La seconde édition des rencontres du réseau « The Connected Past » s’est tenue à Paris le 26 avril 2014. Ce réseau international se consacre aux applications en sciences historiques de l’analyse de réseaux. Sébastien Plutniak, membre de la MAFBO responsable de l’analyse des matériaux céramiques, a présenté un poster intitulé « Archaeology as practical mereology: an attempt to analyze a set of ceramic refits using network analysis tools ». Le poster est disponible ci-dessous, ainsi que le résumé.
The Connected Past second edition event has been held in Paris on Saturday 26th April 2014. This conference was dedicated to network analysis applications in historical sciences (see the programm here). MAFBO team member Sébastien Plutniak, in charge of the pottery analysis in our project, has presented a poster entitled « Archaeology as practical mereology: an attempt to analyze a set of ceramic refits using network analysis tools ». [the poster is avaible here]
After two decades of research there has been renewed interest in the refitting analysis of archaeological objects (Cziesla et al. 1990; Hofman and Enloe 1992; Schurmans and De Bie 2007). This analysis is time-consuming, but the result is the most unambiguous criterion to determine the relationships between archaeological objects (contrary to stylistic features, for instance). This kind of information is obviously relational, but, surprisingly and to our knowledge, no attempt has been made to use network analysis to analyze these relationships (Brughmans 2012). In this paper, we propose some preliminary methodological developments in this direction, in order to combine network and refitting analysis. The discussion is based on ceramic objects discovered in a rainforest rock-shelter, at Liang Abu, East-Kalimantan, Indonesia (Ricaut et al. 2011). Searching for conjoining sherds is usually done to reconstruct a vase, rather than as a heuristic method to understand the dynamics of a site. There is plenty of literature regarding stone tool refitting (Art and Cziesla 1990), but ceramic refitting has been considered less-widely and is notably absent in Shepard’s classical manual (Shepard1956) Focusing on ceramics will lead us to determine the specific properties of refitting relationships according to the type of archaeological object. Such characterizations have already been proposed, such as Bollong’s six points typology (Bollong 1994), but the description was limited to a quantification of each type. We discuss that graph analysis can provide an integrative framework for the different metrics, and that the number of relations can be summed in addition to describing the structure of a set of refitting relations. These structural properties could enlighten taphonomic phenomenons and site formation processes, information which are valuable to the understanding of complex karstic stratigraphy, as found at Liang Abu.