Denmark

**Author(s):** Matzig, DN; Warnock, R; Marwick, B; Riede, F &lt;br&gt;&lt;br&gt;The use of phylogenetic methods for the of study macroarchaeological processes of cultural evolution has long been advocated for, especially for the study of artefact shape. Due to technical limitations, the use of discrete character traits in these efforts prevails, which comes along with a non-trivial set of challenges. However, recent methodological advances in palaeobiology (Parins-Fukuchi 2017; Zhang, Drummond, and Mendes 2021) allow us now to infer Bayesian phylogenies using continuous characters. In this article, we present an exploratory analysis of cultural evolution in the European Final Palaeolithic and earliest Mesolithic (~15.000-11.000 BP) using a time-scaled Bayesian phylogeny based on the outline shape of lithic tools obtained from legacy data via 2D outline based geometric morphometrics. We estimate the rates of diversification and trait evolution in artefact shape, and test whether the recovered patterns correlate with the climatic and environmental changes within this timeframe. While common in cultural evolutionary analyses of language, the extension of Bayesian phylogenetic methods to archaeology represents a major methodological breakthrough.

Cultural Evolution Society 2022

An analysis of European Late Upper Palaeolithic stone tool evolution using a Bayesian phylogenetic framework

cultural macroevolution; bayesian phylogenies; phylogenetic comparative methods; geometric morphometrics; archaeology; late upper palaeolithic; stone tools

### Welcome to the fourth Cultural Evolution Society Conference, CES2022! 

As last year, Underline will support the online component of the conference. The ongoing COVID-19 pandemic is causing a great deal of uncertainty for all of us. Therefore, we seek to offer a hybrid format that will hopefully allow some folks to travel to Aarhus, others to meet up locally or regionally in CES hub-lets, and other yet to participate online. The field of cultural evolution is growing in size and scope; its theories, methods and applications - both academic and applied - are getting ever more robust. Do join us in what promises to be an exciting exchange of ideas and a celebration of our field.

##### ***For SoMe interaction and live tweeting please tag @CulturalEvolSoc and use #CESconf2022***


#### The CES 2022 local organising committee
* Luseadra Joy McKerracher
* Jesper Sørensen 
* Niels Nørkjær Johannsen 
* Benjamin Grant Purzycki 
* Mette Løvschal 
* Felix Riede
* Iza Romanowska 
* Karsten Olsen

Please click on the button bellow if you want to watch the videos

The field of cultural evolution is growing in size and scope; its theories, methods and applications - both academic and applied - are getting ever more robust. Do join us in what promises to be an exciting exchange of ideas and a celebration of our field.

**Author(s):** Matzig, DN; Warnock, R; Marwick, B; Riede, F The use of phylogenetic methods for the of study macroarchaeological processes of cultural evolution has long been advocated for, especially for the study of artefact shape. Due to technical limitations, the use of discrete character traits in these efforts prevails, which comes along with a non-trivial set of challenges. However, recent methodological advances in palaeobiology (Parins-Fukuchi 2017; Zhang, Drummond, and Mendes 2021) allow us now to infer Bayesian phylogenies using continuous characters. In this article, we present an exploratory analysis of cultural evolution in the European Final Palaeolithic and earliest Mesolithic (~15.000-11.000 BP) using a time-scaled Bayesian phylogeny based on the outline shape of lithic tools obtained from legacy data via 2D outline based geometric morphometrics. We estimate the rates of diversification and trait evolution in artefact shape, and test whether the recovered patterns correlate with the climatic and environmental changes within this timeframe. While common in cultural evolutionary analyses of language, the extension of Bayesian phylogenetic methods to archaeology represents a major methodological breakthrough.

