On The Very Idea of Religions

On the Very Idea of Religions (In the Modern West and in Early Medieval China)
Author(s): Robert Ford Campany

Source: History of Religions, Vol. 42, No. 4 (May 2003), pp. 287-319
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/378757
Accessed: 08-09-2017 02:45 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and
extend access to History of Religions
This content downloaded from 142.103.160.110 on Fri, 08 Sep 2017 02:45:29 UTC
All use subject to http://about.jstor.org/terms
Robert Ford Campany O N T H E V E RY I D E A O F
R E L IGIONS ( I N TH E
MO D E R N W E S T A N D I N
E A R LY M E D I E VA L
C H I NA )
Well into the nineteenth century, there “were” only four religions: Chris-
tianity, Judaism, Islam, and a fourth variously named Paganism, Idolatry,
or Heathenism.1 Today a researcher can claim, “We have identified nine
thousand and nine hundred distinct and separate religions in the world,
Some of the positions taken below to some extent resemble those argued by Stephen R.
Bokenkamp in an as yet unpublished paper, “The Silkworm and the Bodhi Tree: The Ling-
bao Attempt to Replace Buddhism in China and Our Attempt to Place Lingbao Daoism,”
of which I received a copy only at a late phase in the writing of this essay. My remarks here
were first delivered in sketch form at a symposium at Harvard University in May 2000, and
then more elaborately at the annual meeting of the Association for Asian Studies in Wash-
ington, D.C., in April 2002, and I am grateful for the opportunity to present them in both
venues. I am also grateful to Bokenkamp, John McRae, and Michael Satlow for critical
comments on an earlier draft of this essay.
1
For an overview, see Jonathan Z. Smith, “Religion, Religions, Religious,” in Critical
Terms for Religious Studies, ed. Mark C. Taylor (Chicago: University of Chicago Press,
1998), pp. 275–80; on the extension of this typology to colonial frontiers, see David
Chidester, Savage Systems: Colonialism and Comparative Religion in Southern Africa
(Charlottesville: University of Virginia Press, 1996), pp. 32, 238– 41; on the mid-Victorian
shift from classification to diffusion and then development as the major tropes—or, as the
Scot John Ferguson McLennan put it in 1863, “the divisions, the movements, and the
progress of mankind”—see George W. Stocking, Jr., Victorian Anthropology (New York:
Free Press, 1987), pp. 165–66. Other early classifications favored types of religion rather than
named religions; e.g., see David Hume, The Natural History of Religion, ed. H. E. Root
(Stanford, Calif.: Stanford University Press, 1957), which deploys a binary “polytheism”/
“monotheism” not very far from what Jonathan Smith has identified as the most basic clas-
sification of religions and groups, “theirs” vs. “ours” (see “Religion, Religions, Religious,”
ç 2003 by The University of Chicago. All rights reserved.

0018-2710/2003/4204-0001$10.00
288 On the Very Idea of Religions
increasing by two or three new religions every day.”2 Whatever shifts in

human life and communities might be related to these juxtaposed classi-
fications, what is clearly paramount is that massive changes have oc-
curred in the criteria and systems of classification themselves and in the
awareness of the classifiers. Taxonomic map is not religious territory.
What is involved when we invoke the categories “religions” or “a re-
ligion” (as distinct from “religion” in the generic sense and from “the re-
ligious,” taxa with which I will not deal here)? To what extent are these
categories helpfully invoked in the study of specific non-Western cultures
and periods—for example, early medieval China?
In approaching these questions I will assume five axioms:
1. Discourse on religions is first and foremost a linguistic affair, what-
ever concepts or theories end up being invoked. We normally focus on
“theories” and “methods” that operate at high levels of abstraction, but at
the working end of religious studies much is decided at the more concrete
level of the language in which descriptions and interpretations are couched
and research questions framed. We must therefore attend closely to that
language and its implications. While some have criticized the category
“religions,” they have typically failed to include scrutiny of the languages
both of modern scholarship and of the other cultures being studied.3
2. Language and concepts are metaphorical in character, in the sense
richly developed in the work of George Lakoff and Mark Johnson over
the past twenty years.4 Underlying the metaphors in which even the most
p. 276, Map Is Not Territory: Studies in the History of Religions [Leiden: Brill, 1978], pp.
241– 42, “What a Difference a Difference Makes,” in “To See Ourselves as Others See Us”:
Christians, Jews, “Others” in Late Antiquity, ed. Jacob Neusner and Ernest S. Frerichs
[Chico, Calif.: Scholars Press, 1985], pp. 15–16); and Chantepie de la Saussaye, Manual of
the Science of Religion, trans. Beatrice S. Colyer-Fergusson (London: Longmans, Green,
1981), chaps. 7–8, which deploys and discusses a variety of classification schemes.
2
The statement is attributed to David B. Barrett, editor of the World Christian Encyclo-
pedia, in Toby Lester, “Oh, Gods!” Atlantic Monthly (February 2002), p. 38.
3
See, e.g., Robert D. Baird, Category Formation and the History of Religions (The
Hague: Mouton, 1971), pp. 134– 42. A refreshing recent exception is Timothy Fitzgerald,
“A Critique of ‘Religion’ as a Cross-Cultural Category,” Method and Theory in the Study
of Religion 9, no. 2 (1997) 35– 47, though I do not by any means agree with some of his con-
clusions. On the other hand, Fitzgerald’s “Religious Studies as Cultural Studies: A Philo-
sophical and Anthropological Critique of the Concept of Religion,” Diskus 3 (1995): 35– 47,
criticizes the use of “religion” mainly for its conceptual fuzziness. I would respond that,
while scholarly usage of the term is certainly vague and ambiguous by turns, the discourse
on “religions” is structured according to certain prominent metaphors that are not fuzzy at
all but that may be problematic in other ways, as we will see below.
4
See especially George Lakoff and Mark Johnson, Metaphors We Live By (Chicago:
University of Chicago Press, 1980), and Philosophy in the Flesh: The Embodied Mind and
Its Challenge to Western Thought (New York: Basic, 1999); George Lakoff, Women, Fire,
and Dangerous Things: What Categories Reveal about the Mind (Chicago: University of
Chicago Press, 1987); and Mark Johnson, The Body in the Mind: The Bodily Basis of Mean-
ing, Imagination, and Reason (Chicago: University of Chicago Press, 1987). Also of note
is Earl R. MacCormac, Metaphor and Myth in Science and Religion (Durham, N.C.: Duke
History of Religions 289
apparently neutral descriptive statements about religions are couched will

be found implications that silently but powerfully determine the ques-
tions we ask and the assumptions we make about the nature of religions.
3. Discourse about religions is rooted in Western language communi-
ties and in the history of Western cultures. As Wilfred Cantwell Smith
famously argued four decades ago, it is not the case that “religion” (in ei-
ther its generic or its specific sense) is simply “our word for” a universally
existent entity or a universally recognized category.5 To speak of “re-
ligions” is to demarcate things in ways that are not inevitable or immuta-
ble but, rather, are contingent on the shape of Western history, thought,
and institutions, Other cultures may, and do, lack closely equivalent
demarcations.
4. Contrary to Cantwell Smith, however, I take it that the helpfulness
of the category “religions” is not to be measured by the extent to which
people in the target culture and era—here, early medieval China—would
have recognized it as one of their own. That premodern Chinese “lacked
a word for ‘religion(s),’ ” has been noted for almost a century now,6 does
not, prima facie, constitute a reason for modern scholars not to use the
term. It would do so only if one assumed that the sole legitimate task of
historical scholarship on religion is to recover and repeat, in the language
of the original documents, earlier people’s claims—an assumption
Jonathan Z. Smith has summarized as “a morality of regard for local in-
terpretations.”7 Even if such a project were desirable, it is impossible for
many reasons, including the fact that the language of research is not the
same as the language of the sources, necessitating translation—hence in-
terpretation—at every turn (a responsibility abnegated by the fantasy that
we might somehow simply present the texts pristine and whole to our
readers), as well as the fact that, at least until time travel becomes pos-
sible, the agendas driving the questions asked and the materials selected
must derive from knowledge communities contemporaneous to the
scholar. As Jonathan Smith put it recently, “ ‘Religion’ is not a native
term; it is a term created by scholars for their intellectual purpose and
University Press, 1976). For an important set of essays that build on metaphor analysis but
also suggest revisions and extensions, see James W. Fernandez, ed., Beyond Metaphor: The
Theory of Tropes in Anthropology (Stanford, Calif.: Stanford University Press, 1991). Par-
ticularly important for my purposes here is the essay by Naomi Quinn, “The Cultural Basis
of Metaphor” in Fernandez, ed., pp. 56–93.
5
See Wilfred Cantwell Smith, The Meaning and End of Religion (New York: Mac-
millan, 1963).
6 W. E. Soothill noted as early as 1919 that the modern Chinese term zongjiao [ ]
was a borrowing from Japanese and that the term had been coined recently in Japanese to
translate the Western “religion” (cited in Smith, The Meaning and End of Religion, p. 58).
7
Jonathan Z. Smith, Drudgery Divine: On the Comparison of Early Christianities and
the Religions of Late Antiquity (Chicago: University of Chicago Press, 1990), p. 52.
therefore is theirs to define. It is a second-order . . . concept.”8 But use of

this category without regard to whether Chinese usages work differently
constitutes a sort of category blindness to aspects of the historical evi-
dence and can enable the illusion that the category is universal and natu-
ral. This, then, leads to my final axiom.
5. That early medieval Chinese discourses lacked one-for-one “ver-
sions” of the Western category “religions” does not mean that they lacked
some usages that are analogous—ones that do something like the same
work, ones invoked in the sorts of contexts in which “religions” would
be invoked in modern Western discourses. Hence we must pay close
attention to two cultural and temporal sets of linguistic usages and their
metaphorical implications, and juxtapose these results.9 Such an inquiry
has the purpose not of ruling out Western usages simply because they are
Western but of clarifying certain aspects of both members of the compari-
son and the nature of the differences between them, as a way of better un-
derstanding both the contours and limitations of the category “religions”
and the contours and limitations of the early medieval Chinese dis-
courses on analogous topics as well as the nature of the fit, or lack of fit,
between the two. To pursue this latter goal is to stop short of affirming,
with Jonathan Smith, that “comparison provides the means by which we
‘re-vision’ our data in order to solve our theoretical problems,”10 at least
if one of our problems is the extent to which our categories match other
cultures’ and the discovery of the difference that a categorical difference
makes to our (mis)understanding of them. To become aware of the pecu-
liar shape and implications of our category “religion” is to see more
8
Smith, “Religion, Religions, Religious,” p. 281.
9
A recent work that, though not on the topic of religion(s), is exemplary in its careful
attention to Chinese metaphors and the differences from Western analogues in their impli-
cations is Jane Geaney, On the Epistemology of the Senses in Early Chinese Thought (Ho-
nolulu: University of Hawaii Press, 2002). Edward Slingerland is also doing important
work in this area.
10
Smith, Drudgery Divine, p. 52. Compare this statement in F. J. P. Poole, “Metaphors
and Maps: Towards Comparison in the Anthropology of Religion,” Journal of the Ameri-
can Academy of Religion 54 (1986): 414–15: “Neither phenomenologically whole entities
nor their local meanings are preserved in comparison”—to which I would reply that if their
local meanings (for my purposes read: categories) are not preserved, neither will ours be
if ours are themselves the other member of the comparison rather than being taken as the
inert, fixed “third term.” It may be, as Smith points out, that sound comparisons are triad-
ically structured (see Drudgery Divine, pp. 33, 51, 86–87, 99, 117), but I would insist that
the “third term,” itself a culturally and historically conditioned variable, is itself likewise
liable to critical modification, revision, and possible rejection as a result of the process of
comparison. Perhaps this is what Smith means to allow for in his more recent comment,
“So classify we must—though we can learn from the past to eschew dual classifications
such as that between ‘universal’ and ‘ethnic’ or the host of related dualisms” (“A Matter of
Class: Taxonomies of Religion,” Harvard Theological Review 89 [1996]: 402), to which I
would simply reply that dual classifications may not, on analysis, be the only ones subject
to rejection or revision.
clearly the ways in which it implicitly shapes not only the answers to our
historical and interpretive questions but also the very form of those ques-
tions and, therefore, the form that any possible answer can take.
on religions as entities (of certain metaphorically

