Lamarck Redux?

In a fascinating review of a recent book on evolution (TNR, Sept 4), Oren Harman suggests that reports of the death of Lamarck, proclaimed in every middle school science classroom for well over a century, may be somewhat exaggerated: Lamarckism “is and isn’t” dead.

Insofar as Lamarckism meant “direct adaptive feedback from the soma to the germ line,” it is “indeed extinct.” If I lose my leg, I won’t have legless children. Yet, there is evidence, he claims, that “a non-genetic trait that is acquired, whether epigenetically (a particular packaging of the DNA) or through learned or copied behavior (such as birdsong, or even washing potatoes), can be transmitted between generations and ultimately into an organism’s biology. This can happen in two ways: an environmental challenge may induce developmental changes that unmask a lot of existing genetic variation, which can then be captured by natural selection; or, since cultural (and epigenetic) variation can move much faster than genetic variation, it may create a holding situation until the ‘right’ mutations show up.”

He explains “epigenesis” through the analogy of recipe and baking: “If DNA is a recipe for a cake, then development is the process of baking. The molecular machinery responsible for the differential expression of genes (baking) is itself made not of DNA but mostly of protein, and it is referred to as the epigenetic system . . . . We have known for a while that the way a cake is baked can affect which ingredients are expressed in the final product, even if they all exist to begin with in the original batter. But if the epigenetic baking apparatus is itself heritable and influenced by the organisms’ interaction with the environment, then the neo-Darwinian divide between nature and nurture collapses, and the incorporation of the organisms’ needs into its biology once again emerges as an important part of evolution.” The environcment is not, on this theory, merely the “agent of selection,” but “actually helps to determine which variants are there to be selected in the first place, too, because of the role it plays in development, or the baking of the cake.” What Harman seems to be talking about is simliar to what Rupert Sheldrake has called morphic fields, though Harman is linking this more strongly to genetics than Sheldrake does (in my recollection).

To illustrate, Harman refers to macaque monkeys on the Japanese island of Koshima. Fifty years ago, researchers used sweet potatoes to lure the monkeys out of the forest to the shore where they were easier to study, and they found that one monkey, whom they christened Imo, washed her potato before eating, and soon all the other monkeys were doing the same. Later, Imo learned how to separate sand from grain by throwing handfuls into the water, where the grain would rise to the surface while the sand would sink. These practices spread from adults to children, and “soon a whole new culture developed: infants who were being carried by their mothers in the ocean became used to the water, and began to dive and even to swim in it. Hungry adult males began eating fish that the fishermen had discarded, and before long a culture of fishing had sprung up. So what began as the clever behavior of one individual blossomed into a new culture for the entire group.”

He suggests the possibility that these kinds of learned behaviors might be “slowly assimilated into the genes.” This is not simple Lamarckism, since “it ultimately depends on genesis; but it is not naive neo-Darwinism either, because random mutations and selection are mediated by the organism’s behavior and development to produce adaptive instincts. New learned behaviors acquired during development (and often only during a well-defined window of time) construct the environment in which genetic variations are selected.”