This paper argues that the belief in and reverence for superhuman
     beings cannot be understood as the chief distinguishing characteristic
     of religious phenomena. The consideration of Buddhism has always been
     central to the discussion of what religion is, and this paper focuses
     on the limitations of the human-superhuman dichotomy as it might be
     used to apply to Buddhist traditions. The argument makes three points:
     a) There are important sects of Buddhism that do not rely on reverence
     for superhuman beings, and the concept "superhuman" is difficult (if
     not impossible) to use in cross-cultural studies because of cultural
     variations in what it means to be human; b) the insistence that
     "philosophies" should be systematically distinguished from "religions"
     is arbitrary and culturally biased; and c) Buddhist doctrines that
     assert that reality is ultimately "nondual" provide the conceptual
     context for understanding superhuman beings in Mahayana, and this
     conceptuality is not consonant with superhuman definitions of
     religion.

                                INTRODUCTION

The definition of religion continues to be a matter of dispute among
scholars, and I suspect that disagreements about the topic will persist so
long as religion is studied in academic circles. This is as it should be.
No area of academic specialization should allow its self-definition ever to
be a settled thing. Still, certain aspects of this discussion can be laid
aside from time to time as our understanding of religious phenomena becomes
more sophisticated.

A primary methodological issue to be decided in attempting to establish a
definition of religion is whether and how any specific etic concept can
allow for a sympathetic, nondistortional understanding of the various emic
religious phenomena within the universe of human cultures. This is an
especially important problem, because Western observers have frequently
exported ideas about religion into foreign contexts where they do not fit
the conceptuality of the peoples they study. Of course, we in the West are
predisposed to believe that religion has to do with the worship of
superhuman beings. Yet, following Durkheim, many scholars have concluded
that the reverence for superhuman beings is not universal among religions,
and they have suggested various conceptualities that are more complex for
defining and studying religions around the world. Those who have argued
against Durkheim's position have suggested that reverence for superhuman
beings is so widespread among the world's religions as to be almost
universal, and it supplies necessary constraints for the study of religion,
which becomes unnecessarily vague if allowed to follow Durkheim's
inclusivist views (Goody 1961; Horton 1960; Spiro 1966). Recently Orru and
Wang (1992) have argued in this journal that the reverence for superhuman
beings is basic even to Buddhism, the classic example of a nontheistic
religion.

In this article I hope to make a case against exclusivist definitions that
find the chief distinguishing marks of religious phenomena in the reverence
for superhuman beings. For reasons that will become clear as we proceed,
the nature of Buddhism as a religious system has been the linchpin in
arguments for nontheistic definitions of religion. I will argue that
theistic definitions of religion cannot account for many aspects of
Buddhism, which is normally considered to be one of the main religions of
the world. The nature of Buddhism as a religion, then, has important
implications for our understanding of the nature of religious phenomena
generally, and particularly for the ways in which we conceptualize our
studies of religion within the social sciences.

                   BUDDHISM AND REVERENCE FOR SUPERHUMANS

In order to establish that religion has no necessary relationship with
superhuman beings, we need only establish that some religious phenomena
exist that do not depend on a relationship between humans and superhumans.
Buddhism certainly has such forms, and these forms are hardly marginal
within the Buddhist tradition at large. Theravada Buddhism has long been
the example most frequently cited as an "atheistic religion" (e.g.,
Cunningham et al. 1991:24-25; Schmidt 1988:10-11). While it might be true
that many people within Theravada (both monks and laity) maintain
relationships with spiritual beings, it is also true that many do not.
Furthermore, many of those who do not worship spirits work strenuously
against the practices of the spiritists who according to the "atheists," or
"nontheists," as I prefer to call them) have imported superstitions into
the originally pure dharma as it was taught by the historical Buddha. For
example, I interviewed a number of monks in Thailand who were vehement in
their opposition to the spirit cults that had grown up around what they
considered to be the pure traditions associated with the historical Buddha
(see also Gombrich and Obeyesekere 1988:15-16). Even though it is a matter
of some dispute among Buddhists, particularly between the Mahayana and
Theravada branches, whether the historical Buddha encouraged the worship of
spiritual beings, the traditional orthodox position of Theravadins (that
the Buddha did not encourage spiritism, but rather a rigorous method of
practice that would lead to insight and the release from suffering) is
certainly a respectable position, both historically and soteriologically.
Around this nontheistic ideology have grown temples, ritual practices, a
sacred canon, pilgrimage sites, an elite corps of priest-monks, the
legitimation of political institutions, and practically everything else
that we would normally consider religious.

Theravada is not the only branch of Buddhism to evince this "atheistic"
attitude. Modern Theravada is the only surviving school of many so-caUed
Hinayana schools from ancient times, most of which stressed the importance
of individual human effort by monks to achieve enlightenment, in sharp
contrast with the reliance on spiritual beings that has been prominent in
other religions, including some branches of Buddhism. Yet, even within
Mahayana Buddhism, which in some cases allows for and encourages a reliance
on superhuman beings to achieve enlightenment), the dependence on
superhuman beings is not universal. For example, Zen is remarkable for the
irreverence some of its practitioners show toward superhuman beings. During
my initiation to Zen study some years ago, the Roshi's first words to me
were "There is no God." In this sect of Rinzai Zen, our practice was based
entirely on sitting meditation, Koan instruction, menial labor, silent
eating, and so forth. Never was there mention or worship of superhuman
beings.

It seems to me that even such a cursory examination of the evidence would
definitively exclude theistic definitions of religion, because the
existence of certain nontheistic Buddhist sects should be sufficient to
demonstrate that the concept "religion" is not necessarily linked with a
reverence for superhuman beings. We might let it pass with this. However,
there is a more fundamental flaw in such definitions: They do not define
their most important term, "superhuman," with sufficient clarity for it to
be applied to Buddhism (or other religions, for that matter) in a way that
would allow us to make sense of Buddhist subtleties. Instead, the use of
"superhuman" forces a distortion of the Buddhist worldview. Consider
Spiro's definition (1966:98):

     Superhuman beings. These refer to any beings believed to possess power
     greater than man, who can work good and/or evil on man, and whose
     relationships with man can, to acme degree, be influenced by
     [activities involving values and ritual].

