A vast amount of empirical and theoretical research on public good games indicates that the threat of punishment can curb free-riding in human groups engaged in joint enterprises. Since punishment is often costly, however, this raises an issue of second-order free-riding: indeed, the sanctioning system itself is a common good which can be exploited. Most investigations, so far, considered peer punishment: players could impose fines on those who exploited them, at a cost to themselves. Only a minority considered so-called pool punishment. In this scenario, players contribute to a punishment pool before engaging in the joint enterprise, and without knowing who the free-riders will be. Theoretical investigations (Sigmund et al., Nature 466:861–863, 2010) have shown that peer punishment is more efficient, but pool punishment more stable. Social learning, i.e., the preferential imitation of successful strategies, should lead to pool punishment if sanctions are also imposed on second-order free-riders, but to peer punishment if they are not. Here we describe an economic experiment (the Mutual Aid game) which tests this prediction. We find that pool punishment only emerges if second-order free riders are punished, but that peer punishment is more stable than expected. Basically, our experiment shows that social learning can lead to a spontaneously emerging social contract, based on a sanctioning institution to overcome the free rider problem.

We report an experiment in which subjects may learn from each other. Specifically, a “queue” of players who are identically informed ex ante make decisions in sequence over two lotteries. Every player except the first in the queue observes (only) his immediate predecessor’s choice and payoff before making his own decision. In equilibrium decisions are identical from the first or second player onwards in all experimental conditions. However, complete adherence to equilibrium play is seldom observed in our experiment. We further analyze our data using a quantal response equilibrium approach and test for behavioral regularities related to base rate fallacy/conservatism bias, social conformity/rebelliousness, and preference for experimentation (preferring the lottery with potentially more information spillover value). Our estimations reveal a consistent preference for experimentation across conditions, and further analysis offers some support to our surmise that this behavioral regularity is due, in part, to an attempt to influence others behind in the queue.

Does social learning and subsequent private information processing differ depending on whether the observer shares the same group identity as the predecessor whose action is observed? In this paper, we conduct a lab experiment to answer this question, in which subjects first observe a social signal and then receive a private signal. We find that subjects put greater weights on the social signal if they share with the predecessor the same group identity that is induced in the experimental environment. We also provide suggestive evidence that such an ingroup-outgroup difference cannot be explained by individuals’ beliefs of the predecessor’s rationality. Moreover, heterogeneous effects of group identity exist in weights given to the subsequent private signal: Compared to when the predecessor is an outgroup, those who have learned from an ingroup predecessor put a greater (smaller) weight on the private signal if it contradicts (confirms) the social signal. We conjecture that such group effects are consistent with the perspective that group identity works as a framing device and brings about certain decision heuristics in the social signal phase, which no longer exist in the private signal phase.

Striking a balance between individual and social learning is one of the key capabilities that support adaptation under uncertainty. Although intergenerational transmission of information is ubiquitous, little is known about when and how newcomers switch from learning loyally from preceding models to exploring independently. Using a behavioural experiment, we investigated how social information available from a preceding demonstrator affects the timing of becoming independent and individual performance thereafter. Participants worked on a 30-armed bandit task for 100 trials. For the first 15 trials, participants simply observed the choices of a demonstrator who had accumulated more knowledge about the environment and passively received rewards from the demonstrator's choices. Thereafter, participants could switch to making independent choices at any time. We had three conditions differing in the social information available from the demonstrator: choice only, reward only or both. Results showed that both participants’ strategies about when to stop observational learning and their behavioural patterns after independence depended on the available social information. Participants generally failed to make the best use of previously observed social information in their subsequent independent choices, suggesting the importance of direct communication beyond passive observation for better intergenerational transmission under uncertainty. Implications for cultural evolution are discussed.

