This blog post reinterprets animal behavior and cultural transmission patterns from the perspective of ‘selfish replicators,’ examining how evolutionary phenomena difficult to explain solely by genes can be reconceptualized.
The debates surrounding 21st-century evolutionary theory cannot be explained without reference to Richard Dawkins’s controversial work, The Selfish Gene. This book presented the most aggressive interpretation of Darwin’s theory of evolution and has garnered both significant support and criticism since its publication. Adaptationists, represented by figures like Daniel Dennett and Steven Pinker, deeply agree with Dawkins’ perspective of explaining diverse natural phenomena through genes, one of the most fundamental units of life. Conversely, scholars including Stephen Jay Gould criticize Dawkins’ theory for being overly fixated on a gene-determinism viewpoint, thereby distorting reality. This paper aims to meticulously analyze the arguments presented in The Selfish Gene. Through this analysis, it will demonstrate that introducing the concept of the ‘selfish replicator’ can partially resolve several criticisms levied against Dawkins’ theory. Furthermore, it will suggest potential modifications to Dawkins’ theory.
In his work On the Origin of Species, Darwin presented the theory of evolution, positing that the form of organisms gradually changes through natural selection. Despite its simplicity, Darwin’s theory explained numerous biological phenomena, quickly establishing itself as the mainstream paradigm in biology and exerting a profound influence on other academic fields beyond biology. Darwin’s theory of evolution is fundamentally built upon two core assumptions:
1. Certain phenotypes confer survival advantages over others (allowing them to produce more offspring),
2. And these phenotypes are inherited by offspring.
Under these assumptions, it is easy to see that over time, the proportion of individuals possessing advantageous traits gradually increases. As in the commonly cited example, giraffes with longer necks could eat leaves more easily than those with shorter necks, and consequently produced more offspring with longer necks. The reason most giraffes today have long necks is explained as the result of the accumulation of this selective pressure.
Dawkins sought to rephrase and restate these two assumptions under the question, ‘Who is the direct target of natural selection?’ His answer, presented in The Selfish Gene, is the gene. A gene is a unit of DNA that ‘separates and recombines at effective frequencies,’ carrying the blueprint for the various substances that produce life phenomena and directly influencing the phenotype. Dawkins argued that natural selection acts on genes, not individuals. To use the giraffe example above, the explanation is that because long-necked giraffes survive better, the frequency of the ‘gene for a long neck’ increases, making it appear that there are more long-necked giraffes. The basic assumptions of Darwinism, as modified by Dawkins, are as follows.
1. The phenotype expressed by a given gene produces more offspring than the phenotype expressed by another corresponding gene.
2. Those genes are passed on to offspring.
This perspective is highly challenging as it drastically reduces the importance previously assigned to the ‘individual’ in traditional evolutionary theory. According to The Selfish Gene, an individual is merely a simple collection of various phenotypes—from eyes, nose, and mouth to muscles, blood vessels, and the brain—and holds little evolutionary biological significance as a single organism. Dawkins emphasized this point by defining the individual in extreme terms, such as a ‘gene-carrying machine’. He appears to interpret various levels of phenotypes—not only the physical phenotypes of animals but also personality, altruistic behavior, and even human culture (the concept of memes)—through the lens of the ‘selfish gene’.
However, the theories of Dawkins and genetic determinists have faced diverse scholarly critiques. Criticism has been raised widely, not only by mainstream evolutionary biologists like Stephen Jay Gould but also by figures outside the scientific community. Among these, two main points stand out as scientifically valid critiques. The first criticism is that genes directly controlling many of the phenotypes Dawkins cited as examples have not actually been discovered. Physical traits like eye color or neck length often correspond relatively clearly to a single gene. For example, human eye color appears as black, brown, blue, or green depending on the dominant/recessive combination of the EYCL1~3 genes. However, phenotypes like animal behavioral traits involve complex interactions between the brain and the entire organism, making it difficult to clearly pinpoint which specific gene is associated with them. There is no research showing that a specific gene induces an animal to adopt a ‘Tit for tat’ strategy, or that a specific gene causes an animal to emit a warning call upon seeing a predator. If a direct correlation between genes and behavior has not been established, the question arises: how is the ‘selfishness’ of genes related to behavior? Therefore, Dawkins’ theory inevitably faces significant criticism for lacking experimental and empirical evidence.
