Naïve group selectionism (see T&R III) is the unquestioning belief that adaptations can evolve at all levels of the biological hierarchy--for the good of individuals, groups, species and even ecosystems--without requiring special conditions. Many people are prone to naïve group selectionism, today no less than in the past. That is why I always caution against it and featured it early in this series of blogs. If multilevel selection theory tells us anything, it is that adaptations at level X of the biological hierarchy require a corresponding process of natural selection at the same level and tend to be undermined by selection at lower levels.
Another form of naivete is just as common but less well publicized--the unquestioning belief that natural selection operates at the level of the gene and that this constitutes an argument against group selection. Call it naïve gene selectionism.
Ever since the rediscovery of Mendel and his peas, evolution has been conceptualized in terms of gene frequency change. Genes codes for traits that are expressed in individual organisms, such as wrinkly and smooth seeds. When a trait increases the relative fitness of the individual bearing the trait, then the frequency of the gene(s) coding for the trait increases in the population.
Group selection is a straightforward extension of this scheme. Genes still code for traits that are expressed in individuals, such as altruism and selfishness as defined by Hamilton (see T&R IX). The main twist to the story is that a trait such as altruism does not increase the fitness of the individual, compared to selfish individuals in the same group. Group selection is required to explain how altruism can evolve despite its local disadvantage, as we saw in the case of the haystack model. When altruism does evolve by group selection, however, the frequency of the gene(s) coding for altruism increases is the total population. How could it be otherwise?
In sexually reproducing species, each individual is a unique collection of genes that will never occur again, no matter how well it survives and reproduces. George Williams made this point in Adaptation and Natural Selection (see T&R IV) and used it to argue that genes are "units of selection" in a way that individuals are not. After all, an entity must persist across generations to be operated upon by natural selection. Genes replicate with high fidelity and persist across generations, whereas sexually reproducing individuals do not.
If sexually reproducing individuals are not "replicators" and "units of selection," then what are they? Depending upon the author, they became variously known as "interactors," "vehicles," and "targets" of selection. These terms reflect the fact that individuals are the entities that actually interact with the environment and each other. Genes differentially survive and reproduce only insofar as they cause individuals to differentially survive and reproduce.
Notice that these distinctions apply to all the standard examples of evolution, such as Mendel selecting his peas, moths adapting to trees darkened by soot, and the beaks of the finches on the Galapagos Islands. Calling genes "replicators" and individuals "targets" doesn't add anything new. As George has told me on numerous occasions, he wrote Adaptation and Natural Selection largely to explain the basic concepts of population genetics to a wider biological audience, not to propose a radical new theory of his own.
So, what's wrong with the "gene's eye view" of evolution if it merely popularizes the standard story of genetic evolution? The error is to suppose that it constitutes an argument against group selection. As I have already shown, what I call "the original problem" (see T&R II) is the standard story of genetic evolution with a single twist. Genes code for traits that are expressed in individuals, but the traits are locally disadvantageous and require between-group selection to evolve. We might need to use terms such as "interactor," "vehicle," or "target" to describe groups in this case, but we don't need to tinker with the concept of genes as "replicators."
Nevertheless, as group selection entered its dark age, the concept of genes as "replicators" and "the fundamental unit of selection" became regarded as a drop-dead argument against group selection. Here is one of hundreds of examples that could be cited, from Richard Alexander in 1979:
In 1966 Williams published a book criticizing what he called "some current evolutionary thought" and chastised biologists for invoking selection uncritically at whatever level seemed convenient. Williams' book was the first truly general argument that selection is hardly ever effective on anything but the heritable units of "genetic replicators" (Dawkins, 1978) contained in the genotypes of individuals.
In this passage, Alexander properly cautions against naïve group selectionism but then unknowingly commits naïve gene selectionism. Williams didn't reject group selection because only genes are replicators, but because (according to his assessment) within-group selection invariably trumps between-group selection. Genes are the replicators regardless of which level of selection prevails!
Richard Dawkins' role in establishing naïve gene selectionism will be the subject of my next installment. Also, a word about the word "naïve" is in order. When an idea proves to be wrong, early adherents are often portrayed as naïve, as if any smart and well-informed person should have known better. That is a naïve rendering of scientific and intellectual discourse. Typically, ideas emerge as wrong only after protracted interactions among smart and well-informed people. It is only in retrospect that the idea becomes "obviously" wrong and continuing adherence can be called naïve.
