Complexity and Disagreement

I’ve just flown over to the US for a few days, and I grabbed a copy of Roger Lewin’s book on Complexity for the flight over. I enjoyed Lewin’s writing style which makes a sometimes bewildering subject accessible.

It has also thrown me into a fairly heady state and the following somewhat rambling post is the result.

I especially liked the way Lewin filled in the characters of the many scientists he interviewed. I was struck by how passionately and sometimes bluntly they disagreed with and sometimes ridiculed each other. In their reported off-hand remarks, I felt they often showed a remarkable talent for leaping to big conclusions about each other’s personalities from relatively small actual factual information. They seemed to enjoy reaching what seemed definitive explanations about each other’s motivations based on what seemed quite flimsy evidence. I was struck by how this contrasted with the rigour with which they pursued their own scientific investigations.

I suppose this sounds like I am finding fault here but that’s not my intention. Indeed, I find myself becoming increasingly appreciative of how diverse are the meanings that we humans make of the data we are presented with, even those of us with reputations for rigorous thought. I was reminded of the slew of comments to Rogers Cadenhead’s post about his dispute with Dave Winer. (I referred to it yesterday) – such a range of different meanings being made out of a limited amount of data.

I’m probably as neurotically attached to people getting along together as the next man. Somehow, I think I need to get over this and allow for a more, well, complex, interpretation of the value of heated disputes. I, for one, learnt some new things when I reflected on the comments I initially strongly disagreed with there. I let myself be changed by the debate. Maybe it’s had that effect of lots of readers, though not necessarily in the same way. If I measure the value of the debate simply by its efficacy in resolving the dispute (the implicit criterion of what I said yesterday), I’d say it was horribly unproductive. But that tends to a rather circular argument in favour of politeness and people learning to agree with each other.

You see, I realise more and more than I’m quite an inconsistent and chaotic thinker. I don’t experience my own mind as a particularly harmonious phenomenon, it is often troubled and muddled. I harbour the smug opinion that I have this in common with rather a high proportion of my fellow human beings.

Ok, now back to Roger Lewin’s book. Here he quotes philospher Patricia Churchland who decided she wanted to understand neurobiology in order to continue her research. She has thought a lot about the human brain and how it works. I think what she says suggests that what we could call “muddled thinking” is not a flaw in our brains but actually what makes them brilliant. A good thing, not a problem.

Churchland tells Lewn:

Nature is not an intelligent engineer… It doesn’t start from scratch each time it wants to build a new system, but has to work with what’s already there… the result is a system no human engineer would ever design, but it is wonderfully powerful, energy efficient and computationally brilliant… Nervous systems evolved, and that makes it difficult for neurobiologists… to look at the wiring diagram and figure out what’s going on…. [Artificial intelligence researchers] tend to approach the problem within the framework of electrical engineering, and with prejudices about how they think brains should process information, instead of finding out what they do.

Churchland advocates a model of the brain as a parallel processing device, which I crudely interpret to mean one that can think several inconsistent thoughts all at the same time:

The nervous system is a parallel-processing device, and this conveys several interesting properties. For a start, signals are interpreted in many different networks simultaneously. Next, neurons are themselves very complex liittle analogue computers. Last, the interactions between neurons are non-linear and modifiable. Real neural networks are non-linear dynamical systems, and hence new properties can emerge at the network level.

So she concludes

When you think about brain activity, it’s correct to think about emergent properties at higher levels that depend on lower-level phenomena in the system.

So perhaps we can think of our apparently contradictory or chaotic thoughts as the lower-level phenomena, and the meaning we make of them or the actions that follow, the potentially more impressive emergent properties.

And I’d argue that what may be true for our own brains may be true for our communities of brains. That our disagreements are part of a bigger more impressive kind of intelligence. What we do together collectively is built not on some linear, consensus, but emerges from our diversity. The plea for politeness may be a way to close down some of our collective intelligence. Sometimes the effort to move forward based on some carefully contrived explicit agreement may actually kill off the potential genius of a group.

3 thoughts on “Complexity and Disagreement

  1. IdeaFestival

    Hyperlinked dreams

    Running a bit against type, Johnnie Moore, who is reading Roger Lewis’ book Complexity, wonders if our understanding of tough problems might be helped with a few more pointed exchanges. Out of disagreement, progress. True. His post reminds me of

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.