Success-biased social learning: Cultural and evolutionary dynamics

Success bias is a social learning strategy whereby learners tend to acquire the cultural variants of successful individuals. I develop a general model of success-biased social learning for discrete cultural traits with stochastic payoffs, and investigate its dynamics when only two variants are present. I find that success bias inherently favors rare variants, and consequently performs worse than unbiased imitation (i.e. random copying) when success payoffs are at least mildly stochastic and the optimal variant is common. Because of this weakness, success bias fails to replace unbiased imitation in an evolutionary model when selection is fairly weak or when the environment is relatively stable, and sometimes fails to invade at all. I briefly discuss the optimal strength of success bias, the complicated nature of defining success in social learning contexts, and the value of variant frequency as an important source of information to social learners. I conclude with predictions regarding the prevalence of success bias in different behavioral domains.     

An experimental comparison of human social learning strategies: payoff-biased social learning is adaptive but underused

Analytical models have identified a set of social learning strategies that are predicted to be adaptive relative to individual (asocial) learning. In the present study, human participants engaged in an ecologically valid artifact-design task with the opportunity to engage in a range of social learning strategies: payoff bias, conformity, averaging and random copying. The artifact (an arrowhead) was composed of multiple continuous and discrete attributes which jointly generated a complex multimodal adaptive landscape that likely reflects actual cultural fitness environments. Participants exhibited a mix of individual learning and payoff-biased social learning, with negligible frequencies of the other social learning strategies. This preference for payoff-biased social learning was evident from the initial trials, suggesting that participants came into the study with an intrinsic preference for this strategy. There was also a small but significant increase in the frequency of payoff-biased social learning over sessions, suggesting that strategy choice may itself be subject to learning. Frequency of payoff-biased social learning predicted both absolute and relative success in the task, especially in a multimodal (rather than unimodal) fitness environment. This effect was driven by a minority of hardcore social learners who copied the best group member on more than half of trials. These hardcore social learners were also above-average individual learners, suggesting a link between individual and social learning ability. The lower-than-expected frequency of social learning may reflect the existence of information producer–scrounger dynamics in human populations.     

Evolution of social learning: a mathematical analysis

Social learning is an important ability seen in a wide range of animals including humans. It has been argued that individual learning, social learning, and innate determination of behavior are favored by natural selection when environmental changes occur at short, intermediate, and long intervals, respectively. Only recently, however, has the hypothesis been examined by means of mathematical models. In this paper, we construct a simple model in which each organism uses one of three genetically determined strategies--it is an individual learner, a social learner or an "innate"--and the three types of organisms are in direct competition with each other. A reduced model, involving only the individual learners and innates, is effectively linear, and we show that by solving the eigenvalue problem of this reduced system we arrive at a good approximation to the global dynamics of the full model. We also study the effect of stochastic environmental changes and reversible mutations among the three strategies. Our results are consistent with the predictions of previous studies. In addition, we identify a critical level of environmental constancy below which only individual and social learners are present.     

The evolution of social learning rules: Payoff-biased and frequency-dependent biased transmission

Humans and other animals do not use social learning indiscriminately, rather, natural selection has favoured the evolution of social learning rules that make selective use of social learning to acquire relevant information in a changing environment. We present a gene-culture coevolutionary analysis of a small selection of such rules (unbiased social learning, payoff-biased social learning and frequency-dependent biased social learning, including conformism and anti-conformism) in a population of asocial learners where the environment is subject to a constant probability of change to a novel state. We define conditions under which each rule evolves to a genetically polymorphic equilibrium. We find that payoff-biased social learning may evolve under high levels of environmental variation if the fitness benefit associated with the acquired behaviour is either high or low but not of intermediate value. In contrast, both conformist and anti-conformist biases can become fixed when environment variation is low, whereupon the mean fitness in the population is higher than for a population of asocial learners. Our examination of the population dynamics reveals stable limit cycles under conformist and anti-conformist biases and some highly complex dynamics including chaos. Anti-conformists can out-compete conformists when conditions favour a low equilibrium frequency of the learned behaviour. We conclude that evolution, punctuated by the repeated successful invasion of different social learning rules, should continuously favour a reduction in the equilibrium frequency of asocial learning, and propose that, among competing social learning rules, the dominant rule will be the one that can persist with the lowest frequency of asocial learning.     

The evolutionary basis of human social learning

Humans are characterized by an extreme dependence on culturally transmitted information. Such dependence requires the complex integration of social and asocial information to generate effective learning and decision making. Recent formal theory predicts that natural selection should favour adaptive learning strategies, but relevant empirical work is scarce and rarely examines multiple strategies or tasks. We tested nine hypotheses derived from theoretical models, running a series of experiments investigating factors affecting when and how humans use social information, and whether such behaviour is adaptive, across several computer-based tasks. The number of demonstrators, consensus among demonstrators, confidence of subjects, task difficulty, number of sessions, cost of asocial learning, subject performance and demonstrator performance all influenced subjects' use of social information, and did so adaptively. Our analysis provides strong support for the hypothesis that human social learning is regulated by adaptive learning rules.     

