Religion (1995) 25, 31–34
Religion as Anthropomorphism: A New Theory that
Invites Definitional and Epistemic Scrutiny
E A. Y
This essay will focus briefly on (1) a definitional and (2) an epistemic analysis of Stewart
Guthrie’s cultural–anthropological theory of anthropomorphism in his book Faces in
the Clouds. In Part I of the essay, I will examine specific definitional claims about
religion that Guthrie advances in chapter 1 (‘The Need for a Theory’) and chapter 3
(‘The Origin of Anthropomorphism’). In Part II, crucial statements in chapter 6
(‘Anthropomorphism in Philosophy and Science’) and chapter 7 (‘Religion as
Anthropomorphism’) raise questions about Guthrie’s epistemic assumptions that in
philosophy and science the objects referred to as anthropomorphic have critically been
known to be errors and have been wisely set aside in the margins of those enterprises,
whereas the objects referred to as anthropomorphic in religion have always been at the
centre of religion. Guthrie employs five theoretical criteria (of observability, simplicity,
generality, fallibility, and probability) to explain why religion always anthropomorphizes.
The essay concludes with a formal question about the epistemic status of
Guthrie’s observability and universality criteria.
Definitional Claims
After stating in the introduction of his Faces in the Clouds (1993, hereafter FIC, pp. 3–7)
that religion should be understood as systematic anthropomorphism; that we anthropomorphize
because guessing in Pascal-like fashion is a good bet; and that anthropomorphism
is familiar, pervasive and powerful, Stewart Guthrie turns his attention, in the
first chapter, to the definitional problem in religion as a way of drawing sharp attention
to the need for a theory of religion. He claims that ‘writers have speculated on the
nature and origins of religion for well over two thousand years but have not produced
so much as a widely accepted definition. Instead, there are nearly as many definitions as
writers. Religion is difficult to define because definitions imply theories, and no good
general theory of religion exists’ (FIC, p. 8).
If no good general theory exists and that is precisely why religion is difficult to define,
then two logical implications seem to follow.1
Guthrie assumes, first, that all definitions
imply theories and, second, that when there are no theories of religion, there will
obviously be no definitions. But it is crucial to ask about the first claim in order to
question the second. Do all definitions imply theories? The logician I. M. Copi argues
otherwise in stating that only one of the five standard kinds of definitions—namely, a
theoretical definition—is ‘tantamount to proposing the acceptance of a theory, and, as
the name implies, theories are notoriously debatable. Here one definition is replaced
by another as our knowledge and theoretical understanding increase’ (Copi 1968,
p. 101). The other four, often referred to as stipulative, lexical, precising, and persuasive
definitions, do not linguistically and semantically imply theories. Only a
theoretical definition of religion (the definiendum) would imply a theory of religion (the
definiens).
Given the Copi clarification, Guthrie’s argument that definitions imply theories
would be strengthened if, and only if, he were willing to modify his original claim and
state that only theoretical definitions of religion imply theories of religion. Such a
modification would also help to explain why Guthrie would be entitled to ignore
(as he does) a variety of stipulative, lexical, precising, and persuasive definitions of
0048–721X/95/010031 + 04 $08.00/0  1995 Academic Press Limited
religion that have been employed in the discipline of religion (Penner and Yonan 1972,
pp. 110–17).
If Guthrie were willing to grant linguistic and semantic status to theoretical definitions
advanced by believers and non-believers alike, he would be in a better position to
understand, explain, and criticize the theories of religion advanced by both, especially in
relation to the problem of anthropomorphism. Believers and non-believers have
formulated theories that provide internal interpretations and external explanations that,
as Robert Segal rightly argues, account for religion in terms of meanings and causes
(Segal 1992, pp. 77–8). Whether or not definitions are theoretical would depend on
whether they were introduced, as Fred Suppe puts it, ‘in the initial formulation of a
theory, theoretical terms are introduced definitionally; for example, in the initial
formulation of thermodynamics entropy was introduced definitionally’ (Suppe 1977,
p. 205).
