18 - From “mathematical dissection of anatomies” to morphometrics: a twenty-first-century appreciation of Charles Oxnard

Summary

Introduction

In a series of books in the 1970s and 1980s, Charles Oxnard, with the help of colleagues and students, propounded a nearly complete biometric methodology for “biomathematical anatomy” in the comparative context, and used it to explore the functional diversity of the primates, including the prosimians, in extraordinary anatomical detail. Today, twenty years later, is a good time to look back at this accomplishment. In retrospect, Oxnard captured the methodological challenges of this domain quite precisely: problems of statistical representation, sample summary, and correlations among structural, functional, and phylogenetic information. To proceed, he selected from a wide range of toolkits that had already been set on the table by others: tools of multivariate ordination, graphical data display, biomechanical engineering, and phenetics. There resulted masterful book-length expositions of methods and comparative findings all across the order Primates, from prosimian to human, from extinct to extant.

One of us (FLB) first encountered Oxnard's work in reviewing the literature for his thesis (Bookstein, 1978) on the measurement of biological shape – the work that, after some vicissitudes, became contemporary morphometrics. Oxnard's 1973 book is subtitled “Some mathematical, physical, and engineering approaches,” and Bookstein learned quickly that if Oxnard had not noted an approach here, either it was not worth noting or it had been invented after 1973. Multivariate morphometrics, Blum's medial (symmetric) axis, cluster analysis, analysis of outlines by tangent lines, optical Fourier analysis, optical stress analysis, Cartesian transformation grids – all were recapitulated in that single short book, as a potential toolkit for graduate students in biomathematics (like Bookstein), in comparative anatomy, or in applications of computer graphics in biology, or for anybody else interested.