It seems that Spiro presumes a Western, empirical, positivist model of
humanness, according to standard academic practice. Since what it means to
be "human" varies videly from culture to culture, however, we cannot take
for granted what it means to be "greater than human." Furthermore, it
cannot be assumed when studying religion in a non-Western culture that a
definition of humanity favoring our own view is warranted, without
prejudicing the case in advance. In order to understand whether Buddhism is
based on a reverence for superhuman beings, we must first understand what
Buddhists consider a human to be. Ultimately we must even consider how
Buddhists understand the nature of being itself. Only in this way can we
hope to understand the Buddhist view of "superhuman beings."

Humanness in the World;view of Samsara

In every culture that has formed against the backdrop of samsara,
individual humans have believed that they find themselves in certain
situations and with certain natural proclivities because of the karma they
have collected over countless former lifetimes. This applies to Buddhist
cultures as well as to several others that originated in the Indian
subcontinent, including Hinduism and Jainism. We might say that this view
accounts for all elements that go into forming human identity, including
both "natural" (genetic?) and "nurtural" (i.e., contextual) elements.
According to this view, every sentient being is responsible for its own
future according to its actions, insofar as every action entails certain
results according to the laws of karma Likewise, the true nature of any
being in existence is most properly understood not merely in reference to
causes that have influenced it since its birth in this particular lifetime,
but according to the entire history of events that have taken place, and
especially decisions that have been made throughout many previous
lifetimes. Such actions and decisions have defined the karmic make-up of
the person in its present form. I use the pronoun "it" in order to
emphasize that any being who finds itself as a human has a history that
certainly includes births as lower forms of life, which might include both
animals and depraved spirits. Decisions and acts that were made in these
previous lifetimes cumulatively determine the character and social location
of any human being. A mark of fully realized humans (that is, of those who
understand their true nature and how they relate to the order of the
cosmos) is that they are aware of all their previous lifetimes, because,
logically speaking, they would have to be aware of their entire history in
order to know themselves. This is true in the case of Gautama, and stories
about his previous births (as monkey, hare, elephant, and human prince, for
instance) are popular sources for instruction about basic Buddhist virtues
(Speyer 1971). However, in none of these stories is Gautama Bud&a ever said
to have been a "superhuman." The karma that Gautama collected so that he
could be born as a prince of the Shakya clan and take the final steps
toward enlightenment was from lifetimes as an animal or a human being. At
his death he had never become anything more than human, except insofar as
he came to realize his true nature, which is the same for all sentient
beings.

While Theravada Buddhists will talk about the Buddha's being
"supermundane," we should not conclude from this that they revere him as a
superhuman being in the same sense described by Spiro. Theravadins believe
the Buddha to have been a very special being. When he was born into this
world during the sixth century B.C.E., it was only the last of many
rebirths he had experienced in his journey through samsara. By this point
he had reached a stage of development that allowed him to do a great number
of very special things, things typically not possible for human beings. For
example, he was born into his mother's womb without the need of sexual
intercourse between his parents, and his mother gave birth to him through
her side with absolutely no pain (Corless 1989:5-6; Warren 1973:38-48).
Just before he died, trees bloomed out of season in order to shower him
with their blossoms (Warren 1973:95-96). However, these events in the life
of Gautama merely occurred near the end of a long string of events that
made up the Buddha's identity throughout his many lifetimes. When the
Buddha died, he ceased to exist as an individuated being, because, having
broken the bond between desire and suffering, he escaped the causal chain
that leads to continued individuated existence and rebirth. Even though
very few beings ever achieve the level of perception that Gautama did,
Theravada recommends the arhat path to those who want to pass into nirvana
at their deaths. After all, to attain a supermundane state in Buddhism is
not so much a matter of having miraculous powers or coming from another
world, but having a level of perception that enables any sentient being to
have insight into the true nature of reality that is always and everywhere
the same (Tambiah 1970:40-41; see also Angutarra Nikaya, quoted in von
Glasenapp 1966: 138). To the extent that the Buddha was a superhuman being,
we are all superhuman, in that our being transcends our humanness. Gautama
is distinct from other humans because he had come to understand this
"supermundane" reality with great clarity and constancy, and because he
could teach others about it (Carter 1982:19-23). Any arhat can come to
understand the same thing without ceasing to be human.

Considering all of this, how are we to understand the word "superhuman" in
a cross-cultural context? Certainly Buddhists claim powers and attributes
for Gautama that modern Western empiricists would not consider a part of
human potential. However, among those people who share in the worldview of
samsara, it is almost universally believed that human beings can do
wondrous things (Tambiah 1970:50; Eliade 1958). How to solve the issue of
cultural prejudice in the taxonomy of religious phenomena, so that the emic
structure of one culture/tradition is not favored over another, is a
problem in the interpretation of culture that cannot be overestimated as we
fabricate an interpretive conceptuality. I would suggest that the category
"superhuman" is inherently flawed and should be abandoned as a etic tool,
because it necessarily exports corollary emic biases from the culture of
the interpreter and would impose an alien, distortional hermeneutic on the
worldviews of cultures that do not share in the emic structure of the
interpreter's culture. This problem has implications not only for the
grosser problems of cross-cultural comparison, but also for the
interpretation of diversity within Western, Abrahamic traditions.

                     IS ATHEISTIC BUDDHISM A RELIGION?

Even those who are aware of these subtleties in Buddhist conceptuality
might persist in denying the relevance of atheistic forms of Buddhism on
the grounds that such sects as Theravada and Zen are not really forms of
religion after all. Such objections could take at least two forms.

Philosophy, Religion, and Nominal Definitions

First, it could be objected that it is necessary to maintain a clear
definitional boundary between "philosophies" and "religions" (Spiro
1966:93; Orru and Wang 1992:58-59). Following this line of reasoning, we
might maintain that a "philosophy" is a system that lends perspective to
human existence through beliefs and practices derived from the efforts of
human reason, introspection, and community discourse. Thus, the nontheistic
aspects of Theravada and Zen would be better understood as "philosophies"
and could be distinguished from the various "religious" forms of Buddhism
that give prominence to reverence for superhuman beings. One could argue
further that any insistence that all forms of Buddhism are religious
renders the argument tautological. There is no reason, it could be claimed,
to assume that every Buddhist sect is "religious" (as opposed to
"philosophical"), unless terms for the definition of religion are
established in advance that allow for atheistic sects to be understood as
instances of religious expression. This, of course, begs the question.
Arguments such as a religion, and all of its forms are therefore religious"
simply won't do.