In the light of our need to be clear about the status of theoretical definitions, one can
ask Guthrie whether the theories he describes that have been developed by believers
(such as Schleiermacher, Otto, Eliade, and Tillich) and non-believers (represented by
Hume, Feuerbach, Freud, Malinowski, Durkheim, Tylor, Horton, Geertz, and Bellah)
really constitute two quite different sets. If so, how can these two theoretical sets imply
the same definitional claims that warrant Guthrie’s assertion that religion is nothing but
anthropomorphism (FIC, pp. 8–38, 62–90)? Are the different theories advanced by
believers compared to non-believers reducible to the common definition that religion
always anthropomorphizes (FIC, pp. 177–204)?
Epistemic Scrutiny
In chapters 6 and 7 of Faces in the Clouds, several important warrants need to be
scrutinized in order to ask what philosophical justifications are foundationally required
to support Guthrie’s epistemic assumption that in philosophy and science the objects
referred to as anthropomorphic have been recognized to be unalloyed errors and have
been wisely set aside in the margins of those enterprises, whereas the objects referred
to as anthropomorphic in religion have always been at the centre of religion (FIC, pp.
152–204). The epistemic difference that Guthrie assumes here is developed into a
theoretical and definitional claim: religion is anthropomorphism because it persists in
perceiving the non-human world in terms of human-like models (FIC, pp. 200–4).
Guthrie states the epistemic matter directly in terms of two standard explanations of
anthropomorphism: familiarity and comfort (FIC, pp. 73–90). Although ‘our conceptions
both of the human and of the non-human world are constantly subject to
revision’, nevertheless ‘anthropomorphism offers the greatest intellectual coherence
possible’, and ‘the search for comprehension that underlies it comprises the same
processes—of economizing, generalizing, ordering, and system-building—as does the
rest of thought’ (FIC, pp. 88–90). Anthropomorphism, like other model-like categories
of the mind, results not so much from a desire to find those that, given some context,
matter most.2
In Guthrie’s words, ‘scanning the world with human-like models, we
frequently suppose we find what we are looking for where in fact it does not exist’ (FIC,
p. 90).
It is precisely these Kant-like epistemic antinomies between the known anthropomorphic
models of the mind and the never-fully-known empirical objects that ‘people
in many fields—literary critics, journalists, philosophers, scientists, and others—call
attention to. As anthropomorphism is chased from one realm it springs up in another.
32 E. A. Yonan
Unmasking instances of anthropomorphism is like stamping patches of a bigger fire
because anthropomorphism stems from an effort broader than itself’ (FIC, p. 92).
As pervasive for Guthrie as anthropomorphism is, it culminates in religion. If all
domains of culture share a common (more than family-resemblance) feature of
ostensible communication with human-like beings through some form of symbolic
action, ‘religion always anthropomorphizes in more generalized, systematized, and
integrated ways’ (FIC, p. 200). ‘Religion not only anthropomorphizes (as everyone
admits) but is anthropomorphism’ (FIC, p. 200).
Guthrie’s definitional claim is embedded in his theory which ‘should correspond to
observation, should be simple, and should be general’ (FIC, p. 201). These theoretical
criteria of (1) observability, (2) simplicity, and (3) generality justify his definitional claim
that religion is what religion does: religion is anthropomorphism because anthropomorphizing
is the only thing all religions share. If anthropomorphizing rests on a persistent
and knowable mistake, then (4) error should be added to Guthrie’s theoretical criteria.
A Pascal-like bet should also be added as (5) a probability criterion to Guthrie’s other
four because it provides an explanatory and argumentative strategy for claiming that ‘If
we are mistaken about anthropomorphism, we lose little, while if we are right, we gain
much’ (FIC, p. 6).
