Talks

Daniel R. Kelly, Purdue University, US:
When is ignorance adaptive in cultural evolution?
Cultural evolution, like natural selection, is a process that gradually accumulates innovation and generates complex design without any foresight or individual designer. Against this theoretic backdrop, I develop and explore a more specific idea found in Henrich (2015) I’ll call the Adaptive Ignorance Thesis: the psychology of cultural innovation is often effective without anyone fully understanding the complex process that psychology supports or even the packages of adaptations that process produces.
I gather together scattered claims and examples Henrich offers in support of this idea, formulate the main line of argumentation advanced in favor of it, and distinguish and assess several versions of the thesis that differ on the role and import they assign to ignorance in generating adaptive design and innovation. For example, a weaker version holds that we can use and reap benefits from innovative cultural artifacts despite our ignorance, i.e. even though the operation and full purpose of those artifacts is causally opaque to us. A stronger version holds that at least some artifacts are able to deliver adaptive benefits because of our ignorance, i.e. some cultural innovations only work if we fail to understand what they are doing, and how.
Finally, I attempt to identify conditions under which each version is likely to hold, especially in light of Henrich’s plausible case that the sophistication found in distinctively human mindreading capacities and artifact cognition evolved to facilitate cultural evolution, and that they do so by helping us better understand different aspects of this gradual, innovative process.

Mathieu Charbonneau, Central European University, Hungary:
Recombination, modularity, and innovation
A central mechanism involved in the generation of cultural variation, and one often described as humans’ special ability to produce cumulative culture, is that of cultural recombination. The recombination process can be understood as the bringing together of existing cultural traditions or of existing cultural traditions’ sub-components into novel, complex cultural composites. Central to the mechanism of recombination is the capacity for cultural traditions to possess modular structures. Indeed, as the recombination process can decompose old cultural traditions and recompose them into new ones, the recombined sub-components must be relatively independent of the traditions of which they are part.
In this paper, I argue that the study of innovation through recombination and its complementary notion of cultural modularity both depend on a deeper understanding of the generative processes involved in social transmission. More specifically, the hierarchical and functional structures of these generative processes play a key role in structuring modular cultural traditions. A strictly informational account is bound to fail to make sense of the recombination process and of modular traditions because both phenomena depend on the capacity of an individual to enact, alter, and learn how to produce public displays. The argument is supported with the case study of the transition between the simple flaking technique characteristic of the Oldowan industry and the more complex ones of the Early Acheulean, a paradigmatic cultural module, and arguably the oldest one.

Francisco Brahm, University of Cambridge, UK:
Cultural evolution and the theory of the firm
What is the purpose of firms (companies)? Several social scientists, mainly from economics, propose that firms are a way to avoid the costs involved in market exchange. However, ultimate causation is unaddressed: a clear explanation for the evolutionary emergence and success of firms is lacking. Using the framework of cultural evolution, in this project we hypothesize that firms are an excellent mechanism to accelerate the  evolution of group-level traits.
Cultural evolution focuses on individual-level traits, that is, traits that can be transmitted
person-to-person. However, the theory hasn’t yet incorporated traits that are generated by organized complexity between specialized individuals, where the whole is more than the sum of the parts. Models and evidence shows that group-level traits have a main evolutionary hurdle, low imitability. This hurdle severely limits their transmission using social learning, rendering group-selection the key evolutionary force. However, the speed of group selection generated by tribe conflicts (or other similar processes) is too slow to account for the hyper-accelerated speed of cultural evolution observed in the aftermath of the industrial revolution.
We argue that two fundamental (and historically non-obvious) characteristics of firms — low exit/death costs (produced by incorporation and limited liability) and limited entry (and ensuing lower social learning costs) — accelerates the evolution of group-level traits, both by promoting innovation in group  traits and by fast-tracking the invasion of those that succeed. Before firms, carriers of group-level traits such as tribes (and their members) had to “die”; with firms, individuals don´t have to, firms die instead (and quite frequently so!). We sustain our claims using the historical account of the birth of modern firms and simple formal models.

Regine E. Stolarczyk, University of Tübingen, Germany:
What’s new in tool behavior:  Using cognigrams and effective chains to detect new and old traits in the Middle Stone Age of Southern Africa and beyond
One of the most striking features of hominid development is an increasing innovativeness in object behavior and a marked flexibility, variability and complexity in producing and using tools. Therefore, it is of utmost importance to study innovations in archaeological remains to understand not only cultural change but also cognitive steps in human evolution. The most common way in Paleolithic research to look at innovations is to identify new tool types and deduce new production techniques by studying tools. However, it is not enough to merely focus on the appearance of new tool types because the tool is just one small part of a behavior.
Innovative behaviors are seldom completely new solutions, but rather mosaics of new and old traits and the innovative components affect diverse parts of object behavior, like tool types, technological processes or mental concepts. If we want to understand cultural change and innovative capacities of hominids, we have to identify this mosaic and evaluate what is really new. Cognigrams and effective chains can be used to detect such innovative traits in tool behavior by reconstructing past behaviors and compare them with each other. The presented study focuses mainly on identifying innovations in the Middle Stone Age of Southern Africa, from new tool types (e.g. retouchers), new artefact groups like organic tools, new production techniques such as polishing bone, and new behaviors like hunting with bow and arrow to so far unknown mental aspects (e.g. initiating and anticipating effects of tools without controlling the effect during its agency).

