May 08, 2003

Elsewhere on the Web

Things spotted today while surfing in between interludes of writing (or, more precisely, trying to write) the book of my dissertation ...

Both Helen Montana and Cosma Shalizi blog on a paper in Nature that uses self-replicating programs that perform logic functions in order to gain "energy." The paper shows that complex functions, involving the coordination of many instructions, can evolve from simple ones, without any need for outside intervention. As Cosma says, the paper might as well have been entitled, "Nature to Creationists: Drop Dead." Lots of interesting stuff; "bad" mutations that turn out to have highly positive consequences when combined with future mutations; variant paths to the same complex function. It's a very nice paper, available here if you have institutional access. And here if you don't (thanks to Scott Martens).

Kieran Healy posts on the bane of professors' lives - plagiarism, and how to spot it. A pity for British intelligence that they couldn't read him before plagiarizing an article from the Internet for their report on Iraq's nasty habits; they might have done a better job in disguising their sources.

Dan Drezner on economics and art markets (permalinks bloggered - it's time to shift over to MT Dan!). Like Dan, I'm all for pushing the economic analysis of art as far as it can go; however, I also think that this kind of analysis reaches its limits relatively soon. Art markets are just plain weird, as this paper by an old professor of mine, John Brewer, makes clear. The value of expensive artwork is heavily influenced by the assessments of art critics and "experts," so that it varies wildly, as a consequence both of critical fads and market manipulations (which often go hand in hand - e.g. Duveen's relationship with Berenson). Because one can't "objectively" evaluate the worth of art, except in very broad terms, it's about as non-transparent a market as you can imagine. At its worst, it's a conspiracy of mutual backscratching and puffery, in which it's in no-one's interests to admit that the emperor has no clothes; i.e. that this or that well-known artist really isn't very good, or that a substantial number of Old Masters are very probably modern fakes.

Posted by Henry at May 8, 2003 07:57 PM | TrackBack

Lacking institutional access, I haven’t read the paper. However, I will give odds that the more interesting consequence - that the genetic diversity of a population has survival value for the individual’s progeny - has not been explored. It kinda makes a mess of the “selfish gene” conception of things.

I took John Koza’s genetic programming class at Stanford. None of this stuff should have come as a shock to anyone whose been paying attention to the artificial life/evolutionary programming people. And, I’ll hazard a guess that it won’t make much difference to the creationists.

Posted by: Scott Martens at May 9, 2003 05:18 AM

My bad. The paper can be downloaded without the benefit of paid institutional access to Nature from

Frankly, I don’t see anything in it that wasn’t in Koza’s class four years ago. The main innovation is the introduction of biology terms like “Phylogenetic depth” and “genotypic space”, and the willingness to say that the results have a direct impact on evolutionary theory, which Holland and Koza have generally been hesitant to do.

Posted by: Scott Martens at May 9, 2003 05:28 AM
Post a comment

Remember personal info?