Are there areas of religion that are non-anthropomorphic? If so, criteria (1) and (3)
would be open to serious criticism. If not, Guthrie wins his formal bet in claiming that
religion is nothing but anthropomorphism. He advances a good guess that anthropomorphism
is always familiar, pervasive, powerful, spontaneous, and even compellingly
evident because it generates models of the nonhuman world in humanlike terms. In any
case Faces in the Clouds is so rich and comprehensive in detail and argument that it will
surely inspire considerable debate in several disciplines for many years to come.
Notes
1 While I agree with Guthrie that defining religion has been a challenging and longstanding
problem and that no one definition of religion has been widely accepted, I disagree with his
definitional assumption that religion is unique in not having a single definition that meets the
various tests of uniform agreement. See Religion Defined and Explained on describing and
classifying specific definitions and theories of religion (Clarke and Byrne 1993). For example, is
it the case that all philosophers agree on a single definition of philosophy? The Institution of
Philosophy: A Discipline in Crisis presents contemporary disagreement and very little widespread
consensus (Cohen and Dascal 1989). Similarly, anthropology has had its own problems of
definition. In a very interesting article on ‘Waddling In’, Clifford Geertz argues that one of the
advantages of contemporary anthropology as a scholarly enterprise is that no one in the discipline
knows exactly what it is (Geertz 1985, p. 623). If one looks closely at these and other related
disciplines, one finds as much definitional diversity as in religion. Moreover, definitional diversity
does not seem to be the problem, especially when Guthrie claims that religion is difficult to
define because all definitions imply theories. Copi and other philosophers of language argue that
only theoretical definitions imply theories. See the arguments for a variety of definitions,
including theoretical definitions, in Definitions and Definability (Fetzer, Shatz, and Schlesinger
1991).
2 I am using the term ‘epistemic’ in the widely influential way Alvin I. Goldman uses it in his book
Epistemology and Cognition, where he states that ‘the multidisciplinary conception of epistemology
is the suitable way of integrating knowledge in many disciplines’ that have widely shared interests
(Goldman 1986, pp. 8–9, 181–4, 378–80).
Bibliography
Cohen, Avner and Marcelo Dascal (eds), The Institution of Philosophy: A Discipline in Crisis?, LaSalle,
IL, Open Court 1989
Clarke, Peter B. and Peter Byrne (eds), Religion Defined and Explained, New York, St. Martin’s Press
1993.
Religion as Anthropomorphism 33
Copi, Irving M., Introduction to Logic, 3rd edn, New York, Macmillan 1968.
Fetzer, James H., David Shatz, and George N. Schlesinger (eds), Definitions and Definability:
Philosophical Perspectives, Boston, Kluwer Academic Publishers 1991.
Geertz, Clifford, ‘Waddling In’, Times Literary Supplement (June 7, 1985), pp. 623–4).
Goldman, Alvin I., Epistemology and Cognition, Cambridge, MA, Harvard University Press 1986.
Guthrie, Stewart E. Faces in the Clouds: A New Theory of Religion, New York and Oxford, Oxford
University Press 1993.
Penner, Hans H. and Edward A. Yonan, ‘Is a Science of Religion Possible?’, Journal of Religion 52
(1972) pp. 107–33.
Segal, Robert A., Explaining and Interpreting Religion: Essays on Issues, New York, Peter Lang 1992.
Suppe, Fred, ‘The Search for Philosophical Understanding of Scientific Theories’, in Fred Suppe
(ed.), The Structure of Scientific Theories, Urbana, IL, University of Illinois Press 1977.
EDWARD A. YONAN is Griswold Distinguished Professor of Religion at Millikin
University, where he has taught since 1980. He is an editor and contributor to Religion
and Reductionism: Essays on Eliade, Segal, and the Challenge of the Social Sciences for the Study
of Religion, Studies in the History of Religions, No. 62 (Leiden, Brill 1994).
Department of Religion, Millikin University, Decatur, IL 62522, U.S.A.
34 E. A. Yonan