The Machine in the Ghost

--Building a brain and a theory of brain

A Theory of Brain

Why do we need a theory?

The idea of a theory of brain may seem odd, or even somehow wrong to many of us. Is the brain something we can even have a theory about? No one has tried to formulate a theory of liver or kidney or ankle, after all, and it might seem a rather silly thing to attempt. Nonetheless, we all know the brain is much more complex than any other organ or system in the bodies of higher animals; it supports behavior that no other structure nature has produced on earth can match. The brain is most special in being the seat of our self and our mind; it may even serve as a channel to some superphysical aspect of our being. Boggling amounts of resources are being devoted to making machines that do various things that the brain does. The brain is therefore a common interest of science, philosophy, and engineering. It certainly deserves a theory.

Philsophers have been pondering abstract theories of mind for millenia. Brain scientists and psychologists generally think of mind, or what the brain does, more in terms of concrete behavior. They are working bottom-up, but remain very close to the bottom. Patricia Churchland describes the attitude of neuroscientists some thirty years ago as follows.

"Theorizing about brain functions is often considered slightly disreputable and anyhow a waste of time--perhaps even 'philosophical'. ...asked about the role of theories in the discipline will likely answer with one or all of the following: (1) The time for theories has not yet arrived, since not enough is known about structural detail.' (2) What is available by way of theory is too abstract, is untestable, and is anyhow irrelevant to experimental neuroscience.' (3) You cannot get a grant for that sort of monkey business.' " (P.S. Churchland, Neurophilosophy, 1989)

Researchers that fit Churchland's description may have been typical at that time, but at least some neuroresearchers recognized the need for theory fairly early on.

"What is conspicuously lacking is a broad framework of ideas within which to interpret these different approaches." (Francis Crick on neuroscience and studies of the brain, Scientific American, 1979)

Thirteen years later, but still almost 20 years before now, we see no progress in defining a theory and even some backsliding of expectations by the same author in the same publication.

"There are many possible approaches to the problem of consciousness. Some psychologists feel than any satisfactory theory should try to explain as many aspects of consciousness as possible, including emotion, imagination, dreams, mystical experiences and so on. Although such an all-embracing theory will be necessary in the long run, we thought it wiser to begin with the particular aspect of consciousness that is likely to yield most easily. What this aspect may be is a matter of personal judgement." (Crick and Koch: "The Problem of Consciousness", Scientific American, 267-3, 1992)

We see here a sort of giving up, a putting off of the quest for a broad theory to an indefinite `someday' in favor of a more traditional isolated analytical approach. Consciousness is but one aspect of mind, and we are not even sure how important it is, overall. Yet even this one aspect seems to be so complex that the leading researchers choose some even smaller aspect of that one aspect, visual consciousness in the case of Crick and Koch, which itself may not even be critical. This approach will surely produce valuable results about the restricted topics, but we are left with the original problem of how to integrate the results with other narrow results to explain overall brain function.

Even quite recently, Marvin Minsky points out a weakness of narrow, low-level neuroscientific research that is not guided by theory.

"When you talk to neuroscientists, they seem so unsophisticated; they major in biology and know about potassium and calcium channels, but they don't have sophisticated psychological ideas. Neuroscientists should be asking: What phenomenon should I try to explain? Can I make a theory of it? Then, can I design an experiment to see if one of those theories is better than the others? If you don't have two theories, then you can't do an experiment. And they usually don't even have one." (Marvin Minsky, interview by Discover magazine, http://discovermagazine.com/2007/jan/interview-minsky, 2007)

Still, the theories described by Chuchland, Crick, Minsky, Dennett and so many others have so far either failed to ground themselves to a level of reality that is low enough to inspire the bottom-up researchers or have given us narrowly-scoped theories that do not guide us in integrating the results of low-level research into a coherent overall picture of brain function. There is not even a foggy consensus of what a theory of brain should look like.

What should a theory of brain look like?

A theory of brain must establish a context for individual aspects of mind such as perception, consciousness, learning, social behavior, language use, aesthetics and creativity. Individual aspects of brain function can only be truly understood in such a context. The theory must allow how each aspect fits into overall brain function, brain development, brain evolution, and the behavior of an organism in its environment. It must also explain how brain function supports each aspect in its overall operation.

The widespread efforts to understand consciousness in recent decades presents a good example of the problems that working without an over-arching theory can produce. So much work and thinking is being done to explain consciousness from so many different perspectives and levels, yet no one has even come up with a clear and generally acceptable statement of what consciousness is.

The biggest challenge for a theory of brain is the huge canyon that separates the science of the brain on one side and the philosophy of mind, psychology, linguistics and other theory-heavy investigations on the other. You hear a lot on the philosophy side about theories of mind, but on the science and engineering side, tens of thousands of researchers are working without even one guiding theory; the most ambitious of efforts have been toward achieving some foggy notion of artificial intelligence, and even that is already a project that has been largely abandoned in favor of achieving relatively small aspects of whatever a group construes intelligence to be. Work on either side of the gap has proceeded with very little insight from the other. A theory of brain will seek to bridge that canyon and allow the data-driven science and the theory-driven philosophy to move toward each other and hopefully meet somewhere in the middle, thereby giving direction to the bottom-up scientific research and directing the abstract concepts of the top-down approaches toward a needed grounding in experimental science.

Our theory of brain will be a rather small set of simple notions that together can do two things for us. One is to explain existing scientific observations and predict behavior. The other is to organize the burgeoning but scattered data being generated on the brain, the nervous system and animal behavior into a coherent system that can integrate or at least coordinate research at various levels and in different fields by supplying a framework to guide neuroscience, behavioral sciences and philosophy toward some common point of synthesis that we hope exists somewhere in the large gap that currently separates them.

The linchpin for our framework must be a target that is reachable by both the top-down efforts and the bottom up efforts. For that middle-ground target, I propose the `concept', a notion that is fundamental in philosophy of mind, linguistics, and psychology on one hand, and a notion that must have a physical correlate in a functional brain model or an emulation of the higher mental behaviors that are based on brain function, particularly the use of language.

|| On to Roughing Out a Theory of Brain