Seek the Original

What is the idea? Well, I happen to think the name “seek the original” is perfect, and says it all, but here's a longer version anyway.

Whenever there's novelty, there's an original, a first instance, and the first is usually the best.

When I put it that way, it strikes me that I'm contradicting The Anatomy of Melancholy (see Nothing New Under the Sun).

… he that comes last is commonly best.

When I see a pair contradictory memes, my instinctive reaction is that one has to be right, the other wrong. Really, though, they're just opposite poles, the two extremes between which the reality falls. Both contain some truth.

Just to have some concrete examples, consider The Lord of the Rings and Neuromancer. Both were so thoroughly imitated that the imitations became whole new genres of fiction—epic fantasy and cyberpunk, respectively—but both originals are still the best. Or, if you like technical examples, how about The C Programming Language and The TeXbook? There have been a lot of books written on both subjects, but you just can't beat getting the information straight from the creator(s).

Sometimes seeking the original even requires seeking earlier versions of a single work. Whitman, for example, kept fooling with Leaves of Grass, so that the final version was quite different from the original. I should admit I've never read the whole final version … but maybe that's part of the point, that it doesn't compel me to read it. The same thing can happen in technical works, too. I happened to notice a great difference between two successive versions of the ODBC specification; it seemed clear that control of the work had passed from the creators to some other group with far less interest in clarity. Of course, if two versions of a work are so radically different, one has to wonder whether the versions should be unbundled and considered as separate works.

It's true there are cases in which the original isn't so great. Sometimes the person with the idea isn't able to execute it properly, just well enough to convey it to others; and sometimes the creator of a piece of software isn't interested in explaining the concepts behind it, or isn't articulate enough.

The biggest exception, though, has to be works rendered in technological media. The first book about C may still be the best, but the first C compiler, if it still exists, certainly isn't … and likewise for the first automobile and first television. I'm not sure what to make of this. Is it a difference between art and technology (or craft), that the first is best in one but not in the other? That would fit with my view of art as novelty. Or is it that I'm comparing apples and oranges? The first copy of Leaves of Grass probably isn't the best, it's probably all old and dusty and worn-out; and maybe, if you're interested in learning from the design of a car, the first automobile is just the thing.

Science is an interesting example to consider, here. The earliest works do have a valuable clarity of concept, but the latest works have advantages, too, in the same way that a modern car does—improved notation and terminology, a broader base of examples, and so on. It's very surprising to me that both principles (or extremes) can exist in the same medium—so much for my theory about the medium determining which principle applies! I guess it has more to do with whether the work is used as a source of concepts or as a tool.