Conformity without majority? The case for demarcating social from majority influences

Abstract

Keywords: conformity decision making frequency-dependent learning majority influence nonhuman animals social learningIn this review, we explore the extent to which the recent evidence for conformity in nonhuman animals may alternatively be explained by the animals' preference for social information regardless of the number of individuals demonstrating the respective behaviour. Conformity as a research topic originated in human psychology and has been described as the phenomenon in which individuals change their behaviour to match the behaviour displayed by the majority of group members. Recent studies have aimed to investigate the same process in nonhuman animals; however, most of the adopted designs have not been able to control for social influences independent of any majority influence and some studies have not even incorporated a majority in their designs. This begs the question to what extent the 'conformity interpretation' is preliminary and should be revisited in light of animals' general susceptibility to social influences. Similarly, demarcating social from majority influences sheds new light on the original findings in human psychology and motivates reinterpretation of the reported behavioural patterns in terms of social instead of majority influences. Conformity can have profound ramifications for individual fitness and group dynamics; identifying the exact source responsible for animals' behavioural adjustments is essential for understanding animals' learning biases and interpreting cross-species data in terms of evolutionary processes.

A reappreciation of ‘conformity’

“…Owing to our emphasis (van Leeuwen et al, 2015) upon the need for observation records in interpreting transmission events, we are delighted to find more detailed analysis of the observation records of the vervet monkeys (van de Waal et al, 2013) in their follow-up paper (Whiten & van de Waal, this issue). Whiten & van de Waal, (this issue) present an analysis of how the number of individuals feeding from the locally preferred food colour correlated with the number of behaviours (handfuls of corn) regarding this same food colour.…”

“…To clarify, Whiten & van de Waal, (this issue) aim to address a subject pertaining to the analysis of conformist transmission that was discussed in van Leeuwen et al (2015) and Aplin et al (2015a). In summary, where van Leeuwen et al (2015) argued for keeping separate the biases of following the majority of individuals versus the majority of observed behaviours, and only reserving the term 'conformist transmission' for the former, Aplin et al (2015a) argued for grouping the biases together under the same term, i.e. 'conformist transmission'.…”

“…correlated with the 'majority of behaviours' performed across the population, thereby adding to a recent debate about how 'the majority' should be operationalized in order to study conformist transmission (see Aplin et al, 2015a in response to ; van Leeuwen et al, 2015). Here, we respond to Whiten & van de Waal, (this issue) by (1) discussing how their arguments against our alternative explanations for their conformity interpretation (as advanced in van de Waal et al, 2013) may be misguided, (2) defending the position that their correlation between the 'majority of individuals' and the 'majority of behaviours' is tangential to the current debate, (3) presenting evidence in favour of our original suggestion to keep reliance on the 'majority of individuals' and the 'majority of behaviours' as two separate learning biases, and (4) realigning the debate between Aplin et al (2015a) and van Leeuwen et al (2015) to focus again on animals' observation records as prerequisite knowledge to interpret their behavioural decisions in terms of learning biases.…”

“…van de Waal et al (2013) interpreted these behavioural adjustments by the immigrants as 'conformity'. In response to this interpretation, van Leeuwen and Haun (2014; also see van Leeuwen et al, 2015) pointed out that although the immigrants might have been guided by inclinations to conform to the majority, alternatively, they might have been guided by other (social) learning biases that are independent of majority considerations. For instance, the immigrants might have been focused on copying particular resident individuals, such as visibly dominant individuals, or indeed any resident individual, precipitated by their immigration-induced stress, anxiety or general state of uncertainty.…”

“…If these experienced individuals turn out to start using strategy A with a 70% likelihood, in fact, one could consider this to be a strong ('disproportionate' in a sense) indication of majority influence (see Haun, Rekers, & Tomasello, 2014). Thus, contrary to the conformist transmission setting, when individuals are experienced, it seems less valid to interpret a copying probability in accord with the relative majority size (here: 70%) in terms of random copying: past experience must be weighted in and perhaps a lower threshold than the majority display accepted as strong evidence for conformity (see van Leeuwen & Haun, 2014).…”