Let me toss out a few more details on the cognitive/neuroscience side to seed the discussion. (Apologies for the length, hope this is helpful.)

The brain works probabilistically. That is, it is constantly identifying likely patterns, likely connections. In particular, the brain works on the basis of Bayesian probability , which entails a built-in mechanism for accepting new data and adjusting probability estimates. Bayesian probabilities are conditional, i.e. based on the following facts ..... it is X% probable that Y is true

When you look at an apple sitting on your table, the brain computes the likelihood that the patterns of light reaching your eyes is associated with an object of not-quite-spheroid shape with an indentation at the top, color varying through frequencies we associate with the word "red" (or "yellow", or "green") and so on.

It's only LIKELY that represents an apple, not certain. You could be looking at a cleverly painted picture, for instance, which gives the impression of a 3 dimensional object. HOW likely it is that those light waves correspond to an actual object is determined in part by OTHER INFORMATION: did you put an apple there recently? Have you touched the source of the reflected light and if so, was it 3 dimensional? and so forth.

Now, there are at least two levels of this sort of reasoning in the brain. One type is represented in artificial intelligence by artificial neural nets. A neural net can be trained to identify images of apples by presenting it with many instances of apple images and also instances identified as not apples. The ANN guesses which it is 'seeing', accepts feedback between what it first computes and the specified identity of the image, adjusts its guesses and repeats this many many times. (Computational complexit theory establishes how many distinct training instances are needed to result in any given degree of accuracy with any given degree of certainty.)

One interesting aspect of ANNs is that the criteria on which the trained ANN identifies images correctly are often not obvious to humans. They may or may not correspond to criteria we would use in conscious reasoning -- but they work, in many cases.

There is good neuroscience to suggest that children learn visual recognition of many objects this way. It's an unconscious process. Young puppies learn that the shiny place on the kitchen floor doesn't represent a dangerous hole in exactly the same way.

However, there is a different way of reasoning probabilistically. That way is represented by, for instance, Bayesian belief nets. Belief nets are a way of combining likelihoods/uncertainties of many different attributes of a situation. Belief nets are usually used to support conscious decisionmaking.

For instance, if I am driving along my usual route to work during commute time and all of the traffic ahead of me turns off at an exit that usually isn't busy, and if the roads are icy that day, I might conclude that there is a high probability of an accident closing the road ahead of me which I happen not to have learned about yet.

The point of all this is that our conscious awareness and reasoning are based on probability assessments we constantly make, mostly through the inputs from the part of the brain that isn't conscious. In fact, there's neuroscience research that suggests a major requirement for intelligence (humans vs. dogs, for instance) is that the 'higher' brain functions suppress a lot of the unconscious processing that goes on constantly. Children with autism find it hard to do this and as a result are overwhelmed with sights, sounds, touch sensations.

So consciousness is a lot light a spotlight, in that it is selective in its focus. The unconscious is constantly computing all sorts of things about our sensory inputs, most of which we never pay attention to. This is the price we pay for being able to learn the calculus, compose music, write blog entries.

The Hard Problem, then, can be thought of this way:
how do we get from computing the probability that our eyes are seeing a 3 dimensional object to computing the probability that we are seeing an apple, with all that 'apple' implies?

The Dennetts of the world believe that all concepts are simply arrived at statistically as a result of our experience in society. And we can see how this might be so about apples. We teach children that the snack they are eating is a slice of apple. We show them books with pictures of apples and apple trees. Over time they eat apples of varying color, texture and taste. From that they develop the ability to judge the likelihood that something they see, hold or taste would be called by others an 'apple'. In this model Grim's television signal is generated inside the TV set.

It's harder to explain how such a mechanism might work to create concepts like 'justice', 'quark' or 'mitochondria'.

Another school of thought believes that there are patterns of meaning that impose themselves on the neural makeup of the brain. These exist in some way separate from the brain's activity and outside of it. In the extreme form, this approach understands concepts to be more 'real' than the individual minds that use them.