As far as it goes, this is an important critique. Yet those who propose
nontheistic definitions for religious phenomena are not looking at things
quite so simplistically. In the first place, any definition is an analytic
(or tautological) statement: Bachelors are unmarried men. Thus, the mere
assertion that philosophical and religious systems are to be distinguished
on the basis of attitudes toward superhuman beings is every bit as
tautological as the position against which it is invoked. In fact, Goody's
(1961:148-155) review of the reasons for adopting a more inclusive
definition of religion (including nontheistic forms) is much more thorough
and convincing than are the reasons he uses to support his Tylorian,
animistic conclusion. Like Spiro, Goody chooses Tylor's position on almost
purely utilitarian grounds: If we define religion in this way, then it's
easy for us to define what we're studying in precise terms.

The problem that confronts us in the definition of religion is how to
escape the impasse of merely asserting that religion is "x," and then
searching for evidence in favor of one's position. Those who have argued
for more inclusive rather than nominal definitions implicitly advocate a
more inductive method, according to which the articulation of theoretical
constructs (including definitions) is part of the process by which we
understand individual cultural phenomena and the nature of religion at
large. Of course, this means that the construction of more inclusive
definitions will continue to be an unresolved matter, as definitional
debates and the empirical examination of cultural data proceed
reciprocally.

I suspect that the urge to establish nominal, exclusive definitions of
religion stems from the need of some sociologists and anthropologists to
define essential subcategories (such as religion) within their disciplines,
as contrasted with others (like me) whose primary interest is in religion
but who use sociological theory and method to direct their research. For
example, Horton's (1960:212) definition of religion as "an extension of
social relationships beyond the confines of purely human society" is
clearly (in the context of his article) an attempt to define religion in
such a way that he can advance a specific hypothesis about social relations
and discourse about spiritual beings. For the purposes of his article, his
definition makes good sense. However, it is quite another matter for Horton
to indict universally those who discuss more inclusive definitions as
"carry[ing] on an endless and entirely barren argument" (1960:201), or for
Spiro (1966:87) to describe such discussions as "interminable (and
fruitless) controversies."

The starting point for an inductive discussion of this topic must be W.
Cantwell Smith's (1962) seminal study, The Meaning and End of Religion. By
extension from Smith's arguments about the history of the meaning of the
term "religion," we can conclude that "religion" is an emic concept that
has had a variety of uses in European languages and cultures over the
centuries since it was first used in its original Latin forms. Only in
relatively recent times has it begun to connote the sorts of things that we
in the comparative study of religion now discuss. There is a simple reason
for this. For most of religio's long history, it had meaning in the
relatively narrow conceptual context of European Christian culture. As
European imperial culture began to encounter the religions of the Americas,
Africa, and Asia, intellectuals in Europe (such as Tylor and Durkheim)
debated ways in which the word "religion" might be used to describe the
variety of cultural forms that they had discovered in the dawning age of
global awareness. For better or worse, studies in comparative religion have
developed as attempts to describe a variety of cultural forms on analogy
with generally accepted paradigms among "the world's religions." The five
most important and least ambiguous expressions of religion in the world
have been Judaism, Christianity, Islam, Hinduism, and Buddhism. To the
degree that anything is called "religious," it must in some way be
analogous to the sorts of forms that are found within these major
traditions. Practically every introductory textbook on religion is
testimony to this fact, and Orru and Wang (1992) add their testimony
implicitly, since they do not try to discredit nontheistic Buddhism, as
such, from being included in the discussion, but rather to show that
Buddhism is not as nontheistic as some interpreters have thought it to be.

Still, the use of the term "religion" to describe certain nontheistic forms
of cultural expression is not merely an accident of history. The intuitions
of early explorers, warriors, and scholars were right when they understood
Theravada to be analogous with Christianity and Judaism, even if some did
not understand that the Buddha is not revered as a god, and even if they
had a condescending attitude toward it as a heathen faith. If I may repeat
a line from an earlier paragraph, the nontheistic forms of Buddhism
(including Theravada and Zen) include temples, ritual practices, a sacred
canon, pilgrimage sites, the reverence for saints, an elite corps of
priest-monks, the legitimation of political institutions, and practically
everything else that we would normally consider a religion to include. Of
course, whether this conclusion is valid depends on an empirical
examination of the cultural facts. We cannot argue this point here.
However, the cluster of such cultural forms under the heading "Theravada
Buddhism" looks sufficiently like other clusters from Medieval Europe,
Bronze Age Egypt, and Trobriand Melanesia that many scholars find it useful
to discuss them all as expressions of "religion," even though some of them
are not linked in any thorough way with the reverence for gods and spirits.

To consider some of these forms as "philosophies," simply because they do
not have the reverence for superhuman beings at their core, is surely a
case of exporting cultural bias. For us in the West, particularly for those
of us who work in institutions so heavily indebted to the European
Enlightenment, the differences between philosophy and religion might be
important. However, the distinction is certainly not a cultural universal,
and we should not insist that it is necessary to distinguish between
"philosophies" and "religions" in cross-cultural studies. There are two
reasons for this. First, intuitions about religion among Westerners are not
as straightforward as some might think (Spiro 1966:91; Horton 1960:211).
Enlightenment thinking has so penetrated religious ideologies in the West
that beliefs about ultimate reality among many "religious" people are
nearly indistinguishable from those thought to be more "philosophical." For
example, many religious people in the West, particularly some liberal
Protestants and Reform Jews, direct their reverence toward the "Ground of
All Being" rather than to individual superhuman beings, and they seem to
have as much in common with their academic philosophical counterparts as
with Jewish or Christian Fundamentalists. Likewise, many of those involved
in the theology of liberation have so adopted Marxist principles that they
work more to reform the social infrastructure than to appeal for
supernatural intervention on their behalf, even though they regard
themselves as the modern inheritors of the biblical tradition. Hence, for
those of us who have studied the last 200 years of Western theology, it is
not a simple intuitive matter to distinguish religion from philosophy on
the grounds that the former deals with superhuman beings.

