6 JULY 2006
hen
scientists describe what they know about consciousness, their knowledge
of the mysteries of the mind often seems rather meager and unsatisfactory
-- which would not have come as a surprise to Lord Kelvin.
The 19th century British physicist is famous for (among other things) declaring knowledge without numbers to barely be knowledge at all.
"When you can measure what you are speaking about, and express it in numbers,
you know something about it," Kelvin declared "But 
when
you cannot measure it, when you cannot express it in numbers, your knowledge
is of a meager and unsatisfactory kind: it may be the beginning of knowledge,
but you have scarcely. . . advanced to the state of science."
Because it is hard to measure consciousness in any meaningful way, it has not been a very fruitful field of scientific study. For many years, it was regarded by many as beyond the scope of science altogether. But the pioneers of the mind's geography have persisted, and a few attempts have been made to devise mathematical measures of mentality.
The latest comes in a new paper in the Proceedings of the National Academy of Sciences by Anil Seth, Eugene Izhikevich, George Reeke and Gerald Edelman of The Neurosciences Institute in San Diego. In that paper Seth and colleagues expand Edelman's theory of consciousness to include a new quantitative measure of consciousness in action.
"In action" is an important part of the new idea, for as the authors point out, consciousness is "a process, not a thing or capacity."
Conscious systems are always embodied in bodies that are embedded in environments, Seth and colleagues note. "Conscious scenes arise ultimately from transactions between organisms and environments, and these transactions are fundamentally processes."
An
essential feature of these transactions is their capacity to both make
distinctions (that is, tune in to particular features of the environment),
and to put the big picture together (that is, integrate the distinct parts
into a coherent whole). With all the wildly complex sensory stimulation
in the world to deal with, picking out the pieces and putting them all
together is a tall order. Which is probably why it took evolution so long
to produce brains complex enough to do it.
The question is, how complex does a conscious brain have to be, and how should you measure that complexity?
In the Edelman theory, the essence of consciousness emerges from what he calls "neuronal group selection." Groups of the brain's nerve cells (or neurons) cooperate to carry out specific mental tasks, like identifying familiar faces or tracking moving objects. These neuronal groups then have to communicate with each other so any one part of the brain knows what another is doing (or seeing, or thinking).
To that end, the brain's architecture has evolved to include an important signaling relay station called the thalamus. It lies at the heart of the system coordinating the higher-order mental process going on in the brain's outer layer, the cortex. Somehow this arrangement allows the brain to pick individual objects out of a scene yet still be aware of the scene in its entirety, with all the various sensory modes (sight, sound, touch, etc.) combined into a unified representation.
Seth and collaborators propose a number that quantifies the cause-and-effect relationships that underlie the consciousness process. This number, called "causal density," is "a measure of the fraction of interactions among neuronal elements that are causally significant" in the processes picking out the pieces and fitting them together
Neural cause and effect is, of course, a complicated game. Sensory input can cause a neuron to fire an electrical impulse (or fire impulses more rapidly), and the effect of those pulses can be to induce even more neurons to fire (or to slow their firing rates down). All this impulse firing encodes information that flows through the brain, telling different parts what other parts are up to, ultimately painting the coherent consciousness picture.
In a system with low causal density, a lot of impulse firing might be going on, but not in a way that produces a well-ordered, useful big picture of the world. (Many TV shows are like that.) On the other hand, high causal density indicates that a lot of distinct activity (dealing with the pieces) is also globally coordinated in a way that allows different parts of the brain to predict what other parts will be doing next. And that is basically what consciousness is all about. Above some threshold level of causal density, consciousness emerges.
Having such a measure of consciousness could have practical uses -- for anesthesiologists, say, or researchers studying the possibility of consciousness in other species. Actually calculating this number, however, poses some serious practical hurdles because of the brain's vast complexity. But Seth and colleagues suggest that statistical methods might be devised to make at least some good approximations possible.
Even so, the scientists realize that no one number, however useful, can actually represent all the complexity of consciousness -- no more, they say, than the gross domestic product sums up all the details of the U.S. economy. But having a number to work with will go a long way toward making consciousness research a respectable part of science -- and it might even help in testing whether the theory behind it is actually right.
E-mail: tsiegfried@nasw.org
