From 2016... Little totalitarian enforcers: 3-year-olds were taught the rules of a game, & when a puppet played the game a different way, they corrected him with normative language (“That’s not how it’s done!”)

Young Children See a Single Action and Infer a Social Norm: Promiscuous Normativity in 3-Year-Olds. Marco F. H. Schmidt, Lucas P. Butler, Julia Heinz, and Michael Tomasello. https://doi.org/10.1177/0956797616661182

Abstract: Human social life depends heavily on social norms that prescribe and proscribe specific actions. Typically, young children learn social norms from adult instruction. In the work reported here, we showed that this is not the whole story: Three-year-old children are promiscuous normativists. In other words, they spontaneously inferred the presence of social norms even when an adult had done nothing to indicate such a norm in either language or behavior. And children of this age even went so far as to enforce these self-inferred norms when third parties “broke” them. These results suggest that children do not just passively acquire social norms from adult behavior and instruction; rather, they have a natural and proactive tendency to go from “is” to “ought.” That is, children go from observed actions to prescribed actions and do not perceive them simply as guidelines for their own behavior but rather as objective normative rules applying to everyone equally.

Keywords: children, cognitive development, cooperation, social cognition, social norms, open materials


Individuals of virtually all social species attempt to influence and control others’ behavior, from threatening aggression to offering mating. But beginning with Homo sapiens sapiens some 200,000 years ago, human social groups—now, cultural groups—created a new form of social control in which the group as a whole communicated collective expectations for individual behavior in the form of social norms. Some social norms regulate, for instance, food distribution or mating and thus reduce interpersonal conflict and foster cooperative group functioning (Boyd & Richerson, 2009; Chudek & Henrich, 2011). But for other social norms, individuals are expected to conform merely to coordinate with other group members or to display their commitment to the group (Hogg & Reid, 2006; Lewis, 1969; Parsons, 1951; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). These conventional norms range from conventional ways of talking, dressing, using artifacts, and preparing food to cultural and religious rituals (Rossano, 2012; Schmidt & Tomasello, 2012; Turiel, 1983).

Young children are born into a world structured by social norms. For the first 3 years, however, they seem to perceive only the expectations that specific other individuals (e.g., their parents) have for their behavior. But from around age 3, children begin to say and do things that indicate a richer understanding of social norms as generic prescriptions and proscriptions coming from something larger than an individual and applying universally to anyone engaging in a certain activity (Nagel, 1986; Rakoczy & Schmidt, 2013; Schmidt & Tomasello, 2012; Smetana & Braeges, 1990). In several recent studies, 3-year-olds were taught the rules of a game, and then when a puppet played the game a different way, they corrected him or even taught him the right way to play it (Rakoczy, Brosche, Warneken, & Tomasello, 2009; Rakoczy, Warneken, & Tomasello, 2008; Schmidt, Rakoczy, & Tomasello, 2011, 2012). They often did this with so-called generic normative language (e.g., “That’s not how it’s done!”). At a very young age, children thus cross over from being targets to being enforcers, and they clearly recognize the generic, even objective, nature of the norms they are enforcing.

All previous studies of young children’s norm learning have exposed them to the norm in a context that suggests the presence of a right way to act. Typically, an adult explicitly teaches children the norm using generic normative language (“This is how it’s done”) and a conventional label (e.g., “This is daxing”; Rakoczy et al., 2008), and the objects are artifacts clearly designed to be used in a normatively prescribed way (Casler, Terziyan, & Greene, 2009; Schmidt et al., 2011). By contrast, in the current two experiments, we explored whether children who see adults performing actions might overinterpret those actions as instantiated generic social norms, even without any teaching, language, or artifacts. Young children have been shown to be promiscuous imitators who overimitate actions irrelevant to an instrumental goal (Lyons, Young, & Keil, 2007; McGuigan, Whiten, Flynn, & Horner, 2007; Nielsen & Tomaselli, 2010) and promiscuous teleologists who overattribute purposeful design to natural kinds (Kelemen, 1999, 2004). In the experiments reported here, we investigated the possibility that they are also promiscuous normativists who overattribute objective social norms when there actually are none.

In two experiments, 3-year-old children saw an adult spontaneously perform a novel action with some materials, and then they saw a puppet perform a different action, with the same materials, that had the same result. This gave the children the opportunity to spontaneously intervene and protest if they perceived the action as normatively wrong (Rakoczy et al., 2008; Schmidt et al., 2011). The modeled action was arbitrary, without obvious purpose, and thus rather open to overinterpretation in terms of the way something is done (as in “This is how we do it!”; Schmidt & Tomasello, 2012; Searle, 1995). To investigate the key question, we manipulated both the manner of demonstration (between participants) and the type of materials (within participants) used in that demonstration. Prior research suggested that children readily learn generic and normative knowledge in both pedagogical and nonpedagogical contexts (Butler & Markman, 2012, 2014, 2016; Butler, Schmidt, Bürgel, & Tomasello, 2015; Butler & Tomasello, 2016; Csibra & Gergely, 2009, 2011; Schmidt et al., 2011; Vredenburgh, Kushnir, & Casasola, 2015); accordingly, children saw the identical action performed by an adult, either (a) pedagogically for their benefit or (b) intentionally or accidentally by a stranger in an incidental observation.

In Experiment 1, each child saw the adult spontaneously fishing objects out of her bag. The adult used a tool, either an artifact (e.g., a human-made object with a hook), from which the child could infer a conventional purpose, or a natural “tool” (e.g., a branch that happened to be usable as a hook) that suggested no conventional purpose. In Experiment 2, we went a step further, stripping away all of the cues—both in the objects themselves and in the social-pragmatic context—that might suggest a norm. We did so by using purposeless junk objects that the adult spontaneously took out of a trash bag (not out of her own bag). The trash bag was filled with junk objects that were incidentally found on the child’s chair in the beginning of the experiment. We predicted that in both experiments, regardless of whether the objects had a conventional purpose, children would infer a social norm from a single intentional action and would thus protest more when the action was pedagogical or intentional than when it was accidental.