Even more importantly for our argument, in South and East Asia the
distinction between philosophy and devotion is simply not the matter for
concern that it is in the West. The articulation of the trimarga in the
Bhagavad Gita, for example, insists that "philosophy" (jnana), work, and
devotion are simply three different aspects of one spiritual path; any
distinction that we draw among them is a product of our delusions. Even
those Hindus who understand the trimarga hierarchically (e.g., Shankara's
conviction that jnana is a superior path to devotion) understand the
overall purpose and goal of these different paths to be the same.
Similarly, although Buddhists might disagree among themselves about the
relative merits of theistic and nontheistic approaches to the dharma, the
all agree that what they share in their respect for Gautama and in their
reverence for the dharma binds them together as parts of a common
enterprise.

The most telling example regarding this distinction is the insistence among
some that Gautama Bud&a taught a philosophy and not a religion. It is easy
enough to understand the hestitation of a scholar to apply the term
"religion" to the Budha's teachings because of his lack of concern with
superhuman beings. However, a similar trepidation should accompany the
attribution of "philosophy" to his teachings. Ignoring for a moment that it
would be just as hard to establish any definition of philosophy to which
all philosophers would agree, or which would apply to every person revered
in the West as a member of the philosophical pantheon, we must surely admit
that the Bud&a has no analogue among Western philosophers who would compare
with him so closely as would Jesus or a Sufi saint. The Bud&a was not,
after all, a man who spent long hours in rational argument (although he did
participate in this, and well), as much as one who devoted himself to a
strict ascetic regimen and to meditational practice. This way of life has
much more in common with the Rosicrucians, Kundalini Yogins, or Thomas
Merton than it does with Aristotle, Aquinas, Hume, Kant, or Wittgenstein.
The same could be said of many masters in the Theravada and Zen traditions.

So, while it might be a difficult matter to define precisely how
nontheistic Budhism is a religious system, it is rather more difficult to
discuss Buddhism (or any of its schools) as a "philosophical" system. The
real problem with fitting Bud&ism into any scientific taxonomy is that of
cross cultural comparison: How are we to use an etic terminology to
describe emic conceptual systems in a nondistortional way? This is no mean
task, but it will simply not do to impose a naive Western dichotomy on
South and East Asian traditions. Scholars who use nontheistic, inclusive
definitions of religion insist that the word "religion" be defined not
merely according to the intuition of Westerners, nor from the need to
establish operative concepts by scientists, but that it be derived
inductively from the careful consideration of ideas, institutions, and
practices as they are understood in various cultures of the world.

Nontheistic Religious Forms as Religious Epiphenomena. Elite vs. Folk
Practice

It is currently in vogue for scholars of religion to pay much greater heed
than in the past to popular ideas and practices that have provided the
social context and material wherewithal for the elites who have theorized
about the nature of reality (e.g., Tambiah 1970,1984; Spiro 1982; Gombrich
and Obeyesekere 1988); and rightly so (see especially Tambiah 1984:7-8). In
popular quarters, religious belief in practically every tradition
(including Theravada Buddhism) seems to be marked by the reverence for
superhuman beings. Some have concluded from this that elite beliefs and
practices that depart significantly from the dominant pattern among popular
groups are secondary religious phenomena, and should not be included among
the raw data from which a definition of religion is fashioned. There is
some meat to this position, particularly given the dependency of elites
upon the masses for their material support and the reciprocal relations
that are established between them in matters of ritual and institutional
affiliation.

Still, even if the social reality of elite religious traditions relies on
the existence of popular belief in "superhuman beings," this is an
empirical matter to be demonstrated, not a matter of definition. For, even
if elites do not exist in the absence of popular traditions that assert the
existence of "superhuman beings," any adequate definition must also take
into account the attitudes of elites who have traditionally maintained
control of dominant intellectual traditions and social institutions.
Morever, certain elites do maintain active relationships with their support
groups that entail the popular embrace of their "nonsuperhuman" belief
systems. For example, Rudolf Bultmann and Paul Tillich were both members of
the Church during their entire professional careers; they taught at highly
respected theological institutions, where they educated several generations
of clergy to service in European and American churches; they preached from
the pulpit regularly in good Protestant fashion; they had a wide popular
readership for their writings; but the "God" who was the center of their
work and lives would be better described as "the Ground of All Being"
rather than as a "superhuman being." To suggest either that Tillich and
Bultmann promoted the worship of superhuman beings, or that their
relationship to modern Protestantism is epiphenomenal, would be a serious
distortion of modern Christianity. While we would not want to say that
Bultmann and Tillich were proponents of a nontheistic religion, their
understanding of God was so sophisticated that to describe it as reverence
for "superhuman beings," or as the extension of social relationships beyond
the merely human world, would be the grossest of distortions.

Similarly, certain elites of Buddhism have traditionally discussed the
Buddha not as a "superhuman," but as a highly evolved human being who had
wonderful things to teach, and his teachings have been the focus of their
traditions (Swearer 1989). Are the monks who preside over the cluster of
beliefs, practices, and institutions of Theravada Buddhism nonreligious
and/or religious epiphenomena, simply because they do not utilize metaphors
about superhuman beings? To follow this line of reasoning to its disastrous
conduction, we would have to conclude either that the Buddha himself was
secondary to Buddhism (since his method for attaining enlightenment was
recommended for and practiced by an elite corps of monks dependent for
their well-being on the donations of lay folk who certainly did not
understand the subtlety of their thought), or that Buddhism itself is only
secondarily a religion (because of the spiritual beliefs of its lay
supporters).

Finally, if our definition of religion does not allow for the inclusion of
the subtlest thinkers and spiritual masters of a tradition, then we must
wonder about its usefulness. Any adequate definition of religion must not
only allow for the most common aspects of religious tradition to be
included within its scope, but it must also provide a conceptuality that
will allow us to appreciate the intellectual subtlety of the greatest
religious thinkers. Otherwise, it would be difficult for us to describe the
functional symbiosis of elite and popular traditions, and there would be
little incentive for us to master intellectual traditions that might
enliven our own elite discourse.