We use a revealed preference approach to disentangle conformity, an intrinsic taste to follow others, from information-driven herding. We provide observations from a series of sequential decision-making experiments in which subjects choose the type of information they observe before making their decision. Namely, subjects choose between observing a private (statistically informative) signal or the history of play of predecessors who have not chosen a private signal (i.e., a statistically uninformative word-of-mouth signal). In our setup, subjects choose the statistically uninformative social signal $34\%$ of the time and, of those, $88\%$ follow their observed predecessors’ actions. When allowing for payoff externalities by paying subjects according to the collective action chosen by majority rule, the results are amplifed and the social signal is chosen in $51\%$ of all cases, and $59\%$ of those who pick the social signal follow the majority choice. The results from the majority treatment demonstrate that conformist behavior is not driven by inequality aversion, nor by strategic voting behavior in which voters balance others who are uninformed. Raising the stakes five-fold does not eliminate conformist behavior; in both treatments, the social signal is chosen nearly $50\%$ of the time. Individual level analysis yields the identification of rules of thumb subjects use in making their decisions.

An increasingly common phenomenon in modern work and school settings is individuals taking on too many tasks and spending effort without commensurate rewards. Such an imbalance of efforts and rewards leads to myriad negative consequences, such as burnout, anxiety and disease. Here, we develop a model to explain how such effort–reward imbalances can come about as a result of biased social learning dynamics. Our model is based on a phenomenon that on some US college campuses is called ‘the floating duck syndrome’. This phrase refers to the social pressure on individuals to advertise their successes but hide the struggles and the effort put in to achieve them. We show that a bias against revealing the true effort results in social learning dynamics that lead others to underestimate the difficulty of the world. This in turn leads individuals to both invest too much total effort and spread this effort over too many activities, reducing the success rate from each activity and creating effort–reward imbalances. We also consider potential ways to counteract the floating duck effect: we find that solutions other than addressing the root cause, biased observation of effort, are unlikely to work.

While humans are highly cooperative, they can also behave spitefully. Yet spite remains understudied. Spite can be normatively driven and while previous experiments have found some evidence that cooperation and punishment may spread via social learning, no experiments have considered the social transmission of spiteful behaviour. Here we present an online experiment where, following an opportunity to earn wealth, we asked participants to choose an action towards an anonymous partner across a full spectrum of social behaviour, from spite to altruism. In accordance with cultural evolutionary theory, participants were presented with social information that varied in source and content. Across six conditions, we informed participants that either the majority or the highest earner had chosen to behave spitefully, neutrally or altruistically. We found an overall tendency towards altruism, but at lower levels among those exposed to spite compared with altruism. We found no difference between social information that came from the majority or the highest earner. Exploratory analysis revealed that participants’ earnings negatively correlated with altruistic behaviour. Our results contrast with previous literature that report high rates of spite in experimental samples and a greater propensity for individuals to copy successful individuals over the majority.

Results from cultural evolutionary theory often suggest that social learning can lead cultural groups to differ markedly in the same environment. Put differently, cultural evolutionary processes can in principle stabilise behavioural differences between groups, which in turn could lead selection pressures to vary across cultural groups. Separating the effects of culture from other confounds, however, is often a daunting and sometimes intractable challenge for the working empiricist. To meet this challenge, we exploit a cultural border dividing Switzerland in ways that are independent of institutional, environmental and genetic variation. Using a regression discontinuity design, we estimate discontinuities at the border in terms of preferences related to fertility and mortality, the two basic components of genetic fitness. We specifically select six referenda related to health and fertility and analyse differences in the proportion of yes votes across municipalities on the two sides of the border. Our results show multiple discontinuities and thus indicate a potential role of culture in shaping stable differences between groups in preferences and choices related to individual health and fertility. These findings further suggest that at least one of the two groups, in order to uphold its cultural values, has supported policies that could impose fitness costs on individuals relative to the alternative policy under consideration.

Knowledge and behaviour are transmitted from one individual to another through social learning and eventually disseminated across the population. People often learn useful behaviours socially through selective bias rather than random selection of targets. Prestige bias, or the tendency to selectively imitate prestigious individuals, has been considered an important factor in influencing human behaviour. Although its importance in human society and culture has been recognised, the formulation of prestige bias is less developed than that of other social learning biases. To examine the effects of prestige bias on cultural evolution theoretically, it is imperative to formulate prestige and investigate its basic properties. We reviewed two definitions: one based on first-order cues, such as the demonstrator's appearance and job title, and the other based on second-order cues, such as people's behaviour towards the demonstrator (e.g. people increasingly pay attention to prestigious individuals). This study builds a computational model of prestige bias based on these two definitions and compares the cultural evolutionary dynamics they generate. Our models demonstrate the importance of distinguishing between the two types of formalisation, because they can have different influences on cultural evolution.

