First posted 24 Oct 2005. I used to be a 'the mind reduces to the brain' guy, whereas now, I'm more of a 'the brain (and everything else) reduces to the mind' guy. The book that this post refers to was my introduction to phenomenology - the study of experience. Thinking of the world in terms of experience has ended up being the biggest shift in how I see things. That shift has taken more than 10 years, from the time of the post below, though.
The Internal World
All we ever have direct access to is our mental states. Those states could be the result of our interaction, via our sense, with the external world; they could be part of a dream; they could come from wires and electrodes stimulating our brains, which sit in vats of syrupy liquid in a world run by computers. Both philosophy and science fiction have wrestled with questions of whether those mental states correspond to or represent an external world. The question is no less interesting (and no closer to being answered) than it has ever been.
But those mental states, that internal world itself, is rich ground for investigation as well. Phenomenology is the study of experience, and a recent book I read was my introduction to it. The book, an atypical mix of novel, philosophical introduction and scientific primer, is called Radiant Cool, and here is what I have taken from it:
I've always taken objectivity to mean the view 'from outside' and subjectivity to be the view 'from inside' a person. But, using a specific visual perception example, even in our internal 'views' we are aware of the difference between some relatively stable mental object, say the image of a chair, and any number of particular angles (or points of view) from which that chair can be seen.
Think about it. As we walk around a chair, looking at it, our mental states do not feel like we're seeing a constantly changing stream of things, but rather that there is a single thing there, which we are viewing from a constantly changing angle. All of this is equally true if we simply imagine walking around a chair, which shows that the physical presence of a material chair isn't necessary. The point, which had never occurred to me before, is that even within our subjective experiences, there is a stability and assumed objectivity of entities rather than just a constant flux. This is despite the fact that the sensory input we receive is literally nothing but an ever-flowing flux. So... there are non-sensory elements to our experiences, even to those experiences that - on the face of it - seem to be purely sensory.
Affordance / superposition
When we think about it a bit more, it becomes obvious that the in-built assumption or experience of objectivity / permanence is not the only non-sensory element to our perceptual consciousness. Everything we see presents itself not just in its bare form but always with a collection of non-sensory baggage - inseparable emotions, associations, anticipations, etc. This property of immediate perceptual objects to 'carry' these other properties has been called affordance or superposition.
The possible bits of baggage that can accompany perception are virtually innumerable. We can model a theoretical mental-state-space of consciousness, with each possibility constituting a binary (either experienced or not) dimension in that space. If each conscious moment were represented by a time 'slice' of this space, then putting them all back to front temporally would create a squiggly line tracking through state-space moment by moment.
In fact, the most important of these non-sensory elements - perhaps even THE defining one for explaining consciousness - is temporality itself. Our mental worlds are built upon it. And I'm not speaking here of memory and anticipation - casting back or forward from the present moment. Temporality is embedded even WITHIN each moment. Each immediate experience can be thought of as having three components: retention, presence and protention. The middle one is the instantaneous snapshot of the flux of sensory input. Retention is the context in which that snapshot sits, grounded in what the previous instant contained. Protention is the immediate projection or extrapolation of what the retained and present elements imply for the next moment.
A simple but insightful example brought this home to me. Think of a musical melody. In any instant, the only SENSORY element of the experience is a single note, yet your experience of that note is inseparable from the pitch, volume, duration and other qualities of the notes immediately before and after it.
Lloyd used diagrams and explanations of simple, synchronous neural networks to show how temporality can be embedded within a cognitive system. If some of the nodes (neurons) in the system encode the previous values of the neurons that receive and 'process' the sensory inputs, and if those time-lagged neurons feed back into the sensory processing ones, the feedback loop ensures that each instant recursively captures all previous instants. Then the 'output' from the sensory-processing neurons constitutes the system's best guess for what the next moment (immediate future) will be. A number of examples showed that this model world works very well at keeping time and tracking the future. But do brains work this way?
Lloyd then moved from toy brains to real ones, using the tools currently at science's disposal to look at the evolution of brain states over time while subjects performed different mental tasks. The results suggested that we can track these states quite well, and Lloyd showed how further empirical testing can pull much of the discussion from the world of the philosopher to that of the scientist. But it was also obvious that further technological advance was necessary to make findings more exact and more robust.
Mental state = Brain state?
Where is this all going? Well, Lloyd and others think that the ideal objective is a deepening of our understanding of both brain states and mental states, an understanding that may well include a one-for-one matching between the two that reduces the mental to the physical, firmly placing the mental world within the natural world. I say 'ideal' objective because we may never be able to reach a neuron-by-neuron level mapping of brain states. If that degree of specificity is required for perfect mental state matching, then we may be stuck with some degree of averaging and clumping of brain and experience states. Still, this could provide the empirical basis for finally dumping Cartesian dualism, the Cartesian theatre, and all the misunderstanding they create.