Bias, learning to walk at the edge of Chaos

This is a guest post by Meika, for which, many thanks.

Cambridge-based researchers recently published a study providing experimental data which supports the idea that the human brain lives “on the edge of chaos”.
Its press release ends:

According to [co-author] Kitzbichler, this new evidence is only a starting point. “A natural next question we plan to address in future research will be: How do measures of critical dynamics relate to cognitive performance or neuropsychiatric disorders and their treatments?”

Well, taking that ‘cognitive performance’ a bit more specifically to include learning, and learning to walk in particular, the following story leads in an interesting direction for us fourculture fans.
Five years ago… a Japanese based research team reported that their eight-legged software robots, controlled by chaotic based learning algorithms, only managed to walk, rather than chaotically flailing their legs about, after they put a ‘weight’ on the robot. As reported in New Scientist:

“Many findings point to the presence of chaotic patterns in general in the human brain,” says Max Lungarella, who researches artificial intelligence at the University of Tokyo. But Kuniyoshi and Suzuki’s approach is still unconventional, he says. “It diverges radically from the traditional way of thinking about intelligence.”

The key thing to pick up here is that the weight on the robot can be isomorphically mapped to any sort of goal: a need, a want, a bias. This weight gives purpose and direction, without which, evolution would never have had the chance to select what became animal life.

In the introduction to Thought Styles, Mary Douglas mentions a letter by Lucien Levy-Bruhl to E. E. Evans-Pritchard in which there is an “afferent bias” that needs explaining in the social sciences.

That book goes on to explore thought styles as a group of four cultures defined by their bias oppositionally, by their perception of nature as: fragile, robust, robust within limits, and totally unpredictable.
One can see that just as a weight gives direction to a flailing software robotic spider above, it seems likely, that any bias or drive can give direction to any learning process based in chaotic algorithms. Evolution will work on those biases, and leave the learning on the spot for the individual. A flailing spider will not survive, a spider that can chase flies, trap flies, or flee birds will have a much better chance at survival. Hunger, the want to fiddle with things, and fear all are powerful motivations that guide our learning.
(Hardwired complex behaviour can only be selected for after this.)
The biases of the four cultures described by Mary Douglas  I suspect work similarly to the basic drives of our animal life. And like religious fervour and even rationality, these cultural biases have arisen in evolutionary processes, which we are only beginning to dimly perceive in our day to day cultural lives. They allow survival by giving impetus to learning by a biological machine that operates at the edge of chaos.
The other interesting thing about the robotic work in Japan is that the weight, the bias is not an impediment, but an aid to learning, if not a necessary precondition for chaotic algorithm based processes in the first place. That is, without our biases we would, perhaps, not be here at all.


L. Levy-Bruhl (1952)  “A letter to EE Evans-Pritchard,” British Journal of Sociology III, no. 2: 117-23.; cited in Mary Douglas (1996) Thought Styles: Critical Essays on Good Taste (London: SAGE publications, 1996).


7 thoughts on “Bias, learning to walk at the edge of Chaos

  1. Meika (or anyone else!)– How does chaos complicate or enrich evolutionary theory in biology? How does the nonlinearity that chaos features interact with mutation/drift/natural selection? Is there a canonical text (or at least something authoritative & comprehensive) on this?

  2. @dk
    “Is there a canonical text (or at least something authoritative & comprehensive) on this?”

    No idea, must be in a state of flux. I’m too deep into making bronze sculptures to keep up to date with chaos and complexity literature.

    However, fractalising the fourcultures (the iteration is in the choices/rejections of/by individuals) would be a very interesting and I suspect a rewarding avenue of research and writing.

    1. Meika, I love these – particularly Spacehead. Now I must look out my old copy of Chariots of the Gods….

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s