Japan

In order to adaptively solve complex problems or make difficult decisions, people must strategically combine personal information acquired directly from experience (individual learning) and social information acquired from others (social learning). The game of football (soccer) provides extensive real world data with which to quantify this strategic information use. I analyse a 5-year dataset of all games (n = 9127, 2012–2017) in five top European leagues to quantify the extent to which a manager’s initial formation is guided by their personal past use or success with that formation, or other managers’ use or success with that formation. I focus on the 4231 formation, the dominant formation during this period. As predicted, a manager’s choice of whether to use 4231 is influenced by both their recent use of 4231 (personal information) and the use of 4231 in the entire population of managers in that division (social information). Against expectations, managers relied more on personal than social information, although this estimate was highly variable across managers and divisions. Finally, there did not appear to be an adaptive tradeoff between social and personal information use, with the relative reliance on each failing to predict managerial success.

Cultural Evolution Society 2021

Cultural evolution of football tactics: strategic social learning in managers’ choice of formation

### Conference Welcome

Welcome to the third Cultural Evolution Society Conference, CES2021. In the conference, which will be held from June 9-11, 2021, about 180 researchers from East Asia, Oceania, Middle East, Europe, North and South America give presentations on their latest research advancing the evolutionary understanding of human and nonhuman animal culture.

### Schedule and Program

[![](https://s3.eu-west-1.amazonaws.com/underline.prod/uploads/markdown_image/1/image/1ef06ffd05f80957b4a5b685f63ee3ba.png)](https://www.chain.hokudai.ac.jp/CES2020/wp-content/uploads/2021/06/fullprogram.pdf)
[![](https://s3.eu-west-1.amazonaws.com/underline.prod/uploads/markdown_image/1/image/12285eb22580949635121d15d5b46325.png)](https://www.chain.hokudai.ac.jp/CES2020/wp-content/uploads/2021/06/abstracts.pdf)

**Program-at-glance**

![](https://s3.eu-west-1.amazonaws.com/underline.prod/uploads/markdown_image/1/image/c0d0d4c8fcdad49d12631574fd966ae8.png)

### About the conference 
 
**Help-desk:** 
Please first visit [Help-desk](https://underline.io/events/120/help-desk) where you can find the instructions and manuals for exploring the conference site. 

**Plenary sessions:**
Five leading researchers give a live talk. It is like a TV show. The audience can submit a question from the chat window. Chat messages are immediately delivered to speakers and moderators.

**Q&A sessions (parallel):**
Each presenter first gives a 2-minute live lightning talk followed by an 8-minute live discussion with the audience. It uses Zoom.

**Discussion sessions:** 
The audience can approach presenters and discuss their research. The sessions are held in a dedicated room in Gathertown.
 
**Gathertown:**
It is a platform dedicated to live virtual interactions.
Gather Town is available for all participants 24 hours during the conference.
You can meet and talk with anybody at a lounge or at a table in Gathertown.
Further Gather Town instruction and how to use short keys please find in our [Help desk page](https://underline.io/events/120/help-desk). 

**Videos and slides:**

* Plenary sessions and the Q&A sessions will be recorded and uploaded after the sessions.
* All the videos and slides are available for 60 days after the conference but only to registered ones. 

**Text-based chatting with the speakers:** 
You can chat with any speaker in the Q&A and discussion sessions. A chat window is available on each speaker’s page, and you can leave your questions there. 

**Announcement to the other participants:**
If you want to make an announcement to the other participants (e.g., “Hey, let’s meet at the lounge in Gathertown from 9pm CEST!), please use twitter (#CESConf2021).



### Code of conduct

See the full details of [the code of conduct](https://www.chain.hokudai.ac.jp/CES2020/code-of-conduct/).
&nbsp; 
 CES is dedicated to providing a welcoming and harassment-free conference experience for everyone, regardless of gender, gender identity and expression, sexual orientation, disability, physical appearance, body size, race, age, religion, nationality, or the presence of or care for dependents. 
We do not tolerate harassment of conference participants in any form. Conference participants violating these rules may be sanctioned or expelled from the conference at the discretion of the conference organizers.