Speaking of wrong ideas, George Williams made a wrong move when he used the uniqueness of sexually reproducing individuals to conclude that only genes have the persistence to qualify as "units of selection." As Elliott Sober pointed out in his 1984 book The Nature of Selection, phenotypic traits have the persistence that George was looking for, regardless of their genetic (or non genetic) basis. Imagine selecting for a trait such as wing length in a laboratory population of fruit flies. Before you begin the selection experiment, you measure wing length for a number of generations and get the same bell-shaped curve. The bell-shaped curve persists across generations, even though each fruit fly is a unique collection of genes that will never recur. Now you begin the selection experiment by allowing only the flies with the longest wings to reproduce. If the trait is heritable, then there will be a response to selection and the bell-shaped curve will shift in the direction of longer wings.
The fact that each fly is a unique combination of genes is beside the point. Moreover, the very existence of genes is beside the point. If phenotypic distributions can recreate themselves generation after generation and respond to selection without the existence of genes, then so much the worse for genes.
This is not just idle speculation. It is likely that stable gene-like entities are the product of evolution, not a pre-condition for evolution. Biological evolution preceded genes and some forms of biological and cultural evolution might still proceed without genes or gene-like entities.
The concept of evolution without replicators is fascinating but shouldn't obscure the simpler and more basic point that I am making in this installment of the T&R series. Even when genes do function as replicators, they have no bearing upon the group selection controversy because they function as replicators regardless of which level of selection prevails. We need to guard against naïve gene selectionism in the same way that we guard against naïve group selectionism.
To be continued...
After writing a bestselling atheist "consciousness-raiser," is it at all surprising that Dawkins now finds his evolution book being prominently linked to atheism in the media mind?
Copyright © 2004 Elsevier B.V. All rights reserved.
Cooperation evolution and self-regulation dynamics in metapopulation: Stage-equilibrium hypothesis
Cang Huia, , Feng Zhangb, , , Xiaozhuo Hanc and Zizhen Lid,
Abstract
The cause, maintenance and significance of cooperation are the key to understand mutual altruism - . Differential models based on mean-field assumption and pair approximation and cellular automaton were built on metapopulation framework to reveal effect of empty patches and multi-behavior strategies on persistence under habitat degradation. (1) Cooperators always excluded under mean-field assumption, showing coexistence of defective and cooperative behaviors impossible in well-mixed population. (2) Metapopulation survive even when colonization rate lower than extinction rate, due to compensation of cooperation rewards to extinction debt. (3) With change of temptation to defect and other parameters, metapopulation can be pure cooperators, pure defectors, and cooperator–defector coexistence (aggregated cooperators encircled by the defectors with relatively fixed borders). (4) Under habitat destruction, including patch isolation and habitat decay, metapopulation remains constant through the self-regulation of cooperator–defector frequencies. The habitat improvement always accompanied contradictorily with behavioral degradation (more cheaters). Results and literature evidences lead to stage-equilibrium hypothesis: multi-element system can maintain and stabilize function and systematic level under environmental stress through self-regulation of element proportions. It emphasizes the homeostatic utilitarian of diversity, besides narrowly and diffusely utilitarian and ethical impetration. This hypothesis compared with Gaia theory, niche construction, ecosystem engineering, and Baldwin effect.
But what if in actuality it is altruism as a biological strategy that enables the growth of the culture that regulates it, and does so differently according to the diversity of biological forms and diversity of their cultural environments? Because altruism has been around much longer than humans, and cultures have arguably been around regulating group behaviors from at least the emergence of bacteria on earth. The problems of freeloaders were certainly around and certainly under similar control, long before humans were here to address them.
Further, if cheating is the result of being genetically selfish rather than altruistic, then if while in a cooperative group selfishness is being equated to cheating, the sanctions applied to cheaters can only temporarily suppress the behavior - and, if cooperation is essentially altruistic, such suppression will never make a true cooperator.
It appears theories here that attempt to predict the dynamics between cheating and altruism are unconsciously being fashioned to fit the definition of cheating to the shape of that theoretical structure. Cheating being defined as rule breaking that gains an unfair advantage over rule followers, the inference being the act wouldn't occur if it didn't produce short term advantage, but is in the end improper because otherwise would be long term advantageous.
So it's not that remarkable that the theories tend to show short term advantage in primary group formations but disadvantage when time has allowed more complex groupings to evolve, The term cheating will otherwise be inconsistent with a different structural pattern. And then what other dichotomy of behavioral traits will account for the complexities of social interaction? Assuming such traits are genetic to begin with.
There are just too many questions about how or why these particular traits required an assumption they were each somehow a structural unit, or that their structures evolved by some sort of tandem yet separate process, for example. There seems to be no room in the process for consideration that they rose from the organism's experiences within groups, rather than somehow being zapped into genetic forms that then became endemic to group formation. Thus no consideration of how experiences turn into traits at all, etc.