There is widespread interest in documenting the amount and geographic
distribution of genetic variation in the human species. This informati
on is desired by the biomedical community, who want a densely packed m
ap of SNP (single nucleotide polymorphism) sites to be used to identif
y genes associated with disease by linkage disequilibrium between sets
of adjacent markers and the occurence of disease in populations, and
to characterize disease-related variation among populations. Anthropol
ogists use genetic variation to reconstruct our species' history, and
to understand the role of culture and geography in the global distribu
tion of human variation. The requirements for these two perspectives s
eem to be converging on a need for an accessible, representative DNA b
ank and statistical database of human variation. However, both Fields
have been using conceptual models that are oversimplified, and this ma
y lead to unrealistic expectations of the questions that can be answer
ed from genetic data.