Some of the recent comments I have received have made me realize that maybe I should re-emphasize some of the very early points I made on this blog. The point of this blog is to blatantly promote a phenotypic view of evolution, and do try to dislodge the dominant paradigm of the gene as the center of evolution. In the discussion that follows it is convenient to use Dawkins as a straw man. My own feeling, based on no evidence, is that most evolutionary biologists accept Fisher as a brilliant founder of modern genetics, and accept his as a very genic view of evolution. Interestingly, Dawkins perspective, working through the lens of Williams, is the logical outcome of taking Fisher’s work to its extremes. So, just as I feel most evolutionary biologists accept Fisher, I feel that that they are deeply uncomfortable with Dawkins, but most of these biologists would have trouble articulating exactly why Dawkins is wrong. Somewhere between Fisher’s deeply mathematical prose and Dawkins polemics something has gone awry. My feelings are that where Dawkins goes astray is very fundamental, and goes all the way back to Fisher. Basically Fisher imagined a genetical world that was a reasonable abstraction for a world in which we had no idea what a gene was, we were at the very beginnings of our understanding of inheritance, and we lacked the computational machinery to do anything more than relatively simple analytical models.
Another reading of Fisher, however, is that quantitative genetics is fundamentally a phenotypic model. The average offspring is the mean of the parents, but the loss of variation due to averaging is recovered in the form of within family variation. We can interpret Fisher’s book is an example of the phenotypic view of evolution that is illustrated using a simple Mendelian model of genetics, but which can be expanded as necessary and as computational power allows. Viewed in this way the phenotypic perspective I am advocating may be more of a descendent of Fisher’s legacy than the more classical genic view.
Just a bit of self promotion, and maybe a bit of motivational speaking (http://gallery4share.com/c/chris-farley-snl-matt-foley.html)
So here are some of the relevant points:
1) Phenotypes create new phenotypes: At first blush this is just a change in perspective. At the risk of setting up a Dawkinsonian straw man, the classic genic view is that genes make copies of themselves, and use phenotypes as a mechanism to protect themselves, and help them survive to the next generation. This is why Dawkins refers to DNA as “immortal coils”. In the phenotypic perspective parent phenotypes create offspring phenotypes using “transition equations”. These transition equations are accepted to be impossibly complex, and so we accept at the outset that the best we can do are approximations. The simplest approximation to a transition equation is probably the heritability of quantitative genetics, or the simple Mendelian math of a Punnett square, however, in many situations it will be useful to add complications ranging from maternal inheritance and indirect genetic effects, to epigenetic effects, and all the way up to cultural effects.
What this change of perspective buys us is that genes are no longer the center of evolution. There are no such things as vehicles and replicators. These are the construct of a fevered mind that deeply misunderstands evolution. Instead, genes are relegated to being a prominent, but certainly not the only, contributor to the transition equation. This leaves the transition equation as an open ended construct that can incorporate new scientific findings. Rather than having to totally reconstruct our understanding of evolution every time we come up with a new mode of inheritance, we simply need to recognize that the transition equation was more complex than we had originally thought, and we need to modify that equation appropriately.
2) Some aspects of our understanding of evolution change with a shift to a phenotypic perspective, but our basic understanding remains remarkably similar. There have been many definitions of evolution, some of which have relied on a genic view. For example, a classic definition is that evolution is change in gene frequency. Re-framing our understanding to a phenotypic perspective demands a careful rethinking of what we mean by evolution. My own definition is evolution is the change in the distribution of phenotypes in a population due to the gain or loss of individuals. This definition is consistent with phenotypically oriented classic definitions, but ends up being more specific in many ways.
Classically there have been four forces of evolution that have been identified: mutation, migration, selection and drift. These have mostly been defined in genetic terms. Thus, drift is often called “genetic drift”, mutation is discussed in terms of change in DNA structure. However, these terms can be defined and discussed in phenotypic terms without reference to the specifics of the underlying mechanisms of inheritance. Clearly migration and selection do not need reference to genes, and our understanding of them really does not need to change at all. From a phenotypic perspective “mutation” need not be genetic change. It can be any change that randomly alters the phenotype of an individual, and that does not correlate with fitness. There is the interesting caveat here, however, that based on our definition of evolution, such random changes do not become “evolution” until they are passed on to offspring. Similarly, drift can be viewed as a change in phenotype frequencies due to the random gain or loss of individuals. With the phenotypic perspective, however, a fifth force must be recognized. This force is easily ignored in the genic world, but cannot be ignored in the phenotypic perspective. This is force is secular environmental change. A lasting change in the environment, such as global warming, can change the distribution of phenotypes directly, and in at least some cases it will be an intergenerational event. For example, global warming is changing sex ratios in some reptiles. If we assume an individuals sex is fixed at hatching, then indeed this change in the distribution of males and females is an evolutionary change by our definition.
I seem to have run into my self imposed thousand word limit, so I will continue this review next week.