sensory information into memories, or how we maintain ongoing storage of those memories, or how we retrieve those memories. Chalmers insists that we can imagine systems that do all these functions and yet do not have subjective experience. In fact, I dare say that the average desktop computer with a permanently recording video camera would be capable of these functions and yet not claim to have subjective experience. It's no surprise that Chalmers thinks this; before he started his work in philosophy, he was a computer scientist and will (like any of us) bring his past experiences with him wherever he goes. If anything, this is a good reason why people should change careers at least once.
The claim is an interesting one, however. Anyone who has watched Ghost in the Shell would find this argument familiar. After all, the massive collection and interaction of data on the (future version of the) internet is what gives rise to the Puppet Master AI. Nonetheless, there is still a step that can be made to close the gap between an explanation of how the brain manages memories and the phenomenon of experience.
We know that the brain's activities largely revolve around memories. A new sensory stimulus is received (a word, perhaps) and is translated into a meaning that is held in one part of the memory. The activity of the brain that stirs in response to the stimulus also stirs an associated memory. This simultaneous stirring of memories allows the brain to recall two memories at the same time. It is a kind of global access to all memories. Each memory, however, provides an interpretive locus for other memories. One “fact” helps define the next “fact” so to speak.
At this point, we need to depart from analytic philosophy and go to psychoanalytic philosophy. Psychoanalysis tells us that the self is formed through interpellation into a symbolic order, when a single sensory experience is connected with a meaning. It is often modelled by the entry of the child into language. From the first word onwards, every other word gains meaning until eventually an entire symbolic order is created and the person (the subject) is formed.
Let me bridge the gap here – the gap between analytic philosophy and continental philosophy. If a signifier (a word, an action, whatever) creates a memory, and it is the collection of all these signifiers and their meanings/signifieds that forms the subject, then do we not have a mechanism by which neurobiology can account for the creation of a self?
Furthermore, it is the self that permits subjective experience. The self is necessary for there to be an “I” that can have an experience. “I” cannot have an experience if there is no “I” to begin with. Therefore, the self must be formed so that experience can happen. The self is formed through the creation of a symbolic order (sounds, sights, smells... all stored as memories in the brain... all attached to meanings).
I would wager that Chalmers' objection to this will be that I have done nothing other than explain how the self is formed. On the contrary, I think that this is a reasonable space in which to think about how the brain allows experience, rather than just sensory processing “in the dark.”