technical paper

**Author(s):** Perez, J; Derex, M; Kandler, A; Fogarty, L As individuals explore a space of potential solutions, they gather many different types of
information. One type of information can be knowledge about how to produce the most efficient
solution they have discovered so far. Another type of information could concern the features that
seem to predict better than average solutions, which can be used to guide the generation of new
solutions. The transmission of the latter, referred to as boundary-rules, is widespread in human
populations. Yet, empirical research in cultural evolution has mainly focused on the former,
assuming that culture accumulates as individuals modify previously discovered solutions. These
two types of social information are likely to have different consequences on how cultural variation
is introduced. We conducted an online experiment with 400 human participants to compare the
effect of transmitting rules versus solutions on the cultural variation introduced by social learners.
We found that social learners who received a rule were more prone to generate new variants than
those who received a solution, and that the variants they generated were less similar to one
another. Receiving a rule also led social learners to be less similar to their demonstrators
compared to receiving a solution. Finally, we found that rules are more often transmitted without
modifications than solutions, but that such modifications have a greater impact on inter-
generational variation. Despite these differences, both transmitting rules and solutions were found
to support cumulative improvement of the productions.

Does the nature of the information received by social learners influence their exploratory behavior ?

**Author(s):** Bourahla Y.; Atencia M.; Euzenat J. Vertical knowledge transmission is expected to enable societies to increase culture fitness over that of a single generation. This has already been demonstrated experimentally for the knowledge that neural network-based agents agents have of their world. However, this approach relies on drastically selecting agents who transmit their knowledge. Here we test if this is necessary. We have designed an agent-based experiment in which agents interact with each other and adapt their knowledge upon interaction failures, so performing horizontal transmission. After interacting for a fixed period the older generation, amounting to half of the agents, dies and the younger half reproduces (without selection). The new generation of agents learns their knowledge from the previous generation without teacher selection (equiprobable chances of knowledge transmission). This showed that knowledge improves across generations with inter-generation transmission. This is in agreement with the previous results. These new results also showed that, when the horizontal transmission period is sufficiently long, there is no need for selecting parents and teachers, for this knowledge to improve. This can be explained by the fact that if culture is widely shared among the population, it will be transmitted without selecting the fittest part of the population.

Inter-generation knowledge transmission without individual selection

**Author(s):** Faessler, L; Lalive, R; Efferson, C Cultural evolution theory implies that social learning can lead cultural groups to differ in the same environment and drive genetic selection. Yet, distinguishing the effect of culture from other confounds is a difficult empirical challenge. To meet this challenge, we exploit the Swiss language border, a cultural border that divides Switzerland without any institutional boundary. Using a regression discontinuity design, we estimate discontinuities in attitudes right at the cultural border. Preliminary results confirm multiple discontinuities related to health and fertility. German speakers are more likely to adopt healthy behaviors and vote against initiatives to share health care costs, while French speakers visit the doctor more often. German speakers report significantly better health but surprisingly do not live longer. Our results provide new evidence that social learning generates cultural variation without environmental variation in health and fertility domains, thus demonstrating the potential for cultural processes to create variation in domains directly related to genetic fitness.

How Culture Affects our Fitness: Evidence from a Swiss Language Border

**Author(s):** Crema, E. R; Shennan, S; Rivas, J; García-Puchol, O. Culture transmission studies have often been applied to investigate the dispersal and adoption of farming and its associated cultural package. The introduction of farming had broader societal consequences, with a profound impact on the demography and the culture of both incumbent and migrant populations, offering a rich research ground where this theoretical field can be explored and further applied.

Archaeologists have been able to reconstruct the population dynamics during this key transitional era by using time frequencies of radiocarbon dates. The radiocarbon record shows an intriguing signature in many European areas: a gap of 200 to 500 years between the last hunter gatherer and first farmers. This gap would either indicate a complete absence of contact between the two communities, or an extremely low population density, insufficient to yield any tangible evidence in the archaeological record. 