                          BUDDHISM AND NONDUALITY

My argument to this point, and in the paper as a whole, is not to provide a
new or refined definition of religion. Those offered by Geertz (1966),
Cunningham et al. 41991:11-32), Schmidt (1988:3-28), and Nielsen et aL
(1988:3-10), for example, might serve as models for what I would propose,
especially insofar as they do not attempt to provide simple bipolar
definitions of religious phenomena. The attempt to isolate religious
phenomena among all social phenomena is a wrongheaded approach to the
problem, since social phenomena of any specific type (economic, political,
philosophical, artistic, or religious) are so interconnected with those of
other types, that finding any hard and fast distinguishing marks among them
is a vain and methodologically ill-advised hope. Thus, I agree with Goody
(1961) in rejecting the "sacred-profane" dichotomy by itself as an adequate
basis for the definition of religion, since it is vague and far too
reductionistic. Yet no other such simple polarity as "human-superhuman"
will serve any better to define the complex set of phenomena that scholars
of religion choose to study. Likewise, I agree with Horton (1960:201) that

     to go ahead with the comparative study of religion while leaving the
     scope of the term undefined is to behave in a self-stultifying way;
     for until some fairly precise criteria of inclusion of phenomena in
     the denotation of "religion" have been given, it is impossible to
     specify those variables whose behaviour we have to try to explain in
     our study.

This does not imply (as Horton does) that the definition should be
fashioned according to a single bipolar scale, or that the human-superhuman
dichotomy will serve as even one useful part of more complex definitional
schemes.

My argument in this paper is that apart from the problems of reduction that
accompany the "human-superhuman" definition, such a taxonomy cannot even
account for important aspects of religion where such a dichotomy might
apply. As such, the "human-superhuman" dichotomy should be either abandoned
or radically revised as a primary means of distinguishing religious
phenomena

There can be no doubt that an important part of religion generally is the
concern of humans to come into relationship with some sort of power, being,
personality, or spirit that is more closely connected with ultimate reality
than are we ourselves. However, in the history of the world's religions, as
people have established their relationships with ultimate reality and
articulated their understanding of it, many of them have not expressed
their understanding in teens of superhuman beings. In the subtle and
artfully crafted statements of many religious peoples around the world,
ultimate reality has been conceived in many different ways. Sometimes
superhuman beings are considered to be the ultimate source of reality, but
elsewhere such beings can even be considered impediments to establishing an
appropriate relationship with being itself, which is the expressly stated
goal of their religious practice.

Nowhere is this clearer than in Mahayana Buddhism's articulation of
nonduality as the fundamental characteristic of all reality. This
orientation to descriptions of the world is so basic to Mahayana Buddhist
thought that it influences not only the elite "philosophical" reflection
about the nature of things in the world, but also the way that superhuman
beings are conceived in those aspects of Mahayana that hold reverence for
superhuman beings to be a fundamental part of Buddhist life and practice.
In order to understand how all Buddhist schools, both theist and nontheist,
are unified by principles more fundamental than whether their practitioners
are theists, we must understand those aspects of Buddhist conceptuality
that define what it means to be a superhuman being, especially for those
schools that revere them. In this way we will be able to see that Buddhists
generally do not understand differences of opinion among themselves about
superhuman beings to be as important as their agreement on principles that
are believed to be indispensable for the liberation of sentient beings from
suffering. We shall also be able to see both that superhuman beings are
relatively unimportant in Buddhist conceptuality as a whole, and that it is
arbitrary to distinguish between "religious" and "philosophical" Buddhist
sects.

Before turning to Mahayans views, let us consider orthodox Theravada views
on the relationship of the Buddha to the central teachings of Buddhism,
which are known collectively as the dharma (Pall dhamma). It is not
uncommon for Westerners to believe that the dharma, like the Gospel or the
Torah, is valid because it derives from a superhuman source. For example,
Orru and Wang (1992:50-56), in their argument against Durkheim's use of
Buddhism to establish the usefulness of sacred-profane for defining
religion, contend that the Four Noble Truths are not sacred in themselves,
but have importance because they are related to the "superhuman" Buddha.
They claim that whatever authority the Four Noble Truths have in Buddhism
derives from their connection with the Buddha himself, especially in his
transcendental form. According to this interpretation, ultimacy in Buddhism
depends on the ultimacy of a superhuman being.

This representation of Buddhism is seriously distortional, even apart from
those issues discussed already. There are important traditions (especially
in Theravada) that report the Buddha as instructing his followers to revere
his teachings more than him. In fact, the implications of two basic
Buddhist principles, "impermanence" and "no-self" (anitya and anatman,
respectively [Pali: anicca and anatta]), have been understood to imply that
the phenomenal appearance of Siddhartha Gautama as a human being is nothing
to be clutched after, because his existence as such was, at best, a passing
sight or illusion. However, the truths that he taught (e.g., the Four Noble
Truths) have a more enduring importance because they can help others to
recognize their own true, impermanent nature (Conze 1980:25-26). Thus, many
Theravadins consider the deification of the Buddha to be an important
problem, because it obscures the centrality of the Budha's teachings, which
they consider to be the heart of the religion (Tambiah 1970:42-46). One way
that Theravadins have conceptualized this problem is to distinguish between
the Buddha's rupakaya ("form body") and his dharmahaya ("dharma body"). His
physical body (or rupakaya) continues to exist only in relics (like
toenails and hairs), which are the object of reverence. His transcendent
spirit is thought to exist in his teachings, or dharma. Thus, to the extent
that the Buddha is a transcendental being, it is equivalent with his
teachings, which include primarily the four noble truths (Reynolds and
Hallisey 1987:325; Corless 1989:36-41).

