Imitation

Imitation is a practical means by which people and other animals can benefit from knowledge they did not discover themselves. It reduces the burden on individuals who cannot know everything, lets information travel without central direction, and can spread useful practices quickly. But it is not automatically wise: the same impulse that helps a community adopt a successful innovation can also sustain bad decisions when people follow others regardless of what happens. The distinction is between learning from others and surrendering judgment to them.

Imitation as a Response to Limited Knowledge

People often use others’ behavior as an efficient guide when direct investigation would be costly. The account’s city examples make this mechanism concrete: a person uncertain whether to take an umbrella can look at what people on the street are carrying, while a driver who finds that the cars on a street-cleaning route remain parked shortly before the deadline can infer that cleaning has been suspended. In such cases, people may benefit from information gathered by others without knowing who gathered it or how.

This kind of imitation is presented as a rational response to cognitive limits, not as proof that individuals have stopped thinking. People can specialize in uncovering information, while others learn from the visible results of that effort. Because the relevant information can percolate through a system quickly, imitation also works without a central authority directing who should learn what. Its usefulness, however, depends on consequences: someone who receives a ticket after relying on bad information has reason to check the rules independently next time. For imitation to persist in the long run, it must work often enough to reward reliance on it.

Innovation and Its Transmission

The examples of the Koshima macaques distinguish invention from the spread of an invention. In the early 1950s, a one-year-old female macaque named Imo began washing sweet potatoes in a creek before eating them. Within a few years, washing had become common among the macaques. Imo later found another way to make food easier to eat: she threw wheat mixed with sand into the ocean, where the sand sank and the wheat floated. Again, fellow macaques adopted the practice over time.

These episodes might seem to place the intelligence entirely in one unusually inventive animal. The account locates the group’s contribution elsewhere: in its adoption of Imo’s useful practices. Invention may begin with an individual, but selecting among possible solutions and spreading one through a community can be collective work. Imitation can thereby accelerate adaptation, allowing a group to benefit from a successful practice without waiting for the same discovery to recur independently across generations.

Intelligent Learning and Blind Conformity

Computer simulations by Robert Boyd and Peter J. Richerson clarify why imitation can help under some conditions and harm under others. Their agents could either try a behavior themselves and observe what happened, or watch someone who had already chosen a behavior. The simulations found that a sizable share of imitators benefited the group, provided that people could stop copying and learn for themselves when the benefits of independent learning became high enough. Following others without regard to results, by contrast, reduced group well-being.

The difference is not always easy to identify from the outside. People rarely describe themselves as mindlessly conforming, and shared behavior alone does not establish whether they have learned intelligently or simply followed the crowd. Two conditions support intelligent imitation: there must initially be a wide array of options and information, and at least some people must be willing to put their own judgment ahead of the group’s. This preserves alternatives for collective comparison and makes it possible for a community to depart from a practice that no longer works.

Overconfidence can sometimes supply the independence that this process requires, although it carries a cost. People tend to overestimate their knowledge and decision-making ability, especially when facing difficult problems, and so may choose badly for themselves. Yet because they give less weight to public information and rely more on their own judgment, overconfident people can resist a negative information cascade or disrupt it. Their dissent makes the shared signal seem less certain, encouraging others to reconsider their own information instead of merely following the prevailing example.

Imitation and Collective Judgment

The book places imitation alongside other ways of combining distributed knowledge. In the search for the lost submarine Scorpion, John Craven gathered people with varied expertise and had each offer separate estimates about possible explanations for the submarine’s fate. Their wagers addressed different particulars, including the submarine’s speed as it sank and the steepness of its descent; Craven used Bayes’s theorem to combine their judgments into an estimate of its final location. This is not imitation in the narrow sense of copying a visible behavior. It illuminates a related principle: a group can make better use of dispersed knowledge when individuals contribute distinct judgments instead of simply converging on one another’s guesses.

The distinction also matters in large collaborations. During the 2003 outbreak of SARS, the World Health Organization brought eleven laboratories into a joint effort to identify the disease’s cause. They shared results, samples, and images and spoke in daily teleconferences. The eventual finding that a coronavirus caused SARS could not be credited to one person or laboratory: the network’s collective work also depended on eliminating other candidates. Together, the researchers reached in weeks what any one lab might have taken months or years to establish. The case shows how sharing information can multiply the value of specialized work without requiring participants to imitate one another’s conclusions.

These comparisons place imitation within a broader account of collective intelligence. It can make valuable information travel rapidly, but it works best alongside independent observation, competing possibilities, and willingness to revise. The Koshima macaques benefited from following a successful innovation; Boyd and Richerson’s simulations show why copying must remain conditional; Craven’s team and the SARS laboratories show other ways distinct contributions can be pooled. Across these cases, the group’s advantage lies not in agreement for its own sake, but in transmitting what works while retaining the capacity to question it.