In this paper we fit a modified Lotka-Volterra model with population and cultural transmission dynamics to the observed time frequency of radiocarbon dates, and we explore possible explanations for the gap, considering cultural and demographic tendencies. Usually, these gaps have been attributed to environmental dynamics. However, our preliminary findings suggest human interaction and demographic processes could have more relevance than previously thought.

Population dynamics and cultural transmission during the introduction of farming in Europe

**Author(s):** Araujo, A G M; Okumura, M. The archaeological record is the only long-term record of human culture. Therefore, researchers interested in understanding cultural evolutionary patterns of long duration would greatly benefit from studying archaeological materials. The Cultural Evolutionary Theory postulates that one can approach cultural changes or stasis through evolutionary terms borrowed (and then adapted) from Biological Evolutionary Theory. More recently, there has been a debate about how to observe fitness in cultural evolution. We believe that the debate about fitness can not advance properly without an assessment of three different modalities of knowledge that were devised by Greek philosophers: doxa, sophia, and episteme. Doxa constitutes common sense constructed in an unsystematic way and that form the fabric of relatively uninformed idea of how the world operates. Sophia can be understood as wisdom, or informed knowledge which comes with experience and old age. Episteme is “real knowledge”, which we can translate nowadays as scientific knowledge. It is radically different from doxa and at the same time transcends sophia in the sense that it uses the scientific method, rather than personal experience, to build knowledge about the world. Here we argue that each of these modes of knowledge involves different cultural transmission processes and have distinct consequences in terms of cultural fitness. We propose that the majority of examples and processes discussed in the Cultural Evolutionary literature are related to doxa.

An Archaeological Regard on the Concept of Fitness in Cultural Evolution

**Author(s):** Herzog, N.; Clark, M. We investigate the socioecological conditions which favor the combination of individual and social learning that manifest in the invention and innovation of technology. First, using an agent-based simulation, we predict how ecological variation, population structure, and the costs and accuracy of social learning can affect cultural transmission patterns and cultural variability. Then, we compare archaeological case studies in the American Great Basin and Southwest to test how environmental variation and population structure influence patterns of cultural variation, invention, innovation, and investment in new technologies over time.

From mind to matter: Modeling the ecology of social learning and its archaeological outcomes

Introduction: The spatial composition of human portraits obeys historically changing cultural norms. It is also affected by cognitive factors that cause greater spontaneous attention to what is in front rather in the back of an agent. Scenes with more space in front of a directed object are both more often produced and judged as more aesthetically pleasant. This leads to the prediction that, in profile-oriented human portraits, compositions with more space in front of depicted agents (a “forward bias”) should be over-represented. Materials and Methods: We analyzed a large dataset (total N of 1,831 paintings by 582 unique identified European painters from the 15th to the 20th century) collected from WikiArt and for which we measured how the free space was distributed around the sitter. Results: In line with a forward bias, painters tended to put more free space in front of, rather than behind, the sitters. We also found evidence that this forward bias became stronger when cultural norms of spatial composition favoring centering became less stringent. Discussion: This study exemplifies both the importance of cognitive factors in shaping cultural productions and how they interact with explicit cultural norms. I will also discuss how this case study –and others- have implications for using large-scale datasets that have become recently available, other types of methods, and specific frameworks.

A Forward Bias in Human Profile-Oriented Portraits

**Author(s):** Julie De Keyzer Assessing aggregate bundles of cultural replicators in the domain of visual art, we look into measurable proxies of cultural movements, c.q. the European Enlightenment. Here we focus on a metric we expect to be correlated with the introduction of the European Enlightenment, namely a trend in the height of the visible horizon. We hypothesize a diachronic lowering of the horizon line to a balanced ‘observer-objective’ perspective, as a proxy for overthrowing a more chthonic perspective in pre-Modern time. We expect this trend to be more pronounced in depictions of secular themes, and in regions that eased themselves away from Catholicism.
We manually compiled a dataset of 1141 European paintings (1500-1900). Using Growth Curve Analysis, we built a linear mixed-effect regression model, with the proportion of the horizon line as the response variable, and the date of conception, regional provenance, and theme (secular vs. religious) as covariates. Individual artists and museum collection were added as random effects.
Our result indicate a curvilinear effect of time on the height of the horizon, coinciding with the Age of Enlightenment, and a lower horizon in profane and less catholic regions. These results suggest that paintings bear a measurable mark of the intellectual history in Europe.