Elizabeth Renner, Mark Atkinson, Christine Caldwell, University of Stirling, UK:
Effect of information source on squirrel monkey performance in two binary discrimination tasks
Human cumulative culture relies on the use of social learning in the acquisition of skills, technologies, and ritual behaviours. Despite the ability of numerous non-human species to learn socially (with various degrees of fidelity), there is a lack of strong evidence for cumulative culture in other species. Examining the ways in which humans and non-humans use social and individual information may provide insights into the mechanisms underlying human cumulative culture. Here, we investigate whether members of a non-human species use information differently when it is acquired individually or socially.
We utilise two simple “win-stay, lose-shift” tasks with squirrel monkeys (Saimiri sciureus). In each task, participants attempt to locate a reward concealed by one of two items. There is no way to know a priori which of the items conceals the reward. We manipulate how monkeys can acquire information about the location of the reward: socially, by seeing a human demonstrator select an item (and locate or not locate the reward), or individually, by itself selecting an item (and locating or not locating the reward). In one task, the objects are presented two-dimensionally on a touch-sensitive tablet, and in the other the objects are three-dimensional cups.
For each version of the task, we analyse the effects of information source (social or individual) and information type (rewarded or unrewarded); we also compare performance on the tablet and three-dimensional versions of the task.

Elena Hoicka, Sarah Rose, Stephanie Powell, University of Sheffield, UK:
Developing a parent report measure of problem solving from 12 to 47 months
Recent lab research shows young children are creative. One-year-olds think divergently (Hoicka, et al., 2016) – that is, they generate many different ideas. Furthermore, 2-year-olds figure out how to use novel tools on their own (Reindl, et al., 2016). However less clear is how children are creative in their daily lives. To examine one aspect of early creativity – problem solving – we created a parent-report measure, the Early Problem Solving Survey (EPSS).
In this exploratory (N=219) study, parents of 12- to 47-month-olds completed the 19-question EPSS. Questions determined children’s propensity to problem solve, e.g., “If a toy breaks, my child tries to fix it before asking for help.” and, “My child needs help completing simple puzzles.” (reverse scored). Five items were removed because over 20% of parents answered “Not Applicable.” One item did not correlate with the remaining average score of all items, “My child figures out how to do things by watching others do them.” so was dropped. The remaining 13 items correlated with the average score of these items, Spearman’s rho>.2, pKR20=0.72. The EPSS correlated strongly with age, Pearson’s r=0.51, p<.001.
A subset of 2- and 3-year-olds from the above study were tested for inter-observer agreement by having a parent complete the EPSS, while an experimenter ran the Unusual Box Test (UBT) – a lab-based measure of divergent thinking. Children’s scores on the EPSS were moderately to highly correlated with the UBT (N=35), Pearson’s r=0.43, p=.011.

Helena Miton, Central European University, Hungary:
New perspectives on social learning strategies and cultural transmission experiments
Social learning strategies (SLS) have been stated to be one of the preferential mechanisms for the spread of innovations and adaptive behaviors. They have been empirically supported by results of cultural transmission experiments – mainly group or replacement experiments – in which information flows through generations of participants busy solving a specific task. Several characteristics of those experiments nevertheless limit the strength of the empirical support for the existence and fitness-enhancing consequences of SLS.
First, the experimental set-ups tend to impose some SLS on the participants, i.e., there are very limited possibilities not to behave accordingly to the SLS studied in those experimental settings. Second, the experimental set-ups usually feed the participants with information that, under ecological conditions, the participants would have to extract from their social environment, which challenges the ecological validity of the results. Third, the set-ups often inform the participants of the expected pay-offs of behavioral variants, contradicting the idea that SLS are adaptive specifically when the pay-offs are unknown. Such limits emerge as consequences of a phenotypic gambit.
How can such limits be overcome? Instead of studying SLS in purely behavioral terms, we favor a cognition-oriented stance towards social learning and suggest revising these research questions accordingly. From this perspective, the conditions under which one given social cue or characteristic of a source of information becomes more important than others becomes a central question for the experimental study of SLS.

Tamar Rosenberg-Yefet, Tel Aviv University, Israel:
Convergent evolution or cultural transmission? Reconstructing the process of technological innovation in Lower Paleolithic societies – The case of the Levallois method
The emergence, transmission and assimilation of technological innovations across a vast geographical and chronological range, such as monumental architecture, pottery, writing and so on, raises a fundamental question concerning the mechanism that enabled such phenomena. Was it invented in one core-area and spread by a process of cultural transmission? Alternatively, was it invented independently in different areas as a solution to similar needs, a process known as convergent evolution?
These two models will be applied here concerning the reconstruction of one of the major processes of technological innovation in stone tool production during Paleolithic times. After producing stone items in several particular ways for over one million years during Lower Paleolithic times in the Old World, humans invented a new and sophisticated way to produce stone items, of predetermined morphology, known as the Levallois method. It is generally accepted that the invention, introduction and assimilation of the Levallois method reflect significantly the cognitive and technological capabilities of Middle Pleistocene humans, as well as a striking case of human cultural evolution. The Levallois method unique history of research and the recently excavated sites attributed to the emergence of the technology during Late Lower Paleolithic times, pose a rare opportunity to examine mechanisms of invention, distribution and assimilation of technological innovations. The Levallois core assemblage from the Late Acheulian site of Revadim, Israel, will serve as a case study for demonstrating what we believe to be a process of cultural transmission of stone tool production innovation some half a million years ago.