While consensus around Darwin’s theory of evolution in its most broad sense has been clearly established, debates within biology around the foundations of selection processes are still going strong. Over the past 60 years in particular, exceptional focus has been placed on what is called the ‘levels of selection problem’.  A solution to this problem will answer the following two questions. First, what is the evolutionary unit that is being selected for in natural selection? Is it ecosystems, species, demes, colonies, groups, organisms, organs, tissues, cells, organelles, nuclei, chromosomes, or genes? Is it a combination of many of these? Second, what causes frequency change in these fundamental evolutionary units of selection?

The Extended Selfish Gene puts forward a solution to this problem. The book comprises the fourth edition of Dawkins’ classic The Selfish Gene (1976), with two chapters from his more technical The Extended Phenotype (1982) tacked onto the end of it. Its core argument is that biological evolutionary processes should be understood by taking a ‘gene’s eye view’ on them. Doing so promises to fix the gene as the fundamental unit of selection and allow for a causal gene-based account of frequency change.

As Dawkins readily concedes, this idea is not a wholly novel as it has its origins in Fisher’s attempt to unify Darwinism with the rise of Mendelian genetics, among other earlier gene-centered views. His project is unique, however, in that it gives a more explanatory account than its predecessors in this lineage by making broad use of two technical concepts: replicators and vehicles.

A replicator is broadly defined as anything that fecund and accurate copies can be made of while durably maintaining structural integrity across generations. In an exclusively biological system, only a gene in competition with its alleles has the right type of abstract, and even potentially immortal, staying power to be considered a replicator. It is important to emphasize that Dawkins is loose in his definition of genes and that he designates them as any portion of chromosomal material that is able to fulfill the role of a replicator. This flexibility has allowed Dawkins’ conception to weather the many radical advances that the science of genetics has undergone since the 1970’s when his book was first published.

By contrast, a vehicle is defined as a temporary manifestation that holistically engages with the environment in order to protect and propagate the aforementioned replicators. Vehicles, ‘survival machines’, or organisms as Dawkins also sometimes calls them in a biological context, may transform, shift, change, or reproduce, but they typically do so with extremely low or even abysmal fidelity. This consequentially designates them as not being replicators in any technical sense of the term. To get a better grasp on what evolution looks like from Dawkins’ perspective, let’s fill this out by getting an example on the table.

From the ‘gene’s eye view’, a jackrabbit is a vehicle that moves its replicators about on the Mojave Desert while doing things like obtaining food and avoiding bobcats. The jackrabbit’s ability to do these organism level actions however, is a direct consequence of the metaphorical interest of its genes, who are selfishly and solely concerned with replicating themselves across generations. This means that when the jackrabbit scurries away from a predator such as a lynx, this should not be seen as a fitness increasing behavior, but rather as an attempt on the part of the jackrabbit’s genes to maximize their ability to replicate themselves and ensure their own survival. While a given jackrabbit organism will typically survive for a mere 5-6 years in the wild, the metaphysical unity of an individual yet simultaneously dispersed gene has already survived that timespan a thousand-fold in a myriad of vehicles and is almost certainly showing no signs of expiring any time soon. It is this drastic resilience that secures the gene’s role as the protagonist in evolutionary history.

One of the main virtues of the ‘gene’s eye view’ is its ability to explain the evolution of altruism. Since the advent of Darwin’s seminal theory, this has been one of the most contentious problems in evolutionary biology. How can altruistic behavior that ultimately decreases an organism’s fitness allow it to outcompete fitter organisms which behave more selfishly? The reproductive cost incurred by the altruist, and the consequential reproductive benefit brought upon its competitors, makes the very emergence of altruism seem utterly paradoxical, despite the fact that the natural world is flush with the trait. For Darwin, this problem cut to the very heart of the theory of natural selection, and threatened to be fatal to it.

In his discussion of the evolution of primitive social and moral qualities in chapter five of The Descent of Man, Darwin entertains a solution that departs from his otherwise rigid view of the organism as the fundamental unit of selection. He speculates that while moral behavior might yield a minor advantage or even a disadvantage in fitness for an individual within a singular tribe, tribes made up of individuals who were willing to sacrifice themselves for the greater good would be more likely to supersede tribes that were made up of individuals who were overwhelmingly selfish. This was taken to be continuous with his theory of natural selection and was the genesis of what would come to be known as group selection theory; this was a commonly held view until the 1960’s, and saw a revival in the 1990’s in the form of multi-level selection theory.