imagined kinds)
The most basic aspect of how religions are imagined in Western discourse
is that they are construed as entities; they are reified. One prominent way
in which Western discourse reifies religions is by the deceptively simple
use of the morphological device of the English suffix “ism” and its
European equivalents. By adding “ism” to a root noun or adjective that
does not yet designate a religion, we form new, abstract entities, and by
adding “ist” we denote things or tendencies that belong to these entities.
The Oxford English Dictionary traces “ism” to the Greek -ismos, a form
for the nominalization of action verbs, and defines one of its chief uses as
“forming the name of a system of theory or practice.” The use of such suf-
fixes, rampant in the study of Chinese religions (where we have the big
three of Confucianism, Daoism, and Buddhism) and elsewhere, amounts
to a kind of shorthand, a convenient way to generalize over vast numbers
of particulars.11 But it is also a sleight of hand, creating in three key-
strokes an entity that, in addition to its sudden existence as a thing among
other things, is further implied to have the property of systematicity and
therefore to be a well-integrated and clearly demarcated whole, such that
aspects or parts of the whole must resemble each other more strongly than
they resemble any outside aspects or parts. This “religion holism,” as I
will call it, or “substantialist fallacy . . . of misplaced concreteness,” as
Timothy Fitzgerald has characterized it,12 has given rise to serious mis-
understandings. As pointed out by Jonathan Smith:
To raise the issue of the setting of early Christianities [note the plural!] is to
ask at the outset the question of comparison and, thereby, to deny any initial
11
One scholar in the field of Chinese thought and religion who has long been admirably
sensitive to such issues, as well as to the unconsidered use of such categories as “science,”
“magic,” and “religion,” is Nathan Sivin. His classic article “On the Word ‘Taoist’ as a
Source of Perplexity” (History of Religions 17 [1978]: 303–30), questioning the vague use
of the taxa “Taoism” and “Taoist,” was years ahead of its time and has yet to be adequately
heeded or responded to even by scholars specializing in “Daoist” studies. In his recent in-
troduction to an older essay by Joseph Needham and Lu Gwei-Djen on the history of Chi-
nese medicine, Sivin returns to the problems posed by such categories, observing, “These
[“isms”] are fixtures in the sorts of history of philosophy that are more interested in dis-
embodied isms than in the activity of particular human beings” (“Editor’s Introduction,” in
Joseph Needham and Lu Gwei-Djen, Science and Civilisation in China, vol. 6, Biology
and Biological Technology, pt. 6, Medicine, ed. Nathan Sivin [Cambridge: Cambridge Uni-
versity Press, 2000], p. 8).
12
Fitzgerald, “A Critique of ‘Religion,’ ” p. 106.
postulation of “uniqueness.” Much will depend on the framing of the issue.

The traditional vague terminology of “Early Christianity,” “Jewish,” “Gentile,”
“Pagan,” “Greco-Oriental,” etc. will not suffice. Each of these generic terms
denote complex plural phenomena. For purposes of comparison, they must be
disaggregated and each component compared with respect to some larger topic
of scholarly interest. That is to say, with respect to this or that feature, modes
of Christianity may differ more significantly between themselves than between
some mode of one or another Late Antique religion. The presupposition of
“holism” is not “phenomenological,” it is a major, conservative, theoretical
presupposition which has done much mischief in the study of religious materi-
als, nowhere more so than in the question of Christian “origins.”13
Without further specification, an apparent entity named by some such

name as “Daoism” seems to exist simply in a kind of contextless stasis.
We can write its history, but the very form of the name suggests that
“Daoism” is one unitary, perduring thing whose permutations we simply
trace through time. One important line of questioning obscured by reify-
ing and essentializing usages is that of how it happened—in an exuber-
ance of detail befitting the intricacy of the subject—that such and such a
group came to live their lives (or some aspect of their lives) differently
by conforming their usage to new dictates, as well as the question of how
it was understood to have happened (and portrayed as having happened)
by people closer to the events. In other words, it is precisely the manifold
ritual, social, institutional, rhetorical, and narrative processes at work be-
hind such nominal forms as “Islamization,”14 “Christianization,”15 “Bud-
dhicization,” “Confucianization,”16 and “Sinification”17 that ought to be
subjects of inquiry. We should ask how, not simply state that, such pro-
cesses occur, but the trope of reification leaves us with few tools for do-
ing so. At least these last-mentioned nominalizations have the virtue of
implying processes by their morphology, a dynamism once suggested by
the Greek root -ismos but now completely drained from our “isms.”
13
Smith, Drudgery Divine, pp. 117–18.
14
Of which an exemplary study may be found in Devin DeWeese, Islamization and Na-
tive Religion in the Golden Horde: Baba Tükles and Conversion to Islam in Historical and
Epic Tradition (University Park: Pennsylvania State Press, 1994).
15
A sensitive study of how dying and death were gradually “Christianized” in medieval
Europe and came to be performed as ritual processes differently than they had been before
“Christianization” may be found in Frederick S. Paxton, Christianizing Death: The Creation
of a Ritual Process in Early Medieval Europe (Ithaca, N.Y.: Cornell University Press, 1990).
16
See Miyakawa Hisayuki, “The Confucianization of South China,” in The Confucian
Persuasion, ed. Arthur F. Wright (Stanford, Calif.: Stanford University Press, 1960), pp.
21– 46.
17 Considerations of the “sinification” of “Buddhism” should now take as their starting
point the pertinent reassessment of the shape of this category in Robert H. Sharf, Coming
to Terms with Chinese Buddhism: A Reading of the Treasure Store Treatise (Honolulu:
University of Hawaii Press, 2002), pp. 1–25.
So pervasive is the habit of reification that we do well to remind our-

selves that “religions” do not exist as things in the world. The pertinent
res include texts, images, and other artifacts; structures such as temples
and tombs; and the people who made, used, or otherwise came into con-
tact with these. Anything else is an idea. So, if “Daoism” or “Buddhism”
are unitary, perduring things, they are so because we, possibly along with
cultural others (though certainly not early medieval “Daoists” or “Bud-
dhists,” since the English language in which it is possible to form the
word “Daoism” did not yet exist), imagine and construct them as such in
the ways we speak, not because they are natural existents we find in the
world alongside the res we characterize as “belonging to” them.
Only slightly less abstract and more metaphorical are two nouns often
attached to the adjective “religious” to form phrases naming the same
purported entities named by the “isms,” or else to the “ists” formed from
these: “tradition” and “system.” When we speak of “the Buddhist tradi-
tion” or “the system of Buddhism,”18 we seem to mean what we mean
when we simply say “Buddhism,” but with emphasis on continuous trans-
mission through time (in the first case) or on principled, organized, delib-
erate, and rigorous coherence (in the second).
To call a “religion,” X, “the Xist tradition,” implies a holism, unity,
and continuity that ought not to be taken for granted. It does not broach
the actual means of transmission; nor the content of what is transmitted;
nor the changes (often dramatic) in what is transmitted; nor the hugeness
of what is forgotten, or lost, or suppressed, or destroyed; nor the contes-
tations over what is transmitted and remembered and who decides. Again,
it is the actual processes and practices of remembering and transmitting
that are obscured by the usage “the Xist tradition”19 and by the way of
thinking about “religions” that the phrase implies.
To speak of “the Xist system” or “the system of Xism” implies—what
else?—systematicity. This habit of thought and speech, rooted in Durk-
heim’s definition of religion and enshrined in Clifford Geertz’s famous
essay first published in the mid-1960s,20 continues despite postmodern
attentions to fragmentation, contestation, and all that is nonsystematic
in culture and religion. It is curious that when Geertz sets out to explicate
the first phrase of his definition of “a religion,” which runs “a system of
18
As opposed to, e.g., “the Buddhist tradition of celebrating the Buddha’s birthday” or
“the Buddhist system of transference of merit.”
19
Good resources for the study of such processes are to be found, e.g., in works such as
Maurice Halbwachs, On Collective Memory, ed. and trans. Lewis A. Coser (Chicago: Uni-
versity of Chicago Press, 1992); James Fentress and Chris Wickham, Social Memory: New
Perspectives on the Past (Oxford: Blackwell, 1992); Paul Connerton, How Societies Re-
member (Cambridge: Cambridge University Press, 1989).
20
Clifford Geertz, “Religion as a Cultural System,” in The Interpretation of Cultures
(New York: Basic, 1973), pp. 87–125; the essay was first published in 1966.
symbols which acts to,” he exclusively devotes himself to what is meant

by “symbol” and says nothing whatsoever about what it means to speak
of a “system” of symbols.21 In 1966, given the predominance of struc-
turalist approaches, this element of the definition could simply be taken
for granted. That is no longer true.
The discourse of religious “traditions” and “systems” is quasi-meta-
phorical at best, since “tradition” and “system” in such formulations re-
main highly abstract. Other forms of speech about “religions” are more
metaphorically rich. My point in what follows is not to urge that we es-
chew metaphor entirely, for that would be impossible, but rather to urge
that we become alert to the evidence-distorting and thought-limiting im-
plications of certain particular metaphors with which we have become
numbingly familiar. (In the last section I suggest alternatives.) All meta-
phors, like categories, highlight certain aspects of things and obscure
others, thus affording us handles on complex, abstract, and unwieldy phe-
nomena. A critique of a metaphor, then, does not consist in showing that
it is somehow “wrong” but in pointing out what it hides and noting the
importance, for certain purposes, of attending to these hidden aspects.
What it hides might turn out to be something well worth seeing, and
sometimes the hiding serves a latent ideology or set of interests, or en-
codes an uninvoked but silently looming model or set of expectations
based on the resemblance implied in the metaphor.
religions are living organisms (often plants)
Taoism: Growth of a Religion22

