In this paper I discuss one possible extension of Richard Lewontin’s proposal in The Triple Helix. After reviewing the theoretical commitments common to discussions that assume we will be able to compute an organism from its genes, I turn to Lewontin’s arguments that we will never be able to compute phenotype from genotype because the genotype specifies an organism’s phenotype relative to a range of environments. The focus of the discussion in this paper, however, is on what might follow if (...) we take seriously the claim that genetic structure does not determine phenotypic structure. The question is: What becomes causally efficacious in an explanation of the development of a heritable trait if genes are not sufficient? Any answer to this question, and even the question itself, is central to an understanding of the types of relations and structures into which humans enter and which they create in an environment. (shrink)