The second criticism is that the concept of ‘memes,’ introduced by Dawkins in the latter part of the book, is overly contrived. While most of The Selfish Gene consists of analyzing and interpreting animal behavior and habits, the claim that cultural elements can also evolve in the same manner as biological replicators appears somewhat abruptly in the latter part of the book. Critics of memes emphasize that the theory of the selfish gene originally deals with biological genes, and applying it directly to cultural or social phenomena without sufficient logical scrutiny is problematic. Crucially, cultural phenomena and biological genes differ significantly in their modes of existence and operation. While copies of cultural content can persist virtually indefinitely in immaterial spaces like the internet, copies of genes disappear when the individual carrying them dies. Furthermore, while infinitely replicating a single original is possible in cultural phenomena, biological reality imposes constraints like food and habitat that make such infinite replication impossible. Despite these differences, it is argued that extending biological theory to cultural phenomena is not valid.
This criticism fundamentally arises because Dawkins conflated the concept of the ‘selfish gene’ in two distinct ways. The first meaning refers to the biological gene itself. In this case, the ‘selfish gene’ denotes ‘a single DNA fragment that acts selfishly’. In this sense, the selfish gene is the cause of various biological phenomena, and thus anything non-DNA cannot be included in this concept. The second criticism raised by meme critics is based precisely on this definition. The second meaning is a concept that completely excludes biological factors, encompassing everything that ‘creates copies of itself and whose replication success rate is altered by the surrounding environment’. While the first meaning of the selfish gene also satisfies this condition, it encompasses various examples presented in the book, such as memes or competitive programs in Robert Axelrod’s simulations. To clearly distinguish these conceptual categories, we will refer to entities belonging to the second meaning as ‘selfish replicators,’ distinct from ‘selfish genes’ which denote the first meaning. From here on, we will reorganize Dawkins’ theory from the perspective centered on the ‘selfish replicator’ and examine the advantages this restructuring offers over Dawkins’ original exposition.
First, let us define the selfish replicator more rigorously. A selfish replicator can first be defined as a ‘collection of copies’. For example, the replicator called the ‘EYCL gene’ can be understood as the total sum of the corresponding DNA fragments present in the cells of all humans worldwide. Since all these copies were replicated from a common ancestor, the characteristics of that ancestor have been identically transmitted. Therefore, defining the collection based on those characteristics is valid. For the EYCL gene, its chromosomal location and nucleotide sequence content would serve as criteria distinguishing this set. Furthermore, a selfish replicator possesses a phenotype. When the existence of a replicator causes a certain characteristic to manifest as a result, we call that characteristic the replicator’s phenotype. While the existence of a replicator does not necessarily guarantee the expression of a phenotype, if the replicator’s presence can be reasonably inferred as the cause when a phenotype manifests, a causal relationship can be considered established. According to this definition, biological phenotypes arising from specific DNA (e.g., the pigment protein produced by the EYCL gene and the resulting eye color change) are phenotypes, as are shared rituals or doctrines among those who possess the replicon ‘belief in religion’. Once a phenotype is expressed, its success in increasing the replicator’s numbers or its path toward extinction is determined by how well it fits the environment. This degree is defined as the replicator’s ‘replication success rate,’ determined by the interaction between the replicator’s phenotype and the environment in which the replica is placed.
Following this definition, we can readily identify numerous familiar concepts as selfish replicators. First, DNA is a clear selfish replicator. DNA can replicate itself identically during an organism’s reproduction, generates specific phenotypes within the organism, and these phenotypes influence the parent organism’s reproductive success in some way. Various cultural elements such as religion, language, and political beliefs also possess the nature of selfish replicators. These cultural elements spread from mind to mind through oral tradition or various media while maintaining a similar structure, and their transmissibility—that is, their replication success—is determined by how well their characteristics align with public tastes or social conditions. Dawkins noted this commonality between genes and cultural phenomena, leading him to coin the concept of the ‘meme’. Beyond the examples Dawkins presented, various other selfish replicators can exist. For instance, Thomas Kuhn’s ‘paradigm’ also possesses the characteristics of a selfish replicator. Belief in a paradigm is selected based on the validity (replication success rate) of the way it explains natural phenomena (phenotype). Paradigms with a high replication success rate are adopted by more scientists and leave more copies—that is, beliefs in the paradigm—in the minds of scientists than paradigms that are less successful. The selfish replicator can be seen as a kind of format that extends evolutionary theory to apply to a very broad range of fields.
