>
> >'Hill climing' in evolution is linked to the idea of 'fit' to an
> >'evolutionary niche'. Richard Levins and Richards Lewontin's 1985 book,
> >'The Dialectical Biologist' gives a good critique of such a description of
> >evolution.
>
> Peter, I have looked for their analysis of Sewall Wright's adaptive
> landscape in this book, and cannot find it. However I do see Lewontin
> making important use of the idea of multiple peaks in what I imagine to be
> his contribution to Griffiths, et al. Genetic Analysis. Here Lewontin seems
> to be making the exact opposite point of Eigen, i.e., populations may
> indeed get trapped on relatively low local peaks. Which seems consistent
> with Lewontin's critique of the panglossian adaptationist programme.
Eigen doesn't deny the fact that species can get trapped on relatively low local peaks - he just thinks that the 'valley' will in general be shallow, and localised. It's not that hill climbing is wrong - its just that 'fitness' cannot be meaningfully measured as a general property of organism. My critique is directed at 'fitness', not hill climbing. The statement that in evolution 'fitness increases' is, according to Levins and Lewontin, largely tautological.
Here's the only paragraph referring to Sewall Wright in the book:
"Finally, the interaction of random and deterministic processes gives results in evolution that are different from the consequences of either type of process acting alone. In Sewall Wright's model, selection alone would lead all local populations to the same gene frequencies, so no selection among populations would be possible. The random drift that arises from small numbers within each population would result in the nonadaptive fixation of genes. The joint effect, however, is to allow variation among local populations, which provides the variability for new cycles of selection in different directions." p. 285
Thus, from the jumping-off point of a 'fitness landscape', they go in essentially the same direction as Eigen - "variation among local populations" in a small population is equivalent to Eigen's "quasi-species". Eigen: "The wild type of a virus, however, is not such a point [i.e. a single point in the fitness landscape] referring to one 'fittest' type; it is rather represented by a broad spectrum including many neutral mutants, which we call a 'quasi-species'."
The key point is not 'one hill' or 'many hills' but 'one dimension' or 'many dimensions'. 'One dimension' remains basically Cartesian - the evolution of a species is the increase of a single determinant - fitness. 'Many dimensions' makes 'fitness' a concept which has to be situated, not general, to be applicable (note, however, that Eigen's model is not yet dialectical, in part because he is working with very simple organisms, virus in fact - Lewontin and Levins includes the additional element of the environment being determined by the organism, which means that the 'fitness landscape' might be changed by the organism.)
Peter -- Peter van Heusden : pvanheus at hgmp.mrc.ac.uk : PGP key available Criticism has torn up the imaginary flowers from the chain not so that man shall wear the unadorned, bleak chain but so that he will shake off the chain and pluck the living flower. - Karl Marx
NOTE: I do not speak for the HGMP or the MRC.