Cognitive demands of sociality: from simple mechanisms to complex behaviour

Individual learning and behaviour are often shaped by the actions of others. For example, individuals might acquire information from others about their environment (Rendell et al., 2010; Leadbeater, ••) while experienced individuals might teach naïve counterparts how to behave in a given context (Thornton & Raihani, 2008). Furthermore, individuals might cooperate with others to achieve personal or common goals, or compete with them over limiting resources (McAuliffe & Thornton, ••; Thornton & McAuliffe, ••). Despite the vast and fruitful body of research into these topics, a key outstanding question is whether social learning and behaviour rely on specialized, higher-order cognitive adaptations, either for social learning or for social behaviour more generally. It was with this broad question in mind that this mini-series of reviews was commissioned. The three contributions ask, respectively, whether social learning is an adaptation to social living (Leadbeater, ••), whether cooperative breeding selects for advanced socio-cognitive abilities (Thornton & McAuliffe, ••), and more generally, how considering both psychology and ecology is essential to understand the mechanisms underpinning social learning and behaviour (McAuliffe & Thornton, ••). A key take-home message from all three contributions is that seemingly complex behavioural and learning outcomes can often arise through relatively simple psychological mechanisms and, relatedly, that similar behavioural outcomes can often be achieved via different cognitive means. The tendency to attribute higher-order cognitive abilities to animals that learn and behave socially has, in part, stemmed from comparisons with humans, who are thought to possess and use advanced social cognition to navigate their social world. For example, some forms of human teaching require the teacher to ‘know what the pupil knows’ (i.e. possess theory of mind) which is a putatively uniquely human capacity (Thornton & Raihani, 2008; Thornton & McAuliffe, 2012). Similarly, the human willingness to voluntarily share resources with others (see Engel, 2011) or to punish cheats (Raihani & McAuliffe, 2012) are both thought to stem from intrinsic fairness preferences (Fehr & Schmidt, 1999), which rely on the ability to compare own payoffs with those of a social partner. It is perhaps natural, therefore, to ask whether apparently related behaviours in other species might also rely on similar cognitive mechanisms. A key emphasized by the commissioned reviews, however, is that behaviours that serve similar functions need not all rely on the same cognitive mechanisms. For example, although human punishment is thought to be motivated, at least in part, by inequity aversion (Raihani & McAuliffe, 2012), the bluestreak cleaner wrasse Labroides dimidiatus shows no evidence for fairness preferences (Raihani et al., 2012a,b) despite being known to punish social cheats (e.g. Raihani, Grutter & Bshary, 2010; Raihani et al., 2012a,b). Similarly, although some forms of human teaching might use – or even rely on – theory of mind, examples of teaching in both humans and non-humans exist that do not necessitate such complex cognition (e.g. Thornton & McAuliffe, 2012). For example, a meerkat helper can teach a pup by provisioning it with increasingly live prey items, giving the pup opportunities to learn handling skills. However, this behaviour need not rely on the helper ‘knowing what the pup knows’. Instead, experimental evidence indicates that helpers adjust provisioning behaviour according to the pup’s begging calls which serve as a reliable proxy for age and therefore prey-handling expertise (Thornton & McAuliffe, 2006). These examples highlight the importance of distinguishing between behavioural function, on the one hand, and proximate mechanism on the other (Scott-Phillips, Dickins & West, 2011; Thornton & Raihani, in press). Behaviour with similar evolutionary functions (e.g. promoting learning or cooperation) might occur via markedly different mechanisms – and functions should therefore not be defined according to mechanisms (Thornton & Raihani, in press). Instead, it may be more helpful to ask why – from an ultimate perspective – some species use cognitively complex psychological mechanisms to achieve outcomes where simpler and less demanding mechanisms might often suffice. A related point is that ignoring ecology may lead to dubious inferences about mechanisms underpinning behaviour. For example, as pointed out by McAuliffe & Thornton (••), the Toppan Best-set Premedia Limitedbs_bs_query Journal Code: JZO Proofreader: Mony Article No: JZO12199 Delivery date: 03 Dec 2014 Page Extent: 3 Journal of Zoology. Print ISSN 0952-8369