ABSTRACT
Determining spatial units (e.g., layers, areas) and the location of archaeological remains within these units is fundamental in archaeology. However, the fuzzy boundaries of spatial units (e.g., sediments) and the uncertainty of objects’ past locations (moved by possible post-depositional perturbations) are obstacles. Refitting analysis of remains has long been used to reconstruct stratigraphic sequences, evaluate the impact of taphonomic processes, and detect intentional behaviour. However, four pending problems are identified and possible solutions presented. (1) Common origin is often determined from similarities in fragments. We show the inherent biases of this approach and present a new method while arguing that only physical refits should be used. (2) Studies often count the number of refits, although this can lead to ambiguous cases of admixture between spatial units. The Topological Study of Archaeological Refitting (TSAR) method is applied and extended to include fragments’ morphometry and distances between their locations. (3) Determining changes and fragmentation in a site formation process is limited by the impossibility of observing its past states. Three solutions are suggested (publishing refitting data, using experimental data, and simulation), fostering their collective development. (4) Current limits in computer-based simulation of fragmentation processes are presented, envisioning perspectives from high-performance computing.