In this history of Taoism I have tried . . . to show the coherence of its develop-
ment and its constant absorption and integration of outside contributions.23
The very first stage of Chinese Buddhism—that tiny exotic plant flowering on
the ruins of the Han empire.24
21
Note, e.g., the prominent use of “religious system,” with all of the attendant holistic
tendencies, such as “A constitution is necessary in order that the parts or elements of the
religious enterprise might function” (is there an actual “religion,” as opposed to Lease’s
petri dish model, that has anything approximating a working “constitution”?) and “any re-
ligion which advances through the first two stages . . . i.e., through the meeting of the fun-
damental needs for which any religion arises, and through the articulation of its essence or
system—will . . . advance to a final stage, . . . the claim by a religion to be able to provide
a total understanding,” in Gary Lease, “The History of ‘Religious’ Consciousness and the
Diffusion of Culture: Strategies for Surviving Dissolution,” Historical Reflections/Réflex-
ions historiques 20 (1994): 466–69, quoted passages on 463 and 469, respectively.
22 The title of the magisterial history by Isabelle Robinet, trans. Phyllis Brooks (Stan-
ford, Calif.: Stanford University Press, 1997); all italics in lists are added.
23 Robinet, p. xv.
24
Erik Zürcher, “A New Look at the Earliest Chinese Buddhist Texts,” in From Benares
to Beijing, ed. Koichi Shinohara and Gregory Schopen (Oakville, Ontario: Mosaic, 1991),
p. 293.
When this southern Buddhism in the course of time was blended with that of
the north, we see at least the full maturity of sinicized Buddhism.25
In the East where it [Manichaeism] flourished at a time when there were no
longer any real Manichaeans to be found in the West.26
Christianity itself contained astrological elements; too many traces of the Hel-
lenistic and Oriental religions . . . were intertwined at its very roots for it
to be able to rid itself of them completely.27
At first it [Buddhism] must have lived on among the foreigners who had
brought it with them from their home countries.28
The Chinese Buddhist texts that on the basis of internal and external evidence
may be ascribed to the “embryonic phase” of Chinese Buddhism.29
As suggested by these few examples, authors often turn to organismic

metaphors for religions when they want to portray their changing for-
tunes over time, their flourishing, or their decline; additionally, organis-
mic metaphors provide a way of imagining a religion’s appropriation of
outside elements.
There are at least three costs to using such metaphors, however.
(1) They locate agency in religion-entities themselves rather than in the
people (whether individuals or groups) who participate in, support, op-
pose, thwart, or otherwise act to shape the nature and fortunes of the pu-
tative religion-entities in question. Religions-seen-as-organisms assume
a life of their own. (2) Seen as organisms, religions take on a tacit tele-
ology; there must be a predictable, fully mature form toward which they
are striving—again quite independently of human agents. (3) Living be-
ings, while radically dependent on their environments, are nevertheless
clearly bounded entities. Even at microscopic levels there are clearly
identifiable, if porous, interfaces where organisms stop and their environ-
ments start. Furthermore, living beings are holistically integrated. For-
eign agents are mostly recognized as such and are quickly broken down
and assimilated or expelled (or else the organism sickens and dies). In-
gested food is transformed into a building block of the organism that is
indistinguishable from others. And, of course, organisms have unique
genetic codes that act as the master blueprints in every cell of the organ-
ism, directing all growth from a uniform structure.
25
Arthur F. Wright, Studies in Chinese Buddhism, ed. Robert M. Somers (New Haven,
Conn.: Yale University Press, 1990), p. 39.
26
Kurt Rudolph, Gnosis: The Nature and History of Gnosticism (New York: Harper-
Collins, 1987), p. 331.
27 Jean Seznec, The Survival of the Pagan Gods: The Mythological Tradition and Its Place
in Renaissance Humanism and Art, trans. Barbara F. Sessions (Princeton, N.J.: Princeton
University Press, 1953), p. 43.
28
Erik Zürcher, The Buddhist Conquest of China, 2d ed. (Leiden: Brill, 1972), p. 23.
29
Zürcher, “A New Look at the Earliest Chinese Buddhist Texts,” p. 277.
In sum, seeing religions as comparable to organisms encourages us to

imagine them not only as entities, but as entities of particular kinds: au-
tonomous agents going about the business of fulfilling their developmen-
tal teleology; living beings that completely transform ingested substances
into parts of themselves unrecognizable from other parts; and clearly de-
marcated, sharply bounded, and holistically and functionally hyperorga-
nized life-forms, every component of which shares the same fundamental
essence as every other part.30 We can see what a close affinity religion or-
ganicism bears to religion holism. It should be self-evident that none of
the features just listed are true of any collective human enterprise.
religions are personified agents
Where Manichaeism failed, Christianity succeeded when Constantine es-

poused it.31
It [Manichaeism] knew how to adapt to Chinese tradition in its missionary
practice.32
Taoism of the middle ages saw itself as universal.33
Christianity offered twice as much cultural continuity to the Hellenized Jews
as to Gentiles. . . . Little need be said of the extent to which Christianity
maintained cultural continuity with Judaism.34
Daoism constructed its detailed bureaucratic arrangement only to transcend it
through meditative unity with the transcendent Dao and to tease it with a
celebration of eccentric immortality.35
The first example aptly illustrates the pitfalls of ascribing agency to

religions-as-persons: the first two clauses of the sentence seem to imply
that the two religions are autonomous actors, one of which failed and the
other of which succeeded; however, the ensuing clause of the very same
sentence attributes the latter religion’s success to a ruler’s policy. In a pas-
sage leading up to the last example, the authors make the point that “Dao-
30
Reading passages based on organic metaphors for religions, one is powerfully re-
minded of the long Western tradition of reflection on the morphology of organic forms,
enshrined in Goethe’s musings on the Urpflanze and perhaps best summarized in E. S.
Russell, Form and Function: A Contribution to the History of Animal Morphology (1916;
reprint, Chicago: University of Chicago Press, 1982).
31
Garth Fowden, Empire to Commonwealth: Consequences of Monotheism in Late An-
tiquity (Princeton, N.J.: Princeton University Press, 1993), p. 76.
32
Rudolph, p. 332.
33
Kristofer Schipper, “Purity and Strangers: Shifting Boundaries in Medieval Taoism,”
T ’oung Pao 80 (1994): 63.
34
Rodney Stark, The Rise of Christianity: A Sociologist Reconsiders History (Prince-
ton, N.J.: Princeton University Press, 1996), p. 59.
35
Meir Shahar and Robert P. Weller, “Introduction: Gods and Society in China,” in Un-
ruly Gods: Divinity and Society in China, ed. Meir Shahar and Robert P. Weller (Honolulu:
University of Hawaii Press, 1996), p. 10.
ism” constructed elaborate bureaucratic hierarchies for the gods; now they
wish to show that “Daoism” harbors anti- or extrabureaucratic types as
well. This sentence is how they resolve the tension. Rather than seeing the
tension between the bureaucratic and antibureaucratic as resulting from
many textual and ritual skirmishes over time by multiple historical agents
with multiple agendas addressing multiple contexts, we are asked to see it
as the work of a single, intentional, but impersonal agent, “Daoism,” that
is teleologically portrayed as doing X only to transcend X by proceeding
to do Y. It is as if “Daoism” is a whimsical child at work on a sand castle
or a painter addressing her canvas. The metaphor utterly obscures the
actual processes that occurred to bring about the presence of such conflict-
ing imagery and figures; it also obscures the interests and agency of the
persons and groups actually responsible for them—and especially in this
case one surmises that struggles over power and status were key contexts
prompting the adoption of one or another idiom, bureaucratic or anti-
bureaucratic, in specific situations.
religions are marketable commodities
Philosophic Taoism had lost some of its appeal. . . . Its ideas had been dis-
cussed incessantly for decades and had lost their freshness.36
religions are armies; their spread and success are warfare
It [Manichaeism] could hold its ground even more successfully and more per-
manently in the East.37
Buddhism was well entrenched in eastern Iranian areas.38
The Buddhist Conquest of China
The assumption of earlier occasional infiltrations of Buddhist elements into
Tibet is an obvious one to make.39
It [Buddhism] must have slowly infiltrated from the North-West.40
These last two metaphors for “religions” are not much more satisfying
than the ones discussed above, but they have their advantages. The mar-
ketable commodity metaphor at least implies that persuasion is involved
in religious life, that some people are trying to win others over to their
36
Wright, p. 38.
37
Rudolph (n. 26 above), p. 331.
38
Hans-Joachim Klimkeit, Gnosis on the Silk Road: Gnostic Parables, Hymns and
Prayers from Central Asia (New York: HarperCollins, 1993), p. 5.
39
Giuseppe Tucci, The Religions of Tibet, trans. Geoffrey Samuel (Berkeley: University
of California Press, 1980), p. 1. This is a secondary metaphor: the use of “infiltrate” to de-
scribe a clandestine military or intelligence-gathering maneuver stems from a primary
metaphor involving liquids and the breakdown of filters designed to exclude foreign agents.
40
Zürcher, Buddhist Conquest (n. 28 above), pp. 22–23.
points of view; by its very structure this metaphor forces us to recall that
there is a human population involved, a clientele of agents making choices,
and that much is at stake in those choices. Indeed, Pierre Bourdieu, in an
exceedingly rare foray into the topic of religion, makes the market meta-
phor the cornerstone of his approach, and the sociologist Rodney Stark
has recently written a deeply market-based analysis of “the rise of Chris-
tianity.”41 The martial metaphor goes further, picturing interreligious or
religious-cultural encounters as pitched battles or guerilla maneuvers.
However, like the metaphors surveyed above, both of these metaphors
hide the agency of actual human beings and the issues at stake in their
struggles, gathering it all up into a vague, collectivized “ism.” What
Bruce Lincoln has recently observed of the treatment of stories classified
under the taxon “myth” applies also to this sort of usage: “Myth is often
treated as an anonymous and collective product, in which questions of
authorship are irrelevant. Lévi-Strauss has done this in a most sophisti-
cated and challenging fashion, treating myth as a logical structure that es-
sentially writes itself, variants being the product of an impersonal process
whereby that structure explores its own variables until exhausting the
available possibilities. Such a view alleviates the frustration of those who
seek authors and ‘original versions’ of mythical texts, but the price for this
is unacceptably high, since it drains agency from the act of narration.”42
There are other key metaphors for religions, which I will mention only
in passing: religions are substances (usually liquid or viscous, with the
common tropes being spread, influence [on which more below], and dif-
fusion); religions are containers (of people, ideas, trends, artifacts, and
values); religions are the contents of cultures, societies, nations, or groups
that are containers (leading to several unhelpful but commonly used
ways of narrating the passage of religions across cultures and societies);
and religions are buildings.
Now, all such usages occur when scholars want to speak at a high level
of generality and abstraction, and at such levels their use is perhaps in-
evitable. Some such constructions, in some contexts, are perhaps quite
harmless or trivial. It is admittedly hard to arrive at alternatives that al-
low one to make generalizations of any kind; metaphors reduce complex,
messy phenomena by analogies to things simpler and more familiar, and
at times we understandably turn to them for the verbal and conceptual
economy they offer. (It is impossible, for example, to imagine writing a
41
Pierre Bourdieu, “Genèse et structure du champ religieux,” Revue française de sociol-
ogie 12 (1971): 295–334; Stark.
42
Bruce Lincoln, Theorizing Myth: Narrative, Ideology, and Scholarship (Chicago:
University of Chicago Press, 1999), p. 149. Compare the similar statement about shifts in
the meanings of words on p. 18.
short encyclopedia entry on the historical relationship between two tra-

ditions without frequent recourse to such constructions.) And in singling
out a few sentences from some colleagues’ works I do not by any means
suggest that I myself have successfully eschewed such expressions in my
own writings, although in my most recent book I do attempt alternative
strategies whose success I leave for others to judge.43 But we use such
locutions at the price of positioning ourselves to forget about some of the
most important human aspects of the historical processes we seek to un-
derstand and portray. That is a very high price to pay.
some early medieval chinese metaphors