&nbsp;
**Harassment can occur in-person or virtually, and includes, but is not limited to:**
* Comments (verbal, text, etc.) that reinforce social structures of domination related to gender, gender identity and expression, sexual orientation, disability, physical appearance, body size, race, age, religion, nationality
* Sexual images in public spaces (physical or virtual)
* Deliberate intimidation, stalking, or following
* Harassing photography or recording (including sharing of presentation materials without consent)
* Sustained disruption of talks or other events (including “Zoombombing”)
* Inappropriate physical contact or unwelcome sexual attention
* Advocating for, or encouraging, any of the above behaviour 

&nbsp;
 **Inclusiveness Committee**:
&nbsp;
 To facilitate implementing the code of conduct and foster a climate of inclusion, we have developed an inclusiveness committee. Below is their contact information:
 **Hiromi Matsumae**:	matsumae.hiromi.g@tokai.ac.jp
 **Patrick Savage**:	psavage@sfc.keio.ac.jp
 **Sarah Mathew**:	Sarah.Mathew@asu.edu
 **Gmail:** inclusivenessCES2021@gmail.com
 See [more details about reporting the harassment.](https://www.chain.hokudai.ac.jp/CES2020/code-of-conduct/)

### Organizing committee

**Honorary Conference Organizer:** [Kenichi Aoki (Meiji University)](https://sites.google.com/site/kenichi1aoki/home) 

**Chair:** [Masanori Takezawa (Hokkaido University)](https://lalalarakko.github.io)

**Scientific Committee:** [Kenichi Aoki (Meiji University)](https://sites.google.com/site/kenichi1aoki/home), [Yasuo Ihara (the University of Tokyo)](https://researchmap.jp/read0209180?lang=en), [Hiromi Matsumae (Tokai University)](https://sites.google.com/site/hiromimatsumae/eng), [Naoko Matsumoto (Okayama University)](https://researchmap.jp/nmatsumoto_okayama-u), [Masako Myowa (Kyoto University)](https://myowa.educ.kyoto-u.ac.jp/en), [Hisashi Nakao (Nanzan University)](https://www.hisashinakao.com/), [Patrick E. Savage (Keio University)](https://sites.google.com/view/comp-music-lab/home)

**Local Organizing Committee:** [Yutaka Horita (Teikyo University)](https://sites.google.com/site/yhoritapage/), [Masanori Takezawa (Hokkaido University)](https://lalalarakko.github.io/), [Ayaka Takimoto (Hokkaido University)](https://lynx.let.hokudai.ac.jp/~takimoto/), [Kohei Tamura (Tohoku University)](https://researchmap.jp/7000017854)

**Inclusiveness Committee:** [Sarah Mathew (Arizona State University)](https://www.sarahmathew.net/), [Hiromi Matsumae (Tokai University)](https://sites.google.com/site/hiromimatsumae/eng), [Patrick E. Savage (Keio University)](https://sites.google.com/view/comp-music-lab/home)


![](https://s3.eu-west-1.amazonaws.com/underline.prod/uploads/markdown_image/1/image/967110a00d200a6b20d6f4fa6556ba13.png)

For attending the conference, you first need to register and pay fees from [the link here](https://www.chain.hokudai.ac.jp/CES2020/registration/). After the payment of the registration fees, you will receive a ticket, which is necessary for accessing the conference site. After the payment of the registration fees, you will receive a ticket, which is necessary for accessing the conference site

Registration fee

technical paper

We conducted the largest multiple-iteration retelling study to date with two different studies that 1) test how happiness and sadness are transmitted across retellings, using participant-generated stories and 2) make an exploratory comparison between the transmission of happiness and sadness and disgust, risk, embarrassment, using researcher-generated stories. We recruited 12,840 participants and examined data from 19,086 retellings over three iterations. Using a novel Bayesian model that tracks changes over multiple retellings, both studies reach the conclusion that retelling preserves the story’s degree of happiness and sadness with a high degree of precision, even when length shrinks and aspects of story coherence and rationalization deteriorate. This finding partly goes against Bartlett’s (1932) influential work that stresses increased rationalization in serial reproduction. Our exploratory comparative study found that embarrassment is also well transmitted, while disgust and risk were poorly transmitted. We propose a story-emotion model of retelling. We suggest that in a happy or sad story, people process the story by encapsulating the events and details into an overall story emotion. In retelling, the recalled story emotion provides the starting point for reconstruction and guides the selection, invention, assembly, and ordering of the story elements. Hence, we posit that thematic or story emotions, especially happiness and sadness, operate as anchors for remembering and retelling stories.

