Dynamics of spreading of ceramic techniques and style: actualist comparative data and agent-based modelling – DIFFCERAM

In order to construct the agent-based model, protocols of investigations have been elaborated which are common to the different ethnographic fields.
The first questionnaire is a «traditional« questionnaire whose objective is to collect the necessary informations for reconstructing the technical and stylistic changes at an individual level. On the basis of these informations, it will be possible by aggregation to reconstruct the rates of adoption of each trait in the course of two or three generations.
The second questionnaire is aimed at obtaining empirical data on the source of influence at the origin of the borrowings. It includes three parts:
- A/ the first part includes a «generator of names« which aims at establishing the network of contacts of each artisan.
-B/ the second part includes a «translator« of names given by the interviewed subject. It is a series of questions aimed at understanding the characteristics of the contacts of ego.
-C/ the third part gathers questions aimed at assessing, on the one hand the tresholds of activation of the actors, and on the other hand the reasons which can lead to imitate behaviours of the others.
The third questionnaire aims at collecting informations on the resistance to change on the basis of field experiments. The hypothesis is that this resistance varies depending on the degree of expertise of the craftspeople and the nature of the ceramic trait. It can be measured with a study of a) the apprenticeship of the technical and stylistic ceramic traits, b) the degree of expertise and the understanding of the properties of the technical task by the craftspeople.

The expected results are:
- to establish and to explain the curves of diffusion observed on the different ethnographic fields (first group of regularities) by precising the socio-economic contexts into which the ceramic traits have diffused or not diffused;
- to establish the structure of the networks of advices linking the craftspeople and the reasons which lead them to imitate the behaviour of the others (second group of regularities)
- to assess the resistance to change depending on the nature of the ceramic traits thanks to the comparative analysis of the apprenticeship modalities and transmission of the ceramic traits, the degree of expertise of the craftspeople and the influence of the cultural representations on the understanding of the properties of the technical tasks.

The construction of the agent-based model will be based on the second group of regularities and on the effects of resistance to change in order to assess the ensemble of mechanisms explaining the first group of regularities.
The simulation will enable to create virtual diffusion curves which will be compared to the empirical diffusion curves. Validity of the mechanisms will be thus tested.
The most elegant result one can obtain will be a model with a unique ensemble of explanatory mechanisms whose specific combination and specificity vary depending on the context and the type of trait whose diffusion has be explained.

Research perspectives have to be considered depedning on the disciplines:
- in archaeology: researches on the cultural dynamics in the changes observed on the longue durée;
- in sociology: researches on the mechanisms underlying the diffusion of behaviours depending on the sociological composition of the communities;
- in psychology: researches on the cognitive mechanisms underlying the adoption of new behaviours.

- to come -

In archaeology, it is possible to identify interactions between groups on the basis of diffusion of technical or stylistic traits. However, there is no frame of reference based on empirical data, which enable us to interpret such diffusion in terms of modalities of transmission. As a result, it is difficult to distinguish, for example, between demic and cultural diffusion, which prevents in the long run restitution of the history of cultural dynamics.
In order to overcome this problem and enrich archaeological interpretations, the aim of this project is to build up a frame of reference related to the process of diffusion of technical and stylistic ceramic traits. Building this frame of reference will be done with the help of a multi-agent model on the basis of actualist data in order to reconciliate modelling and empirical approaches. For this purpose, the conditions into which techniques and ceramic styles are transmitted and diffused, will be explored. The hypothesis is that these conditions may correspond to regularities which can be expressed in the following verbal form: “In C (socio-economic context), IF i (interactions) and IF n (nature of the trait), THEN d (diffusion)”. The empirical data shall be collected on several terrains so as to make the parameters vary. For this purpose, comparative ethnoarchaeological inquiries shall be conducted in five different countries - India, Kenya, Ethiopia, Cameroun, and the Philippines. The explanatory mechanisms and the nature of the ceramic traits shall be studied in psycho-social terms by characterizing the methods of learning and transmission of pottery traditions.
The set of actualist data will then feed a multi-agent model which will enable us to simulate and explore the conditions into which process of diffusion takes place.
Three disciplines will be involved: social anthropology, joining both ethno-archaeologists and anthropologists to collect the empirical data, sociology for formalising the regularities, and experimental psychology to understand the mechanisms underlying the regularities.
The results expected are: a) a comparative analysis of the different socio-economic contexts in which ceramic traits did or did not diffuse ; b) a comparative analysis of how ceramic traits are learned and transmitted ; c) a modelling of the conditions in which ceramic traits are diffused (contexts, interactions, nature of the trait) to simulate the observations made on a micro-scale and see their emergent effects on the populational level – of several generations – and bring regularities to light.
These regularities, bearing on the context into which interactions prompt transmission, shall be used in the future as frame of reference for interpreting diffusion process in terms of interactions (the unknown variable), and subsequently the dynamics at work in the evolution of cultural traits and cultures.