Do we find anything in early medieval Chinese discourses remotely anal-
ogous to Western patterns of discourse on “religions,” and if so, what
predominant metaphors are at work and what are their implications?
Below I provide a mere sampling of statements drawn from a small
number of texts; this is an area that warrants much more research, and
my findings are preliminary. Along the way I will also comment on how
Western translators have dealt with such passages.
1. founder or paragon synecdoche

The names, partial names, or titles of founding or paradigmatic figures
are sometimes used synecdochally to refer nominally to what in Western
discourse would be called an entire “religion” or “tradition.” For ex-
ample, Wei Shou [ ] (506–72), compiler of an official history of the
Toba Wei (the Wei shu), included a section on religion titled Shi Lao zhi
[ ]—usually translated “Treatise on Buddhism and Daoism,” but
the Shi and Lao of the title are truncations of the names of ‡akyamuni
and Laozi. A similar usage occurs in the title of what we might term a
“lineage” or “school” of texts, ideas, and practices, Huang Lao [ ],
short for Huangdi [ ], the Yellow Thearch (a figure of ancient myth),
and Laozi, putative author of the Daodejing and (in some circles) cosmic
deity. This type of synecdoche is also seen not in titles but in the context
of ongoing discourse:
Literally, “the ruination and rejection of Buddha,” but the context clearly re-
quires that we understand fo as gathering up and nominalizing the whole set of
43
See the brief discussion in Robert Ford Campany, To Live as Long as Heaven and
Earth: A Translation and Study of Ge Hong’s Traditions of Divine Transcendents (Berke-
ley: University of California Press, 2002), pp. 6–8.
This content downloaded from 142.103.160.110 on Fri, 08 spread one02:45:29

Sep 2017 pica short
UTC
phenomena we would habitually call “Buddhism” (Leon Hurvitz translates,

“The Buddhist suppression . . .”).44
“[The Duke of ] Zhou and Confucius are identical with the Buddha, the Buddha
with [the Duke of ] Zhou and Confucius. This is merely to name them with re-
spect to outer and inner.”45
Chinese in 2. “way” or “path” (dao [ ] and its compounds)
Level 1s is
scaled to 90% In early medieval Chinese discourse, probably the most ubiquitous way
of nominalizing what we would call “religions” was to speak of one or
multiple “ways” or “paths”—one or more dao [ ]. I begin with the trea-
tise, in dialogue format, known as the Mouzi lihuo lun [ ]
(“Master Mou’s Treatise for the Removal of Doubts”), by an unknown
author.46 When, as rarely in this text, it is a matter of the foreign what-
we-would-call-religion nominalized, and it is uncertain what its particular
practices, values, or scriptures are, the term used in every case is “the
dao—way or path—of Buddha” (fodao [ ]). The term first occurs in
the question: “If the dao of Buddha is so venerable, why did not Yao,
Shun, the Duke of Zhou, and Confucius practice (xiu [ ]) it?”47 (John P.
Keenan renders fodao here as “the Way of the Buddha”; Erik Zürcher
simply translates it as “Buddhism,” an understandable choice, but one that
masks the Chinese metaphor and its difference from the Western “ism.”)48
Elsewhere the interlocutor asks why, since the people who constituted
the intellectual and cultural paragons of society at the time—the “forest
of classicists” (rulin [ ], ru being a designation often translated as
“Confucians” [a habit that merits reconsideration])49—did not regard the
44
The Chinese text with Hurvitz’s translation appear in Leon Hurvitz, trans., “Wei Shou,
Treatise on Buddhism and Taoism: An English Translation of the Original Chinese Text of
Wei-shu CXIV and the Japanese Annotation of Tsukamoto Zenryu,” in Mizuno Seiichi and
Nagahiro Toshio, Unko sekkutsu no kenkyu, 16:68–69 (Kyoto: Kyoto University Research
Institute for Humanistic Sciences, 1951–56).
45
Sun Chuo [ ] (ca. 300–380 c.e.), Yudao lum [ ] (“Essay in Clarification
of the Path [or Way]”), T 2102, 52:17a. Here and throughout, texts in the Chinese Buddhist
canon are referred to as follows: the number following the letter T indicates the serial number
assigned to the text in the Taisho edition (Tokyo: Taisho shinshu daizokyo, 1924–35), and
the numbers following the comma indicate the volume and (after the colon) the page, reg-
ister(s), and (in some cases) line numbers of the passage in question.
46
On the uncertain provenance of this text, see Zürcher, Buddhist Conquest, pp. 13–15;
for a recent translation and study, see John P. Keenan, How Master Mou Removes Our
Doubts: A Reader-Response Study and Translation of the Mou-tzu Li-huo lun (Albany,
N.Y.: SUNY Press, 1994).
47
T 2102, 52:2b26.
48
Keenan, p. 79; Zürcher, Buddhist Conquest, p. 265.
49
For a recent and well-informed discussion of the significance of this term in early
texts, see Michael Nylan, The Five “Confucian” Classics (New Haven, Conn.: Yale Uni-
versity Press, 2001), pp. 2n., and 364–65.
“dao of Buddha” as venerable during his visit there, Master Mou holds it
in such high esteem.50
The term dao is also used to summarily nominalize multiple “ways” in
the following passage:
“In both the daos it is a single ‘intentionless action.’ Why then do you discrim-
inate and rank them, saying they [the daos] are different?”51
Or, in the modern Western idiom: both Daoism and Buddhism employ
the concept, value, and terminology of “intentionless action” (the famous
wuwei ); why then do you assert that Daoism and Buddhism are different
paths and that the latter is superior? The implied author, Master Mou,
goes on to pose analogies with the uses of the terms “vegetation” and
“metal”: things may belong in common to these genera, but they differ at
the level of species. He then clinches the analogy with this line: “If this
is so of the myriad things, how can daos alone [be different]?” (Keenan
renders dao in the first question above as “teachings,” while rendering
the latter one as “doctrines,” both of which hide the Chinese metaphor
implicit in dao and set up a too-easy equivalence between it and the fa-
miliar Western tendency to reduce religions to “doctrines.”)52
In Wei Shou’s treatise we find such usages as the following:
“This, then, was the modest beginning of the influx of the Way of the Buddha.”53
“In the time of Emperor Huan, [Xiang Kai] spoke of the Way of Buddha, the
Yellow Emperor, and [Laozi].”54
“[He] had always honored the Way of Buddha.”55
Of course, fodao was also the expression of choice for denoting more
specifically the path to enlightenment established by the Buddha, a set of
teachings and practices more delimited than the more general usage seen
50
T 2102, 52:5c12.
51
T 2102, 52:6b15.
52
Keenan, pp. 157–58.
53
Quoting the translation of Hurvitz (n. 44 above), pp. 26–27.
54
Again quoting ibid., p. 46.
55
Once again quoting ibid., p. 66.
above, where fodao is clearly being used to nominalize the entirety of

what we in English would refer to as “Buddhism.”56
In the year 340 c.e., the officials Yu Bing [ ] and He Chong [ ]
debated the issue of whether the Buddhist sangha was autonomous with
respect to the polity.57 Both He, defending the pro-Buddhist position, and
Yu, arguing that monks were obliged to perform obeisance to the ruler,
use the term shendao [ ] (divine path, or path to divinity, or way of
spirits) in both the singular and plural to nominalize bodies of practice
that seem analogous to what is meant by “religion(s).” (Zürcher, again ig-
noring the metaphoric structure of such an expression, renders shendao as
“spiritual doctrine.”)
“The myriad quarters differ in their customs; their shendao are hard to distin-
guish.”58
“Moreover, from its first appearance in Han times down to the present, al-
though the Law [see the section below on this metaphor] has alternately flour-
ished and decayed it has not been spoiled by bogus and wanton [practices]. As
a shendao it has lasted longer than any other.”59
Again, as in the case of fodao, these uses of shendao are exceptional.

The term’s more standard meanings in religious discourses include the
paths of rebirth as spirits as opposed to humans,60 the way of serving
spirits, a way of characterizing a religious path or method as superior,61
or the inscription-lined pathway leading to a prominent person’s tomb.
In the fifth-century Celestial Master scripture Inner Explanations of
the Three Heavens (Santian neijie jing [ ]), which offers a
mythic “history” of what we would term “religions” in China up until its
56
A computer-assisted search of the Chinese Buddhist canon turns up over 7,300 in-
stances of the term fodao, the overwhelming majority of which exhibit this narrower usage.
57
For the political as well as ideological background of the debate, see Zürcher, Buddhist
Conquest (n. 28 above), pp. 106–10, and for a translation of the documents see pp. 160–63.
See also Helwig Schmidt-Glintzer, Das Hung-Ming Chi und die Aufnahme des Buddhismus
in China (Wiesbaden: Steiner, 1976), pp. 53 ff.; and Tsukamoto Zenryu, A History of Early
Chinese Buddhism: From Its Introduction to the Death of Hui-yüan, trans. Leon Hurvitz (To-
kyo: Kodansha International, 1985), 1:340 ff.
58
T 2102, 52:79b26.
59
T 2102, 52:80a3–5. The statements are repeated in another compilation, T 2036,
49:520c–521a.
60
As in statements such as: “The paths of humans and of spirits are different” (T 2122,
53:521b13), and “Among [those on] the paths of ghosts and spirits there is also eating, but
one cannot attain satiety” (T 2082, 51:792c14).
61
As in the statement, “The way of Buddha (fodao) is a divine way (shendao)” (T 2121,
53:81b).
own time, religious plurality is similarly a matter of various daos, and

