We are proud to present 15 exciting presentations in poster format:
1. Mark Atkinson1, Mónica Tamariz, Elizabeth Renner, Christine Caldwell, 1University of Stirling
The effect of different task goals and population turnover on cultural accumulation
Experimental studies of cultural evolution can be broadly divided into two categories based on the way the task is framed. In some, participants are instructed to maximise some consistent measure of task success, guided by the output of a preceding participant (e.g. in Caldwell & Millen’s, 2008, study of spaghetti tower building). In others, they are instructed to reproduce the output of a previous participant as accurately as possible (e.g. Kirby, Cornish & Smith’s, 2008, study of artificial language learning). The different task goals have different effects on the accumulating culture: in the former, the outputs typically become more elaborate as participants attempt to outperform their predecessors; in the latter, the outputs typically simplify under a pressure for learnability. We directly compare the effects of these alternative task goals on the cultural evolution of an otherwise similar behaviour: building structures using spaghetti and modelling clay. Beginning single diffusion chains with the same seed structures, participants are instructed to either build as tall a structure as possible, or to copy the previous structure. We also compare social iteration (copying from others’ output) with individual iteration (copying from your own), to assess the effects of population turnover. Together, these manipulations may provide insights into characteristic signatures of ratcheting cumulative culture, through differences in the patterns of change observed across conditions (e.g. in terms of the directionality of trends of change, variation in typical modifications, and between-chain diversity and convergence).
2. Elisa Bandini & Claudio Tennie, University of Birmingham
The role of Individual and social learning in primate tool-use
Primates, especially Great Apes, are considered the most proficient and creative tool-users in the animal kingdom. Despite their extensive behavioural repertoire, the cognitive mechanisms responsible for these behaviours remain largely unknown. Whilst some researchers assume that social learning drives the appearance of tool-use across several individuals, we argue that many of these behaviours may instead be within each individual’s spontaneous capabilities: i.e. they are within a species’ Zone of Latent Solutions (Tennie et al. 2009). Although individual learning may be responsible for the re-emergence of the behaviours across individuals, social learning also plays a role, through processes such as stimulus enhancement, the behaviour spreads quickly throughout a community. However social learning is not required for the behaviour to emerge across individuals. To understand how tool-use arises in both human and non-human primate communities, both theories must be tested. We present the results of the first experiments on the effect of individual and social learning in chimpanzee and macaque tool-use behaviours in order to shed light on the evolution of material culture in the primate lineage.
3. Charlotte Brand1, Gillian R. Brown, Catherine P. Cross, University of St Andrews
Sex differences in risk-taking underpin sex differences in social learning in an adult human population
Understanding when individuals choose to learn socially or asocially is crucial for understanding when and how cumulative culture emerges. A tendency to learn asocially has been found to correlate with traits such as risk-taking and boldness across a broad range of species. We investigated whether sex differences in human risk-taking also result in a sex difference in asocial learning. Male and female participants (n=88) were given a computer-based task that involved constructing a virtual spaceship. Participants were given the option of using social or asocial information to complete the task. In Condition 1, social information was ‘risky,’ in that choosing it could lead to large increases or decreases in the participant’s score. Asocial information was ‘safe,’ in that it guaranteed the participant’s score remained the same, or increased slightly. In Condition 2, social information was ‘safe’ and asocial ‘risky’, and in a control, both options were safe. We modelled the likelihood of choosing asocial information using Bayesian MCMC methods. Women exhibited a strong preference for using social information when asocial information was risky, whereas men’s preference for asocial information remained consistent across conditions. This is the first evidence that sex differences in risk-taking underpin sex differences in human social learning.
4. Jennifer Cook, University of Birmingham
Dopamine and social learning in humans
Much is known of the role of dopamine in “individual learning” – learning from one’s own personal experience of the world. For example, it is known that individual learning varies as a function of naturally occurring genetic variation affecting the dopamine system, and that individual learning can be artificially manipulated using drugs that affect the dopamine system. However, whether dopamine plays the same role in social as in individual learning is a question that remains under-explored. In this set of studies we investigated the effect of a) naturally occurring genetic variation in the dopamine system and, b) artificial changes in dopamine function (via administration of the drug Methylphenidate) on a measure of social learning.