Serial Reproduction of Narratives: Happiness and Sadness Survive While Coherence Suffers

Conformity has been defined as a bias to copy the majority. This definition may not match recent research suggesting that social-learning strategies flexibly guide who, what and when to copy. This study therefore investigates the flexibility of frequency-dependent social-learning strategies across 3 domains: when learning asocial skills, social norms and cooperation. 316 participants played an economic game with a payoff matrix designed to imitate the learning of either asocial skills, social norms or cooperative norms. Participants saw the choices made by a group who had previously played the games. We investigated whether the participant would adjust to this frequency-dependent social information based on (i) whether the group learned a similar or different game-type to the participant and (ii) the reliability of this similarity information. The similarity information could be reliably correct, uninformative, or reliably incorrect. This range reflects instances when traits suggesting similarity on observable traits may not predict who is actually similar to oneself on unobservable traits. The participants adjusted their frequency-dependent social-learning strategies to both the similarity and reliability information when learning an asocial skill and a social norm, though performed better when learning from reliably similar others. The participants adjusted more to similar others when learning to cooperate, which suggests a strong in-group preference when learning cooperative norms. Together, these results contrast the view that conformist learners are biased to ‘always follow the majority’ and instead suggests that the decision to conform remains flexible to various social cues about who we may be conforming to.

Biases or balancing act? Investigating the flexibility of conformity

In cultural evolution, solutions have always been found by humans and transmitted to other humans. Advances in artificial intelligence have created systems that appear to bypass human cultural evolution. AlphaZero beat Go masters without learning from human play. Consequently, the behavior demonstrated by such algorithms can avoid the biases shown by humans and maintained via social learning. In this work, we explore the hypothesis that humans can learn novel strategies from algorithms via social learning and that these strategies can be further culturally transmitted to other humans. 
We ran a preregistered online transmission chain experiment where we mix human and machine players and measure the performance of human-only and hybrid chains. We use a planning task where humans are known to show a suboptimal decision strategy. Participants solved the task in chains, with each participant seeing the solution of the previous participants. In half of the chains, we replaced a single participant with an algorithm that used a strategy opposite to humans. We found that people immediately playing after the algorithm benefited from it, but people further down the chain reversed back to their biased strategy and lost what was gained via hybrid social learning. We discuss the boundaries of successful hybrid social learning and the possible mechanisms behind our findings.
We address with this work a novel hypothesis in cultural evolution, namely that artificial agents can actively contribute to the discovery of new solutions and we present transmission chains as a tool to investigate hybrid cultural transmission.

Hybrid social learning in human-algorithm cultural transmission

The present research investigates how psychological mechanisms and social network structures generate patterns of cultural change and diversity. The two psychological mechanisms studied here are cultural drift and indirect minority influence; the former is parameterized by an error rate and the latter by a leniency threshold. The patterns of cultural change are examined in terms of magnitude (small vs large), speed (gradual vs rapid), and frequency (frequent vs rare). Diversity and polarization in a society are examined in terms of global cultural variation (inverse Simpson index) and local neighborhood difference (Hamming distance). Key findings are that in networks with high connectivity or local community structures (complete, scale-free, random, and modular networks) cultural drift can produce a rapid, large, and rare pattern of cultural change (punctuated equilibrium), whereas in lattice or small world networks, it produces a more gradual change pattern. Indirect minority influence robustly produces a gradual, small, and frequent pattern of cultural change (gradualism) across various network structures. When cultural change occurs in social networks that have a modular community structure, indirect minority influence generates a regime of cultural diversity whereas cultural drift generates a polarized regime. Finally, cultural drift and indirect minority influence generate distinct tipping points for social change in different network structures, but prediction of whether and when cultural change emerges is difficult at tipping points in both cases.