central to the scripture’s agenda—as we will see below—is to narrate the
history of these daos’ interrelationships so as to clarify their respective
statuses and identities (and so as to privilege the one championed by the
scripture’s authors). We find such statements as the following:
...
“At this time, he [Lord Lao] issued the Three Ways to instruct the people of
heaven. . . . At that time, the rule of the Six Heavens flourished and the Three
Ways and Teachings were put into practice.”62
These “Three Ways” are not the “Confucianism, Daoism, and Buddhism”
familiar from textbooks on Chinese religions over the past century but
“the Great Way of Intentionless Action” (wuwei dadao [ ]),
“the Way of Buddha” (fodao [ ]), and “the Great Way of the Pure
Contract” (qingyue dadao [ ]).
Later, after narrating the Buddha’s birth, the text observes:
“At this, the Way of Buddha flourished once more.”63
Then, in its version of the most common story of how the “way of Bud-
dha” was introduced to China—that of the dream of Han Emperor
Ming—the scripture observes:
“His officials interpreted this dream to mean that this was the perfected form of
the Buddha, so they sent envoys into the Western Kingdoms to copy and bring
back Buddha scriptures. Then [or: because of this] they built Buddha stupas
and temples, and so [the Way of Buddha? Buddha stupas and temples?] cov-
ered and spread across the Central Kingdom, and the Three Ways intermingled
and became confused. As a result, the people became mixed and disordered;
62
HY 11096, 1:3a. Here and throughout, texts in the Zhengtong daozang (the Daoist
canon of the zhengtong reign period [1436– 49]) are cited by their serial number in Wang Tu-
chien, ed., Combined Indices to the Authors and Titles of Books in Two Collections of Taoist
Literature, Harvard-Yenching Sinological Index Series no. 25 (Beijing: Yanjing University,
1925), abbreviated as HY, followed by fascicle and folio page numbers. The translation
given here is that of Stephen R. Bokenkamp, Early Daoist Scriptures (Berkeley: University
of California Press, 1997), p. 209, except that I have capitalized differently to conform with
passages below and have added “and Teachings” in the last phrase. An alternate rendition
of the last clause would be: “and the Three Ways were taught and put into practice.”
63
HY 1196, 1:4b, slightly emending Bokenkamp, Early Daoist Scriptures, p. 212.
those of the Center mingled with outsiders, and each had his own particular
object of veneration.”64
There is an odd vagueness as to the intended subject of the second sen-

tence: perhaps it is the “way of Buddha,” or perhaps the “Buddha stupas
and temples,” that are said to have “covered and spread across the Cen-
tral Kingdom.” In any case, flourishing, covering and spreading an area
(perhaps), intermingling and becoming confused: these are the actions,
and the only actions, attributed to daos.65
Elsewhere, in a passage lamenting people’s tendency to continue “re-
vering” or “upholding” (feng [ ]) daos for which there is no longer any
need, we read:
“Today, though there are some who revere the “Way of Five Pecks of Rice,”
there are others who uphold the “[Way of ] Intentionless Action” and the “Way
of Banners and Flowers,” which follows the Way of Buddha. All of these [de-
viant ways] are old matters of the Six Heavens. All have been abolished!”66
Finally, when this scripture wants to indicate that multiple “ways”

have a common source, it resorts to a different metaphor commonly used
in Chinese discourses for this purpose: that of trunk or root versus
branch.
...
“Now the three Ways are but different branches extending from the same
root. . . . These three Ways are equally methods of the Most High Lord Lao,
though they differ in their teachings and transformative effects. All three find
their source in the true Way.”67
What are the implications of the dao metaphor? Although many scrip-
tures of Celestial Master, Shangqing, and Lingbao provenance personify
dao as an ultimate cosmic deity or force with wishes and commandments
for humanity, such is not its sense in the contexts under survey here; it is
used rather to nominalize things that seem analogous to what we would
64
HY 1196, 1:5b, partially quoting but emending Bokenkamp, Early Daoist Scriptures,
pp. 214–15, to provide a more literal reading.
65
I note what seems to be an unexpected whiff of Tillichian “ultimate concern” in the
unusual expression youshang [ ] in the last line of the passage—the normal sense of
you being “to concern oneself with” and that of shang being (here) “uppermost.”
66
HY 1196, 1:7a, modifying Bokenkamp, Early Daoist Scriptures, p. 218, in the direc-
tion of increased literalness.
67
HY 1196, 1:9b, slightly modifying Bokenkamp, Early Daoist Scriptures, p. 222.
call “religions.” It does so by imagining them as paths. Imagined as the

objects of human agents’ actions, paths may be issued, set forth, laid
down, upheld, followed, strayed or deviated from, or lost,68 or the wrong
path may be taken; imagined as agents, they may deviate or be correct,
they may flourish or decline, or they may remain distinct or become in-
termingled and confused. These are very weak senses of agency, and they
are nonorganic. “Path” metaphors are, however, rather holistic in at least
two senses: a path, unless broken, runs continuously from beginning to
end, whether it divides or rejoins, and it is not possible for an individual
to walk-practice (the double sense of a common verb in such contexts,
xing [ ]) more than one path at a time unless the paths have merged to
form one. But note, finally, that people’s relation to daos is not one of pas-
sive containment, membership, or sheer belonging. People seek, travel,
follow, abandon, or deviate from daos, rather than simply being con-
tained in them; the verbs are verbs of doing, not copulae.69
3. “law,” “method[s],” or “regulations” (fa [ ])

Another common nominalizing idiom, used more often to refer to what
we would term “Buddhism” than to “Daoism” (but also used for the lat-
ter as well), is the use of the term fa [ ], alone or (like dao) in com-
pounds. In translations of imported scriptures this term was (like dao)
often employed as a technical equivalent of the Sanskrit dharma, but in
contexts such as the ones collected here its use is clearly more general-
izing than that, referring to the sum total of teachings, communities, in-
stitutions, and practices associated with the Indian sage.
All of the following instances are drawn from Wei Shou’s sixth-century
treatise. In each, it clearly seems that fa is meant not in the limited and
rather technical sense of the teaching attributed to the Buddha’s discovery
and teaching but in a much broader sense approximating what is meant by
“Buddhism” in modern discourses.
“Hereupon the essence of the Law was greatly manifested in the Middle
Plain.”70
68
For example, the expression midao [ ], meaning “to lose one’s way,” “a mis-
guided way,” “a path taken by those who are misguided,” “a path of confusion,” etc., de-
pending on context, appears almost a hundred times in the Taisho canon.
69
The path metaphor, fundamental to Buddhist discourse, is so richly developed in that
tradition that some have called for its appropriation as a cross-cultural category, in part as
a corrective to the tendency to use Western-derived categories to analyze non-Western so-
cieties but not the reverse. See Robert E. Buswell, Jr., and Robert M. Gimello, “Introduc-
tion,” in Paths to Liberation: The Marga and Its Transformations in Buddhist Thought, ed.
Robert E. Buswell and Robert M. Gimello (Honolulu: University of Hawaii Press, 1992),
pp. 1–9.
70
Hurvitz (n. 44 above), p. 50.
“He both loved the Yellow Thearch and Lao[zi] and held highly the Law of
Buddha.”71
“Throughout the time of persecution of the Law it [a stone vihara with an im-
age over the tomb of the monk Huishi] still stood whole.”72
“When the Law of Buddha was suppressed. . . .”73
The reader will further recall the statement already extracted above from
the Inner Explanations of the Three Heavens, a reminder that fa was used
not only to nominalize the repertoire of practices and understandings im-
ported from India and Central Asia:
“These three Ways are equally methods [or laws] of the Most High Lord Lao.”
The fa metaphor is synecdochal, reducing the totality of aspects of

what we would call a “religion” to one aspect. But that aspect is not
creedal, as it typically is in the West, but rather praxeological, a set of
norms or regulations—implying, first and foremost, regulations concern-
ing what to do. Very seldom is agency of any kind attributed to fas, and,
to my knowledge, they are never personified.
4. “the teachings of x” (x [ ] )
At some point in the modern era, probably (in Chinese) as a back-formation
from a Japanese neologism that, like the Japanese shukyo (Chinese zong-
jiao [ ]), was created to translate “religions” and its equivalents in
other European languages, writers of Japanese and Chinese began to use
expressions of the form “X jiao” to denote what Euro-Americans were
calling “religions.” Thus, dokyo/daojiao [ ], literally “the teaching[s]
of or about the dao,” was used in contexts where “Daoism” would be used
in Western discourses; similarly bukkyo/fojiao [ ] for “Buddhism” and
(less commonly) rujiao [ ] for “Confucianism.” Now, as Cantwell
Smith and others noted long ago, premodern Chinese discourse almost
completely lacks this formulation used in this way. In the Chinese Bud-
dhist canon, for example, one finds over four thousand instances of the
juxtaposition of the terms fo and jiao, but more than ninety-nine percent
71
Modifying ibid., p. 52. We also see “founder synecdoche” here.
72
Ibid., pp. 62–63.
73
Ibid., p. 69.
of these simply mean something like “the Buddha taught” or “the Bud-
dha’s teaching [that].” Only with extreme rarity do such compounds seem
to gather up and nominalize everything that one might mean in Euro-
American discourse by a term such as “Buddhism,” and even when one
does find such cases, the ambiguity of the syntax usually permits other,
nonreifying readings.
One such instance appears in the perhaps early fourth century Treatise
for the Rectification of Unjust Criticism (Zhengwu lun [ ]) by an
unknown author:74
75
Arthur Link translates: “Buddhism guides men by means of compassion

and love,”76 taking the string fojiao as the compound subject of the sen-
tence and rendering it in its modern sense. But, taking the string as sub-
ject and verb, one could alternatively translate: “The Buddha taught and
led by means of compassion and humaneness.” Even if one insists on
reading fojiao as a nominal compound, it would surely be better to render
the sentence along the lines of “The Buddha’s teaching leads by means
of compassion and humaneness.”
Western translators sometimes render jiao as “doctrine” (or, as we
have seen, they render a compound meaning “the teaching of X,” even
more distortively as “Xism”). But the meaning of the term emphasizes
not—unlike “doctrine” (with its basis in doxa)—the attitude of the “be-
liever” but the source of the teaching, the one who taught it, or else (in
cases like mingjiao [ ] “the teaching of names”77) what it is a teach-
ing about. Jiaos are not personified, and only seldom is any agency met-
aphorically attributed to them (although the Zhengwu lun statement
might constitute an exception); normally teachers are portrayed as bring-
ing about certain effects in people by means of teachings, rather than
teachings somehow acting of themselves.
5. metaphors for the interrelationships among the things

