Shape analysis using the thin‐plate spline: Neanderthal cranial shape as an example

Abstract

This paper describes a new geometric method for illustrating and quantifying biological shape difference. The technique is discussed in some detail, and is illustrated by applying it to the problem of characterizing Neanderthal cranial shape. The method of thin-plate splines uses a mathematical model based on the bending of a hypothetical steel plate in order 1) objectively to generate a D'Arcy Thompson-style deformed grid that illustrates the shape contrast between two forms, and 2) to quantify the shape difference by breaking it down into a series of components based on scale.Results confirm the presence of some features traditionally attributed to Neanderthals, but some "classic'' Neanderthal features do not in fact characterize the Neanderthal sample. Traits may have been misidentified because of failure to take into account differences in absolute size, use of a misleading frame of reference, and interpretation of one aspect of a large-scale change as a localized feature. In particular, the important Neanderthal "specialization" midfacial prognathism, defined as forward displacement of the nasal region and the tooth row relative to more lateral facial structures, does not appear to be a typical Neanderthal trait. The uniqueness of Neanderthals appears to have been exaggerated, and may be related to Boule's peculiar, flawed reconstruction of the skull of La Chapelle-aux-Saints. The method of thinplate splines is a powerful technique, capable ofproviding a new and insightful perspective on morphological problems in physical anthropology. Physical anthropologists have long been concerned with measurement of biological shape and the quantification of shape differences. Over time, more and more effective methods for this task have become available. Most recently, the field of geometric morphometrics has offered a variety of new techniques, which yield much richer information on shape difference than was possible in the past. One of these techniques, the method of thin-plate splines (Bookstein, 1989, 19911, seems particularly well suited to the kinds of problems commonly approached by physical anthropologists. The thin-plate spline generates useful illustrations of shape difference, and allows one rigorously to address the issue of geometric scale: to ascertain which shape differences are large-scale, involving coordinated changes across broad areas of the form, and which are localized to more limited regions. This article outlines the technique in some detail, using a well known paleoanthropological problem as an example.The thin-plate spline is one of a series of new geometric techniques for assessing shape difference. Geometric morphometrics goes beyond traditional shape analysis by gaining more information from the landmarks under study. Not only information on chords and angles is included, but information regarding the relative positions of the landmarks is retained as well. Reviewing 0 1996 WILEY-LISS, INC

Quantitative analysis of Neanderthal temporal bone morphology using three‐dimensional geometric morphometrics

“…Where the zygomatic suture landmarks were included (step 1), an inferior position of the lateral margin of the glenoid fossa and, to a lesser extent, of the inferior aspect of the zygomatic suture, contributed to the separation of Neanderthals from modern humans in both PCA and CVA. As was also found by Yaroch (1996), it is the more inferior placement of these structures relative to the porion that results in an apparent elevated position of the external auditory meatus relative to the glenoid fossa and the zygomatic arch in Neanderthals. This shape difference between Neanderthals and modern humans was no longer observed when the two zygomatic suture landmarks were removed.…”

“…These results are in agreement with previous multivariate analyses of craniofacial measurements (Stringer, 1974, 1989, 1992; Bräuer and Rimbach, 1990; Bräuer, 1992; Turbón et al, 1997), which found Neanderthals to be widely separated from modern humans. However, they disagree with the geometric morphometric study of the differences between Neanderthals and modern humans by Yaroch (1996), which found the differences between Neanderthals and modern humans to be minimal based on two‐dimensional coordinates of cranial landmarks.…”

Ancestry Estimation in Forensic Anthropology: Geometric Morphometric versus Standard and Nonstandard Interlandmark Distances

“…Geometric morphometric methods have been widely used in biological anthropology to explore morphological shape differences in various craniofacial regions and for classification purposes within forensic anthropology . Slice and Ross recently developed a program, 3D‐ID, for ancestry estimation using landmark data and GM methods .…”

An Application of the Relative Warps Analysis to Problems in Human Paleontology - With Notes on Raw Data Quality

“…As was emphasized in the introduction, an additional goal of this paper is to raise a few methodological questions with regard to the use of 2D representations of 3D objects (i.e., pictures of skulls) as well as previous applications of geometric morphometrics in paleoanthropology (Yaroch, 1996).…”