If we take both positions as hypotheses (rather than as articles of faith) how could they be tested?

Through assessing probabilities. ;-) Pinker thinks that, based on the mechanisms already identified, it is likely that we will be able to identify the complex interactions associated with forming and using more abstract concepts. A Platonist thinks that it is most likely that concepts have true being which we participate in, by our nature as thinking creatures.

There's a third possibility. And that is that consciousness is the result of the brain's activity as a complex, adaptive system. Complex adaptive systems exhibit order and behaviors that cannot be predicted from the specific detailed behaviors of their interacting, individual parts, but which nonetheless are real characteristics of the system as a whole. In software, agent-based models attempt to capture key aspects of such complex systems for study and prediction -- very different from artificial neural nets or Bayesian belief net computation.

In this hypothesis, consciousness and identity are both dependent on the brain and also not reducible to individual chains of chemical neurotransmitters in limited brain regions.

It seems to me to be very likely that hypothesis 3 is the best fit to what we know at this time of how the brain works. But that's not the whole story, because each of us is an element in the much more complex adaptive system that is our universe.

And at the very core of that universe, if quantum mechanics is correct, lies probabilistic behavior, not determined behavior. Quantum particles cannot be specifically located in both time and space simultaneously, rather must be described in terms of probability distributions associated with their location and energy state.

So where does that leave us? If I build an agent-based model with, say, 1000 agents in it, each agent having a determined set of behaviors, the resulting system-level behaviors are very hard to predict. Now think of a system with more-than-trillions of elements interacting, each of which is probabilistic in its behavior instead. The complexity of such a system is vastly vastly greater than my little agent model.

And that's the world we live in. In fact, that's the brain we use -- the brain of a three-year-old child has about 10,000 trillion connections among neurons, many of which turn out not to be needed (just as a trained ANN often ignores certain connections over time). An adult has 'only' an estimated 1,000 to 5,000 trillion such connections.

We've learned a lot about many of those connections, how they work chemically, how in groups they do the sorts of things that dogs and birds can do. But my own guess is that we will find the higher functions are system-level behaviors.

So I could reword Grim's question as, are our minds system-level behaviors of the brain or are they behaviors associated with the larger universe as a whole?

-- Molon, on lunch break LOL

Thank you, Molon. That was informative and interesting.

One of the branches of the discussion, I think, is this: If (to use your terms) the consciousness of the mind is a associated with the larger universe as a whole, is our experience of it accidental? In other words, do we have consciousness because we were meant to have it, or because we evolved brains that happened to be able to pick up on something ambient?

It happens that there is a parallel question for those who think that consciousness is just something that arises from systemic processes. Is it an accident (and perhaps a flaw)? What should we do with it?

Molon, I don't think your question about purpose or accident is answerable, but having a brain that evolved through random processes to pick up things ambient and having a brain meant to pick up things ambient are not necessarily mutually exclusive. I recommend Finding Darwin's God by Kenneth Miller. There is also a writer for First Things who has written several articles about the compatibility of science (specifically biological evolution) and faith. Unfortunately, his name escapes me at the moment. Either way one answers your question, the answer is ultimately a matter of faith, and I don't believe it can ever be resolved scientifically.

Grim:

In other words, do we have consciousness because we were meant to have it, or because we evolved brains that happened to be able to pick up on something ambient?

Better yet, do all of us really have "consciousness", or at least an equal capacity for it?

While we tend to assume consciousness is an all-pervasive experience, so long as one is not asleep or dead, the fact is that we seem to able to intelligently interact with familiar objects and environments without being truly conscious of them, so that our consciousness is actually shut off for a large part of the day. As a child I used to sleepwalk, and I could perform complicated actions in my sleep which could not have been assisted in any way by consciousness.

It may be a mistake to assume that consciousness is an inherent property of the brain, any more than the things we see, smell, and taste are part of our brains.

Good point, Glen.

Thanks, Grim. I'm busy at the moment - but still interested.