42 



T.W. HOLLIDAY AND S.E. CHURCHILL 



(indicated by circles). Only 2 of the 9 (22%) LUP specimens (Barma 

 Grande 2 and Bichon 1 ) even fall in the region of overlap between the 

 recent Sub-Saharan Africans and Europeans, and none of the 

 Mesolithic sample does. In contrast, there is a tendency for the EUP 

 specimens (indicated by triangles) to fall among the Sub-Saharan 

 Africans and outside of the European sample range. This specific 

 result is said to be indicative of a relatively recent African origin for 

 the earliest modern Europeans, and is discussed in detail elsewhere 

 (Holliday, 1995, 1997a). What is of most interest to this chapter is 

 that the Gough's Cave specimen, despite possessing some 'non- 

 typically' European traits, falls squarely among the Europeans in 

 multivariate space, albeit toward the more linearly-built end of the 

 distribution. 



Discussion 



Gough's Cave 1 is relatively unremarkable with regard to stature and 

 body mass; he is small, yet similar to all European samples, save the 

 heavier Neandertals. However, his body shape poses some interest- 

 ing contrasts which need to be further explored. Among recent 

 humans, clear differences in body shape manifest themselves among 

 geographically-dispersed samples. In terms of relative sitting height, 

 for example, some Australian Aboriginal and Sub-Saharan African 

 groups evince mean relative sitting height indices as low as 47.0, 

 while at the other extreme, many Inuit (Eskimo) samples evince 

 mean indices of around 54.0 (Eveleth & Tanner, 1976). What this 

 means is that among some of the more tropically-adapted groups 

 worldwide, the head, neck and trunk comprise less than half (ca. 47% 

 or less) of a person's stature. Yet another way of looking at this is that 

 among these groups, the lower limb accounts for more than half (ca. 

 53+%) of a person's standing height. By way of contrast, among the 

 Inuit and other cold-adapted groups, the head, neck and trunk make 

 up well over half (ca. 54+%) of the average person's height, while the 

 lower limbs make up considerably less than half (ca. 46% or less). 



The explanation for empirical patterns such as the above is that 

 they are due to climatic selection, and more specifically, reflect the 

 adherence of recent humans to the ecological 'rules' of Bergmann 

 (1847) and Allen (1877). These rules state that within a widespread 

 species of warm-blooded animals, those in colder regions will tend 

 to be heavier (Bergmann's rule) and evince shorter extremities 

 (Allen's rule) than do their more tropical conspecifics. Theoretically, 

 it is argued that we find this pattern because animals in cold regions 

 minimize their surface area: volume ratio (SA:V) in order to better 

 conserve body heat, since heat loss occurs through the skin (i.e., the 

 animal's surface). On the other hand, heat loss in hot environments 

 may be facilitated by increasing relative surface area. Changes in 

 body size and shape can drastically affect the SA:V ratio, as dis- 

 cussed in Ruff (1994) and Holliday (1995). 



But do these rules apply to fossil humans as well, or is this an over- 

 extention of biological uniformitarianism? Limited fossil data suggest 

 that prehistoric human populations were characterized by 

 ecogeographical clines that were perhaps even steeper than those 

 one finds today (Trinkaus, 1981, 1991; Stringer, 1989; Ruff, 1991, 

 1993, 1994). For example, the Kenyan Nariokotome Homo erectus 

 skeleton (KNM-WT 15000) is said to be characterized by 'hyper- 

 African' body proportions (Ruff and Walker, 1993), while European 

 Neandertals are characterized by an extremely cold-adapted mor- 

 phology (Trinkaus, 1981, 1986; Holliday, 1995, 1997Z>; Churchill, 

 1998). 



