About This Site
> In General
Books and Stuff
Spoilers and Polluters
Art as Novelty
Patterns of Artistic Output
Seek the Original
Nothing New Under the Sun
SelectionHere's a nice puzzle I heard in electronics class. If a resistor has a silver band, so that it's guaranteed to be within ten percent of its marked value, what's the probability that it's actually within five percent? If the resistances followed a uniform distribution, the probability would be 1/2; and if they followed a normal distribution, the probability would be higher; but in fact the probability is zero … because all the resistors that were within five percent were marked with a gold band, not a silver one! In other words, the process isn't “make good resistors”, it's “make lots of resistors and pick out the good ones”.
The same idea, generalized by allowing the concept “resistor” to vary, is what I call the principle of selection.
If you want to make good Xs, make lots of Xs and select the good ones.
Selection is somewhat counterintuitive. Normally, if you want to make good Xs, Xs of a certain quality, you just work carefully and make nothing but good Xs. If you can't think of a way to make good Xs reliably, selection, a.k.a. quality control, is essential; but even if you can, selection is still worth considering. It can often produce more and better Xs than direct creation.
There are plenty more examples of selection in the context of manufacturing and quality control—notably, integrated circuits—but mostly I'm interested in it in the context of art. (Sadly, I don't know the correct art jargon for it.) Let's start with some simple examples, in various media.
In those examples, selection is absolutely necessary, because there's a distribution channel with limited capacity. We can get slightly different examples by deforming that condition in one way or another.
That brings me to my next point. So far I've talked about what kinds of things are selected, but not about who the things are selected by. In all the examples above (except the last), the people who perform the selection are professionals, editors and the like, but certainly amateurs can select things too. Even me!
Actually, my books have been selected three times, because first they had to be selected for publication. That kind of multilayered selection is fairly common.
Occasionally, I like to break a work into parts, and then pick and choose from among the parts. Consider, for example, what I did with The Matrix Reloaded. That's an unusual case, but the same kind of thing happens every time I pick out an excerpt to go in an essay. I especially like the passage from Cryptonomicon that I included in Thoughts About Stephenson, but there are plenty of examples in other places, too. A lot of good ones are filed under Quotations.
Lat. excerptum < excerpere, to pick out : ex-, out + carpere, to pluck.
Now, before I break a work into parts, it exists as a whole, but before that there was another time when it existed as parts, namely, when it was being created. The creator of a work typically has an excess of parts, some that are variants of one another, all performing the same function, and others that are purely optional. As a result, ve has to pick and choose from among the parts … and that gives us yet another kind of selection.
A particularly clean example of selection of parts is The Top 5 List. The way that that site works is, first, a topic is announced; second, on the order of a hundred different contributors submit entries; and third, an editor picks the best five (or so) and makes them into a list. The whole is pretty much just the sum of its parts, but so what, it still demonstrates the power of selection.
I imagine the same kind of thing goes on internally at The Onion … probably they have a group meeting where everyone throws out random ideas, then they pick the best ones and go off and make articles and things out of them. In other words, they brainstorm.
Now let me step back for a moment and summarize what I've been doing. In this essay, there were three things I wanted to accomplish. First, I wanted to capture the slightly peculiar meaning that I attach to the word “selection”. (In other words, I wanted to define urticator.selection.) That was taken care of within the first two paragraphs. Second, I wanted to stir up your mind with innumerable examples of selection in different contexts. That, too, has been taken care of, I hope. Finally, I wanted to write down the following paragraph.
As we've just seen, brainstorming is a way of applying the principle of selection within the context of ideas. But, it's a way, not the only way—and there is, in fact, another way that is far more important. Everyone, every day, goes around throwing out ideas, and then picking the “best” ones by choosing which ones to remember and repeat. In other words, memes operate by selection! But then, that shouldn't come as a surprise. Genes, to which memes are analogous, also operate by selection: selection by Nature, or, if you prefer, natural selection. (Natural selection isn't the same thing as evolution, but it's close; see Evolution for details.)
That's really the end, the rest is just gravy.
o March (2000)
@ September (2004)