I’m going to really kill two topics with one post here. One of the suggestions I received was to discuss debates in biology, and this is one of them (sort of) as well as two falsehoods. In other words, the debate is over which falsehood is more useful! Dawkins is a proponent of the gene-centric model of evolution while Mayr was a proponent of the organism-model of evolution. I, personally, think both of these models have merit in specific contexts, but not when used as falsehoods. When looking to explain a specific organism’s evolutionary history, we should use the organismic level model which allows for easier explanation of “spandrel” attributes arising from complex genetic interactions which result from other selective pressures. One example of this is the “floppy ears” and color patterns present in domestic dogs as a result of selection for “tameness.” The human capacity for religion may also be a spandrel resulting from our capacity for Complex character selection often results in complex effects on an organism impacting a wide variety of genes. When looking at specific allelic distribution in a population, however, the genetic model allows for statistical analysis of specific alleles within a population to evaluate if selection is occurring within a population.
Wow, so, they’re both totally useful, but they’re often understood as falsehoods. Animals don’t evolve, nor do genes. Populations evolve. A group of organisms is subjected to selective pressure resulting in the genetically influenced differential survival based upon genotype. The offspring of that population, as a result, has a different frequency of certain characteristics. So, ultimately, what is the UNIT of selection? It is not a single organism, but a subgroup of organisms within a population due to some genotypic and/or phenotypic similarity.
Stephen Jay Gould made the point that the genes themselves are not “visible” to selective pressures, merely their phenotypic manifestations. Mayr criticizes the gene-centered model claiming that whole organisms with all the interactions of genes and their resulting phenotypes are the selected units. These claims, however, ignore how much the genetic sequences organisms actually impact the phenotype including behavior and morphology. While the environment certainly plays an additional role, there is some “buffering” from this in the genome, and the degree of buffering is subject to selection as well! In a sense, the inclusive genome (including epigenetic modifications) is what is “ultimately” selected for or against, but it is selected at the whole-organism level and only visible when comparing two populations! Further compounding this is the fact that not all organisms obey the “whole organism” model such as highly communal species of insects and mammals.
Both the “whole organism” and “genocentric” models, however, miss the point that evolution is not something which is measured within an individual organism, or even within a population at a given time; it is measured between populations separated through time OR space or both (I know, time and space being completely distinct from one another is a physics falsehood, but I’m not covering that can of worms).
Evolution occurs through a lack of gene flow between populations. How these populations are separated does not matter. In order to “see” evolution, you must observe populations which are somehow distinct from one another. A sample of genes or phenotypes from one population must be compared to a sample from a separate (somehow) population. While the genes are a vital component of the phenotype, not all genes play a role in the phenotype of the organism in question. (the placental formation genes present in males comes to mind) This does not mean this gene will not be selected for in the future (in the female offspring of that male). In this sense, whole-organism selection fails to predict the maintenance of these genes within males while genocentric selection succeeds. Genocentric selection fails to account for the survival of byproduct traits which (only upon additional mutations) later serve as a means of selection. Noam Chomsky argued this may account for the origin of language in humans. Other byproduct traits include feathers, sweat glands, exoskeletons in early aquatic arthropods, and so forth. Another name for “spandrels” in this sense is “preadaptations.”
These debates were, and continue to be legitimate debates within science in the sense of “at what level are traits selected.” Both of these, however, are often misunderstood to be saying that “animals or genes evolve.” This is not at all what is being stated in either of these. Evolution can be the RESULT of selection, but the selected characters do not evolve.
Note: I am assuming Gould, Lewontin, Mayr, and Dawkins were not saying “animals” or “genes” evolve, but that these were the targets being acted upon by selective forces as this is what I gathered from their books. The phrases “animals evolve” or “genes evolve” are, in the sense of single organisms, falsehoods. Gene frequencies and phenotypes within populations change, and the unit of selection (the whole organism, the gene, or the entire population) is irrelevant to what is actually “evolving.”
Also of note: there are a number of other views on this other than that of Gould/Lewontin, Mayr, and Dawkins. Most prominent of these other views is the D.S. Wilson/E.O. Wilson multilevel selection view which can be found here.