Cultural evolution in European Paintings

 Cognitive processes play a central role in our understanding of the sophisticated forms of group coordination and culture found in humans. While explicit aspects of processes such as memory, attention, language, metacognition and mentalising often are highlighted, a significant role is carried out by the more implicit processes and internal cognitive biases, guiding how cultural information is perceived, stored, represented, and reproduced to the world. Research has, for instance, studied how language or semiotic communication systems across iterations can evolve to reflect emergent structure, in which e.g. complexity (size and structure) may be reduced and systematicity increased as a result of particular cognitive constraints. In other words, this research has indicated how we can use experimental transmission chain designs and latent cognitive biases to observe the evolution of increasingly meaningful and symbolic perceptions and systems. In this talk, I will present a study using a drawing task situated in a linear social transmission paradigm. The study was originally designed to be implemented online due to the covid19 pandemic. Participants (N=170) were asked to copy 12 different drawings, and rate their trial confidence, in four different conditions which varied in the degree of demand on memory. The resulting drawings were analysed for complexity, error, submitted to repeated rating of conventionality by a set of raters, and the relationships between transmission and complexity, conventionality, and error were analyzed with Bayesian modeling. Moreover, the drawings were analysed qualitatively for low-level perceptual formation and developments, e.g. the Gestalt principles. The results indicated overall that across iterations the drawings decreased in complexity and error and became more conventional as a function of memory demands, while drawings without memory demands remained complex and unconventional. The perception and re-presentation of drawings also revealed a partial role of our low-level perceptual system in terms of preferences for certain compositions of lines following the Gestalt principles. The results are in line with previous research, finding that latent cognitive biases are introducing systematic distortions and guiding emergent structure towards cultural priors and expectations. The findings suggest that more research is needed to carefully dissect the role of implicit processes as the workhorse of human cognition, i.e. perceptual and object recognition processes and latent cognitive biases in memory, in order to understand how they interact with the more explicit and reportable processes of cognition to form cumulative processes of social learning.

Transmission of drawings: the role of memory and latent cognitive biases in a social transmission task

 Multiple iterated language learning studies show that artificial languages evolve over generations towards regularity. This trend has been explained as a reflection of the learners’ biases. We test whether this learning bias for regularity is modulated by culturally acquired knowledge, specifically by familiarity and literacy. We constructed musical languages consisting of signals comprising 6 notes denoting meanings of different shapes and colours. We had three languages, with different levels of structure (irregular, semi-regular and fully regular). In addition, for comparison, we added a fourth language in which signals were spoken words. We trained (1) musically-literate musicians (who can read and write musical notation), (2) musically-illiterate musicians (who are familiar with music, but cannot read and write musical notation) and (3) non-musicians on the four languages, with a 50% bottleneck, meaning they were only exposed to half the meaning-signal pairs. They were then asked to produce signals for the full set of meanings. We measured to what extent they reproduced (a) the original signals and (b) the level of regularity in the original languages. Our results suggest that familiarity helps learn and reproduce the signals of a language, but literacy is required for regularity to be picked up and maintained. This in turn indicates that, by modulating human learning biases, literacy may play a role in the evolution of linguistic structure.

Premium content

An analysis of European Late Upper Palaeolithic stone tool evolution using a Bayesian phylogenetic framework

Downloads

Next from Cultural Evolution Society 2022

Does the nature of the information received by social learners influence their exploratory behavior ?

Similar lecture

Cultural Lineages and Rank Freedom