ideology, not a religion, while acknowledging that Nazism resembles a religion in several
respects. There have been similar controversies about whether Soviet Marxism is a
religion or is only analogous to a religion in some ways.
Spiro's proposal goes as follows: “I shall define religion' asan institution consisting of
culturally patterned interaction with culturally postulated superhu
end p.407
man beings' ” (1966, 96). On his view, interactions include both “activities which are
believed to carry out, embody, or to be consistent with the will or desire of superhuman
beings or powers” and “activities which are believed to influence superhuman beings to
satisfy the needs of the actors” (97). Because interaction thus understood requires only
belief in superhuman beings and not their reality, religious actors can, odd though it
sounds, interact with superhuman beings that do not exist. Assuming that the Wagnerian
gods are for Nazism window dressing rather than objects of serious belief, Spiro's
definition will not count Nazism as a religion. Nor will it classify atheistic and wholly
naturalistic varieties of Marxism as religions.
The obvious objection to Spiro's definition is that it is too narrow. Like Kant's definition,
it proposes defining conditions that religions without superhuman beings in their
ontologies, such as Theravada Buddhism, fail to satisfy. Hence, it does not provide a
necessary condition for being a religion. Spiro is aware of this objection and has a good
deal to say in response to it. According to one line of defense he offers, we must
distinguish between the teachings of atheistic philosophical schools and the beliefs of a
religious community. Even though the pure philosophical Theravada of the Pali canon is
atheistic, we always find it in traditional societies coupled with a belief system that is
committed to superhuman beings, such as the nats of Burma and the phi of Laos and
Thailand. Hence, “it cannot be denied that Theravada Buddhists adhere to another belief
system which is theistic to its core” (1966, 94). But not all Theravada Buddhists do
adhere to such a theistic belief system. So this line of defense has the awkward
consequence that Theravada Buddhism is, for some people, part of their religion but is
not a religion when it is found in its purest form.
However, Spiro seems to rest more weight on another line of argument. He insists that a
definition of religion must satisfy a criterion of intuitiveness. For him, at least, “any
definition of religion' which does not include, as a key variable, the belief in superhuman—I won't muddy the metaphysical waters withsupernatural'—beings who
have power to help or harm man is counter-intuitive” (1966, 91). Belief in superhuman
beings will, therefore, be a defining condition of religion according to any definition that
is intuitively adequate for Spiro. His appeal to a criterion of intuitiveness indicates that he
takes the task of defining religion to involve more than merely framing a definition that
will prove useful in anthropological research. In addition to satisfying this pragmatic
constraint, which will require cross-cultural applicability of a definition, an adequate
definition must also analyze or reflect the concept of religion the anthropologist brings to
the study of religious phenomena.
A comparison of the definitions proposed by Geertz and Spiro reveals two significant
kinds of disagreement between them. The two definitions are not even coextensive. Pure