How do the Gough's Cave 1 specimen and his contemporaries fit 

 into this apparently climatically-driven geographical patterning? In 

 order to address this question adequately, we must have at least some 

 understanding of what the pattern in Europe was before the early 



Holocene, i.e., what was the temporal pattern of body proportions in 

 the European Pleistocene? In other words, were there temporal trends 

 in body shape during this time period? The answer is an emphatic 

 'yes'. There is actually more temporal variability in body shape in 

 Pleistocene Europe than there is spatial variability in the world today. 



We begin with the European Neandertals. They exhibit a clearly 

 cold-adapted physique, including low brachial and crural indices, low 

 limb/trunk ratios, extremely large femoral heads and wide trunks. 

 Those who succeed them in the region, however, hominins differen- 

 tially referred to as the 'Cro-Magnons' or the Early Upper Paleolithic 

 (EUP) humans, exhibit the opposite pattern - high brachial and crural 

 indices, high limb/trunk proportions, relatively smaller femoral heads 

 and narrower trunks. Succeeding the Cro-Magnons are the Late Upper 

 Paleolithic (LUP), and subsequent Mesolithic populations. These 

 later two samples have, in this analysis, been divided at the Pleistocene/ 

 Holocene boundary, with the Mesolithic sample (including Gough's 

 Cave 1 ) being restricted to the latter epoch. This division may be 

 biologically insignificant, however, since for virtually all analyses - 

 univariate, bivariate or multivariate, the LUP and Mesolithic samples 

 more closely resemble each other than they do any other group, fossil 

 or recent (see also Holliday, 1995, 1997a). 



Combined or separate, the real question of interest is what was the 

 pattern of body shape in LUP and Mesolithic humans? Importantly, 

 the 'shared' morphology of these two samples (including the speci- 

 men of interest) is in some regards paradoxical (Holliday, 1999). 

 Late Upper Paleolithic and Mesolithic specimens retain the high 

 brachial and crural indices of their presumed ancestors, the 'Cro- 

 Magnons'. Yet unlike the Cro-Magnons, they tend not to possess 

 relatively narrow trunks, relatively small femoral heads, or high 

 limb-trunk ratios. Gough's Cave 1, as a general rule, follows this 

 pattern. Like his contemporaries, his brachial and crural indices are 

 near the upper extreme of the recent European sample. Likewise, as 

 with others from his time period, his limb/trunk proportions are 

 within the European range, although his values for HL/STH, RL/ 

 STH and particularly his TL/STH indices are somewhat more ex- 

 treme than those of average Europeans today. Recall that he falls on 

 the 75th, 85th and 94th percentiles, respectively, of the recent 

 European male sample for these traits. For the other traits (relative 

 femoral head size and relative body breadth), however, he falls very 

 near the recent European mean, and is distinctly different from 

 recent Africans. In multivariate space, however, he lies within the 

 European scatter, and beyond the range of recent Sub-Saharan 

 Africans, as do his Mesolithic contemporaries and the majority of 

 the LUP sample. By way of contrast, the EUP sample tends to cluster 

 more closely with the recent Africans. 



Both in scientific articles (Frayer, 1992; Frayer et al., 1993) and 

 the popular press (Shreeve. 1995), it has been pointed out that the 

 retention of high brachial and crural indices among Late Upper 

 Paleolithic and Mesolithic humans is problematic for Trinkaus' 

 (1981) argument that these indices reflect elevated gene flow (or 

 population dispersal) from Africa associated with the origins of 

 modern humans. After all, these workers argue, the glacial cold of 

 Europe should have modified, at least by the end of the Pleistocene, 

 any previously incoming population toward a more cold-adapted 

 morphology. Yet with regard to brachial and crural indices, the LUP 

 sample have an even more extreme (almost 'hyper-tropical') mor- 

 phology than their EUP forebears. 



As pointed out in Holliday (1999), this argument shows the 

 problems that can arise when single traits are studied in isolation 1 . In 



'We can, for the sake of argument, consider the brachial and crural indices a single trait, 

 since they tend to covary, and are likely influenced by the same gene complexes. 

 Likewise, they are almost certainly influenced by the same environmental factors. 