Gender role ideology, i.e. beliefs about how genders should behave, is shaped by social learning. Accordingly, if perceptions about the beliefs of others are inaccurate this may impact trajectories of cultural change. Consistent with this premise, recent studies report evidence of a tendency to overestimate peer support for inequitable gender norms, especially among men, and that correcting apparent ‘norm misperception’ promotes transitions to relatively egalitarian beliefs. However, supporting evidence largely relies on self-report measures vulnerable to social desirability bias. Consequently, observed patterns may reflect researcher measurement error rather than participant misperception. Addressing this shortcoming, we examine men's gender role ideology using both conventional self-reported and a novel wife-reported measure of men's beliefs in an urbanising community in Tanzania. We confirm that participants overestimate peer support for gender inequity. However, the latter measure, which we argue more accurately captures men's true beliefs, implies that this tendency is relatively modest in magnitude and scope. Overestimation was most pronounced among men holding relatively inequitable beliefs, consistent with misperception of peer beliefs reinforcing inequitable norms. Furthermore, older and poorly educated men overestimated peer support for gender inequity the most, suggesting that outdated and limited social information contribute to norm misperception in this context.

An essential feature of human progress is the use of different modes of learning so agents acquire the appropriate behaviour to survive in a changing environment. Learning may result from agents who discover new knowledge on their own (individual learning), or imitate the behaviour of others (social learning). Social learning is less costly than discovery, but imitation might yield no benefit. Early theoretical models of a population consisting of purely individual and purely social learners found that both types are present in an evolutionary equilibrium. However, the presence of social learners did not provide any improvement to the average population fitness. Subsequent research showed that the presence of social learners could improve the average population fitness, provided that the pure characterisation of the agents’ learning is relaxed. We return to the pure conceptualisation of agents to challenge an assumption in the early work: agents were guaranteed enough resources to perform their desired learning. We show that, if the resources an agent receives are uncertain, this turns social learning into a source of fitness improvement at the population level. Perhaps counter-intuitively, uncertain provision of resources prompts an increase in the proportion of the population that pursues the costlier individual learning activity in equilibrium.

Humans have adapted to an immense array of environments by accumulating culturally transmitted knowledge and skills. Adaptive culture can accumulate either via more distinct cultural traits or via improvements of existing cultural traits. The kind of culture that accumulates depends on, and coevolves with, the social structure of societies. Here, we show that the coevolution of learning networks and cumulative culture results in two distinct pathways to cultural adaptation: highly connected populations with high proficiency but low trait diversity vs. sparsely connected populations with low proficiency but higher trait diversity. Importantly, we show there is a conflict between group-level payoffs, which are maximised in highly connected groups that attain high proficiency, and individual level selection, which favours disconnection. This conflict emerges from the interaction of social learning with population structure and causes populations to cycle between the two cultural and network states. The same conflict creates a paradox where increasing innovation rate lowers group payoffs. Finally, we explore how populations navigate these two pathways in environments where payoffs differ among traits and can change over time, showing that high proficiency is favoured when payoffs are stable and vary strongly between traits, while frequent changes in trait payoffs favour more trait diversity. Our results illustrate the complex interplay between networks, learning and the environment, and so inform our understanding of human social evolution.

For at least three million years, knapping stone has been practiced by hominin societies large and small, past and present. Thus, understanding knapping, knappers, and knapping cultures is fundamental to anthropological research around the world. Although there is a general sense that stone knapping is inherently dangerous and can lead to injury, little is formally, specifically, or systematically known about the frequency, location, or severity of knapping injuries. Toward this end, we conducted a 31-question survey of modern knappers to better understand knapping risks. Responses from 173 survey participants suggest that knapping injuries are a real and persistent hazard, even though a majority of modern knappers use personal protective equipment. A variety of injuries (lacerations, punctures, aches, etc.) can occur on nearly any part of the body. The severity of injury sustained by some of our participants is shocking, and nearly one-quarter of respondents reported having sought or received professional medical attention for a flintknapping-related injury. Overall, the results of this survey suggest that there would have likely been serious, even fatal, costs to knappers in past societies. Such costs may have encouraged the deployment of any social learning capacities possessed by hominins or delayed the learning or exposure of young infants or children to knapping.