Emergent Patterns of Cultural Change: Modeling Psychological Mechanisms and Social Network Structures

Population size has long been considered an important driver of cultural change. Results from population genetics, however, demonstrate that it is not the census population size, N, but the effective size of a population, N_e, that determines important evolutionary parameters. This effective population size governs the rate of change in a population due to drift and affects genetic variability both within and between populations. Here, we systematically examine effective population sizes for traits that evolve culturally, i.e. through processes of innovation and social learning, rather than genetically. We use different modeling approaches to investigate how cultural N_e and levels of diversity depend on the way traits are learned, population connectedness and social network structure. We show that one-to-many and frequency-dependent transmission can temporally or permanently lower effective population size compared to census numbers. Contrary to expectations in the literature, migration does not increase N_e and a process of cultural exchange among groups increases only the less meaningful inbreeding effective number but not the variance effective number. While network density in random networks leaves N_e unchanged, scale-free networks tend to decrease and small-world networks tend to increase N_e compared to census numbers. Finally, for one-to-many transmission and different network structures, effective size and cultural diversity are closely associated. For connectedness, however, small amounts of migration and cultural exchange result in highest diversity independently of N_e. Our results highlight the importance of effective population size for cultural evolution and show that inferring N_e requires detailed knowledge about underlying cultural and demographic processes.

Effective population size for culturally evolving traits

The use of two or more tools in combination - Associative tool use (ATU) – has played a crucial role in human cultural evolution and is also part of chimpanzee tool cultures. Research into the cognitive processes underlying ATU produced evidence for the spontaneous invention of some forms of ATU in some non-human primate and bird species. In contrast, developmental studies with children have been sparse. Whether young children are also able to spontaneously invent different types of ATU and at what age this ability emerges is still unknown. This study aimed to start filling this gap in the literature with two experiments with samples of European, predominantly White preschoolers (E1: N = 66, 3y 6m - 4y 9m; E2: N = 119, 3y 0m - 6y 10m) who were administered a series of novel, non-verbal tool-use tasks measuring three different types of ATU: Tool set, Metatool use, and Sequential tool use. Children had 3 min per task and did not receive any demonstrations. Results from both experiments showed that from 3 years of age, children were able to spontaneously invent the three types of investigated ATU behaviours, suggesting that no immediate social learning was needed for the emergence of these behaviours. E2 showed that performance significantly increased with age. However, success rates were low, demonstrating that the spontaneous invention of ATU in novel tasks with brief exposure is challenging for preschoolers. The tasks add to the pool of tests that can be used to study problem-solving in humans and non-humans.

Young children spontaneously invent different types of associative tool use behaviour

A species is deemed cultural when there exists behavioural variation across populations that can be at least partly attributed to social learning. Understood in this way, culture has been found in all species of great apes. However some ape behaviours have additionally been claimed to be culture-dependent, implying that copying social learning mechanisms underlie their acquisition (as it occurs in human culture). If that was the case, these ape behaviours should have become locally restricted as a consequence of the unavoidable copying errors accumulated during their transmission between connected individuals. Based on this logic we developed the "Method of Local Restriction" in order to identify potential cases of culture-dependent forms in great apes which distribution is compatible with acquisition via copying (i.e. these behaviours are restricted to a single ape population). After conducting an exhaustive literature search across ape species, populations and behavioural domains, we found that the vast majority of great ape behaviours re-appear in unconnected populations and therefore do not necessarily rely on copying to be acquired. However, we found 25 locally restricted behaviours, three of which were locally unique to a single population of a single great ape species. These behaviours represent the best available candidates of culture-dependent forms in great apes, which reliance on copying social learning should be tested in targeted experiments. Overall our results suggest that great ape cultures are sustained by different learning mechanisms from those at play in human cultures, namely by individual learning mediated by non-copying social learning mechanisms.