5. Maria Coto-Sarmiento & Simon Carrignon, Barcelona Sumpercomputing Centre
Exploring the dynamic of changes: An Agent Based Model to understand the amphorae production patterns in the Roman Empire.
The goal of this study is to analyze the cultural dynamics among amphora workshops in the Roman Empire. Specifically, we focus on the evolution of the production of olive oil amphorae found in Baetica (currently Andalusia) from 1st to 3rd century AD. In particular, we analyze a set of measures among different kinds of amphorae shapes from different workshops to quantify the dynamic of changes.To achieve this goal multivariable methods was used to classify each amphorae workshops. These methods allow to know if there were differences on the pattern productions among workshops. Specifically we want to identify the origin of these changes and if these changes were produced by cultural reasons depending on the spatial distance and other cultural constraints. As hypothesis, we propose that spatial distribution of pottery workshops is the main influence of the making techniques processes. Therefore we create a simple Agent Based Model using concepts borrowed from Cultural Evolution Studies and Genetic Programming. This model allow us to test different cultural processes of vertical transmission and cultural accumulation; and quantify which one of those process explain better the distributions and patterns revealed during the data analysis. This study aims to better understand the cultural processes acting among the workshop of Baetica during the Roman Empire and explain the nature of the patterns and differences observed.
6. Emmanuel De Oliveira1, François Osiurak, Nicolas Claidière, Jordan Navarro, Mathieu Lesourd & Emanuelle Reynaud, 1EMC Laboratory, Lyon
Physical Intelligence Does Matter to Cumulative Technological Culture
Cumulative technological culture refers to the systematic progression of cultural objects through generations of individuals. The social intelligence hypothesis suggests that this phenomenon is based on our socio-cognitive skills. An alternative hypothesis is that cumulative technological culture also crucially depends on physical intelligence, enabling people to understand and improve the tools made by predecessors. To test these hypotheses, two hundred participants were organized in twenty microsocieties (plus ten more for a control group) and were asked to complete tests to measure their physical and social intelligence. Then, they were asked to build a paper airplane that could fly as far as possible in a micro-society paradigm. We manipulated the way participants could interact between them with an observation condition – they could watch each other, but couldn’t interact verbally – or the teaching condition – participants could help each other to build the best airplane. We demonstrated that learner’s physical intelligence is a stronger predictor of cumulative technological performance than social intelligence, either in the observation condition or in the teaching condition. Moreover, cumulative performance is only slightly influenced by teachers’ physical and social intelligence. To sum up, social intelligence might play a more limited role than physical intelligence in cumulative technological culture.
7. Natalia Dutra, Lynda Boothroyd, and Emma Flynn, Durham University
Do children use social information in a collective social dilemma?
Conformity and pay-off biased strategies are two of the proposed mechanisms that may have stabilised cooperation within human groups (Lamba, 2014). One way of testing this hypothesis is to investigate these behaviours in children. We investigated the adoption of two types of social information in a public goods game by 53 six- and seven-year-old Brazilian children. They were split in groups and played two rounds of a public goods game (PGG1 and PGG2). Each child received five tokens, and had to decide anonymously how many s/he would donate to the group. The total amount was multiplied and distributed equally between the participants. In PGG2, the players had access to two types of information from PGG1, before making their decision: the contributions of both the majority of the group and the player with the best pay-off. Children only differed in the tendency to not conform to the group in their PGG2 contributions when they were already coordinated with majority in PGG1. There was no evidence of the use of social information, especially conformity, contradicting previous findings on children’s conformity to peers, although this can be attributed to the anonymity of contributions and the conflict of interests induced by the game.