5. so imagined
A full treatment of the subject under discussion here would include at-
tention to metaphors for the relationships among metaphorically con-
strued, plural “religion”-like entities in Chinese discourses as well as
among “religions” in Western discourses. Here I will merely sketch a
74
See Zürcher (n. 28 above), p. 15; Arthur E. Link, “Cheng-wu lun: The Rectification
of Unjustified Criticism,” Oriens Extremis 8 (1961): 136–65; and Tsukamoto (n. 57 above),
pp. 178–79.
75
T 2102, 52:8c17–18.
76
Link, p. 160.
77
On the sense of this term in the early medieval period, see Zürcher, Buddhist Con-
quest, pp. 86–87.
few preliminary observations on another topic on which more research is

needed.
As noted above, one common Chinese metaphor for showing the co-
belonging of two bodies of tradition is the organic imagery of root and
branch. Other, nonorganic metaphors work similarly to say three things
at once: (1) the two (or more) things are in some ultimate sense really the
same; (2) their differences are merely matters of relative location or
function; and (3) one of the things in question is usually suggested to be
the superior member of the dyad, triad, and so on—in other words, al-
though the assertion of sameness appears benign or generous, often the
metaphor effectively casts the interrelationships among the plural things
as hierarchical.
One of these other metaphors portrays the relationship between two
bodies of teaching as one of “inner” and “outer,” usually with the under-
standing that inner is the hierarchically superior position. We already
saw this dyad in one of the statements illustrating “founder synecdoche”
quoted above:
“[The Duke of ] Zhou and Confucius are identical with the Buddha, the Buddha
with [the Duke of ] Zhou and Confucius. This is merely to name them with re-
spect to outer and inner.”78
The language of that which is “outside the realm (or the quarters)” versus
that which is “inside the realm,” fangwai [ ] and fangnei [ ], per-
vaded early medieval polemical discourses, and here it was the outer po-
sition that was conceded to be superior. This dyad is used in Sun Chuo’s
[ ] (ca. 300–380 c.e.) Yudao lun [ ], and we find it sprinkled
throughout the apologetic writings of the Buddhist monk Huiyuan and of
his anticlerical interlocutors.79 This terminology had a pedigree stretch-
ing back to a passage in the “Inner Chapters” of the Zhuangzi (ca. 320
b.c.e.), in which Confucius, here a mouthpiece for Zhuangzi, character-
izes people who pay no attention to proper ritual and custom, freeing
themselves from convention and taking the fashioner of things as their
78
T 2102, 52:17a, emphasis added.
79
On the Yudaolun, see Zürcher, Buddhist Conquest, pp. 132–33; Schmidt-Glintzer (n.
57 above), pp. 59 ff.; and Arthur E. Link and Tim Lee, “Sun Cho’s Yü-tao-lun: A Clarifi-
cation of the Way,” Monumenta Serica 25 (1966): 169–96. Examples in Huiyuan’s and his
interlocutors’ discourses include T 2102, 52:30b6 (“One who has gone forth from the
household is a ‘guest from beyond the realm’ ”—cf. 75a20), 34c20, 84a11, and 84b8.
Zürcher (Buddhist Conquest, p. 98) observes that a noted monk was characterized as a
“gentleman from beyond the world” (fangwai zhi shi )—not because he was of foreign ori-
gin (he was not) but because he was in touch with things “from beyond the realm.” The
hierarchical effect is very clear in such statements as “How could matters from beyond the
realm possibly be embodied within the realm?” (79c5).
companion, as wandering “beyond the realm” or “beyond the guidelines”

(fangwai ) as opposed to inferior types such as Confucius, who here con-
fesses himself doomed to remain “within the realm” (fangnei ).80 Early
translators of Buddhist sutras often used fangwai to characterize the goal
of Buddhist practice.
Another early medieval strategy for asserting that two traditions, de-
spite many surface-level differences, are at a deeper level actually the
same was to speak in terms of “traces” or “footprints” ( ji [ ] or [ ]).
Sun Chuo, for example, asserts that both sets of sages, “Confucian” and
“Buddhist,” were alike “awakened” or “enlightened” ( jue [ ]) and that
although they left different “traces” in their teachings and practices, “that
by which they left traces” (qi suo yi ji zhe [ ]) was the same
state of “awakenedness.”81 The visible differences in teaching and prac-
tice are thus made to seem trivial—though they were hardly trivial to the
authors who left records of debates between rival proponents. Once
again the pedigree for this strategy stretches back to the Zhuangzi.82
Finally, various binary classifications—left-right, yin-yang, life-
death—were employed simultaneously to link and hierarchically distin-
guish traditions. The Celestial Master Scripture of the Inner Explanations
of the Three Heavens offers a cosmologically and mythologically rich ex-
ample of this trope. The scripture begins by characterizing the pneumas
(qi [ ]) of the Central Kingdom (i.e., civilized “China”) as yang and
those of the outer barbarian kingdoms as yin; people living in yin areas,
we are told, require extremely strict prohibitions. Later we read:
Laozi is the lord of living transformation; ‡akyamuni is the lord of transforma-

tion by death. As a result, Laozi was born from his mother’s left armpit and is
lord of the left. The left is the side of the yang breaths that govern the Azure
Palace with its Registers of Life. ‡akyamuni was born from his mother’s right
armpit and is lord of the right. The right is the side of yin breaths and the black
records of the Registers of Death. In this respect the differences between the
teachings of Laozi and ‡akyamuni are those between the laws of left and right
[ ].83
Despite this clear dichotomy, the same scripture, as we saw above, asserts
that “each path in the end returns without distinction to the True Way.”
80
The passage appears in the sixth chapter; for alternate translations, see Burton Wat-
son, Chuang Tzu: Basic Writings (New York: Columbia University Press, 1964), p. 83; and
A. C. Graham, Chuang-tzu: The Inner Chapters (London: Mandala/HarperCollins, 1991),
p. 89. For a later Buddhist author’s comment on this hierarchical taxonomy, which explic-
itly traces it to the Zhuangzi, see T 2126, 54:247a.
81
T 2102, 52:17a13–14. See also Zürcher, Buddhist Conquest, pp. 91, 133.
82
See Graham, p. 133; and Campany, To Live as Long as Heaven and Earth (n. 43
above), p. 201.
83
HY 1196, 1:9b; Bokenkamp, Early Daoist Scriptures (n. 62 above), pp. 222–23.
The “way of Buddha,” having been safely classified as inferior yet nec-
essary for softening the tough nature of barbarians in outlying lands, may
be ascribed to the same originating force as this scripture’s own tradition.
6. on “belief ” and “believers”: an excursus

It is well known that Western discourses on “religions” commonly take
belief—an inward assent to doctrinal propositions or else lists of such
propositions themselves—as their key defining feature. Hence “belief ”
and “faith” become synecdoches for religions in general, and participants
are summarily labeled as “believers.” This complex of assumptions is
strikingly absent from Chinese discourses, at least in the early medieval
period. For one thing, the actual topics of debates among rival proponents
for the most part concern practices and values, not propositions or doc-
trines and not people’s inner attitudes toward these (which is not to say
that matters of mental attitude were ignored in Chinese religious texts and
practices).84 And for another, when the term xin [ ]—the closest ana-
logue to “believe” or “belief ”—is used, it usually connotes not assent to
propositions but trust or confidence in a teaching, method, or path. On
this point, the following passage from Mouzi lihuo lun is instructive:
, ,
“You slander the divine transcendents, repress the wondrous and anomalous,
and do not believe [or trust] that there is a dao of not dying. Why do you be-
lieve that only by the dao of Buddha can one attain deliverance from the
world? [or, why do you trust exclusively in the way of Buddha as a means for
delivering oneself from the world?]”85
84
An easy way to confirm this statement is to scan the content of the questions posted
to “Master Mou” by his interlocutor in the Mouzi lihuo lun. They include such matters as
these: How can you speak so differently from Confucius and still take our indigenous clas-
sics seriously? Why are the Buddhist scriptures so lengthy when compared with the Chi-
nese classics? Why did the Buddha’s body have thirty-two marks? Why must monks shave
their heads and practice celibacy, practices that go against the value of filiality (in that fil-
iality dictates that one’s body is the legacy of the family and must not be willfully injured
or diminished, as well as that one must produce lineage heirs)? Why must monks wear
such strange clothing and beg for their food? Why must Buddhists value renunciation and
giving over the accumulation of resources and taking pleasure in sumptuousness? Why
does Buddha prohibit the eating of meat while permitting the eating of grain (contrary to
one understanding of longevity regimens at the time)? The only strictly doctrinal question
that I can find in the treatise is the one concerning rebirth.
85
T 2102, 52:6b27–28. In my judgment, Keenan (p. 161) errs in his translation of both
key phrases: “refuse to believe in a way to avoid death [emphasis added]” misses the syn-
tactical force the verb you [ ] (here, “that there exists”) has in the Chinese; “believe that
only the Buddha Tao can save the world” is a possible translation, but in such contexts
dushi [ ] usually indicates not what “the dao of Buddha” will do to the world but what
the practitioner can do for himself by means of “the dao of Buddha.”
Even, as here, when it seems to be a question of “belief,” what is indi-

cated by the term is not affirmation or denial of a set of propositional
doctrines but confidence in one way of attaining an ultimate goal and
lack of confidence in another way.86
Finally, when it comes to speaking of the “members” of “religions,” or
“believers” (as Western discourse tends to frame things), in early medi-
eval Chinese there is a suffix, zhe [ ], which works somewhat similarly
to the English-ist and its equivalent European forms. But the crucial dif-
ferences are that zhe follows only verbs or verbal phrases (so that noth-
ing equivalent to the simple “Buddhist” or “Daoist” is possible) and that
there is no invocation of “belief ”—the range of verbs is much wider. In
sum, the language tends to emphasize practice or some mode of active
participation rather than either simple membership in a container-like set
or assent to a set of core doctrines.
What can we conclude from this survey?
1. We do find a tendency in early medieval Chinese texts—especially
in certain types of contexts (as will be seen in the next section)—to refer
nominally to entities that seem to correspond roughly to the ones named
“religions” in Western discourse. This nominalization implies reifica-
tion, though the degree of reification is perhaps less than it is in Western
discourse.
2. Only seldom are these entities metaphorically pictured as agents;
normally it is people (individuals or groups) who are spoken of as doing
things with respect to the entities. With very rare exceptions, verbs of ac-
tion attributed to the entities are intransitive verbs and connote things
like flourishing, declining, or spreading.
3. Although we do find nominalization, weak reification, and meta-
phorical construction of general entities roughly corresponding to our
“religions,” the metaphors are different and carry different implications.
We might speculate that the differences correspond to, or translate into,
differences in how people have actually participated in what we would
call their “religions” in the two contexts of early medieval China and
early modern Europe, in much the way that Lakoff and Johnson suggest
Westerners live in time differently than they otherwise might because
they imagine time as money. It is one thing, for example, to imagine that
one is traveling on a way, another thing to imagine that one is comprised
in a container; it is one thing to picture a way or a teaching “spreading”
86
Even in many Western contexts, one suspects that such a statement would hold true.
Paul Veyne, Did the Greeks Believe in Their Myths? An Essay on the Constitutive Imagi-
nation, trans. Paula Wissig (Chicago: University of Chicago Press, 1988) exemplifies an
approach to the vexed question of “belief ” that is commensurate with the complexity of
people’s actual relations to the stories and assertions current in their cultures, a welcome
improvement over the on/off toggle-switch approach usually taken.
or “flowing” across a terrain, another to imagine an army fighting a war

on hostile territory.
4. The way in which daos, and so on, are spoken of in Chinese texts
by no means implies that they are total, thought-encompassing “systems”
in the sense of conceptual frameworks, unlike the way in which “reli-
gions” have often been pictured in Western discourse. Chinese texts as-
sume that what we term “religions” are fully commensurable and easily
mappable one to another, even when they are as sharply differentiated as
fangwai versus fangnei. Nor do the Chinese metaphors imply the high
degree of holistic integration implied by the Western organic metaphors.
The dao metaphor does, however, entail a degree of teleology—not so
much with regard to the inevitable direction of the history or develop-
ment through time of the dao itself as with regard to the goal of the prac-
titioner who “practices” or “walks” it.
5. We would be better served by translations (when they are possible)
that preserve the metaphors structuring Chinese discourse than ones that
directly map a term such as fodao, “the way of Buddha,” onto the English
“Buddhism” and its Western-language equivalents.
We must go further, however, to consider the typical sorts of contexts
in which such nominalizing, reifying, metaphorical constructions are
called for in the first place.
contexts of reification and metaphorical construction

