What is Natural Selection?

The truth is there are a variety of definitions out there in the mountainous literature on evolution, and there is no consensus, even among biologists, about how Darwin’s central concept should be defined. One reason is that natural selection does not refer to a thing, or a discrete “mechanism.” It’s really a metaphor – what I call an “umbrella category” – that directs our attention to a fundamental aspect of the history of life on Earth.

Given the misunderstandings that are evident in various writings about evolution, it’s worth re-stating Darwin’s basic idea. In a nutshell, he posited that the history of life on Earth has involved a very long, trans-generational process characterized by both continuities and “progressive” (and sometimes regressive) functional developments over time. Moreover, both the continuities and the changes that have occurred in the course of evolution have been the ultimate result of a causal dynamic that is internal to the process itself. It was not imposed from outside.

In essence, this causal dynamic is a process in which the outcomes (in terms of survival and reproduction) in each generation of living organisms are determined in situ by the functional relationships and interactions that occur between organisms and their specific environments. Both the organism and its environment are important players in this dynamic, and it is absolutely wrong to say that inanimate environments do any “selecting”. Even as a metaphor, this is misleading. Likewise, it is onerous to say that something is “selected for.” It implies premeditation.

Darwin characterized this dynamic as “natural selection,” but he well understood (despite his sometimes flagrant rhetoric) that natural selection is not a concrete mechanism, and it does not actually do anything. In fact, it was based on an analogy with artificial selection by animal breeders. Nor did he claim that natural selection was the exclusive agency of evolutionary change; he was well aware of the complexities.

A crucial point about Darwin’s theory, which is often overlooked by his critics, is that it rests on the fundamental assumption that life is a contingent and often precarious process (a “struggle for existence” as Darwin put it) and that “earning a living” (and reproducing) in the “economy of nature” is the basic vocation for all life forms. In other words, in evolution there is no free lunch.

Given this premise, and given the well-documented fact that living systems can vary greatly in their functional capabilities – their ability to earn a living in a given environment – it follows that there will be differential success in surviving and reproducing. So natural selection refers to the survival/reproduction outcomes in each generation, including both the continuities and the changes -- the weeding in as well as the weeding out. (I’m partial to biologist Theodosius Dobzhansky’s distinction between “normalizing” or stabilizing selection, positive selection, and negative selection.) Darwin also adopted the Malthusian assumption of relentless population growth, which greatly intensifies competition for the means of subsistence, but this assumption is not essential to the theory and is not always the case in nature.

So why is there no “standard definition” of natural selection. In part, no doubt, the very subtlety of the idea challenges our efforts to provide a simple one-sentence definition. But partly too, I suspect the differences reflect varying degrees of bias among those who are strongly pro-natural selection and those who would wish to minimize or even reject Darwin’s theory. The pro-Darwinians often speak about natural selection as if it were an active, even omnipotent selecting agency. Thus, paleontologist George Gaylord Simpson asserted that "The mechanism of adaptation is natural selection....[It] usually operates in favor of maintained or increased adaptation to a given way of life." Similarly, biologist Ernst Mayr informed us that "Natural selection does its best to favor the production of programs guaranteeing behavior that increases fitness." And, in his discipline-defining volume Sociobiology, Edward O. Wilson assured us that "natural selection is the agent that molds virtually all of the characters of species."

On the other side of the fence are anti-Darwinian theorists like biologist Robert Reid, in his book Biological Emergences, who claims that natural selection is “irrelevant” to the explanation of complexity in the natural world. Evolution is an internally-driven, “emergent” process, he tells us, and natural selection is mostly “obstructionist”. At best it may play a minor, “fine-tuning” and “stabilizing” role by “weeding out” unfit variants. Likewise, biologist Lynn Margulis and her son and co-author, Dorion Sagan, while not hostile to natural selection, nevertheless downplay its role in making their case for “symbiogenesis” as a creative agency in evolution. In their book, Acquiring Genomes, they speak of natural selection as “a strictly subtractive process.”

In other words, some theorists see natural selection as the “creator” while others see it as the “executioner”. It’s analogous to a situation in which two critics, viewing an old-fashioned black and white photograph, get into an argument about the role of the negative. One critic asserts that, since the paper used for the print is white to begin with, only the black portions can be attributed to the influence of the negative. On the contrary, the other critic argues, the negative is only responsible for the white portion of the print, since that is where the negative acted to block the developer light from passing through to darken the print. In fact, the negative controls both the light and dark portions of the print by virtue of its ability either to block the light or allow it to pass through. And so it is also with natural selection. To repeat, natural selection both weeds in and weeds out evolutionary novelties, and gives a pass to prior developments that still work.

This formulation can be illustrated with a textbook example of evolutionary change -- "industrial melanism." Until the Industrial Revolution, a "cryptic" (light-colored) species of the peppered moth (Biston betularia) predominated in the English countryside over a darker "melanic" form (Biston carbonaria). The wing coloration of B. betularia provided camouflage from avian predators as the moths rested on the trunks of lichen-encrusted trees, an advantage that was not shared by the darker form. But as soot blackened the tree trunks in areas near growing industrial cities, in due course the relative frequency of the two forms was reversed; the birds began to prey more heavily on the now more visible cryptic species and overlooked the darker form.

The question is, where in this example was natural selection "located?" The short answer is that natural selection encompasses the entire configuration of factors that combined to influence differential survival and reproduction. In this case, an alteration in the relationship between the coloration of the trees and the wing pigmentation of the moths, as a consequence of industrial pollution, was an important proximate factor. But this factor was important only because of the inflexible resting behavior of the moths and the feeding habits and perceptual abilities of the birds. Had the moths been subject only to insect-eating bats that use "sonar" rather than a visual detection system to catch insects on the wing, the change in background coloration would not have been significant. Nor would it have been significant had there not been genetically based patterns of wing coloration in the two forms that were available for "selection" in the two forms. (Later studies concerning the additional influence of air pollution can be left out of the discussion for our purpose.)

Accordingly, one cannot properly speak of "mechanisms" or fix on a particular "selection pressure" in explaining the causes of evolutionary change via natural selection. One must focus on the interactions that occur within an organism and between the organism and its environment(s), inclusive of other organisms; natural selection is about adaptively significant changes in organism-environment relationships. But this begs the question: What factors are responsible for bringing about changes in organism-environment relationships? The answer, of course, is many things. It could be a functionally-significant mutation, a chromosomal transposition, a change in the physical environment, a change in one species that affects another species, or it could be a change in behavior that results in a new organism-environment relationship. In fact, a whole sequence of changes may ripple through a complex pattern of relationships. For instance, a climate change might alter the ecology, which might induce a behavioral shift to a new habitat, which might encourage an alteration in nutritional habits, which might precipitate changes in the interactions among different species, resulting ultimately in the differential survival and reproduction of alternative morphological characters and the genes that support them. (An excellent illustration of this causal dynamic can be found in the long-running research program in the Galápagos Islands among "Darwin's finches" by the husband and wife team, Peter and Rosemary Grant.)

The bottom line is this: It is the functional effects or consequences of various organism-environment pattern-changes, insofar as they may impact on differential survival, that constitute the "causes" of natural selection. Another way of putting it is that causation in evolution also runs backwards from our conventional view of things; in evolution, functional effects are also causes. It is an iterative process. To use Ernst Mayr's (1965) well-known distinction, it is the "proximate" functional effects which result from any change in the organism-environment relationship that are the causes of the "ultimate" (transgenerational) selective changes in the genotype, and the gene pool of a species.

Thought for the Day: “When the words are confused, the mind is also” -- Seneca