Stuart Kaufmann (In: "Origins of Order") draws some interesting conclusions
from mathematical models of "evolving" entities on fitness landscapes. His
book is think, complicated, and technical. But, I think it's worth the
trouble. what I was able to extract from it was this:
1) If you are way way down in low-fitness land it pays to make wild leaps in
design space; after all, anywhere is better than where you are.
2) As you move towards a (local or overall) maxima it becomes less and less
benificial to make blind leaps and more and more benificial to take baby steps.
Both these conclusions are echoed by Dawkins and are pretty standard. However,
the next points are interesting:
3) Species will drift away from local maximas DESPITE selection pressure
driving them upwards. How much and how often this happened in the computer
models dependeded on which variables were used. The reason is that variation
continues (ie mutations still arise) even though the "best" solution has
already been found. This drift is benificial in some circumstances becuase it
allows species to drift out of sub-optimized local maximas into the global one.
4) Even at the global maxima there is still random drift causing the species to
wander around in slightly sub-optimized space.
OK, so why did I put this here?
Assume the "fitness landscape" is reality. Your mental model of it (or your
"meme-complex", or whatever term you fancy) is a point wandering on this
surface. All of us as a "culture" are millions of points orbiting each other
(some farther away than others, obviously ;))...in effect we are a mental
species.
None of our perceptions is the same. As a group we approximate an optimum but
as individuals we don't ever actually sit there. We kind of drift in the
vicinity. In addition, we have no idea if we are at a local or global optimum
(likely the former) which means it's still valuable to send out "scouting
parties" into reletvly sub-optimal "meme-space" in search of the proverbial
passage to the perfect truth, or at least "better truth".
I like this image becuase I think it makes obvious several erroneous
assumptions:
1) We don't all have the same concept of reality. This is really not an
assumption anyone makes. But, when we talk about "objective reality" we some
times slip into the habit of opposing it with "subjective reality" when we
should be speaking of "subjective realities"; we each have our own.
2) If, for a moment, you actually had the fitest, most optimal meme-structure
in your mind, I don't think you'd know it. I'll call this the "Eureka!"
fallacy, based upon Archemedies. If you held the perfect mental map in your
mind, the likelyhood is you would drift from it almost instantaneously. We are
searchers by nature, and it is a vain hope that one day our minds will fall
into the correct "mental hole" like a golf ball, whereupon we will all say
"Eureka!"
3) Objective reality, the fitness landscape, is changing, and I might even
add: rapidly. I don't mean that the Gravitational constant of the speed of
light is variable. I mean that what is "significant"; what passes from
sensation through the filters of perception into conciousness is different than
it used to be. How crippling ws dyslexia in the Middle Ages? How important
now is the ability build caluoses? What fitness is; what a "good life" or a
"life weel lead" are has changed and is changing.
So take that image of a million meme-complexes searching mountianous terrain
for the summit and start to move the landscape. Some parts slow, some parts
fast.
And here we are, buzzing in circles like bees, looking for the optimum.
I agree with David. There will always be a dichotomy between the maps and the
geography. We are always in the process of approximating the geography in ever
more precise detail.
We will never run out of things to study becuase life itself creates more
complexity. In the process of learning we create. In the process of creating
we change what we are looking at; creating the opportunity to learn more...
Reed
konsler@ascat.harvard.edu