One of the great mysteries of science is the origin and the nature of consciousness. In The Emperor's New Mind, Roger Penrose suggested that consciousness was "non-computable", that is, so complex that it could not be adequately described by any executable algorithm. However, as our understanding of the brain and its behavior has progressed, computer scientists have become more and more adept at modeling the brain (what Ray Kurzweil has called "reverse engineering" the brain) and surprises are now becoming expectable.
Two months ago I wrote a post, based on a Scientific America article, summarizing some of the data on the Neurobiology that is involved in the development of the sense of the Self. I noted the existence of self-reflective areas in the brain:
Debra A. Gusnard of Washington University describes one such area [involved in integration and synthesis-SW], the medial prefrontal cortex, an area located in the cleft between the hemispheres of the brain , directly behind the eyes (where most people imagine their "self" to reside.) The medial prefrontal cortex may act as an integrating and synthesizing structure. Of note, this area of the brain has a high concentration of neurons called spindle cells, which are thought to be very involved in information processing. [Recent work suggesting the computational ability of single neurons may be more robust than we have recognized will need to be factored into the mix as time goes by.]
It is worth noting that there is a significant area of overlap between such a hypothesized brain structure and the psychoanalytic concepts of "object constancy", "ego structures" and "ego boundaries"; these concepts are relevant to determining the boundaries between self and not-self (or self and object.)
Furthermore, there appear to be specific areas of the brain that respond to our experiences and our traits and determines how they fit into our sense of ourselves. For example, a novel experience triggers areas of the brain that are self reflective, that is, they think about our self involved in the event. [Emphasis not in the original-SW]
I also noted this:
The discovery of "mirror neurons" is another piece of the puzzle. We have cells in our brains that react to seeing others having experiences as if they were happening to us.
Now comes news that seems like a major step forward in modeling a theory of consciousness:
Robot Demonstrates Self Awareness
Dec. 21, 2005— A new robot can recognize the difference between a mirror image of itself and another robot that looks just like it.
The researchers, a team led by Junichi Takeno at Meiji University in Japan, are using an evolutionary approach to the formation of neural networks (analogous to "mirror neurons") which can learn to differentiate between itself and another robot, even if the other robot is identical to it. At the moment the robot is ~70% accurate in determining whether it is looking at itself in a mirror or another robot. This is an impressive feat which is hugely significant.
One of the most important steps necessary for consciousness to exist is the ability to differentiate self and other. It is the earliest differentiation an infant has to make. One of the tests of individuation is the mirror test; can an animal (besides man) differentiate its own mirror image from the image of another animal. Very few animals have unequivocally passed the mirror test; it is a necessary precondition for the ability to self-reflect, to think about oneself and one's behavior and how it affects an independent other person. Without the ability to self-reflect there can be no sense of agency, ie the sense that one can effect one's environment.
The next, and more crucial question, is at what point does such differentiation become consciousness. My puppy (the one who keeps me humble) is clearly conscious in some way. He learns and has the ability to effect changes in his environment. (If he stands at the door and yips, he knows I will jump up and take him outside; he reinforces such behavior in me both positively, with "rewards" outside, and negatively, with "punishments" inside; the difference between the same object as reward and punishment is not completely clear to him yet, but he is now "getting it" 80% of the time.) However, in order for us to consider the puppy to have true consciousness, he would have to be able to reflect on his behavior, and use such self-reflection to plan for future events; at this point he gives no indication of doing anything more than reacting to his environment, albeit in ways that are usually designed (by nature, evolution?) to please his family.
It is that ability to self-reflect that is the sine qua non of consciousness; recognizing the difference between oneself and another is a necessary, though not sufficient, ability in order to attain consciousness. It may well turn out that Penrose was correct, that we cannot construct an object (computer, robot, artificial brain) that has the attribute we call consciousness; however, we may well be able to construct an object which has all the necessary prerequisites to allow for consciousness to emerge on its own.
Recent Comments