8. Carmen Granito, Durham University
The evolution of art styles: How socio-demographical context can shape visual arts
Why do characters in Baroque paintings look more like flesh and blood humans than those in medieval altarpieces? Why are Chinese peripheral “small traditions” simpler than the urban “big traditions”? Why was Renaissance an art boom? Historians and anthropologists of art have provided rich descriptions of how styles change over time and space, but a convincing theoretical framework to explain it has not yet been produced. My project aims to identify some causal factors driving those changes and their effects on art. In particular, following sociolinguistic and cultural evolution studies on language change and graphical communication, I will focus on how socio-demographical factors (population size, degree of contact/isolation, intended audience size) can influence stylistic choices. Experiment 1 will focus on the communicative aspects of art and the role of intercultural contact in driving style towards or away from iconicity/symbolism. Laboratory micro-societies will perform a Pictionary-like task in isolation or contact conditions; I expect drawings to become symbolic in isolation and iconic in the contact condition, based on the idea that talking to strangers elicits referential transparency. The styles of real artworks will then be compared between small isolated (esoteric) communities and large high-contact (exoteric) populations. The expected correlation between stylistic symbolism and social esotericity will be mapped onto population phylogenies to investigate causal relationships. Finally, Experiment 2 will explore the aesthetic and creative aspects of art, and how population structure can influence stylistic complexity and innovation. Participants will perform a decoration task in different social conditions, and innovation rate and graphical complexity will be analysed.
9. Nardie Hanson, University of Birmingham
Arboreal postures elicit hand preference when accessing a hard-to-reach food goal in captive bonobos (Pan paniscus)
Complex arboreal, and in particular, suspensory postures may elicit a preference for the strongest limb to be used in postural support in large bodied primates. However, it has been suggested that for chimpanzees fishing for ants in arboreal postures it is ambilaterality rather than a preference for a particular hand that was selected for (Marchant and McGrew 2007). We recorded hand preference of captive bonobos manipulating a food goal during terrestrial and arboreal postures in a symmetrical environment. When accessing the food goal in the arboreal position the bonobos adopted demanding and predominantly suspensory postures. There was no population level preference for manipulating the goal with either hand in either the terrestrial or arboreal positions. However, four out of seven individuals demonstrated a significant preference when manipulating the goal (two were left handed, two were right handed) in the arboreal position compared to one individual in the terrestrial position. This suggests that in a symmetrical arboreal environment individuals are able to use their preferred or strongest limb for postural support, resulting in a bias for the opposite hand for manipulation. However, the hand that is preferred for postural support differs between individuals. More data is needed for different environments and species to fully understand the influence of the demand of arboreal postural support and environmental complexity on hand preference.
10. Elena Miu, University of St Andrews
Cumulative culture and cultural diversity: Insights for the Social Learning Strategies Tournament II
Human culture is unique in its cumulative aspect – each generation builds on the achievements of others, thus allowing for the accumulation of an astounding amount of culture, which persists in the population for considerably longer in our species. Following Rendell et al. (2010), a computer tournament was organized to investigate which strategy would perform best in a complex, changing environment that incorporated cumulative culture. The strategies could specify how to use social or asocial learning, as well as a third option which allowed them to refine an already known behaviour, thus simulating the incremental improvement characteristic of cumulative culture. We investigated the evolution of culture in this context in terms of the number of known and performed behaviours, the evenness of the distribution of these behaviours in the population, and their persistence in the population over time. Results show that cumulative culture leads to a drastic decrease in cultural amount and evenness, and to a significantly higher persistence as the population converges on the refined behaviours. These results mirror empirical data from programming contests that showed diversity collapsing as populations focused on leading solutions, which raises the question of how cumulative culture in humans might interact with other factors, such as population structure, to lead to its current astonishing diversity
11. Bruce Rawlings, Durham University
Establishing predictors of learning style in humans and chimpanzees: Who are the innovators and who are the social learners?