A particularly devastating attack on group selection theory came in a series of papers by W. D. Hamilton (1964) in which he introduced what would come to be called ‘Hamilton’s rule’; a formal proof of kin selection and defense of the organism level view. ‘Hamilton’s rule’ is neatly captured in the formula rb > c, where r is the coefficient of relatedness, b is the benefit to the recipient, and c is the cost to the actor. While Hamilton’s actual proof involved extensive mathematics and is beyond the current scope, his understanding of altruism nevertheless affords a highly intuitive understanding.

The most straightforward way for an organism to get its genes passed on is for it to increase its personal fitness, or its own potential for reproductive success. The more roundabout way of doing this, however, is for it to increase its inclusive fitness, or the potential reproductive success of organisms that it shares genes with. In making use of the gene as a metaphysical unity, the concept of fitness can be expanded beyond the bounds of an individual organism. So, when a worker bee plunges its apitoxin laced stinger into an intruder to its hive and subsequently violently disembowels itself, its dying without having directly passed on its genes through its offspring should not necessarily be seen as evolutionary disadvantageous. This is because its relatives in the hive have been protected and are now more likely to reproduce and spread their shared genes due to this act of martyrdom. This elegant solution accounted for a variety of puzzling phenomenon such as co-operative breeding and reciprocal altruism. It was truly groundbreaking.

While Hamilton is in many ways the champion of The Extended Selfish Gene, Dawkins deviates from his program by shifting the cost-benefit analysis of ‘Hamilton’s rule’ away from the level of the organism that has a dependency relationship to its genetics, directly onto the individual gene. In doing so, Dawkins abandons personal and inclusive fitness, both of which are bound to the organism level and gives an explanation of altruism that is inspired by Hamilton, exclusively from the ‘gene’s eye view’.

Over the course of the book, Dawkins gives clear-cut analyses of a variety of parental, sexual, and aggressive interactions, and tries to show how the ‘gene’s eye view’ provides the best explanation of them. These were fresh challenges at the time to the dominant positions of his contemporaries in evolutionary biology, and many of them are still in remarkably good shape over forty years later. His use of game-theory as a way to analyze the evolution of social behavior and explain the formation of evolutionarily stable strategies, all while foregoing complex mathematical formulas, was highly innovative. In addition to this, one of Dawkins’ many strengths lie in his extraordinary ability to forecast discoveries in evolutionarily biology. One particularly striking example of this can be found in his explanation of what would come to be termed the ‘green beard effect’. This is a form of selection in which a gene, or a very small group of genes, bring about a phenotypic trait that is identifiable by other individuals and causes changes in their behavior that are based on this trait. While this was originally a mere thought experiment that was meant to serve Dawkins’ explanation of altruistic behavior, since its original publication, the effect has been discovered in numerous species such ants, mice, and slime molds.

The real force of Dawkins’ program, however, becomes most strikingly clear with his coinage of memes late in the book. While genes are seen as the fundamental unit of biological evolution, memes are correspondingly seen as the fundamental unit of cultural evolution. Take a seemingly discreet cultural unit such as the recipe for salsa doña into consideration. I both know it by heart and have it written down in the back of my copy of The Extended Selfish Gene. However, the recipe will undoubtedly survive my own inevitable death as well as the eventual decomposition of the book, just as it predated these instantiations in forms such as notepads, books, websites, and my friend who shared it with me. The salsa doña recipe is meme in so far as it is not bound to any particular ephemeral vehicle and it functions as a replicator in a cultural context.

