Would Someone Please Check If This Man's Pupils Are Fixed and Dilated?

RE: sandwalk.blogspot.com.

Links to websites that don't let you BACK out of them are terribly annoying. Recommend at least a warning be posted.

From the article:

Cleland, for example, doubts that Darwinian evolution, the core of the NASA definition of life, is essential. "I think those arguments are weak," she says. She envisions alien microbes filled with enzymes but lacking genes. The enzymes build more enzymes and the microbes split in two. They couldn't evolve through Darwinian evolution, because they wouldn't have genes. But they might still change, as their environment changed. Cleland doesn't claim any evidence that such things exist, but she argues that scientists can't rule them out.

I don't understand why this doesn't count as "Darwinian evolution". Certainly, in the historical sense of "Darwinian" — i.e., closely based on the ideas of Charles himself — the specific storage mechanism for genetic information doesn't matter, since Darwin himself didn't know about specific mechanisms of heredity.

Suppose an enzyme can build a copy of itself out of ingredients present in the environment. Maybe it's made out of amino acids, maybe out of RNA, but it can copy itself. The inherent wiggliness of molecules will make this copying process imperfect: not every new enzyme will look exactly like the one which produced it. Some fraction of the time, a descendant enzyme will have a different amino acid (or ribonucleotide) sequence, or a different geometric arrangement. These altered forms might be able to reproduce themselves, producing (a) copies of the altered form, (b) copies of the original form or (c) additional new varieties.

Now, suppose the environment changes: the intensity of ultraviolet light goes up, or more sulfur is present, or something. This may change the ability of the original enzyme to reproduce itself, and perhaps one of the altered versions will do a better job. If that altered version produces enzymes which look like the original, it won't last, because the original enzyme can't cope with ultraviolet light, but if the altered version produces itself a significant fraction of the time, it might be able to stay around, producing lots of little baby enzymes.

So: non-random survival of randomly varying replicators. In what way does this not fit into the general framework of evolutionary biology? Granted, the distinction between genotype and phenotype may be altered, but I expect that a definition of "evolution" which eliminates this case would also exclude organisms which we universally agree are alive.