Biological Conceptions of Race 461breeding populations, all of which are very similar. Yet, because there can be slight
differences among the populations that make up a race, race membership must be
defined with reference to many characters. There must also be statistically signif-
icant differences in the mean values of the traits used to define race membership.
Second, as a general rule, geographical races areallopatric populations; they are
populations that are separated from one another by a geographic barrier such as
a mountain range, desert, or large body of water. When such populations remain
reasonably isolated over time, geographic differentiation is likely to occur. Some-
times, however, geographical races aresympatric populations— viz., populations
that share the same geographic range but are, in large part, reproductively iso-
lated from other conspecific populations for some other reason. An example of
sympatry can be found in some human populations, such as the Amish, that are
largely reproductively isolated due to certain cultural and/or religious practices.
The geographical race concept has been used for defining both human and
nonhuman race. In its application to humans, it was often assumed that the
members of a race are similar to one another, and differ from the members of
other races, with respect to certain observable characteristics (skin color, hair
type, bone structure, etc). It was also presumed that these traits are heritable and,
thus, that there are average differences among the races at the genetic level as well
[Dobzhansky, 1953; 1955; Garn, 1961; Coon, 1963]. In the early days of molecular
genetics, researchers focused largely on differences in the distribution of blood
types (e.g., ABO, Rh, MNS, and Rhesus, etc.) as a measure of underlying genetic
differences. Though no group was exclusively of one blood type, biologists found
average differences in the distribution of blood groups among human populations
from diverse geographic regions. Because such populations are also observably
different, a number of biologists and anthropologists took such data as providing
support for the existence of human geographical races.
Before discussing some of the problems with the geographical concept, it is worth
mentioning that it avoids many of the problems faced by the typological concept.
One reason is that, unlike the typological concept, the geographical concept is a
population concept. Geographical races are defined in terms of resemblance within
and variation among populations, rather than in terms of properties possessed by
the individuals that make up a race. A second reason is that, according to the
typological concept, kinds are defined by properties possessed by all and only the
members of a kind. Though the geographical concept also defines races in terms
of shared properties, such properties need not be universal nor unique. As noted
above, there must simply be statistically significant differences among the traits
used to individuate geographical races. Third, the typological concept requires
that the defining properties are intrinsic and explanatory. The geographical con-
cept, on the other hand, simply defines races in terms of clusters of properties
without assuming that such properties are due to a common intrinsic causal prop-
erty (or set of such properties). Indeed, there is often an implicit assumption that
external causal factors explain the clustering.
The geographical race concept was eventually abandoned for reasons originally