Both social (learning from others) and asocial (behavioural innovation) learning are crucial learning strategies for many animal species, including humans. Social learning affords rapid and efficient knowledge acquisition, while asocial learning provides directly reliable information about the environment. Importantly, also, both social and asocial learners are essential for the development and diffusion of new cultures within populations. The learning strategies individuals adopt has received much scientific attention; research shows across a range of animal species individuals differ in their propensity to employ social or asocial learning when facing novel problems. Despite this, little is known about what factors may shape these individual differences. Thus the question remains: Who are the innovators and who are the social learners?My PhD project is investigating the extrinsic (social network properties) and intrinsic (personality) factors predict the propensities for innovation, or social acquisition, of solutions to puzzlebox tasks adopted by human children and chimpanzees. This research will reveal key underlying factors behind the origins of human culture, as well as about behavioural innovation and social learning at the individual and population level across humans and chimpanzees. Currently, I have collected personality, social network and learning strategy selection for 282 children. I will present some of the early findings in this poster, where for example, certain personality traits seem to predict whether children preferred to solve the puzzleboxes socially and asocially, and, within social and asocial learners how they went about doing so. I will also outline future analyses directions and present the study outline for the chimpanzee part of the project (commencing summer 2016).
12. Eva Reindl, Sarah R. Beck, Ian A. Apperly, Claudio Tennie, University of Birmingham
Young children copy cumulative culture without action information
We present two studies investigating whether 4- to 6-year-old children are able to copy cumulative culture in a construction task; we also investigate what kind of social information children need to do so. In study 1, 34 children between 4 and 5.5 years were asked to build a construction from sticks and plasticine that was as tall as possible. Children in a baseline condition were simply asked to build something tall. Children in an action demonstration condition had the chance to observe the experimenter build a tower possessing features (in terms of height and shape) going beyond what children in a baseline group reached individually. Results showed that the demonstration group built taller constructions than the baseline group. Crucially, some children in the action demo group copied the complex technique shown, suggesting that 4- to 5-year-old children are able to copy cumulative culture in a technological task. In study 2, we included an endstate-only demonstration. We replicated the findings from study 1 and, crucially, found that young children are also able to copy technical cumulative culture in the absence of action information.
13. Elizabeth Renner, Mark Atkinson, Christine Caldwell, University of Stirling
Social learning, individual exploration and the use of positive and negative information: investigating the mechanisms underlying cumulative culture in children and nonhuman primates
Many human cultural traits could likely only have arisen through repeated episodes of social learning. Such traits have yet to be convincingly demonstrated in nonhumans, but we lack both an explanation for why this would be the case, and an insight into the potential for such traits to develop in other species. It is therefore necessary to investigate how humans use social information compared to other species, to determine how beneficial each finds social learning relative to individual exploration, and how different types of information may have different effects in humans and nonhumans. We present the results of a large-scale developmental study involving 2- to 5-year-old children playing a tablet-based game, in which their ability to acquire a win-stay lose-shift strategy is assessed in a binary discrimination task, with subsequent transfer test involving a three-way discrimination. In addition to age, we consider the effects of social learning compared to individual exploration on task success. Following previous research (involving nonhuman participants, e.g. in starlings, Templeton, 1998), we also compare the effects of different types of information: positive, negative, and mixed. Finally, we introduce our analogous investigation into the effects of these different types of information (social and individual, and positive and negative) carried out in squirrel monkeys.
14. Damian Ruck, University of Bristol
Principal values in two decades of global cultural change
Cultural values in a society evolve with time. But what factors influence the cultural trajectory of a society? This study aims to determine which factors explain the patterns of global cultural change in the last 25 years. Including to what extent historical factors play a role in cultural change? Using the World value Survey (WVS), we extract principal components of variation in the WVS data without a-priori assumptions based upon the survey design about categories or country-level norms. Using a Bayesian hierarchical framework with horse shoe priors we are able to handle the abundance of missing data and create a simple explanatory model of observed cultural change using socioeconomic, political and historical covariates. Using these methods we can ask naïve questions about the data, such as:1)How does cultural change vary between countries?2)Which country level covariates correlate with change in cultural values (e.g. wealth, inequality, historic religious environment)So by identifying important correlations in the data we can make calibrated predictions even for countries that are not well sampled in the WVS. Therefore, we can make a tentative forecast regarding the results of the upcoming 7th wave of the WVS.
16. Ramiro M. Joly-Mascheroni, City University London
Beyond Humans: Contagious Yawning in Primates elicited by a non-human agent: an android