Two success-biased social learning strategies

I compare the evolutionary dynamics of two success-biased social learning strategies, which, by definition, use the success of others to inform one’s social learning decisions. The first, “Compare Means”, causes a learner to adopt cultural variants with highest mean payoff in her sample. The second, “Imitate the Best”, causes a learner to imitate the single most successful individual in her sample. I summarize conditions under which each strategy performs well or poorly, and investigate their evolution via a gene-culture coevolutionary model. Despite the adaptive appeal of these strategies, both encounter conditions under which they systematically perform worse than simply imitating at random. Compare Means performs worst when the optimal cultural variant is usually at high frequency, while Imitate the Best performs worst when suboptimal variants sometimes produce high payoffs. The extent to which it is optimal to use success-biased social learning depends strongly on the payoff distributions and environmental conditions that human social learners face.     

Network-based diffusion analysis: a new method for detecting social learning

Social learning has been documented in a wide diversity of animals. In free-living animals, however, it has been difficult to discern whether animals learn socially by observing other group members or asocially by acquiring a new behaviour independently. We addressed this challenge by developing network-based diffusion analysis (NBDA), which analyses the spread of traits through animal groups and takes into account that social network structure directs social learning opportunities. NBDA fits agent-based models of social and asocial learning to the observed data using maximum-likelihood estimation. The underlying learning mechanism can then be identified using model selection based on the Akaike information criterion. We tested our method with artificially created learning data that are based on a real-world co-feeding network of macaques. NBDA is better able to discriminate between social and asocial learning in comparison with diffusion curve analysis, the main method that was previously applied in this context. NBDA thus offers a new, more reliable statistical test of learning mechanisms. In addition, it can be used to address a wide range of questions related to social learning, such as identifying behavioural strategies used by animals when deciding whom to copy.     

The effect of friend selection on social influences in obesity

We present an agent-based model of weight choice and peer selection that simulates the effect of peer selection on social multipliers for weight loss interventions. The model generates social clustering around weight through two mechanisms: a causal link from others’ weight to an individual's weight and the propensity to select peers based on weight. We simulated weight loss interventions and tried to identify intervention targets that maximized the spillover of weight loss from intervention participants to nonparticipants. Social multipliers increase with the number of intervention participants’ friends. For example, when friend selection was based on a variable exogenous to weight, the weight lost among non-participants increased by 23% (14.3 lb vs. 11.6 lb) when targeting the most popular obese. Holding constant the number of participants’ friends, multipliers increase with increased weight clustering due to selection, up to a point. For example, among the most popular obese, social multipliers when matching on a characteristic correlated with weight (1.189) were higher than when matching on the exogenous characteristic (1.168) and when matching on weight (1.180). Increased weight clustering also implies more obese “friends of friends” of participants, who reduce social multipliers.     

Age-dependent social learning in a lizard

Evidence of social learning, whereby the actions of an animal facilitate the acquisition of new information by another, is taxonomically biased towards mammals, especially primates, and birds. However, social learning need not be limited to group-living animals because species with less interaction can still benefit from learning about potential predators, food sources, rivals and mates. We trained male skinks (Eulamprus quoyii), a mostly solitary lizard from eastern Australia, in a two-step foraging task. Lizards belonging to ‘young’ and ‘old’ age classes were presented with a novel instrumental task (displacing a lid) and an association task (reward under blue lid). We did not find evidence for age-dependent learning of the instrumental task; however, young males in the presence of a demonstrator learnt the association task faster than young males without a demonstrator, whereas old males in both treatments had similar success rates. We present the first evidence of age-dependent social learning in a lizard and suggest that the use of social information for learning may be more widespread than previously believed.     

Individual differences in learning behaviours in humans: Asocial exploration tendency does not predict reliance on social learning

A number of empirical studies have suggested that individual differences in asocial exploration tendencies in animals may be related to those in social information use. However, because the ‘exploration tendency’ in most previous studies has been measured without considering the information-gathering processes, it is yet hard to conclude that the animal asocial exploration strategies may be tied to social information use. Here, we studied human learning behaviour in both asocial and social two-armed bandit tasks. By fitting reinforcement learning models including asocial and/or social decision processes, we measured each individual's (1) asocial exploration tendency and (2) social information use. We found consistent individual differences in the exploration tendency in the asocial tasks. We also found substantive heterogeneity in the adopted learning strategies in the social task: Nearly one-third of participants used predominantly the copy-when-uncertain strategy, while the remaining two-thirds were most likely to have relied only on asocial learning. However, we found no significant individual association between the exploration frequency in the asocial task and the use of the social information in the social task. Our results suggest that the social learning strategies may be independent from the asocial exploration strategies in humans.     

