About This Site
> Related Topics
The Extended Phenotype
Convergence in Biology
EvolutionSince I've touched on the subject of evolution in many different places, I thought I'd write a little essay to tie them all together; and then, in the process, I realized I actually had something to say about it. Now I get to kill two birds with one stone!
When I say “evolution”, you probably think of biological evolution and the stupid controversies that surround it. That's all well and good … biological evolution is an important subject, well worth learning about, and also a key component of the scientific worldview. However, it's not the main thing I want to talk about here … I'm far more interested in evolution as a general principle. Biological evolution may be important, but the principle of evolution has a real and tangible presence in daily life. We are affected by it—we interact with, and live within, systems that run by evolution—and we can make effective use of it.
Now I guess I'd better explain the principle.
If there are things that reproduce, or that can be reproduced, and if they are subject to cycles of variation and selection, then the things will change over time so as to increase whatever quality the selection is based on. Given sufficient time, the quality will be maximized.
In biological evolution, the things are genes, the variation comes from sexual reproduction and mutation, and the selection is for things that survive long enough to reproduce. But, the things don't have to be genes, they can also be other things, like memes, human artifacts, or computer data. In that last case, the code that manipulates the data is (or was) called a genetic algorithm … an unfortunate name, since genes have nothing to do with it. I'd call it an evolutionary algorithm.
In human affairs, the principle of evolution goes by the name “trial and error”. That's how I arrived at my current routine, for example: by trying different things, observing the results, and incorporating beneficial changes into the basic routine.
Another name the principle goes by is “hill climbing”. To see where that comes from, first we have to think about the space of all possible variations. The essay Different Kinds of Sameness will give you the right idea; if it helps, strategy space is also a space of variations. In most cases, the space of variations is monstrously infinite-dimensional, but here we want to imagine that it's two-dimensional, a horizontal plane. With that done, we can construct a curved surface by defining the height above each point of the plane to be the quality of the corresponding variation. That gives us a nice topographic map, the graph of the quality function.
Now we can see how evolution looks like hill climbing. We start with a bunch of whatever things are evolving … we can think of that as a cluster of points on the quality surface. Variation makes the points divide and scatter, then selection gets rid of the low-lying ones; the net effect is that the cluster seems to move uphill.
The same mental image makes it easy to understand one of the most important characteristics of evolution, which is that it finds local maxima, not global ones. In some sense that's a weakness, but a weakness you understand is really just a fact to take account of … that's a more general statement of the principle Don't Fight Your Mind. Anyway, since the cluster of points moves in the local uphill direction, and not necessarily toward the nearest or largest hill, evolution is also sometimes called “gradient following”.
That's everything I wanted to say about the principle of evolution; all that's left is to point you toward related thoughts and essays.
Although I didn't have much to say about it here, biological evolution has come up in a few places, and is certainly related to evolution in general. If you're completely unfamiliar with it, I recommend the book The Blind Watchmaker as a good starting point. As for essays, well, there's Genetic Takeover, which is about a particular theory of how life might have begun, and also the essay Vision and its subessays, which mention (in passing) some aspects of biological evolution.
The essay Evolutionarily Stable Strategies and its subessays are about strategies, which are things that can be implemented in either hardware (genes) or software (memes). For more about hardware and software, see The Grand Analogy.
Since the big difference between the words “meme” and “idea” is the emphasis on the fact that memes replicate and evolve, pretty much everything about memes is also indirectly about evolution. And, I've written a lot about memes! Most of it is filed under Memes, as you'd expect; the rest you should be able to find by following links. If you want a starting point, you might try Memetic Engineering; that's a favorite of mine.
Finally, let me clarify something I said earlier: “we interact with, and live within, systems that run by evolution”. For the systems we interact with, I was thinking of things that carry memes, like newspapers and the internet; and for the systems we live within, I was thinking of superorganisms.
If you've read all this, and still not had enough, I saved a bit of evolution humor for last.
Age of Transportation, The
Different Kinds of Sameness
In Other Contexts
Separation Effect, The
@ September (2004)