Abstract

Although Correspondence Analysis as a method has been in use for over 50 years, its use in archaeology has been patchy. Its use in anglophone circles grew in response to a key paper in World Archaeology (Bølviken et al. 1982). Subsequent publications in the Scandinavian literature introduced multiple CA (Madsen 1988) and other methods. In the 1980s and early 1990s several bespoke CA packages for archaeology were written (IAStats, Mv-Arch, Bonn Seriation package [WinBASP]). More recently, however, R has become the dominant system for CA in archaeology.

Developments in archaeological theory from the late 1980s onwards led to a strong “anti-mathematical” strain in archaeological work. In recent years, however, this stance has become less prevalent and the occurrence of CA in archaeology has grown. In general, however, the vast majority of applications have been “standard” CA. More recently, multiple CA has seen more applications and there are a small number of applications of the use of bootstrapped and spatial CA.

Building on the seminal work of Mike Baxter (1994, 2003), this paper will review of the application of CA in archaeology in more recent years. It will also look at missed opportunities, strengths, weaknesses and possible directions in the future.

Baxter, M. J. (1994). Exploratory Multivariate Analysis in Archaeology. Edinburgh: Edinburgh University Press. (reprinted by the Percheron Press in 2015).

Baxter, M. J. (2003). Statistics in Archaeology. London: Arnold.

Bøviken, E., E. Helskog, K. Helskog, I. M. Holm-Olsen, L. Solheim and R. Bertlesen (1982). ‘Correspondence Analysis: an alternative to Principal Components.’ World Archaeology 14: 41–60.

Madsen, T. (1988), ed. Multivariate Archaeology. Numerical Approaches to Scandinavian Archaeology. Moesgård: Jutland Archaeological Society.