Thoughts on the Utility of a Bivariate‐splitting Approach to Olmec Ceramic Data Interpretation*

Abstract

A recently resolved debate centred on the interpretation of a major set of INAA data that was based on Olmec ceramics from Mexico. In an attempt to answer questions arising from the debate, this paper discusses the effect of numbers of samples chosen for data interpretation. It also presents interpretations of the ceramic data set based on a bivariate data-splitting approach and compares the results of this with the multivariate analysis approaches employed by the initial publishers of the data. and their laboratories are closed. We would therefore add the plea that there be a revival of the publication in papers of group means and standard deviations, or group means and their percentage errors, as was common in the first decades of archaeometry, together with the safe long-term storage of data, as is practised at MURR.

Detecting modes in low-dimensional archaeological data

“…These boundaries differ slightly from those in Hancock et al (2008) whose data set differed. These groups could be inspected for multimodality with respect to other variables but this is not pursued here.…”

“…Their conclusions concerning the movement of early Olmec ceramics were disputed by Stoltman et al (2005) who analysed subsets of 756 and 944 samples. The debate this triggered is not of concern here except to note that in what might be considered as a 'coda' to the original analyses Hancock et al (2008) reanalysed the data using a bivariate-splitting approach that broadly supported the conclusions of Blomster et al (2005).…”

“…Hancock et al (2008) propose an approach, similar in spirit to recursive subdivision, which they term bivariate-splitting. In their application to Olmec ceramic data, beginning with a 760 by 32 data matrix, they initially subdivide the data set on the basis of Ca, with the next subdivision based on Cr and so on.…”

“…It was noted earlier that the method of Hancock et al (2008) could be interpreted as the sequential search for multiple modes in a set of data. The set of 944 samples is used here to illustrate aspects of analysis germane to the present paper.…”

Where East Meets West: The Formative in Mexico’s Central Highlands

“…The on-going debate over the Olmec as ''mother '' or ''sister'' cultures (Blomster et al 2005a;Flannery and Marcus 2000;Flannery et al 2005;Hancock et al 2008;Lesure 2004;Neff et al 2006a, b;Sharer et al 2006;Stark 2007;Stoltman et al 2005) has resulted in methodological deliberations on how to develop reliable data sets that can advance our understanding of what the presence (or absence) of this style and its variations in different parts of Mesoamerica actually means in terms of human interaction.…”