Dawkins repeatedly strains with varying success to structurally equivocate the ‘gene’s eye view’ with something like a ‘meme’s eye view’. He does this despite the fact that genes are clearly discreet units that are subject to Mendel’s laws of genetic transmission, and memes are non-particulate, lend themselves to blending inheritance, and are subject to no know comparable laws. Nevertheless, the subsequent impact of memes on the study of cultural evolution cannot be overstated and Dawkins’ generalized Darwinism, which is derived from two relatively simply concepts is vast in its scope. It was a major contribution to the study of both biological and cultural evolution and he even speculates at one point that all evolutionary systems have the replicator/vehicle structure at their core, including extraterrestrial life that might be discovered in other galaxies.

Throughout the book, Dawkins writes with an enviable clarity that makes the material accessible to the layperson without sacrificing the precision demanded by the specialist. His ability to simultaneously address both of these audiences without constantly drawing their attention to one another is frankly impressive. Given this, it is perhaps not surprising that the ‘selfish gene’ as a metaphor has received a significant amount of blowback over the years from both audiences, but for distinct reasons. The application of intentional language to genes has led to understandable confusion and charges of metaphorical thinking run amok by Dawkins’ specialist audience. He repeatedly attempts to sway his readers away from this misunderstanding, but as soon as he does this, he further indulges the metaphor in a way that is almost always more confounding than instructive. In earlier editions he defended this choice, but in the forward to the most recent edition he repents it for the confusion that it caused and all but admits that using it was a mistake. At the same time, much of his lay audience has taken Dawkins’ program to be one of toxic ethical nihilism, and this planted him right in the thick of the culture wars. Part of this comes with the territory of being a public intellectual who writes about evolution, but the charge was predictable and could have been at least partially mitigated given how often Dawkins provocatively runs biological and psychological conceptions of altruism together, and explicitly antagonizes religious people.

The more pointed critiques of the ‘gene’s eye view’, however, have turned on two distinct interpretations of it. The first takes it to be an ontological claim about how biological evolution really takes place. The second takes it to be a novel heuristic for best interpreting evolutionary processes. Dawkins advocates the former at times in The Extended Phenotype and the later more repeatedly in The Selfish Gene. This ambivalence is at least partially due to his conviction that even if something like multi-level selection theory were adopted or turned out to be true, this perspective would be best made sense of by, and would likewise reducible to, the ‘gene’s eye view’. I will briefly sample both flavors of critique in turn.

Genes which have been found to incur enhanced transmission over others while generally reducing the fitness of the overall organism, have served as one of the ground floor vindications of the ontological interpretation of the ‘gene’s eye view’. These so-called ‘outlaw’ genes are thought to force the adoption of genic selection because their uniquely deleterious role within the larger system requires genes and their effects to be considered in isolation. Many such as Maynard Smith & Szathmáry (2000), however, have framed this not as a case of intra-genomic conflict, but rather as one of extra-genomic conflict in which ‘outlaw’ genes are fundamentally exogenous to the organism, making the relationship less akin to one of genic selection and more to one of parasite and host. The problem here is that this turns on an elastic and ultimately unjustified metaphysical preference for framing the phenomenon in one way rather than another. The more warranted and immanent attack on this justification by contrast is rooted in the fact that ‘outlaw’ genes necessarily reference genomic organization in gaining their intelligibility (Okasha 2006). If ‘outlaw’ genes only exist by making use of the ‘lawful’ structure of the organism from the start, then they seem to be plainly in support of multi-level selection theory.

The ‘gene’s eye view’ has also been repeatedly challenged for its treatment of gene frequency analysis. Sober (1984) for example has argued that because gene frequencies are typically calculated on the basis of the organism, the ‘gene’s eye view’ obscures its traditional meaning in so far as it compels a potentially irrelevant gross extant calculation. So, when an organism becomes obese or emaciated due to a mutation or some other environmental factor the frequency of a given gene will also be seen as increasing or decreasing, regardless of that organism’s fitness and completely independent of phenotypical presentation, which is what is normally being explained. While the organism does not always have to be the focal unit in evolutionary theory, it is definitely the one that is most often taken up for explanatory purposes, and obscuring this would hold severe consequence for the science.   

