from Jennie Erin Smith's review of Mark S. Blumberg's Freaks of Nature and what they tell us about development and evolution:
In 1998 the Spanish biologist Pere Alberch died in his sleep at the age of forty-three, nine years after having written a remarkable essay called "The Logic of Monsters".
Blumberg uses the subjects of this book--conjoined twins, men without penises, two-faced cats, Alberch's salamanders, and people born without legs--to advance separate but related arguments. the first is that the development of an embryo entails complex chemical ad physical processes involving temperature and gravity, along with proteins, toxins and myriad other molecules, of which DNA is but one. These "elaborate and complex, tortured and convoluted" processes, Blumberg laments, were intimately explored for centuries, then sidelines ever since the discovery of DNA. Yet there are myriad forms for which no gene can take credit. Two-headed ducks are the result not of mutations but of being jostled in utero; temperature, not chromosomes, will determine the sex of a crocodile; the free martin, a sterile and hermaphrodite cow, results from hormone having passed over from a male twin.
Nor, Blumberg's second argument goes, should genes be blindly invoked in such behavourial matters as learning to walk, since "there is no need to hardwire that which is learned through experience". Blumberg's monsters . . . are able to thrive thanks not to hardwiring but the opposite--innate flexibility--and discover ways to walk expertly on their hands, or coordinate graceful movements with a conjoined twin.
It was Pere Alberch who reconsidered the Darwinist position; the "infinite" external forces of natural selection had to be constrained, Alberch believed, by internal rules of development. Monsters, Alberch argued, offer a window on to these rules, as "they represent forms which lack adaptive function while preserving structural order. There is an internal logic to the genesis and transformation of such morphologies, and in that logic we may learn about the constraints on the normal".