Even from the brief survey just conducted, it is evident that Cantwell
Smith cannot have been right to say that “Fundamentally it is the out-
sider who names a religious system. It is the observer who conceptual-
izes a religion as a denotable existent.”87 Such conceptualizations do not
depend on whether the speaker is a religious “insider” or “outsider,” as I
have argued elsewhere in the case of “theoretical” analyses of ritual.88
However, Cantwell Smith was astute to note that “religion as a sys-
tematic entity, as it emerged in the seventeenth and eighteenth centuries,
is a concept of polemics and apologetics”89 and that “religions” tend to
be given names and treated as entities when cultural boundaries are
crossed and when multiple, rival, or new traditions are encountered. It is
a likely hypothesis, argued by Cantwell Smith and ripe for further re-
search, that the tendency to nominalize and reify “religions,” daos, jiaos,
and so on, and to conceive of them in metaphorically specific ways are
87
Smith, The Meaning and End of Religion (n. 5 above), p. 129.
88
Robert Ford Campany, “Xunzi and Durkheim as Theorists of Ritual Practice,” in Dis-
course and Practice, ed. Frank Reynolds and David Tracy (Albany, N.Y.: SUNY Press,
1992), pp. 197–231; repr. in Readings in Ritual Studies, ed. Ronald L. Grimes (Upper Sad-
dle River, N.J.: Prentice-Hall, 1996), pp. 86–103.
89
Smith, The Meaning and End of Religion, p. 43.
most in evidence where there is heightened awareness of religious plu-

rality and difference—and therefore also, very often, religious rivalry
and competition for resources, patronage, and prestige, with attendant at-
tempts to classify and narrate so as to bring some conceptual and rhetori-
cal order to the confusing field of players. Early medieval China and
early modern Europe were two such contexts.
In a social context in which there is but one predominant way of doing
the important things of life, religious practices themselves, to some ex-
tent, define or create their own communities. They do so for this simple
reason: people must come together to do these practices and to learn how
to do them; there is, in other words, a relatively low quotient of “imag-
inedness” in these communities.90 Even in early times, such social envi-
ronments were probably much rarer than we often think, for most human
groups are at least aware of others on their margins who do things differ-
ently. In contexts where such differences become acute, where religious
plurality is not only evident but also the locus of some particular prob-
lems, nominalizations and reifications (one or another “ism,” fodao, or,
at an even more abstract, generic level, “religions”) begin to be invoked.
In early medieval China, as probably in most other such contexts, it was
the attempt to negotiate such differences that created the need for reifi-
cation in the first place.
The point is simple but bears elaboration. Metaphorical expressions
that gather up multiple texts, ideas, practices, and persons and picture
them as a “path” or a “teaching” are used in contrastive situations, where
a difference is being encountered and negotiated. This is so even in a text
such as Wei Shou’s Shilao zhi, where we have the story told of how the
alien teachings and practices of the ‡akya sage were introduced to the
Central Realm—how the foreign practices and ideas came to be imported
into “China,” and how “Chinese” responded. It is much more obviously
the case in a treatise that takes the form of the Mouzi lihuo lun, with its
litany of requests for explanation and justification of strange, foreign
practices and understandings, or a scripture like the Santian neijie jing, a
cosmogonic-mythic narrative reframing of religious difference. Absent a
startling difference that demands to be accounted for, the nominalizations
and metaphoric imagery of such texts would be largely unnecessary; the
texts arise at the boundaries and borders between one set of teachings-
practices and another, and they are always framed from the point of view
of someone who, even when favoring one side over the other, writes as
if it is possible to weigh both on the same scale and implies that two (or
more) particular things are members of a common genus. Such situations,
90
In the sense specified by Benedict Anderson, Imagined Communities: Reflections on
the Origin and Spread of Naturalism, 2d ed. (London: Versa, 1991).
of course, involve contestation, and authors’ stances are not neutral, even
when, as we have seen, they assert the ultimate nondistinction of two or
more “paths.”
One other point bears making here, though its development must await
another context: such discourses as these are analogous to the “compara-
tive religion” born in early modern Europe and by now exported around
the globe. As modern scholars go about the work of “comparing reli-
gions,” they ought to ask whether people in the contexts they study en-
gage(d) in any analogous practices—where, once again, “analogous”
will mean not “their version of our X” but “a Y that, in this or that con-
text, performs something like the same function as our X does (or did) in
its own context.” In a surprising number of cases, the answer will turn
out to be in the affirmative. If myth is ideology in narrative form, and
scholarship is myth with footnotes,91 then other people’s myths will bear
being placed on a par with the myths of the writers’ own traditions. It
may even be found that modern Euro-Americans are not the only ones to
have developed writing practices analogous to footnotes.
Cantwell Smith was doubly wrong, then, when he concluded that there
were no analogues to “religions” in premodern China and that the reason
was an actual emphasis, in China and elsewhere, on interiority, on “faith”
(essentially an attitude of mind-heart-soul) rather than mere “religion.”
On the one hand, the Chinese debates were largely about how to do
things, not about the unseen contents of minds and hearts. On the other
hand, there are analogues to Western discourse on “religion(s),” and they
are occasioned, as Cantwell Smith himself might have predicted, by con-
frontations with difference.
Students of the history of Western discourse on “religions” have re-
peatedly noted that it, too, arose in a context of innovation, diversity, and
fresh contact (often in colonial situations) with foreign ways. In the
West, to speak of one “religion” is also to imply its distinction and dif-
ference from (and also partial similarity to) other species in the same ge-
nus. So much could also be said of the Chinese terminology of dao and
jiao, even when these are used in the singular in phrases such as “the dao
of X,” at least weakly implying a distinction from daos of Y, Z, and so
on. But Western discourse on “religions” is strongly contrastive in an-
other sense as well: to name a “religion” in Western discourse is to imply
a strong sense in which it is a “religion” as opposed to other, non-“reli-
gious” kinds of things. This type of contrast is largely absent in China.
The reasons for this profound taxonomic difference are well worth inves-
tigating, but they would take us beyond the scope of this essay, involving
as they do the shape of “religion” as a generic category, the history of the
91
See Lincoln (n. 42 above), p. 209.
ways in which it has been differentiated from other categories of phe-

nomena, and the ramifications for that history of the institutional clashes
between church and state in European societies.92
In Western discourses, “religions” are, relatively speaking, “like-us,”
whereas “unlike-us” are the “other” categories of “magic,” “supersti-
tion,”93 “witchcraft” and “heresy,” always implied to be different kinds of
things from “religions.” (“Popular religion” is always a borderline cate-
gory—it is religious but is the kind of religion least like “ours.”) This
sort of contrast, too, is largely absent in Chinese discourse, which speaks
of “deviant daos,” “the dao of the left,” “licentious sacrifices,” and so on,
without implying that such daos or such sacrifices are another kind of
thing than daos or sacrifices proper.
Modern, nontheological Western discourse on “religions” is known for
its apparent or attempted neutrality with regard to its objects of dis-
course, at least relative to the Chinese cases mentioned here, where every
party to the discourse has a clear, and a clearly religious, ax to grind. But
that is hardly to suggest that Western discourse on such matters is value-
neutral, and the interestedness shows itself in an unexpected way—in the
matter of the construction and maintenance of Western disciplinary and
academic boundaries. In the West, in other words, defining “Daoism”
may be crucial not only to some practicing Daoists (since Western rec-
ognition may play a crucial role in legitimating or altering the shape of
communities and practices under study) but also when it comes to estab-
lishing who in the academy is qualified to speak about it, just as defining
“religion” is crucial when it is matter of deciding who may speak about
it and in what terms (one thinks of the old and still-ongoing debate about
the sui generis nature of “religion”). One has only to scan any recent an-
nouncement of academic job openings to see the taxonomies (“Daoism,”
“East Asian Buddhism,” etc.) at work. Our discourses on the “isms”
structuring our collegiate curricula, job searches, publication lists, jour-
nals, and conference panels are not value-neutral, and they make a real
difference in the distribution of academic prestige.
92
When the problem of what we could call “religion vs. the state” was debated in early
medieval China—a context in which no such locution was ever developed, since the insti-
tutional context in which it would have made sense did not exist—the debate did take the
form of arguing whether two spheres or realms are involved and whether they should re-
main distinct (fangwai/fangnei ), but the crux of the issue always came down to a specific
set of protocols or practices (should monks be compelled to bow to rulers during court
ceremonies?), and the language does not suggest that practice of the dao is a fundamen-
tally different category of activity than other areas or forms of life. The question was usu-
ally whether one may pursue such self-cultivational activities at court (thus justifying
monks’ refusal to bow to rulers) or only in private settings.
93
It is noteworthy that the modern Chinese expression usually used to translate this
term, mixin [ ], partakes of the path metaphor and essentially means “misguided
trust,” implying that one’s faculty of trust, confidence, or belief has taken a wrong way.
from religion-entities to repertoires and imagined

