Genotype, Phenotype, Schmenotype
The distinction between “genotype” and “phenotype” is an artificial one that obfuscates understanding past a certain point. As Dawkins points out in his selfish gene argument, from the standpoint of the gene, the gene is the phenotype and the organism is the genotype. This is not to say that we should go overboard and anoint the gene as supreme. “Genotype” and “phenotype” are concepts. From a complex system’s standpoint, they are two frozen snapshots (stages) in an ongoing autocatalytic cycle. Other stages between could be singled out and studied (e.g. ontogenesis), but we are not good at conceptualizing dynamic processes and prefer to look at relatively stable forms. We forget that these stable forms are a part of the autocatalytic process, which is ongoing.
Dawkins make a great case for looking at genes as agents undergoing evolution by natural selection (NS). But this does not mean that the “phenotypic” agents are not also doing the same. In multicellular organisms, there is not just one phenotype, but two: the single cell, and the whole organism. Actually, even this is a simplification. There are many phenotypes ranging from the chromosome level to expressed proteins to cells, organs, immune system, et al. Each of these systems can be looked at as phenotypic agents with all the others acting as the genotype.
There is an asymmetry between these systems in that they are organized hierarchically. Higher levels have been built on lower ones, historically speaking. But once catalytic closure is achieved, the question of which came first (chicken or egg) becomes a red herring.