How pleasant to know Mr. Crowther

How pleasant to know Mr. [Crowther]
Who has written such volumes of stuff
Some think him ill-tempered and queer
But a few think him pleasant enough.

I can't say I know who those would be, though Mr. Lear would undoubtedly have been worth knowing.

Mr. Crowther, flack for the Disco. Inst., is not nearly as funny as Edward Lear, alas. In his latest semiliterate scrawling at the Disco. blog, he mangles a paper about directed evolution of RNA enzymes. Actually, given that papers are hard to read, Crowther decided to mangle a press release about the paper. After mistaking the story for a comedy script, Crowther admits: "I couldn't write a funnier script if I tried." Sensing that he may have made a category error, Crowther then projects the mistake on the scientists, writing, "Sadly, these guys just don't get the joke."

After misunderstanding and mischaracterizing the research Crowther harps on the authors' use of the term "predetermined," and concludes:

The authors sum it all up very nicely.

This beautifully illustrates what about evolution is random and what is not.

Which it does, if Crowther would have quoted the rest of the press release:

This beautifully illustrates what about evolution is random and what is not. While the end point is predicted by the selection pressure—i.e., the decreasing concentration of ingredients determines that enzymes will evolve to cope with decreased concentration—the actual mutations that allow this are completely random and cannot be predicted at the outset—i.e., if you bought an “evolution machine” and ran the same experiment, your end product would be an enzyme that could cope with low concentrations too, but the mutations that it acquired to do this might be different.

Crowther seems to believe that any human involvement at all makes the entire system non-random. But the authors' point is to distinguish between the random mutations which accrete in their RNA strands, and the directed process which selects useful mutations and discards the harmful ones. It doesn't matter whether that selection is performed by a machine built by scientists, or if it is performed by a pigeon-breeder, by the mating decisions made by individuals in a population, or by a flood drowning some individuals and not others. None of those selections are random, but the origins of variability in the population are random.

The paper makes this distinction very nicely, and Crowther would be more pleasant to spend time with if he'd think more about what he reads, and less about how to be ill-tempered and queer.