#460 — The Overreach
In 1896, the American painter Abbott Handley Thayer published a paper in The Auk describing a principle of animal coloration that would come to bear his name. Many animals are dark on top and light on the bottom. Under natural overhead illumination, the upper surface receives direct light and the lower surface falls into shadow. If the animal's coloration reverses this gradient — dark where light falls, light where shadow falls — the two effects cancel and the animal appears flat, robbed of the three-dimensional shading cues that make solid objects visible. Thayer demonstrated this with a set of painted decoys: countershaded birds virtually disappeared against their backgrounds while uniformly painted birds stood out.
Thayer's law, as it came to be known, was correct. In 2009, Hannah Rowland and colleagues published the first rigorous experimental confirmation in Proceedings of the Royal Society B, using artificial prey deployed in the field. Countershaded prey survived significantly longer than uniformly colored prey, even when both groups matched the average background color. The mechanism Thayer identified in 1896 — gradient cancellation under directional lighting — was real.
But Thayer did not stop at countershading. Over the following two decades, he extended his principle to a universal claim: all animal coloration serves concealment. He painted elaborate demonstrations showing that flamingos disappear against tropical sunsets. He argued that peacock tails are camouflage when seen from behind, against forest backgrounds. He published Concealing-Coloration in the Animal Kingdom in 1909 with his son Gerald, a lavishly illustrated volume that applied the concealment principle to virtually every animal alive, including those whose coloration seemed designed for the opposite purpose.
Theodore Roosevelt, an experienced field naturalist, published a detailed rebuttal in the Bulletin of the American Museum of Natural History in 1911. His objection was empirical: he had observed skunks, flamingos, and brightly colored butterflies in the field, and they were conspicuous. The coloration of a skunk was not a failed attempt at concealment. It was aposematism — a warning. The coloration of a peacock was not invisible from the right angle. It was sexual display. Roosevelt's critique was accurate, but the damage it did was not confined to Thayer's overreach. It cast doubt on the entire framework. For decades after, countershading itself was treated as a dubious idea — aesthetically appealing but not rigorously demonstrated — because it could not be cleanly separated from the universalism that had discredited its author.
In 1962, the British ecologist V.C. Wynne-Edwards published Animal Dispersion in Relation to Social Behaviour, a book arguing that animals voluntarily restrict their reproduction for the good of the group. Populations that bred without restraint would exhaust their resources and go extinct; populations whose members showed restraint would persist. Group selection, in this account, was the dominant force shaping social behavior. The examples were vivid: bird flocks taking off together as a census mechanism, lekking behavior as a way to count the local population, territorial behavior as a resource-allocation strategy operating at the group level.
George C. Williams demolished the argument in 1966 with Adaptation and Natural Selection. His point was simple and devastating: within any group that practices reproductive restraint, a selfish individual that breeds without limit will leave more offspring. Those offspring inherit the tendency to breed without limit. One generation of individual selection overwhelms any group-level benefit. Wynne-Edwards had identified a real problem — populations do regulate their size — but attributed it to a mechanism that could not survive its own internal dynamics.
The wreckage was extensive. For thirty years after Williams, any invocation of group-level effects in evolutionary biology was treated as naive. David Sloan Wilson spent his career arguing that multilevel selection — the mathematically rigorous successor to Wynne-Edwards' intuition — was formally equivalent to kin selection (Price's equation makes this explicit). The mathematics had been available since George Price's 1970 paper. But the taint of Wynne-Edwards' overreach made the mathematics difficult to hear. The phrase "group selection" became a signal that the speaker did not understand modern evolutionary theory, even when the speaker was making a mathematically valid point about between-group dynamics. The intuition was sound. The overextension made the intuition unutterable.
In 1956, the American psychologist Leon Festinger published When Prophecy Fails, documenting a UFO cult whose members predicted the world would end on December 21, 1954. When the prophecy failed, most members did not abandon the belief. They intensified it, claiming that their faith had saved the world. Festinger used this and subsequent experiments to develop the theory of cognitive dissonance: when a person holds two contradictory cognitions, the psychological discomfort motivates them to reduce the inconsistency, typically by modifying the less entrenched belief.
The core finding was robust. Hundreds of studies confirmed that people do change attitudes to reduce inconsistency between behavior and belief. The twenty-dollar experiment (Festinger and Carlsmith 1959) remains one of the most replicated results in social psychology: subjects paid one dollar to lie about a boring task rated the task as more interesting than subjects paid twenty dollars, because the small payment was insufficient justification for the lie and the resulting dissonance was resolved by adjusting the attitude.
But the framework expanded. By the 1980s, cognitive dissonance had been invoked to explain consumer purchasing decisions, political allegiance, religious conversion, addiction, educational resistance, therapeutic outcomes, and organizational behavior. Every case of attitude change became a potential dissonance reduction. Every failure to change became insufficient dissonance. The theory's explanatory scope grew until it was difficult to identify a behavior it could not accommodate. Daryl Bem proposed self-perception theory in 1967 as an alternative account that predicted the same results without positing any internal discomfort. The debate lasted decades — not because the evidence was unclear, but because dissonance theory had become broad enough to absorb its own counterevidence. The core finding — that insufficient justification leads to attitude change — remained valid. But the framework had expanded past the point where it could be distinguished from its alternatives.
The pattern is specific. It is not the failure of a wrong idea. It is the failure of a right idea taken too far.
In each case, the original finding was correct. Countershading is a real and measurable survival mechanism. Group-level dynamics do affect evolutionary trajectories. People do change attitudes to resolve contradictions. The mistake was not the discovery. The mistake was the scope.
The overreach does not merely fail at the margin. It retroactively contaminates the core. When Thayer's flamingo claim was debunked, countershading itself became suspect — not because the evidence had changed, but because the principle could no longer be evaluated independently of the person who had pushed it past its evidence. When Wynne-Edwards was demolished, multilevel selection became untouchable — not because the mathematics was wrong, but because the name was ruined. The overclaim does not sit alongside the original finding. It metastasizes backward through it.
This is not a problem of persuasion or reputation management. It is a structural feature of how claims propagate. A principle extended without limit becomes indistinguishable from a principle with no content. If everything is countershading, countershading explains nothing. If everything is dissonance reduction, dissonance reduction predicts nothing. The overreach does not add to the original insight. It dissolves the boundary that made the insight meaningful.