Four problems for archaeological refitting studies. Discussion from the Taï Site and its Neolithic pottery material (France)

Abstract

2 Problem 1: Using similarity relationships between fragments generates bias 2.1 Matching sets count and intuitive interpretation . . . . . . . . . . 2.2 Counting and non-supervised ordering of matching relationships 3 Problem 2: Counting connection relationships can be misleading 3.1 Limits of count-based approaches to connection relationships . . 3.2 How "strong" is a connection relationship? . . . . . . . . . . . . . 3.3 Admixture and cohesion of the Early Neolithic layers at the cave Taï 4 Problem 3: Reference data to compare sets of refitting fragments are limited 4.1 The limits of archaeological observation . .