Gene as the unit of natural selection

Gene as the unit of natural selection is the book’s central evolutionary framework: selection changes the frequencies of replicating genetic units, whose persistence across generations can outlast the bodies that carry them. Its argument depends on a deliberately practical definition of a gene, and on keeping that evolutionary claim distinct from both a simple account of development and a moral prescription for human behavior.

Defining the Gene

The term does not mean only a cistron, a sequence bounded by the symbols that mark the start and end of a protein-chain message. Crossing-over can split a cistron, just as it can divide a longer or shorter sequence; genetic units overlap and have no visibly separate physical boundaries on the chromosome. For evolutionary purposes, the book adopts a definition associated with G. C. Williams: a gene is any portion of chromosomal material that can potentially persist for enough generations to serve as a unit of natural selection. The criterion is copying-fidelity, or longevity in the form of copies, rather than a fixed molecular length.

The definition is therefore a matter of degree. A short sequence is less likely than a long one to be split at any particular crossing-over, so it can remain intact across more generations. Larger stretches can qualify when they are tightly linked and rarely separated. In butterfly mimicry, for example, a cluster of genes affecting colour, shape, spot pattern, and flight rhythm may travel together so reliably that the cluster behaves like a single gene; intermediate forms are rarely found in nature, though they can occur when large numbers of butterflies are bred in a laboratory. The book’s gene is thus neither necessarily one cistron nor an entire chromosome, but the largest practical unit able to retain the qualities required of a replicator.

Those qualities are longevity, fecundity, and copying-fidelity. A gene is not literally everlasting: even a cistron can be cut by crossing-over, and the physical DNA molecule has a short life. What may endure for evolutionary time is the pattern reproduced in many copies, potentially for a hundred million years and distributed among many bodies. By contrast, an individual body is fleeting, and the combinations of genes assembled in it are repeatedly reshuffled by sexual reproduction. Genes change partners as they pass through successive bodies; their continuity lies in the copies, not in the persistence of one organism or one DNA molecule.

Selection, Competition, and Selfishness

Natural selection, in its general form, is the differential survival of entities. For that difference to have lasting evolutionary effect, the entities must occur in many copies, and some copies must persist long enough for their relative success to matter. The book argues that small genetic units meet these conditions more effectively than individuals, groups, or species. Its account treats genes as rivals for corresponding places in future generations: a genetic unit that increases its own representation at the expense of its alleles tends to persist. In this specific sense, the gene is described as the basic unit of selfishness.

That language is shorthand, not a claim that genes think or pursue plans. Blind natural selection can produce outcomes that look purposeful: when the book says that genes are trying to increase their numbers, it means that genes whose effects lead to more copies tend to become more common. The distinction matters because the vivid terms “selfish” and “ruthless” can otherwise be mistaken for conscious motives or ethical judgments. The argument concerns what has evolved, not what people ought to do. Indeed, it warns that biological nature offers little support for a society built on generous cooperation, and presents human cultivation of altruism as a response to, rather than a moral endorsement of, evolutionary selfishness.

Nor does selection at the gene level imply that genes operate as independent builders. A modern survival machine contains many thousands of genes, and the construction of a body is an intricate cooperative process: one gene may affect different parts of the body, each part may be influenced by many genes, and the effect of any one gene depends on its interactions with others. A master gene may regulate a cluster, while the external environment is indispensable to development. Saying “gene for long legs” or “gene for altruistic behaviour” is therefore a convenient way to describe a gene that, other things being equal, shifts a developmental outcome; it does not mean that one gene produces the trait by itself.

From Inheritance to Behaviour

The framework also distinguishes ultimate evolutionary influence from immediate control. Genes shape survival machines and nervous systems, but nervous systems make moment-to-moment decisions; the book compares genes to primary policy-makers and brains to executives. Learning, prediction, and simulation give brains increasing scope to direct behavior, and human beings can even refuse to have as many children as they could. Yet a behavior can evolve only if genes influencing the development of nervous systems make that behavior more successful, in terms of representation in the gene pool, than rival genes do.

The honeybee hygiene experiment illustrates how such inherited behavioral differences can be investigated without reducing a complex action to a single undivided trait. Some strains remove infected larvae affected by foul brood; susceptible strains do not. Hygienic workers must locate the diseased grub, uncap its wax cell, pull the larva out, carry it through the hive entrance, and discard it. In W. C. Rothenbuhler’s crosses, first-generation hybrid hives were non-hygienic, although the hygienic genes remained present recessively. After back-crossing the hybrids with a pure hygienic strain, daughter hives showed three patterns: complete hygiene, none, or an intermediate response in which workers uncapped cells but did not remove the larvae. Rothenbuhler inferred separate genetic contributions to uncapping and removal; by uncapping cells himself, he confirmed that some apparently non-hygienic bees could perform the removal stage. The example demonstrates how a coordinated behavior may comprise separable inherited components, while the gene-centered claim remains about differential success across generations.

A parallel distinction applies to strategic behavior. A decision to fight or refrain from fighting can be represented as a cost-benefit calculation, although the animal need not consciously perform one. John Maynard Smith’s evolutionarily stable strategy formalizes how a behavioral policy persists when, given what most of the population does, no alternative strategy can do better. Such strategies connect individual actions to evolutionary outcomes without treating genes as conscious commanders or animals as deliberate theorists. The explanatory bridge is the different success of heritable influences on behavior.

Scope and Interpretive Significance

The gene-centered view does not make organisms irrelevant; it changes the scale at which evolutionary persistence is traced. Organisms are the temporary vehicles through which genetic units are copied, while the copies can survive repeated changes of body and chromosomal combination. This perspective also allows genetic effects to reach beyond the body containing the gene. Under the Central Theorem of the Extended Phenotype, an animal’s behavior tends to maximize the survival of the genes influencing that behavior, whether those genes are in the body acting or in another body affected by the action. Manipulation of another organism can therefore count as an extended phenotypic effect, not because the manipulated organism’s genes have moved, but because the manipulating genes alter the surrounding world in ways that favor their propagation.

The book’s argument is consequently both forceful and qualified. It identifies the gene as a useful practical unit because it can exist in many copies, retain a recognizable pattern across generations, and vary in success relative to alternatives. At the same time, it emphasizes that gene boundaries are not absolute, genetic units interact, and development depends on a web of genetic and environmental causes. Its “selfish gene” is an evolutionary description of differential persistence, not a miniature agent and not a command for human conduct. The lasting significance of the concept lies in that change of perspective: bodies and behaviors remain real, but their evolutionary history is read through the replicating units whose copies continue after any one body has disappeared.