Misunderstanding graphs: The confusion of biological clade diversity diagrams and archaeological frequency seriation diagrams

Abstract

Graph perception involves the accurate decipherment of (often quantitative) data displayed in visual form. Because graph style may reflect discipline-specific tradition, similar graph styles in distinct disciplines can be subject to misinterpretation. Both archaeologist James A. Ford and paleobiologist Stephen Jay Gould confused spindle diagrams representing archaeological frequency seriation and paleontological clade diversity analysis as displaying the same kinds of data and representing the same processes. Similarities between the two kinds of analysis are, however, limited to the use of the same graph style—spindle diagrams—to illustrate the history of frequencies of things. The kinds of frequencies differ in two ways between the two disciplines; frequencies are of low-level Linnaean taxa within a clade representing a higher taxon in paleobiology, and are frequencies of artifact specimens within each of several types in archaeology. Further, frequencies are absolute in clade diversity and relative in frequency seriation. Clade diversity analysis, as practiced by Gould and colleagues, is a time-series analysis that requires knowing the age of taxa prior to analysis of the shape of the spindle diagram. Frequency seriation in archaeology involves ordering multiple collections of artifacts that share at least some types; ordering is based on similar frequencies and a presumed unimodal frequency distribution, and the order is inferred to be a chronology. Different analytical assumptions and goals result in discipline specific rules of graph decipherment, though each of the two kinds of analyses can be performed in each of the two disciplines.