Le Pacifiquede 5000
2000 avant le present/The Pacific from 5000 to 2000 BP
rent approach. What we need to look at is how the process of phenotypic change itself takes place. What effect does climate and diet have on phenotype? Are Polynesians really so skeletally homogeneous? The evidence, limited though it is, suggests not, yet it has been comfortable until now to assume that they are, primarily for the purpose of gaining larger sample sizes.
To my knowledge there have been no features found in isolation, which can be claimed to be distinctively Polynesian. What is needed then is a survey of the extent of so-cal- led "Polynesian"features in populations outside Polynesia (and outside the Pacific) to see whether they occur in the same complex (i.e., pentagonal-shaped head with rocker jaw), or if indeed there are any features which can be said to be distinctively Polynesian. What part have environmental and cultural influences had to play in moulding the ske- letal morphology of the Polynesians?
Geographic and temporal factors
Sample sizes in general in the Pacific are small, and because of this, museum skeletal material with very little geographic or temporal provenance has been conflated with archaeologically excavated remains, in order to maximise samples. However, even within the so-called homogeneous regions of Polynesia, evidence is increasingly poin- ting to diversity within as well as between island groups.
My analysis of infracranial skeletal remains from siteTo-At-36on Tongatapu (van Dijk 1993), when compared to sites To-At-l & 2 on the same island, illustrated that even geographically close sites may produce highly significant differencesin metric measu- rements. Whether these are due to temporal, status, dietary or actual genetic differences is unknown.A hypothesis to account for such regional variation within islands has been proposed (van Dijk 1991)that looks at the effect that cultural and environmental factors (for example dietary differences between groups, and active selection for large body size in many Polynesian societies) may have had on a population's phenotype. A simi- lar pattern of morphological variation within an island group has also been found in Fiji between the Sigatoka material (dated to 2000BP) and later Fijians (Visser 1995).
In cases where skeletal samples from the same island or island group have been diffe- rentiated geographically or temporally in statistical analyses, they often do not cluster togetherbut form close relationships with other groups. Does this reflect biological dif- ferences or is there a problem with the methodology? Results like these are generally not addressed. Pietrusewsky (1989:242) in a comparative study of Lapita skeletal mate- rial with other Pacific populations, found a significant difference between archaic