According to this way of understanding the importance of the Buddha and
dharma within Buddhism, it would be inappropriate to say that the dharma
depends for its validity on its origin with a superhuman Buddha. Yes, the
achievements of the Buddha stand as evidence of its effectiveness as a
method for attaining enlightenment. Yet the dharma is effective (say
Buddhists) because it is true, not because the Buddha spoke it from
superhuman heights. In fact, if forced to point the causal arrow, we would
probably be more correct to say that the Buddha became a "refuge" because
he came to understand the dharma, rather than the other way around, since
it is said that the dharma is discovered periodically by Buddhas who then
bring its truths to suffering beings (Gombrich and Obeyesekere 1988:17;
Carter 1982:21-22). In any case, the successful use of the dharma by
thousands of monks over the past two-and-a-half millennia would serve as
sufficient proof that the dharma has a special, sacred quality. This is why
upon his death the Buddha recommended that members of the Sangha not pay
him excessive reverence with garlands and offerings, but instead rely on
their insight with the dharma as their "rude:

     So Ananda, whether now or after my decease, whoever you are, you must
     remain as islands to yourselves, as defences to yourselves with the
     Dharma as your island and the Dharma MS your defence, remaining
     unconcerned with other islands and other defences. If you ask the
     reason for this, then know that whether now or after my decease,
     whoever remain an island to themselves, as defences to themselves,
     with the Dharma as their island and the Dharma MS their defence, not
     concerning themselves with other islands and other defences, such ones
     are the foremost of my questing disciples (Williams 1989:11; see also
     Warren 1973:95-110)

Even more significant than these Theravada views are those in Mahayana
about the relationship of the Buddha and superhuman beings to the dharma.
Within Mahayana the word dharma has an expanded meaning that includes the
description of reality as ultimately nondual. Since it is within Mahayana
that the reverence for superhuman beings is most prominent within Buddhism,
an important issue to be addressed is how to understand the nature of
superhuman beings as part of a world that is understood to be lacking all
duality, such as the distinction between human and superhuman beings.
Thinkers throughout the Mahayana traditions have considered this problem
with considerable subtlety, especially as a response to the Theravada
critique that the inclusion of theistic ideas is a departure from the
teachings of Gautama. In their consideration of this topic, many Buddhist
thinkers have developed a sophisticated vocabulary to describe their
position, including such concepts as pratitya-samutpada ("dependent
co-arising") and sunyata ("emptiness, openness, no-essence"). In every
Buddhist school from Pure Land, to Zen, to the Tibetan Gelukpas, and among
men (such as the Dalai Lamas) who are regarded by the masses not only as
"philosophical" experts but also as the chief votaries or incarnations of
certain superhuman beings, the argument of many Buddhist thinkers has been
that the reverence paid to gods and spirits is only a vehicle to the
realization of nonduality and the consequent liberation from suffering.
They would contend that this goal is the same for all Buddhists. Whether
one approaches it through superhuman beings or through nontheistic
discipline is not a matter of fundamental significance. Thus, those members
of Mahayana who revere superhuman beings typically consider themselves part
of the same tradition and enterprise as those who do not.

One way in which Mahayana thinkers have explained the relationship of
nondual reality to the distinctions apparent among various phenomena has
been in terms of the trikaya ("triple body") of the Buddha. Confronted, as
they were, with the apparent paradox (on the one hand) of revering the
person of the historical Buddha and all of his previous incarnations, while
(on the other hand) accepting the principles of anitya and anatman,
Mahayana thinkers devised a conceptual system that would allow them to
revere individuated sacred beings while granting such beings no final
absolute, or transcendental status. Through the doctrine of the trikaya,
they identified three aspects according to which reality can be apprehended
and understood, but Mahayana Buddhists believe that the true nature of
reality is the same throughout all three of these realms. Still, the saving
power of the dharma (as a set of Buddhist teachings known through dual
consciousness) has been and continues to be manifest in only two of these
realms. (On trikaya see Corless 1989:35-41; see also Williams
1989:167-184.)

First, the nirmanakaya (or "manifestation body") of the Buddha refers to
his human form which was observable to humans of his time through normal
powers of perception and observation. Since the appearance of the Buddha as
Gautama was such an important occasion for the propagation of the dharma to
beings in need of salvation, Buddhists all over Asia revere his appearance.
They bow down in front of images of him; they make pilgrimages to places
where he was born, where he taught, and where he died; and they revere
those body parts (such as toenails and hairs) that have been preserved in
special shrines over the centuries. In all of this reverence for the
historical Buddha, we can say that people revere sacred images, places,
objects, holidays, and (above all) his teachings. Yet Shakyamuni himself is
regarded by the popularly accepted masters of the tradition not primarily
as a superhuman being (although he is sometimes regarded as such), but
rather as a very highly evolved person who recognized his true nature and
the nature of suffering; these are the same for all sentient beings.

The second body, the sambhogakaya ("enjoyment or bliss body"), refers to
that aspect of reality which is populated by gods and spirits. It is
Mahayana belief that just as the dharma has been spread through human
agents such as Gautama Shakyamuni, it is also helped along by the
intervention of spiritual powers from beyond the world of normal human
perception. According to some, Mahayana beliefs in superhuman beings such
as celestial bodhisattvas and buddhas qualify it as a religion. However,
within Mahayana at its broadest scope, belief in gods and spirits is
nothing more than the recognition of beings that are commonly accepted
throughout Asia as operating in the world. These beings are not the source
of all reality; they are not the creators of this world, nor do they find
fulfillment in their mode of being as gods (Gombrich and Obeyesekere
1988:15-17). Rather, gods and spirits as understood within Mahayana are a
part of conditioned reality, and they are in as much need of enlightenment
as are their human counterparts. In fact, it is generally recognized that
sentient beings cannot make significant karmic progress as gods, but must
become human to make further progress (Tambiah 1970:40). So, while it is
true that Buddhists often worship enlightened divine beings because of
their ability to give good things to people, it is also the case that
deluded gods receive instruction from enlightened humans about superior
modes of consciousness that can only be understood through the teachings of
Buddhism. In some cases even bodhisattvas of the highest rank are reported
to have received instruction from human beings who understood the dharma
better than they (viz. Vimalakirti Nirdesa Sutra, contra Horton 1960:211,
on the dependency of humans). Again, the teachings of Buddhism are
important to Buddhists, as are the "skill in means" (upaya) through which
these truths are made convincing to people. Superhuman beings are important
only to the degree that they are able to manifest the truths of the dharma.

