Materialistic intelligence

On this week's The Skeptic's Guide to the Universe, a listener wrote in saying that her debate opponents, essentially, argue for the irreducible complexity of human-level intelligence--that some miraculous force is required to explain human intelligence; i.e., the old materialism/dualism debate. The Skeptics did a great job in pointing out the logical errors in the arguments the questioner was encountering. However, it got me thinking about what arguments she can use for her position--i.e., that there can be a scientific basis for intelligence materialistically arising simply from the interaction of our neurons, that we are "more than the sum of our parts". She might find some discussion of the phenomenon of emergent behavior helpful. This concept results from the observation that many systems exist where the high-level behavior is not immediately apparent from the properties of the low-level components. That is, it is not inherently apparent from the properties of quarks or electrons that the aerodynamics of a plane's wing or the display of high-level information on my computer screen should be implicit, yet these emergent properties are the sum of trillions of base-level interactions.

The predominant example is probably the social insects: No study of the abilities of individual termites would ever suggest the complex collective accomplishments they are capable of when taken by the thousands or millions. No individual possesses significant intelligence; there are no orders from the queen; there is no psychic hive mind that makes decisions. But when each termite makes individual decisions based on inherent genetic instructions and environmental perceptions including chemical signals, complex behaviors emerge, such as the optimization of path-length to food resources or orientation of mounds relative to sun and airflow. Interestingly, in analogy to the apparent "obvious" gap between humans and animals, there is an "obvious" gap between collective insects such as termites and individual insects with no socialization--there don't seem to be a lot of semi-social insects. Again analogous to human/animal cognition, I suspect this gap is in a large sense perceived or imaginary, caused by factors such as: 1) The enormous evolutionary success of the two endpoints, with the semi-social middle not being terribly advantageous. 2) My own, and, I suspect, the general public's, ignorance of these middle groups. There are probably fossilized pre-social insects, insects that only form loose colonies, insects such as locusts with intermittent emergent behavior that might not strictly be called "socialization." But I'm much more familiar with social or asocial types, which are understandably more often covered by science media. This somewhat overlaps with: 3) Our human nature of categorizing things distinctly, leaving little room for grey-area exceptions (and, in the case of intelligence, our hubris).

It doesn't seem to be at all implausible that the firing of 100 billion neurons with an average of 7000 interconnections each would result in emergent properties not dissimilar in some respects to a learning or chess-playing computer--especially since this seems to be what is observationally suggested. There is still plenty of room for speculation about the rules of how this behavior emerges, what the specific properties and processes are, and the philosophical implications thereof, hell, it's possible some mechanism other than emergent properties may ultimately be determined to be responsible for human cognition. But regardless, there are certainly plausible scientific explanations which are the basis of current research, no metaphysics required, which shoot down suggestions of miraculous requirements as well as the evolution of eyes or flagellum shoot down irreducible complexity.

Some resources that discuss emergent behavior, some directly, some tangentially:

Douglas Hofstadter's Gödel, Escher, Bach
Much of Richard Dawkins, especially perhaps The Selfish Gene
I believe E. O. Wilson covers some of this, perhaps in The Insect Societies
Richard Feynman played around with ants and discusses, among other things, their path optimization in Surely You're Joking, Mr. Feynman!
(And here's PZ Myers discussing ant trails that stumped Feynman, and the Nature paper solving the mystery.)

Eliezer Yudkowsky discusses emergence in posts such as "Reductionism" in his excellent blog series on taking the mysterious out of quantum mechanics, which is very enjoyable for many other reasons, as well.

Finally, I enjoyed Jeff Hawkins' On Intelligence, which attempts to materialistically describe the emergence of intelligence from the brain. I'm not a neuroscientist (neither is the author), but, while his specific models and mechanisms may be wrong, his gestalt concepts of emergent mechanistic intelligence struck me as pretty plausible (rather than, say, Roger Penrose's The Emperor's New Mind, which implies that quantum effects are necessary to explain intelligence, which seems unlikely).