Selectivity in social and asocial learning: investigating the prevalence,effect and development of young children's learning preferences

Culture evolution requires both modification and faithful replication of behaviour, thus it is essential to understand how individuals choose between social and asocial learning. In a quasi-experimental design, 3- and 5-year-olds (176), and adults (52) were presented individually with two novel artificial fruits, and told of the apparatus'' relative difficulty (easy versus hard). Participants were asked if they wanted to attempt the task themselves or watch an experimenter attempt it first; and then had their preference either met or violated. A significant proportion of children and adults (74%) chose to learn socially. For children, this request was efficient, as observing a demonstration made them significantly quicker at the task than learning asocially. However, for 5-year-olds, children who selected asocial learning were also found to be highly efficient at the task, showing that by 5 years children are selective in choosing a learning strategy that is effective for them. Adults further evidenced this trend, and also showed selectivity based on task difficulty. This is the first study to examine the rates, performance outcomes and developmental trajectory of preferences in asocial and social learning, ultimately informing our understanding of innovation.     

A social learning model of conflict and cooperation in human societies

Patterns of conflict and cooperation both within and between societies may be related to the degree of cultural similarity within and between the same societies. A simple model of social learning is used to predict patterns of conflict and cooperation in hypothetical societies that differ in the roles of relatives and nonrelatives in the enculturation of children. The model is illustrated by comparing its predictions to known differences in the patterns of conflict between males inpatrilocal and matrilocal societies.     

Comparison between perfect information and passive–adaptive social learning models of forest harvesting

This paper compares perfect information and passive–adaptive social learning models of forest harvesting using a simple Markov chain model for land-use dynamics. A perfect information model assumes that landowners know true utility values of forest conservation and harvesting. In contrast, a passive–adaptive social learning model assumes that landowners do not know true utility values and they learn these values by their past experiences and by exchanging information with others in a society. We determine conditions under which the same consequences expected from perfect information and passive–adaptive social learning models. We found that the outcome from a perfect information model resembles that from passive–adaptive social learning model only when the perfect information model incorporates little discounting for future values. The stability analysis of landscape dynamics predicts a cyclic overexploitation of forest resources in a passive–adaptive social learning model with short-term memory, while instability of landscapes is never expected in a perfect information model. We discuss the role of discounting the future and discounting the past in the context of forest management.     

Training vervet monkeys to avoid electric wires: Is there evidence for social learning?

Before being released into a large park enclosed by an electric fence, a wild‐caught vervet group (Chlorocebus aethiops) had to learn to avoid electrified wires in a smaller cage. During this training, we observed the group continuously for 12 consecutive days to investigate if social learning was involved in the learning process. Results showed that all monkeys received an electric shock (average=2.5 shocks/individual). Most contacts with the wires occurred during the first few days of training and the vervets were never observed to come into contact with the electric fence in the 18 months after their release into the large park. This suggests that the vervets learned to avoid the electrified wires by trial‐and‐error learning. It is possible that local and stimulus enhancement may have played a role, but we could not carry out the necessary control experiments to quantify the role of these components. Observational conditioning of fear can be ruled out, however, because the vervets did not show fearful behavior toward the wires. Zoo Biol 24:145–151, 2005. © 2005 Wiley‐Liss, Inc.     

Applying evolutionary models to the laboratory study of social learning

Cultural evolution is driven, in part, by the strategies that individuals employ to acquire behavior from others. These strategies themselves are partly products of natural selection, making the study of social learning an inherently Darwinian project. Formal models of the evolution of social learning suggest that reliance on social learning should increase with task difficulty and decrease with the probability of environmental change. These models also make predictions about how individuals integrate information from multiple peers. We present the results of microsociety experiments designed to evaluate these predictions. The first experiment measures baseline individual learning strategy in a two-armed bandit environment with variation in task difficulty and temporal fluctuation in the payoffs of the options. Our second experiment addresses how people in the same environment use minimal social information from a single peer. Our third experiment expands on the second by allowing access to the behavior of several other individuals, permitting frequency-dependent strategies like conformity. In each of these experiments, we vary task difficulty and environmental fluctuation. We present several candidate strategies and compute the expected payoffs to each in our experimental environment. We then fit to the data the different models of the use of social information and identify the best-fitting model via model comparison techniques. We find substantial evidence of both conformist and nonconformist social learning and compare our results to theoretical expectations.