Ultimate reality is the realm of the third body, the dharmakaya ("dharma
body"). This is the sphere of nondual thinking, of immediate spontaneous
perception, and of infinite consciousness. The Buddha's true nature, and
the nature of reality in general, is perceived correctly only in the
dharmakaya, because all reality is, according to Mahayana thought, nondual.
Although the true nature of things as such (including the Buddha and
ourselves) can be known, it can never be expressed in language or in any
other form of symbolic representation, because all language and symbols are
dual by nature. The most that can be said of even Buddhist traditions is
that they are constructed within a conceptual universe that is aware of its
own fundamental incoherence. Buddhism would claim that it represents the
best means of moving people beyond themselves into the world of nondual
awareness that allows for the only accurate perception of reality. Yet it
is also frequently said that those who are truly enlightened must move
beyond even Buddhism itself, since it, too, is a dualistic form of
consciousness. It is in this conceptual context that we must understand the
famous Zen adage: "If you meet the Buddha, kill him."

Still, Buddhists are fully cognizant that it is necessary to work with
people as they are, if the truths of Buddhism are ever to have effect in
transforming people's consciousness. Furthermore, nearly all people, as we
find them in the world and in culture, think, act, and live in a world of
duality. Thus, in order to focus people's concentration on those things
that "tend towards edification," it is necessary to emphasize certain
people, places, images, temples, dates, teachings, and even superhuman
beings as means of salvation. Of course, once one is fully enlightened, the
illusory nature of even the most sacred symbols becomes apparent. Anyone
who has spent time studying Buddhist thought knows that no being is truly
superhuman (or subhuman, for that matter!, because all being is nondual.

Consideration of the trikaya, especially in conjunction with the preceding
discussion, discloses that naive Western concepts about superhumans will
lead to spurious conclusions about the nature of beings as they are
perceived within the worldview of samsara in general, specifically
regarding the status of superhuman beings who are properly understood only
against a background of emptiness and nonduality. Furthermore,
distinguishing nontheistic Buddhist sects as "philosophies" or as otherwise
nonreligious is entirely arbitrary, since the trikaya, as well as other
Buddhist doctrines, are meant to make sense of the unity of Buddhist
traditions and institutions which take a variety of approaches towards the
realization of nondual consciousness. That is, the fact that Zen or even
Theravada does not stress the reverence of superhuman beings is irrelevant
for Buddhists in establishing it as a qualitatively different sort of
enterprise from other sects within Buddhism, because their common purpose
in attempting to realize the dharma overrides any minor difference that
might otherwise distinguish them. Hence, since Mahayana Buddhists
themselves consider it a matter of little importance whether one of their
sects reveres superhuman beings, it is not arbitrary for scholars of
comparative religion to conclude that Zen and other nontheistic sects
within Buddhism are examples of religious phenomena that have no intrinsic
relationship with spiritual beings.

          IMPLICATIONS FOR THE SOCIAL SCIENTIFIC STUDY OF RELIGION

Allow me to expand on three points implicit in the foregoing discussion.
First, religion is nothing if it is not intentional. What people believe
about the world, about ultimate reality, and about their own religion is
indispensable for the characterization of religious phenomena Within the
many emic universes that help to make up the world's religious traditions,
there are many subtle conceptual systems that must be understood in their
own terms before any attempt can be made to fabricate an etic conceptuality
that will account for the details of all of them in a comparative contest.
Diversity among religious conceptualities about the nature of ultimate
reality (with which all religions are concerned) is simply too great to
allow for the simplistic conclusion that religion has to do with those
aspects of human culture that make reference to gods and spirits. To define
religion in this way would exclude some of the most influential religious
ideas and practices the world has known. The Bud&t tradition is among the
moat important of these examples, but many other religious phenomena could
certainly be noted. Hammering an etic device onto religious phenomena may
facilitate the construction of operational models, but it will not help us
to understand the subtlety of human belief and behavior as observed among
religious people from time immemorial.

Second, one of the most important theoretical advances to be made within
the social sciences during the last several decades has been the
recognition that observers of culture must deal honestly and fairly with
the integrity of worldviews incommensurate with their own. Scientists
should be objective, but they should not be imperial Corollary with this
has been the realization that the integrity of worldviews is denied
implicitly when differences among them are collapsed, or discounted as
"accidental," for the convenience of a comparative paradigm. Thus, many
theorists now consider the recognition of "difference" among cultural
expressions to be at least as significant as the definition of similarity
and commonality. Differences among religious traditions in the way that
ultimate reality is conceived and approached comprise one of the most
important defining features of religious phenomena in human culture. To
collapse such diversity and difference into the pat "religion has to do
with superhuman beings" relies only on the most parochial of prejudices
regarding religion, and it shows no awareness of the subtle forms of
thought about ultimate reality and values that have been articulated among
philosophers, poets, theologians, shamans, and gurus in the world's
religious traditions.

We scientific observers of religious phenomena would be wise not to use
definitions that establish simple essences or functions as the defining
characteristics of religions, but to recognize certain general categories
within which there is widespread disagreement among religious peoples
(e.g., Nielsen et al. 1988:1-19). In the case under consideration, it would
be better not to define religion as necessarily dealing with gods, spirits,
or other superhuman beings, but rather to consider the ways in which
religious people approach and conceive of ultimate reality. There is
certainly no dear agreement among Theravada Buddhists, Mahayana Buddhists,
Advaita Vendantists, Sufis, Fundamentalists, Christian Feminists, and
Shamans about how this is to be done. What makes the study of religion so
utterly stimulating to the theoretical mind is the challenge of how to
think about seemingly irreconcilable differences such as these. Having once
considered the broad range of traditionally recognized religious thinking
about such matters (i.e., in the traditions recognized as paradigmatic,
listed above), we are then hard pressed to show why Marxists and American
Civil Religionists should be excluded from consideration by the taxonomists
of religion, since they too, after all, have ideas about the nature of
ultimate reality (e.g., as dialectically material) that are set in a
context of ritual, sacred space, and so forth. Whether such movements as
Marxism should be understood as religious is certainly a question too
complex to be argued here, but the consideration of such matters should not
be excluded in advance by a too-narrow theoretical perspective. It is
through the investigation of such marginal cases that refinements in
taxonomy always advance. Even if the consideration of marginal cases only
helps us to understand traditionally recognized religious systems more
clearly, then expansive definitions will certainly have served a purpose.

