Our data come from Yellen’s 1977 book on the Dobe Ju/’hoansi settlement patterns
and camp structure. That research was undertaken as part of the long-term Harvard
University Kung Bushman ethnology and human ecology project. Yellen (1977, pp.
61–63) discusses the strengths and weaknesses of this data set, including the poten-
tial influences of his anthropological team and of nearby Bantu settlers. He con-
cludes, nevertheless, that “[h]ad neither Bantu nor anthropologists been there,
I believe that the basic pattern would have been the same but less time would have
been spent at the permanent waterhole.” This statement leads us to believe that we
can rely upon the camp location data but that the occupational times may be some-
what less accurate.
We measured the distances between camps on Yellen’s map in the order in which
they were occupied, and we drew the lengths of the camps’ occupations from the
published tables. We analyzed the geographic distances between camps in the order
in which they were occupied to see if they conformed to a power-law distribution
of step lengths predicted by a Lévy flight. We also analyzed the lengths of camp
occupations to see if they conformed to a power-law distribution as they would if
they were the wait times at the turning points of a Lévy flight.