Because the selfish replicator is a much more comprehensive concept than the selfish gene, it has the advantage of allowing specific biological phenomena to be interpreted in various ways beyond the perspective of DNA influence. Provided the replication mechanism can be clearly explained, it offers a far clearer and more flexible interpretation than the ‘selfish gene’. For example, the ‘beaver dam’ case introduced in Chapter 13 of The Selfish Gene is more plausibly explained by the selfish replicator. It is difficult to view the specific structure of a beaver dam as being directly dictated by DNA. However, it is clear that beaver dam structures appear in a consistent form within the habitat shared by a single clan, and that their structure changes based on the environmental conditions of the habitat. Furthermore, beavers construct massive dams over multiple generations, and it is observed that dams built by later generations replicate and reproduce the undamaged parts of dams built by previous generations. Therefore, it is more plausible to infer that a specific ‘beaver dam-building method’ is a selfish replicator possessing the ‘dam structure’ phenotype, evaluated for replication success based on the criterion of ‘sturdiness’, and replicated to subsequent beavers through the structure of existing dams, rather than explaining it solely through DNA influence. The Tit for Tat strategy, central to Chapter 11 of The Selfish Gene, is also difficult to explain solely through biological gene influence. Tit for Tat is a behavioral strategy corresponding to ‘an eye for an eye’: initially cooperative, it immediately retaliates if the opponent defects, but then reverts to cooperation. Akselrod and Hamilton demonstrated through simulations that this strategy is highly successful evolutionarily and frequently observed in nature. However, evidence is insufficient to attribute this directly to genes. Even in human cases, young children who start hitting each other often cannot easily stop. Evolutionary psychology utilizes such observations of infant behavior to confirm instincts imprinted in genes. Considering this, the core element of Tit for Tat—‘tolerance’—suggests it is not inherent in genes. Therefore, Tit for Tat can be explained not as something naturally occurring, but as a type of selfish replicator that spreads through observing others’ actions or through the process of learning its advantages oneself. Of course, phenomena related to biological phenotypes—such as the peacock’s male plumage or the manipulative effects of snail flukes—can be fully explained by the selfish replicator concept of the gene, following Dawkins’ logic.
The selfish gene and the selfish replicator are distinct concepts with clearly delineated levels. The Selfish Gene is highly significant for proposing the possibility of the selfish replicator, but its failure to clearly distinguish the two concepts somewhat undermines the logical structure of its argument. Dawkins presents biological examples to prove that the selfish gene is the agent of evolution and natural selection, and based on this, introduces the concept of the meme. However, a superior explanatory approach would first clearly establish ‘why the selfish replicator is the agent of evolution and natural selection,’ then explain how observable phenomena, such as animal behavior, arise from this selfish replicator. This approach eliminates the necessity for ‘biological genes’ to be the cause of specific behavioral patterns, thereby freeing the theory’s validity from dependence on biological reductionism. This enables a shift in perspective that moves away from ‘gene determinism’. While this perspective does not directly resolve the first criticism, it can provide an opportunity to broaden research directions beyond merely seeking genes linked to phenotypes, as exemplified by the beaver case. The second criticism is naturally resolved by introducing the concept of the selfish replicator, as previously examined.
Of course, this shift in perspective does not prove the truth of Dawkins’ theory. Identifying the selfish replicator involved in each animal behavior may be easier than finding the gene, but it remains a challenging task. Given the concept of the selfish replicator is overly broad, attempts to apply it to individual cases one by one can encounter various problems. Nevertheless, the selfish replicator provides a solid foundation for expanding Darwin’s theory of evolution, which originated from biological theory, into a highly general and universal theory. In this sense, Dawkins’ work serves as a crucial bridge connecting Darwin and the selfish replicator, making its significance by no means small.