In addition to this, Gould (2001) has maintained that the ‘gene’s eye view’ confuses book-keeping with causality, meaning that while evolutionary changes on any level results in changes that will always be able to be accounted for on the genetic level, this actually tells us nothing about why these changes come about. While this critique rightly exposes the way in which ‘gene’s eye view’ pays no heed to one of the major components of the ‘levels of selection problem’, namely giving a causal account of frequency change, it is also potentially too generous to it. While changes on the organism and group level will typically hold consequences on the genetic level, cases of non-genetic or epigenetic inheritance, in particular those involving behavioral imprinting, imitation, or even cultural transmission, will leave genes unchanged and therefore do not lend themselves to the ‘gene’s eye view’ (Jablonka 2001).

Over and beyond ontological considerations, Sober and Lewontin (1982) have contended in their heterosis argument that the ‘gene’s eye view’ is not even heuristically relevant. This is because both genotypic fitness as well as a gene’s phenotypical expression are both highly contextually dependent, and the atomized nature of the ‘gene’s eye view’ in almost every case necessarily leaves out the majority of the relevant picture (Okasha 2006). This makes it largely ineffectual as an explanatory tool. Although obviously Dawkins has a sophisticated understanding of evolutionary biology and surely doesn’t subscribe to beanbag genetics, he does at times seem to be puzzlingly committed to something like this holding heuristic utility.

In wrapping things up, although the ‘gene’s eye view’ has clearly been subject to some devastating critiques, there is no answer to the ‘levels of selection problem’ that has been left unscathed at this point. Dawkins’ program has historically been a major contributor to driving the debate forward, and within a limited range, it has been remarkably fruitful in explaining a highly diverse set of evolutionary phenomenon. The Extended Selfish Gene in particular contains work that cannot be avoided by anyone who is broadly interested in evolutionary theory or is seeking out a more fine-grained understanding of the development of the levels of selection debate. Dawkins’ arguments are elegant, technical, and innovative, and it is hard to ask for much more than that. However, because these books have respectively had such a profound influence on evolutionary biology, it should be pointed out that The Extended Phenotype should undoubtedly be read in its entirety and the decision to include only an abridged version of it in this edition is difficult to make sense of. Rather than bringing the reader deeper into the technical aspects of the theory, it leaves them stranded with a rather confusing ahistorical half measure. This edition therefore cannot be recommended because the reader will incontrovertibly gain more from engaging with both The Selfish Gene and The Extended Phenotype respectively as separate books.

 

Bibliography:

Darwin, C. (1998). The descent of man. Amherst, NY: Prometheus Books.

Dawkins, R. (1976). The selfish gene. Oxford: Oxford University Press.

Dawkins, R. (1976). The selfish gene. Oxford: Oxford University Press.

Gould, S. J. (2001) ‘The Evolutionary Definition of Selective Agency, Validation of the Theory of Hierarchical Selection, and Fallacy of the Selfish Gene’, in R. S. Singh, K. Krimbas, D. Paul, and J. Beatty (eds.) Thinking about Evolution: Historical, Philosophical and Political Perspectives vol. 2, Cambridge: Cambridge University Press, 207–29.

Hamilton, W. (1964). “The genetical evolution of social behaviour. I”. Journal of Theoretical Biology. 7 (1): 1–16.

Hamilton, W. (1964). “The genetical evolution of social behaviour. II”. Journal of Theoretical Biology. 7 (1): 17–52.

Jablonka, E. (2001) ‘The Systems of Inheritance’, in S. Oyama, P. Griffiths, and R. D. Gray (eds.) Cycles of Contingency, Cambridge MA: MIT Press, 99–116.

Okasha, S. (2008). Evolution and the levels of selection. Oxford: Clarendon Press.

Maynard Smith, J., & Szathmáry, E. (2000). The origins of life: From the birth of life to the origin of language. Oxford: Oxford University Press.

Sober, E. (1984). The nature of selection: Evolutionary theory in philosophical focus. Cambridge: MIT Press.

Sober, E. and Lewontin, R. C. (1982) ‘Artifact, Cause and Genic Selection’, Philosophy of Science 49, 157–80, reprinted in R. N. Brandon and R. Burian (1984) (eds.) Genes, Organisms, Populations, Cambridge MA: MIT Press, 109 – 32.

 

© 2018 Charles Beasley

 

Charles Beasley, PhD graduate student in philosophy, Humboldt Universität zu Berlin