The Method of Local Restriction: In search of potential great ape culture-dependent forms

Many features of Human cultural evolution, such as technology (Kolodny et al., 2015), language (Carr et al., 2016), and scientific knowledge (Lehman, 1947) seem to exhibit unbounded evolutionary dynamics; growing in complexity, introducing new innovations, and having no obvious bounds. Evolutionary systems exhibiting unbounded evolutionary dynamics are commonly referred to as being open-ended (Bedau et al., 1998). The creation and analysis of systems that exhibit open-ended evolutionary dynamics is an open problem in the field of Artificial Life (Bedau et al., 2000). Motivated by the observations that nature exhibits unbounded evolution, with the ongoing generation of adaptive novelty and complexity, Artificial Life researchers want to create open-ended evolutionary systems in artificial media (e.g. computer simulations, robots, ... ). The goal of achieving open-endedness in artificial evolutionary systems has led to the formalisation of measurements for unbounded evolutionary dynamics - these measurements take the form of evolutionary activity statistics, and are often called the “ALife Test” for unbounded evolutionary dynamics (Bedau et al., 1998). This test has been applied to the fossil record (Bedau et al., 1998), the patent record (Bedau et al., 2019) and artificial evolutionary systems (Channon, 2003). The activity statistics enable researchers to take any evolving system over time and assess changes to diversity, alongside measuring the introduction and adaptive persistence of new components in the system. Despite having numerous datasets, models, and evaluation methods, the field of cultural evolution is yet to effectively determine whether any given component of a culturally evolving system is bounded or unbounded. Determining what species exhibit open-ended cultural evolution, and in which behavioural domains, should be a key focus of cultural evolution going forward, and the the ALife test provides us with a consistent approach for doing this.

Insights from Artificial Life: Measuring and Classifying Open-Ended Evolutionary Dynamics

Previous work utilizing Seshat: Global History Databank, a major resource for studying patterns of sociocultural evolution in world history, has sought to explain how and why complex societies rose, spread, and, at times, collapsed. Utilizing a large empirical dataset, we have explored the factors that drive social complexity and a host of related social features, from agricultural productivity to ideological developments to governing institutions. 

Here, I will summarize these findings and highlight our recent work trying to uncover the main causal forces behind the evolution of military technologies, a key factor in and indicator of intense interstate competition. I will discuss how this helps us not only understand the evolution of technology generally but also carries far-reaching implications for social dynamics more broadly. I conclude by previewing current efforts within the Seshat project linking these insights on interstate conflict to an exploration of internal competition and how this research can help shed light on not only the rise and continued development of states, but their periodic declines and falls as well.

Uncovering the Impact of External and Internal Competition on Long-Run Social Dynamics

The majority of the developmental literature about the role of outcomes and intentions in moral evaluations has been conducted on English-speaking children and focused on harm and property transgressions (Hilton & Kuhlmeier, 2019). We tested instead a Spanish-speaking sample of 5- to 11-year-old children from Colombia and Spain (N=123) employing moral scenarios involving disloyalty and unfairness.
The outcome-to-intention shift in judgements of transgression severity was moral domain-dependent in Colombian but not Spanish children. Colombians judged failed intentional transgressions more severely than accidental transgressions in case of disloyalty, but not (yet) in case of unfairness. In line with cultural group selection (Richerson & Boyd, 2008), it makes evolutionary sense that children’s sensitivity to intentionality develops earlier within the moral domains (e.g., loyalty) privileged by their own cultures (e.g., collectivistic culture).
Regarding punishment severity, Spanish but not Colombian children began to punish failed intentional transgressions of both moral domains more severely than accidental transgressions around 8 years of age. Moreover, while punishment severity decreased with increasing age for both unfairness and disloyalty in Spanish children, in Colombian children the downward pattern was observed only for unfairness. This might indicate that selective concerns become more pronounced with development because of culture-directed learning processes.
Finally, neither Colombian nor Spanish children enjoyed engaging in punishment. Colombian children in particular even anticipated punishment to feel worse than how it actually felt during and after punishment allocation. This suggests that retribution is unlikely to be the primary motive for children’s punishment behaviour (Carlsmith et al., 2008).

Cultural evolution of football tactics: strategic social learning in managers’ choice of formation

Downloads

Next from Cultural Evolution Society 2021

Serial Reproduction of Narratives: Happiness and Sadness Survive While Coherence Suffers

Similar lecture

Biases or balancing act? Investigating the flexibility of conformity

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES