Sunday, January 13, 2013

Feedback and Convergence

E.F. Schumacher, in A Guide for the Perplexed, contrasts two kinds of problems: convergent and divergent. The first category includes those where people find it easy to solve a problem because they are agreed upon the goal. Technological problems are of this nature; it is the better mousetrap problem. There is no disagreement on the values involved: no part of the team working on the project represents mice or holds that a household filled with them is better.

Divergent problems are those where at least two camps are present and the problem concerns values. Divergent problems are ideological in character. To solve them requires not to find a better mechanism, physical or social, but transcending the conflict by appeal to a value higher than each side holds. Finding those transcending values is very difficult. Hence divergent problems are always with us.

One difference between these two is that convergent problems lend themselves more easily to isolation in experimental settings—and to experimentation. Therefore feedback can be rapid and its results can be recursively introduced into a modified draft solution. If the man leaping from the top floor of a 30-story building dies, maybe his arms—or maybe his wings—weren’t strong enough. Back to the drawing board—after the funeral. Experimenting with value systems is much, much more difficult. Entire communities must be isolated, its members cooperating willingly. And how do we know that the experimental community was large enough to be representative? Did it have an urban as well as a rural component? Did it have industry as well as agriculture? And did it have enough of each currently valued ethnicity? And what about the duration of the experiment? And how do we measure its success? By economics? Reproduction? Education? Crime rates? Some want more, some less of any or all of these. And is there a single matrix measuring happiness that could be applied, a matrix each side would accept?

Testing value systems by feedback throws some light on the problem I touched upon in my recent post on “Virtue and Time”—and on my stoic comment about the slow grind of God’s mills. Virtue is an assertion of value by action. But it is a value—finding its expression to a large extent in a society. Its long-term consequences are almost infinitely complex—whereas Boeing 787’s electrical system problems will be nailed quite soon. If virtue or lack of it could be measured rapidly—never mind its legal expressions, which never capture the whole phenomenon—we would rapidly begin to converge on that seemingly unreachable Good Society. 

No comments:

Post a Comment