Unselective Overimitators: The Evolutionary Implications of Children's Indiscriminate Copying of Successful and Prestigious Models

Abstract

Children are both shrewd about whom to copy-they selectively learn from certain adults-and overimitators-they copy adults' obviously superfluous actions. Is overimitation also selective? Does selectivity change with age? In two experiments, 161 two- to seven-year-old children saw videos of one adult receiving better payoffs or more bystander attention than another. Children then watched the adults perform unnecessary actions on novel transparent devices. Children preferred the adult who received greater payoffs or bystander attention when asked questions like "Who do you think is smarter?" but overimitated both adults' unnecessary actions equally. Although older children overimitated more, unselectivity was consistent across ages. This pattern hints at a plausible adaptive function of overimitation: acquiring rarely demonstrated behaviors by practising them immediately.

Selective copying of the majority suggests children are broadly “optimal‐” rather than “over‐” imitators

“…The findings provide direct evidence that adaptive learning biases are implemented more flexibly than previously thought, and are substantially altered by both the social context (unanimity of demonstrators) and the type of actions demonstrated (causally relevant vs. irrelevant). As expected, we found that the previously reported pervasiveness of “over‐imitation” (Chudek et al., ; Horner & Whiten, ; Lyons et al., ; McGuigan et al., ; Nielsen & Tomaselli, ) is substantially diminished in the more real‐world situation of non‐unanimous demonstrations, and that majority‐biased copying did not extend to majorities who performed irrelevant actions, despite being detected in all instances where the majority performed a causally efficient task solution. Rather than representing a “puzzling” and “mindless” peculiarity of human imitation, or a “copy‐all, correct‐later” strategy (Chudek et al., ; Whiten et al., ), our data suggest that the occurrence of so‐called “over‐imitation” instead fits with the operation of a highly flexible, selective, and adaptive high‐fidelity copying mechanism in our species.…”

“…We also anticipate that had the causally irrelevant action in our study encompassed more substantial efficiency costs, we would have observed lower rates of irrelevant action copying and faster rates of erosion over time; a suggestion consistent with the findings of Keupp, Bancken, Schillmöller, Rakoczy, and Behne (). Varying the ratio of majority versus minority demonstrators who performed the irrelevant action (for example 25:1 instead of 3:1), would also plausibly affect the rate of erosion, as would manipulating the relative age (Wood et al., ), group membership (Oostenbroek & Over, ), or status (McGuigan, ; although see Chudek et al., ) of the demonstrators. Examining the interaction of different types of learning biases in irrelevant action copying is an area ripe for future research.…”

“…. This phenomenon, dubbed "over-imitation" (Lyons et al, 2007), has received much attention in recent years, being replicated in several cultures (Nielsen, Mushin, Tomaselli, & Whiten, 2015;Nielsen & Tomaselli, 2010), and reported to increase with age into adulthood (McGuigan et al, 2011;Nielsen & Tomaselli, 2010) and to be impervious to cues of prestige or success (Chudek, Baron, & Birch, 2016).…”

“…Most previous "over-imitation" research has involved the demonstration of a single sequence of behavior (i.e., the target behavior is performed unanimously) to an observer (for exceptions see, e.g., Chudek et al, 2016;McGuigan & Robertson, 2015;Nielsen & Blank, 2011). However, real-world learning often involves observing multiple individuals behaving differently.…”

“…However, human imitation has also been described as “surprisingly unselective” or “mindless” (Whiten, McGuigan, Marshall‐Pescini, & Hopper, ), and susceptible to behavioral “inefficiency” or “cost” (Lyons, Young, & Keil, ), following numerous reports that both children and adults often blanket copy even those parts of an action sequence that are manifestly causally irrelevant to obtaining the instrumental goal (e.g., Horner & Whiten, ; Kenward, Karlsson, & Persson, ; Lyons et al., ; McGuigan, Makinson, & Whiten, ; McGuigan, Whiten, Flynn, & Horner, ; Nielsen & Tomaselli, ). This phenomenon, dubbed “over‐imitation” (Lyons et al., ), has received much attention in recent years, being replicated in several cultures (Nielsen, Mushin, Tomaselli, & Whiten, ; Nielsen & Tomaselli, ), and reported to increase with age into adulthood (McGuigan et al., ; Nielsen & Tomaselli, ) and to be impervious to cues of prestige or success (Chudek, Baron, & Birch, ).…”