232 HODGES.MEAGHER.NORTON.McBAIN.AND SROUBEK nced th can begin to 1&C as.2009:Et ney c are too weak o rmity (ie an in ine trend towand comfused to give the answer but the eed to e municate them ap iately in that context (Tetlock 2008)have e erved behavior similar to the SF have 【ested people's al int raction from othe choices i d mo reward influenc tal nu d by more thar ting to be ac nd to he like yu engage in bothc ergence and divergen and th e..a het erarchy).such tha what are often accounted a vhich individuals in both Asch and SFI situations vary thei tiating frus ting situati afth people seem quite willing to 1956 others in hat the but the evidence s that in 2010).The little willir cipand he twith the nly ch the most e ints for givin the co simnly fitt sch-type pro (2010)and Ham friends (i.e.,high trust,high social lidarity rs,as predicte H to t and Matsuda.19s5:TakaneSogon.2008:Walam sag that chil challenge for any to th and Harris's study indica is as follows:(a)E kplain why an SFI effec and Tor mascllo' they can se ofter when the ur in th Asch situation with a e point hich the child ha nick th rrect length of a b nvert the n to show that the dyna array of si a fo nstead of a dra ic dis in Asch) oly so as the ment with Some Asch dile cma odisagrecing answers in an SFI dil is no rable evid over a wide ons to con e or diverge from the red rather than as singular and categorical.Within and e in s tuations (c. mimic ma that both Heath 200:Hod values-pragmatics theory or any other theory can address theSecond, our argument is that, both in the SFI situation and in the Asch (1956) dilemma, the pragmatic context is crucial, such that the actions of people in these situations are generally more nu￾anced than a conformity–independence dichotomy can begin to capture. In both the SFI situation and the Asch dilemma situations, people do something surprising, not because they are too weak or too confused to give the correct answer but because they are sensitive to larger, more complex truths and relationships and the need to communicate them appropriately in that context (Tetlock & Mitchell, 2010). Third, the SFI effect complicates the common view that behav￾ior in social interactions and group contexts is driven by normative and/or informational influences. The substantial number of incor￾rect, disagreeing answers we observed suggests that actions are motivated by more than wanting to be accurate and to be liked. Motivations cannot be reduced to goal seeking. Participants’ ac￾tions appear to be guided by multiple values that are in cooperative tension (i.e., a heterarchy), such that what are often accounted as errors at one scale of analysis may reflect more complex, nuanced ways of negotiating frustrating situations when viewed at a larger scale. In both the SFI situation and Asch’s (1956) dilemma, participants are confronted by a discrepancy between what they see and what they hear, but the evidence suggests that in both cases, participants work to speak in a way that respects the truth of what they see and hear while acknowledging social solidarity and trust with their peers and the experimenter. Recent studies in developmental psychology reinforce the weak￾nesses of normative and informational accounts for giving incor￾rect answers in Asch-type situations and provide compelling evi￾dence of the importance of truth and the sensitivity to pragmatic constraints. Corriveau and Harris (2010) and Haun and Tomasello (2011) have both done ingenious studies with young children (3– 4 years) that present them with Asch-type dilemmas. The results are remarkably similar to those observed in adults and provide dra￾matic evidence that children trust their own vision and are willing to dissent from unanimous majorities of peers or adults. Results in Corriveau and Harris’s study indicated that 76% of 4-year-olds and 58% of 3-year-olds always answered correctly. Overall, agreement with wrong answers was 20% in Corriveau and Harris’s experi￾ments and 34% in Haun and Tomasello’s, and both sets of exper￾iments found that agreement with wrong answers declined over trials. Furthermore, both found strong evidence that children were not confused about what was true regarding the sizes they were judging. The most dramatic effect was when the task involved a game in which the child had to pick the correct length of a bridge to allow a bunny to cross a river and retrieve a prize for the child: Children never erred, always dissenting from the majority (Cor￾riveau & Harris, 2010, Experiment 2). If people, children espe￾cially, are fearful of humiliation, exclusion, or retribution, why do they openly dissent (and increasingly so as the experiment unfolds) an overwhelming majority of the time? Something deeper than normative and informational influence is operating to guide choices. Fourth, decisions to converge or diverge from the judgments and choices of others are better described as dynamical and multilay￾ered rather than as singular and categorical. Within and across tasks, there is ongoing variability and tension. Studies in anthro￾pology (e.g., Richerson & Boyd, 2005) reveal that both conver￾gence and divergence are crucial to the formation and evolution of culture, but evidence indicates that, generally, people use individ￾ually acquired information and experience in making decisions rather than doing what the majority of other people are doing (e.g., Eriksson & Coultas, 2009; Eriksson & Strimling, 2009). Claidière and Whiten (2012) concluded that evidence for what they called strong conformity (i.e., an increasing trend toward agreement or behavior matching) was particularly weak when people are moti￾vated to be accurate. Efferson, Lalive, Richerson, McElreath, and Lubell (2008) have even observed behavior similar to the SFI effect we have described. They tested people’s willingness to use social information to guide their own choices in a game, where learning from other’s choices increased monetary rewards, and found that 30% of the time, participants did not follow the major￾ity, failing to maximize their monetary outcomes. Ultimately, individuals engage in both convergence and divergence, and this variability can necessarily only be revealed over time. Therefore, future work is needed to reveal the potentially informative ways in which individuals in both Asch and SFI situations vary their responses across time. Fifth, people seem quite willing to consider the viewpoints and information provided by others in guiding their own choices, but they exercise considerable epistemic and ethical vigilance in doing so (Sperber et al., 2010). They show little willingness to follow the lead of others blindly or completely (Hodges, in press). Partici￾pants’ concern for acting in ways that realize multiple goods rather than simply choosing the most efficient route to a predetermined goal suggests that pragmatic warrant takes precedence over any concerns they might have for simply fitting in or for acting independently. Future studies can explore what occurs when trust and social solidarity are varied in SFI and Asch situations. Would friends (i.e., high trust, high social solidarity) conform less in an Asch situation than strangers, as predicted by Hodges and Geyer (2006)? Would such an effect be moderated by cultural pragmatics (e.g., Matsuda, 1985; Takano & Sogon, 2008; Williams & Sogon, 1984)? The challenge for any alternative account to the values￾pragmatics account is as follows: (a) Explain why an SFI effect occurs at all, (b) explain why it is especially strong when a concern for truth has been primed, (c) explain why people rarely disagree when they can see but disagree substantially often when they cannot see, and (d) explain the agreement and the disagreement that occur in the SFI situation and in the Asch situation with an integrated set of dynamics or processes. The point of our studying the SFI situation was not to produce some clever new effect but to invert the Asch situation to show that the dynamics underlying it are operative across a broad array of situations, including ones where conformity seems a foregone conclusion, instead of a dra￾matic discovery (as in Asch). The dynamics of convergence and divergence are pervasive and deep. Specific phenomena such as agreement with wrong answers in an Asch dilemma or disagreeing answers in an SFI dilemma must be set within the larger dynamics of convergence and diver￾gence. There is now considerable evidence over a wide range of domains, from social anthropology to developmental psychology to social psychology, that divergence is as pervasive as conver￾gence in situations (e.g., imitation, mimicry, majority influence) where the emphasis has been almost solely on convergence (Berger & Heath, 2008; Hodges, in press; Strigul, 2009). Whether values-pragmatics theory or any other theory can address the This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 232 HODGES, MEAGHER, NORTON, MCBAIN, AND SROUBEK