Getting to the Point: Defining, Reconstructing and Investigating Shape Through a Procrustean Protocol

Abstract

Geometric morphometric methods can be employed in numerous archaeological subfields. Their application is, however, complicated by practical considerations. Notable challenges include determining the density of coordinate points and the reconstruction of damaged objects. To explore these issues, a three-dimensional template of the human os coxae was created. This paper presents that template and details reproducible procedures to estimate appropriate coordinate point number and placement as well as impute missing data. It emerges that three-dimensional scans can be conveniently produced and optimal coordinate point density will vary depending upon the hypotheses being tested. Similarly, following a simulation study, it was found the best approach to imputing missing points will depend upon the level of missingness and can have important implications for the identification of statistical signals, such as structural modularity. In summary, in addition to an open-access digitisation template, a flexible protocol is presented which can be adapted to define, capture and reconstruct shape variation to investigate a diverse spectrum of hypotheses in archaeological materials (e.g., bones, lithics, sediments).

Keywords

• geometric morphometric methods
• digitisation template
• procrustes superimposition
• os coxae
• data imputation
• structural modularity