This past December, science writer David Dobbs published an essay (2013) in the online magazine Aeon ( that purports to explain why the ‘selfish gene’ concept is outmoded and should be retired.  It elicited a good deal of commentary, and in early March, Aeon published responses (Sapolsky et al., 2014) to the original article from four individuals (two scientists, a genetic counselor, and a philosopher) as well as additional comments by Dobbs.  For those who are interested in this controversy, responses to the original Dobbs article were also posted elsewhere by Richard Dawkins (2013) and Jerry Coyne (2013a, b).  Below, I provide a sense of the arguments of Dobbs, the tenor of the criticisms of Dobbs’s piece, and selected other critiques of the gene-centric approach to evolution.

Dobbs’s thesis. In the original article, Dobbs argued that evolutionary change is not just about mutations in genes.  He devotes several paragraphs to discussing the transformation of grasshoppers into locusts and extols the importance for this transformation of epigenetic mechanisms, which are associated with changes in gene expression (transcription of a DNA sequence into messenger RNA followed by translation of RNA into polypeptide chains) and do not involve mutations (changes in nucleotide sequence).  To further illustrate this theme, he also discusses the development of different casts of social wasps and honeybees based on epigenetic as opposed to genetic differences.  Dobbs mistakenly appears to regard these phenomena as both a difficult challenge to the framework Dawkins described in The Selfish Gene (1976) and a revelation to the general public.

Previous criticisms of Dobbs’s thesis. As noted by several of the respondents invited by Aeon to comment on the Dobbs piece, as well as by Dawkins and Coyne, such phenotypic transitions orchestrated through alterations in gene expression do not upend the gene-centric view of evolution espoused by Dawkins (and others) who clearly noted his awareness of the relevance of variation in gene expression to the determination of phenotype in The Selfish Gene.  Dobbs also fails to acknowledge that the potential to undergo such morphological transformations in response to stimuli arising in the environment of an organism is likely to be affected by standard nucleotide changes in relevant genes and their associated regulatory regions (i.e., the genes for which expression increases in response to the external factors that elicit the bodily changes). So, while the grasshopper-locust transition in individual organisms per se is not an example of evolution, population level changes in the probability of such transitions could evolve by the standard genetic means that Dawkins discusses.

Dobbs also suggests that the importance of changes in gene expression to the determination of phenotype have not been described for the general audience.  As noted above and in Dawkins’s own response to Dobbs, he directly discussed this issue in The Selfish Gene.  More recently, other authors writing for a general audience have in fact covered this territory.  For example in a 2011 issue of The New Republic, Judith Shulevitz reviewed a book by Richard Francis on this precise topic.  In Origins: How the Nine Months Before Birth Shape the Rest of Our Lives (2010), Annie Murphy Paul also discusses epigenetics.  Unfortunately, these treatments of epigenetics suffer to some degree from some of the same blindspots that characterize the article by Dobbs.

Dobbs also focused a portion of his article on the phenomenon of genetic assimilation, also known as the Baldwin effect, whereby beneficial phenotypic alterations initially generated through epigenetic mechanisms become ‘stabilized’ through subsequent mutation and selection.  As noted by some of the commenters/critics, this phenomenon still ultimately relies on evolution via the standard process of mutation and selection and is in any event not of sufficiently widespread occurrence to overturn the standard paradigm.

Sapolsky, in his commentary on the original article, noted that Dobbs moved the source of action and phenotypic control from individual genes to the entire genome.  He (Sapolsky) then suggested that Dobbs failed to give appropriate attention on the role of environmental factors in influencing gene expression, including in the grasshopper-locust transition. This point is an important one, but Sapolsky might also have mentioned the related but arguably deeper point that contrary to the claims of some supporters of the gene-centric view, the information required to build and maintain an organism resides not just in the DNA of the genome but also in the environment.  Essentially, the genome represents a collection of recipes (a preferable metaphor in comparison to a set of blueprints, as some would have it) that are most likely to provide success, i.e. survival and generation of progeny, if the raw materials and other environmental factors present during the relevant evolutionary history (and that favored particular alleles at any given locus) continue to be present in roughly comparable quantities or ways.

Of particular recent interest, the microbes that normally reside on and inside of us are among the aspects of the environment that are essential for the construction of an optimally functioning human body and consequently can be viewed as one source of the information that complements what is deposited in our own genes.  Since microbes have their own genes, it is fair to assert not merely that human genes (nuclear and mitochondrial) do not contain all of the information necessary to build a human organism but that human genes do not even contain all of the genetic information necessary to build a human organism.  Of course, variations in our genomes may in turn influence which microbial species we harbor and in what quantities, although that influence is constrained by what microbes are in the local environment.

Non-Dobbsian challenges for the metaphor and perspective of Dawkins.  One complexity in discussing the view of evolution promoted by Richard Dawkins is that his overall notion of evolution is not always well-represented by the metaphor that he popularized and that in turn made him widely known.  The Selfish Gene considers numerous complexities and subtleties of evolutionary mechanisms and processes.  The metaphor of the ‘selfish gene’ appears to have the effect on some readers of stripping away all of this intricacy.  For example, Dawkins clearly acknowledges the frequent influences of genes on one another (epistasis) in terms of both magnitude of expression (i.e. gene product synthesis) and quality and quantity of functional effects (mediated primarily by those encoded gene products), yet a significant fraction of those aware of the “selfish gene” metaphor appear not to recognize these aspects of the picture of evolution painted by Dawkins.