communities: two summary proposals
Instead of thinking and speaking of religions as entities, how else might
we think and speak of them? I close with two brief suggestions.
It is at least worth exploring whether such alleged things as “Daoism”
and “Buddhism” are helpfully seen as “imagined communities,” in the
felicitous phrase Benedict Anderson has applied to the similarly abstract
entities known as “nations.” Anderson notes that all communities of
larger than face-to-face-contact size are “imagined” in the following
sense: “the members of even the smallest nation will never know most of
their fellow-members, meet them, or even hear of them, yet in the minds
of each lives the image of their communion.”94 Anderson immediately
notes that the important question is that of how—to use his language, in
what “styles”—this imagining takes place. He writes:
Javanese villagers have always known that they are connected to people they
have never seen, but these ties were once imagined particularistically—as def-
initely stretchable nets of kinship and clientship. Until quite recently, the Jav-
anese language had no word meaning the abstraction “society.”95 We may
today think of the French aristocracy of the ancien régime as a class, but surely
it was imagined this way only very late. To the question “Who is the Comte de
X?” the normal answer would have been, not “a member of the aristocracy,”
but “the lord of X,” “the uncle of the Baronne de Y,” or “a client of the Duc de
Z.”96
Taking our cue from Anderson, we would search our texts for indications
of the imagined communities to which they refer. Such communities
might or might not be as general as “the way of Buddha” (fodao); I sus-
pect that they will often be more particular. They would also vary ac-
cording to situation and interlocutor, just as ethnic identifications are
known to vary according to whom a subject is speaking and what the
topic and context of discussion are.97
Furthermore, the word “refer” as used three sentences ago invites the
misunderstanding that so general an imagined community as fodao
94
Anderson, p. 6. For a recent critique of Anderson’s approach to ethnicity, albeit one that
does not affect my argument here, see Frank Proschan, “Peoples of the Gourd: Imagined Eth-
nicities in Highland Southeast Asia,” Journal of Asian Studies 60 (2001): 999–1001.
95
But did it have any expressions that operated analogously, in contexts where we, or
where contemporary Javanese, might invoke such a term?
96
Anderson, pp. 6–7.
97
See Michael Moerman, “Ethnic Identification in a Complex Civilization: Who Are
the Lue?” American Anthropologist 67 (1965): 1215–30, and “Being Lue: Uses and
Abuses of Ethnic Identification,” American Ethnological Society, Annual Spring Meeting,
Proceedings (1967), pp. 153–69.
somehow just exists and retains unity and coherence independently of

references to it. We should think of the coherence of such imagined com-
munities as something repeatedly claimed, constructed, portrayed, or
posited in texts, rituals, and other artifacts and activities, rather than as
simply given. Much of this claiming concerns the past: the importance of
retrospective selection, organization, and classification by latecomers as
they tell the stories of communities they are in the process of imagining,
highlighting certain aspects of the past and creatively forgetting others,
cannot be overstated.98 Processes of the (again often retrospective) con-
struction of lineages and the selection and arranging of scriptural canons
are places where the process of community-imagining can be observed
especially clearly. As we observe such processes at work, we will notice
common touchstones, things referred to again and again—certain words,
figures, stories, or texts—but how these are portrayed, used, and inter-
preted may vary so dramatically that the mere notation of references to
them gains us very little. This is why the antiholistic use of the plural, as
in Jonathan Smith’s and others’ references to “Christianities,” is a much
more textually and historically accurate scholarly practice, especially if
we are to avoid inadvertently inventing new imagined communities our-
selves. Above all, one thing is surely clear: there is no core “essence”
that could constitute whatever coherence such communities do display.
Our discussion of the contexts of nominalization and reification fur-
thermore shows the importance of “others” against whom an “our X”
(where X will metaphorically be portrayed as a tradition, group, way,
etc.) can be demarcated. We will often observe strategies of community-
imagining at play in texts stemming from contexts of close contact with
perceived others. It may be that some imagined “others” are strictly nec-
essary for the claiming of an “own” identity and coherence.99 For, absent
a perceived plurality of communities, ways, and so on, there is little oc-
casion for nominalizing, reifying, or otherwise picturing an imagined
community as a thing with certain properties.
Another way to think and speak of religions, rather than treating them
as fully integrated systems and as containers into which persons, ideas,
practices, and texts may be fit without remainder, is to imagine them as
repertoires of resources. Ann Swidler has recently shown in considerable
empirical detail that people relate to elements of their culture in this way,
as tool kits or repertoires used variously by individuals in negotiating
98
See recently the provocative comments in Willi Braun, “Amnesia in the Production of
(Christian) History,” Bulletin of the Council of Societies for the Study of Religion 28
(1999): 3–8, comments on 5.
99
Again, Moerman’s work (“Ethnic Identification,” “Being Lue”) on the relativity of
ethnic-identity claims to the conversational situation and the presence of certain types of
interlocutors with certain interests is quite significant in this regard.
their lives.100 Important questions for research include not only what is
in a given repertoire but also how and in what circumstances any given
piece in the repertoire is performed on some occasions and by some ac-
tors but not others.101 Swidler has found of cultural repertoires—and I
see no reason why the same may not be said of religious repertoires—
that they are organized around certain concrete “scenes or situations of
action,”102 often narrative in nature, and that people avail themselves of
multiple scenes as they negotiate their lives, even when these scenes
carry contradictory implications regarding a particular area of life (such
as, in Swidler’s case, love), because each scene is especially good for de-
ciding or talking about one particular aspect of that area of life and no
one scene suffices for all of it. Swidler finds of cultural repertoires that
they are not accessible to everyone in the same degree and that people
use different amounts of culture even when they have equal access to it;
people use culture more in situations of flux or novelty, when their lives
are uncertain—another statement that could easily be extended to how
people use their religions.103 A repertoire may contain different and in-
deed contradictory models of certain areas or aspects of life because
these models answer different sets of questions; people resort to these
models in their discourse about meanings and values even when they re-
ject certain implications of each model as implausible, in part because
each model describes something about the real constraints of life and in-
stitutions or, rather (more correctly), about the lines of action individuals
pursue in the context of those constraints and institutions.104
All of this runs contrary to the Geertzian emphasis on culture as all-
encompassing ethos and on religions as “cultural systems”; Swidler
shows us agents using culture’s repertoire in complex, varying ways on
various occasions, shifting the cultural framing of a problem in mid-
discourse.105 It also runs counter to the tendency to think of religions as
“conceptual systems” (Émile Durkheim), “systems of symbols” (Geertz),
or “theoretical schemes” (Robin Hortin) outside of which “members” of
said religions cannot think.106 And if we imagine religions as repertoires
100
Ann Swidler, Talk of Love: How Culture Matters (Chicago: University of Chicago
Press, 2001).
101
These ideas are introduced in ibid., pp. 24 ff.
102
Ibid., p. 34.
103
See esp., ibid., pp. 52 ff. and 99 ff.
104
See esp., ibid., pp. 132–33.
105
See esp., ibid., p. 79. Compare the post-Geertzian formulations of Robert Hymes,
Way and Byway: Taoism, Local Religion, and Models of Divinity in Sung and Modern
China (Berkeley: University of California Press, 2002), pp. 5–12. Hymes, too, adopts the
metaphor of repertoire.
106
A tendency helpfully analyzed and criticized by Terry Godlove in his essay, “In
What Sense Are Religions Conceptual Frameworks?” Journal of the American Academy of
Religion 52 (1983): 289–305.
used by people in these ways, we may even begin to deconstruct the gap
posited by the modern study of religion between itself and its objects—
“the difference between those who sufficiently transcend culture and his-
tory [and religion(s)] to perceive the universal (and scientific) in contrast
to those who remain trapped in cultural and historical [and religious] par-
ticularity and are therein so naturally amenable to being the object of
study.”107 If we imagine religions and cultures as repertoires, then every-
one—not merely those who study religions but also those who partici-
pate in them—is potentially in the position of bricoleur, syncretist, and
comparativist.
In sum, my argument is not that we should cease speaking of religions
in cultures that lack an analogous vocabulary because they lack that vo-
cabulary; for, as I have shown, matters are not that simple, and even a
culture as different from the modern West as that of early medieval
China, with its situation of new religious imports and plurality, generated
analogous usages. But, if we are to go on speaking of religions, we
should at least find new metaphors for doing so. If possible, the new
metaphors should avoid picturing religions as really existent things in the
world; as organisms; as hard-sided, clearly demarcated containers of
people and things; and as agents, because picturing them in all these
ways falsifies the actual state of things and skews our research questions
in unfortunate ways. Religions do not exist, at least not in the same way
that people and their textual and visual artifacts and performances do.
And when religions are metaphorically imagined as doing things, it be-
comes harder to see the agents who really and nonmetaphorically do
things: people.
Indiana University
107
Catherine Bell, Ritual: Perspectives and Dimensions (New York: Oxford University
Press, 1997), p. 259.

On The Very Idea of Religions

Uploaded by

Copyright:

Available Formats

On The Very Idea of Religions

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

On The Very Idea of Religions

Uploaded by

Copyright:

Available Formats

On the Very Idea of Religions (In the Modern West and in Early Medieval China)

Author(s): Robert Ford Campany

ç 2003 by The University of Chicago. All rights reserved.

increasing by two or three new religions every day.”2 Whatever shifts in

apparently neutral descriptive statements about religions are couched will

therefore is theirs to define. It is a second-order . . . concept.”8 But use of

on religions as entities (of certain metaphorically

postulation of “uniqueness.” Much will depend on the framing of the issue.

Without further specification, an apparent entity named by some such

So pervasive is the habit of reification that we do well to remind our-

symbols which acts to,” he exclusively devotes himself to what is meant

religions are living organisms (often plants)

Taoism: Growth of a Religion22

As suggested by these few examples, authors often turn to organismic

In sum, seeing religions as comparable to organisms encourages us to

religions are personified agents

Where Manichaeism failed, Christianity succeeded when Constantine es-

The first example aptly illustrates the pitfalls of ascribing agency to

religions are marketable commodities

religions are armies; their spread and success are warfare

short encyclopedia entry on the historical relationship between two tra-

some early medieval chinese metaphors

1. founder or paragon synecdoche

This content downloaded from 142.103.160.110 on Fri, 08 spread one02:45:29

phenomena we would habitually call “Buddhism” (Leon Hurvitz translates,

“[He] had always honored the Way of Buddha.”55

above, where fodao is clearly being used to nominalize the entirety of

Again, as in the case of fodao, these uses of shendao are exceptional.

own time, religious plurality is similarly a matter of various daos, and

“At this, the Way of Buddha flourished once more.”63

There is an odd vagueness as to the intended subject of the second sen-

Finally, when this scripture wants to indicate that multiple “ways”

call “religions.” It does so by imagining them as paths. Imagined as the

3. “law,” “method[s],” or “regulations” (fa [ ])

“When the Law of Buddha was suppressed. . . .”73

The fa metaphor is synecdochal, reducing the totality of aspects of

Arthur Link translates: “Buddhism guides men by means of compassion

5. metaphors for the interrelationships among the things

few preliminary observations on another topic on which more research is

companion, as wandering “beyond the realm” or “beyond the guidelines”

Laozi is the lord of living transformation; ‡akyamuni is the lord of transforma-

6. on “belief ” and “believers”: an excursus

Even, as here, when it seems to be a question of “belief,” what is indi-

or “flowing” across a terrain, another to imagine an army fighting a war

contexts of reification and metaphorical construction

most in evidence where there is heightened awareness of religious plu-

ways in which it has been differentiated from other categories of phe-

from religion-entities to repertoires and imagined

somehow just exists and retains unity and coherence independently of

You might also like