Before 1966, biologists casually explained animal behavior as serving the good of the group or species. Individuals sacrificed themselves for the group; resources were conserved for long-term survival; aggression was limited to preserve the population. This framework made intuitive sense: populations that worked together would outcompete populations of selfish individuals.1
But George Williams identified a logical problem: even if group-level cooperation is advantageous, individual-level cheating is more advantageous. If most individuals cooperate for the group's benefit, one individual who defects (cooperates minimally while reaping group benefits) has higher personal fitness. That selfish gene spreads; cooperation erodes. Group-level selection is possible only under specific conditions (group extinction when cooperation fails, rare migration between groups) that are easily violated. Gene-level selection—where genes spread because they make individuals reproduce more—is far more powerful.2
This reframing from group-selection thinking to gene-level thinking is among the most important paradigm shifts in biology. It explains apparent altruism not as genuine sacrifice for the group but as selfishness at the genetic level (kin selection, reciprocal altruism, sexual display). It predicts conflict where group-selection thinking predicted harmony (parent-offspring conflict, sexual jealousy, deception).3
But the tension persists: humans do show genuine group-level cooperation, self-sacrifice for group benefit, and enforcement of cooperation norms. Is this evidence that human evolution involved group selection, or are gene-level mechanisms sufficient?4
Modern evidence suggests gene-level mechanisms are sufficient: reciprocal-altruism networks, coalitional loyalty, reputation effects, and cultural enforcement can all produce cooperation that looks group-level without requiring group-level selection. But human behavior also shows patterns suggestive of group-selection history: willingness to die for abstract group concepts (nation, ideology, team), belief that individuals should sacrifice for collective good, moral intuitions favoring group welfare.5
Williams vs. David Sloan Wilson on Modern Group Selection
Williams (1966) argued persuasively that group selection is rare and gene-level selection is ubiquitous. Wright (following Williams) treats gene-level selection as the default explanation.6 But David Sloan Wilson has argued that the aversion to group-selection thinking has become dogmatic, and that human-level group selection (evolution of norms, enforced cooperation, group-beneficial traits) may be more important than gene-selectionists admit.7 The tension reflects a real empirical question: are human group-level behaviors best explained through individual-level mechanisms (reciprocal altruism, reputation, kin altruism) or through group-level mechanisms (group extinction, inter-group selection)?
Gene-level selection paradigm shift has implications for moral philosophy: if organisms are "designed" by selection for genetic selfishness, then selfishness is natural and morality is a thin veneer. But the actual implication is more subtle: organisms are designed for genetic selfishness, but the strategies that maximize genetic success include cooperation, honesty, fairness, and sacrifice in certain contexts. Morality isn't opposed to nature; it's part of nature's solution to cooperation problems. This reframes the evolutionary basis of ethics from "morality is contrary to our nature" to "our nature includes genuine capacity for moral behavior because cooperation was fitness-enhancing in the EEA."8