Finally, if we in the scientific West can take the integrity of religious
conceptuality seriously, and if we consider especially the case of
Buddhism, we might reflect more profoundly about the relationship between
"words" and "things." There is no discrete part of reality that can simply
be labelled "religion." Nor does the fabrication of operational definitions
create it. Religion exists as an aspect of human culture to the extent that
the interpreters of culture can use the term to make sense of human
societies. As academic researchers, we should use the term "religion" only
in ways that increase the subtlety of our descriptions. Thus, as we use the
term in academic discourse, our definitions must be continually revised as
we come to understand more and more about the phenomena that we study in
its name.

There is creative power in words, and in theorizing we bring worlds into
being. As we construct operational models of religion, we must be careful
not to destroy the worlds we intend to observe by creating new ones that
are intolerant of subtle differences. primary goal of religious studies is
certainly to develop models that do full justice to the emic diversity of
religious phenomena As a religious system, Buddhism presented probably the
earliest articulation of how to deal with this problem evenhandedly through
the conceptuality of the trikaya. In addition to helping articulate the
relationship of superhuman beings to the truths of the dharma, concepts
like the trikaya have also helped to make sense of the diversity of beliefs
and practices in the multiethnic, international Buddhist world, which has
included forms of shamanism, amulet spiritism, tantric practice, silent
meditation, as well as highly rational philosophical speculation. If
Durkheim and other inclusivists have had any weaknesses in their use of
Buddhism, those weaknesses were not found in their conclusions about the
nature of religion. Rather, they (and all of us) have failed to recognize
that Buddhists are not merely objects for our study, but co-equal partners
in the effort to establish a conceptuality adequately sophisticated to
allow for the appreciation of diversity and paradox among the world's
religions. Buddhist intellectuals have recognized the diversity of belief
and practice within their own traditions, and they have crafted a subtle
conceptuality that allows them to make sense of core doctrines without
belittling the diverse attitudes (superstitious and otherwise) that give
Buddhism its rich texture. We do neither Buddhism nor the comparative study
of religion any service by using conceptual tools that flatten this
diversity into a convenient operational package.

In his attempt to move interpreters of Buddhism away from the too
predominant bias of interpreting Buddhism as an ideal tradition of pure
teachings, Tambiah (1984:7) has written:

     to confine the study of "religion" to the doctrinal beliefs and
     philosophical constructs is, as Wilfred Cantwell Smith has argued, an
     unfortunate rationalist Enlightenment legacy that both unduly narrows
     and pauperizes the phenomenon. For me, Buddhism is a shorthand
     expression for a total social phenomenon, civilizational in breadth
     and depth, which encompasses the lives of Buddhist monks and laymen,
     and which cannot be disaggregated in a facile way into its religious,
     political and economic realms as these are currently understood in the
     West.

Although Tambiah is arguing a case somewhat different from this one, his
remarks about Buddhism in particular caution us against using categories
that would "narrow and pauperize" the study of religion in general. There
can be no question that reverence for "superhuman beings" does play a
significant role in the consciousness of many religious people on the face
of the globe, but since it does not include all religious consciousness,
especially not the most profound insights of the acknowledged masters of
most traditions, we should be careful not to universalize from the piety
shown to gods and spirits and say, "This is religion."

                                 REFERENCES

Carter, John Ross 1982 The threefold refuge in the Theravada Buddhist
tradition. Chambersburg, PA: Anima Books.

Conze, Edward 1980 A short history of Buddhism. London: George Allen &
Unwin.

Corless, Roger J. 1989 The vision of Buddhism: The space under the tree New
York: Paragon House.

Cunningham, Lawrence S., John Kelsay, R. Maurice Barineau, and Heather Jo
McVoy 1991 The sacred quest: An invitation to the study of religion. New
York: Macmillan.

Eliade, Mircea 1958 Yoga Bollingen series 56. New York: Pantheon.

Geertz, Clifford 1966 Religion as a cultural system. In Anthropological
approachs to the study of religion, edited by Michael Banton, 1-46. New
York Praeger.

Gombrich, Richard and Gananath Obeyesekere 1988 Buddhism transformed
Religious change in Sri Lanka Princeton: Princeton University Press.

Goody, Jack 1961 Religion and ritual: The definitional problem. The British
Journal of Sociology 12:142-164.

Horton, Robin 1960 A definition of religion and its uses. Journal of the
Royal Anthropological Institute 90:201-226.

Nielsen, Niels C., Jr., Norvin Hein, Frank E. Reynolds, Timothy Erdel, et
al. 1988 Religions of the world, 2nd ed. New York: St. Martin's.

Orru, Marco and Amy Wang 1992 Durkheim, religion, and Buddhism. Journal for
the Scientific Study of Religion 31: 47-61.

Reynolds, Frank and Charles Hallisey 1987 Buddha. Encyclopedia of religion,
vol. 2, edited by Mircea Eliade, 319-333. New York: Macmillan.

Schmidt, Roger 1988 Exploring religion, 2nd ed. Belmont, CA: Wadsworth.

Smith, Wilfred Cantwell 1962 The meaning and end of religion New York:
Macmillan.

Speyer, J. S., trans. 1971 The jatakamala garland of birth stories of
Aryasura. Delhi: Motilal Banarsidass.

Spiro, Melford E. 1966 Religion: Problems of definition and explanation. In
Anthropological approaches to the study of religion, edited by Michael
Banton, 85-126. New York: Praeger.

1982 Buddhism and society: A great tradition and its Burmese vicissitudes,
2nd ed. Berkeley: University of California.

Swearer, Donald K., ed. 1989 Me and mine: Selected essays of Bhikkhu
Buddhadasa. Albany: SUNY Press.

Tambiah, S. J. 1970 Buddhism and the spirit cults in northeast Thailand.
Cambridge: Cambridge University Press.

1984 The Buddhist saints of the forest and the cult of amulets: A study in
charisma, hagiography, sectarianism, and millennial Buddhism. Cambridge:
Cambridge University Press.

von Glasenapp, Helmuth 1966 Buddhism--A non-theistic religion. New York:
George Braziller.

Warren, Henry Clarke [1896] 1973 Buddhism in translation. Passages selected
from the Buddhist sacred books and translated from the original Pali into
English. New York: Atheneum.

Williams, Paul 1989 Mahayana Buddhism The doctrinal foundations. London:
Routledge.