Regarding the metaphor, in addition to having the ironic effect of undermining to some extent the depth of Dawkins’s evolutionary thinking for some of those aware of his ideas, it also could be argued to be too simplistic, depending on precisely how the term ”selfish” is interpreted.  While it is true that all genes succeed through replication, so-called “selfish genetic elements” (SGE), such as meiotic segregation distorters, are distinguishable from other genes by virtue of their replication at the expense of the fitness of the host organism.  So, the longer-term reproductive prospects for SGEs are dominated by genic level selection whereas the extents of reproductive success for other genes depend primarily on selection at the level of the cell or the whole organism.  It is the SGEs that most fully deserve the label “selfish.”

Due to the vast range of the topic, I will highlight just a few examples of additional criticisms of the gene-centric framework for evolution.  Samir Okasha (2006) notes two major problems for the purely gene-focused view of evolution: 1) parents of some species can transmit traits to progeny by means other than genes alone, such as cultural mechanisms or behavioral imprinting, and 2) it can be unclear how to count genes to assess gene frequency in some cases where individual organisms are not the clear units generating progeny.

For example, some plants that are composed of ramets that are chimeric, i.e. different cells have different genotypes.  For such organisms, the genic accounting required for the gene-centric explanation of evolution becomes intractable.  Similar complexities can apply to insect colonies that are initiated by multiple or multiply-mated queens creating a colony of multiple genetic lineages.

Okasha also pointed out that the gene-centric perspective has been presented in some cases as a description of evolutionary reality and in some cases as merely a heuristic device to think more clearly about evolutionary phenomena. In The Extended Phenotype (1982), Dawkins promotes the heuristic interpretation.  Nevertheless, as discussed by Okasha, either interpretation of the gene’s-eye view of evolution runs into difficulties in some biological circumstances.

Peter Godfrey-Smith (2009) also discusses some complexities and limitations of the gene-centric perspective.  For example, crossing over can occur within the boundaries of genes as typically conceived in molecular biology or as conceived by Dawkins as units of phenotype determination and replication.  He also argues that the assumption that genes compete only with allelic variants at the same locus is challenged by such entities as transposons, genetic elements that can move around a genome.  A third point made by Godfrey-Smith is that some discussions of the evolutionary role of genes are more accurately regarded as discussion of the evolutionary role of organisms characterized in terms of genes.

Newer phenomema not considered by Dobbs or Dawkins. Finally, it is worth noting that new biological discoveries continue and continue to enrich our ideas about the mechanisms of evolution.  Susan Lindquist’s lab at the Whitehead Institute and MIT has been a particularly fertile source for uncovering new phenomena and generating new concepts of evolutionary mechanism.  One fascinating finding from Lindquist and her colleagues (Rutherford and Lindquist, 1988) is that in Drosophila the chaperone protein Hsp90 (which facilitates the folding of proteins and may be more critical for proteins with new mutations to achieve functional conformations) can obscure the phenotypic effects of genetic variants that would otherwise substantially alter morphology.  If Hsp90 function is reduced through mutation or environmental inputs, the phenotypic consequences of these genetic variants can become manifest and subject to selection.  In some cases, the phenotypic effects of these genes can become independent of Hsp90 function even if fully restored through the effects of continuing selection.  These phenomena could facilitate evolution of traits that might otherwise be much less likely to evolve.

More recently, Halfmann et al. (2012) have demonstrated that prions can serve as more than unusual infectious agents causing pathology.  A prion is a protein that can adopt conformations that result both in recruitment of other copies of the same protein into the same conformation and into a growing complex based on direct physical association.  What Halfmann, Lindquist, and their associates demonstrate is that proteins exhibiting prion-type behavior exist in a significant fraction of wild strains of the yeast Saccharomyces cerevisiae and can influence cellular phenotypes, sometimes in ways that are beneficial.  The transition to prion behavior can be influenced by changes in environmental conditions.  In a future commentary, I hope to explore more fully how this novel pathway for trait transmission might contribute to evolutionary change.


Dobbs D. Die selfish gene, die. December 3, 2013.

Sapolsky R, Hercher L, James K, Dupre J, Dobbs, D. Dead or Alive? March 11, 2014.

Dawkins R. Adversarial journalism and the selfish gene. December 6, 2013.

Coyne J. David Dobbs mucks up evolution, Part I. December 5, 2013.

Coyne J. David Dobbs mucks up evolution, Part II. December 6, 2013.

Dawkins, R. The Selfish Gene.Oxford University Press, Oxford, 1976.

Shulevitz, J. Lamarck’s Revenge. The New Republic. 2011 Aug 18. – See more at:

Paul, Annie Murphy.  Origins: How the Nine Months Before Birth Shape the Rest of Our Lives. Free Press, New York, 2010.

Okasha, Samir.Evolution and the Levels of Selection.Oxford University press, 2006.

Dawkins, Richard. The Extended Phenotype. Oxford University Press, Oxford, 1982, 1989.

Godfrey-Smith, Peter. Darwinian Populations and Natural Selection. Oxford University Press, New York, 2009.

Rutherford SL, Lindquist S. Hsp90 as a capacitor for morphological evolution.  Nature. 1998 Nov 26;396(6709):336-42. PubMed PMID: 9845070.

Halfmann R, Jarosz DF, Jones SK, Chang A, Lancaster AK, Lindquist S. Prions are a common mechanism for phenotypic inheritance in wild yeasts. Nature. 2012 Feb 15;482(7385):363-8. doi: 10.1038/nature10875. PubMed PMID: 22337056; PubMed Central PMCID: PMC3319070.