2019FHAU: Journal of Ethnographic Theory 9 (2): 439–460
THEMED COLLECTION
THE TURN TO LIFE, PART 2
Life, domesticated and undomesticated
Ghosts, sacrifice, and the efficacy of ritual practice
in early China
Michael P U E T T , Harvard University
This article explores classical Chinese conceptions of and practices surrounding life and vitality. Focus is given to the practice of
sacrifice as well as the rejections of sacrifice among millenarian movements in Chinese late antiquity. My argument is that an
engagement with this material challenges many of our understandings in the anthropology of religion concerning ritual, sacrifice, and interpretation.
Keywords: Chinese religions, sacrifice, ritual, millenarian movements, ghosts, sincerity
In the second century of the common era, a number of
millenarian movements began emerging across the Han
empire. In the Taiping Jing, a document associated with
one of these movements, the era—self-proclaimed as
“late antiquity”—was seen as one overrun with ghosts.
Errors from the past had been accumulating to such a
point that sacrifices were only empowering the ghosts
and increasing their numbers. Death (yin) was overcoming life (yang), with the ghosts killing the living with
all of their power:
Those in late antiquity have again inherited and carry
on the small errors of middle antiquity, and they increasingly make them into ever greater errors . . . When
it comes to summoning the dead, ghosts are not able to
come and eat constantly, and yet the sacrificial offerings
were nonetheless greatly increased, thereby exceeding
the proper standards. Yin grows and overcomes yang.
No one knows which ghostly and spiritual creatures
repeatedly come to gather together and eat, indulging
themselves and having their way, acting like dangerous
thieves and killing people without end. When they kill
a person, [the ghosts] see an increase in the service
[i.e., an increase in the sacrificial offerings] and see no
punishments. Why should they not continue [killing
the living] with all their strength? As a result, pernicious
energies grow daily. It all turns back and attacks the giver
of the sacrifices. (Wang 1992: 46.53)
Sacrifices intended to control the ghosts were in fact only
increasing their dominance. The ghosts not only flocked
to consume the ever-increasing number of sacrifices but
also killed the living in ever-greater numbers—thus resulting in yet more ghosts and yet more sacrificial offerings. The imbalance between the dead and the living
was becoming ever more dangerous:
Living humans are yang, ghosts and spirits are yin . . .
Therefore, when yin triumphs, the ghostly creatures
join together to create horrors so profound that no
words can describe it. This is called the arising of the
yin, and the decline of the yang. It causes rule and order
to be lost and endangers the living. (Wang 1992: 46.50–
51)
These cautionary words were purportedly spoken by
a “Celestial Master,” sent down by a moral deity, Heaven,
to save humanity. Similar claims were made by the
Celestial Masters, a millenarian movement that began
HAU: Journal of Ethnographic Theory. Volume 9, number 2. DOI: http://dx.doi.org/10.1086/706073
© The Society for Ethnographic Theory. All rights reserved. 2575-1433/2019/0902-0018$10.00
Michael PUETT
when the high god Laozi, himself a hypostatization of
the cosmos as a whole, gave revelations to warn of a
coming apocalypse caused in part by the empowerment
of ghosts through human sacrifice.
In both of these cases, a moral deity provides revelations to save the living from the growing power of ghosts.
And among these revelations were calls on humans to reduce the power of ghosts by either reforming sacrifices or
ceasing them altogether.
The emergence of millenarian movements seeking to
reject sacrificial practices and instead focus on sincere
belief in a higher, moral deity is hardly an uncommon
phenomenon in our historical and ethnographic record.
But in the case of early China we have a wealth of material on both these sacrificial practices and the millenarian movements that rejected them, material that
should be brought into our larger discussions. In particular, as we will see, the debates surrounding ghosts and
sacrifice were intimately involved with the work of controlling, enlarging, or enhancing life—a form of work
that opens up interesting possibilities in our anthropological framing of issues of ritual, sacrifice, ancestor
worship, cosmology, and belief.
Ghosts, interpretation, and anthropology
Late antiquity was not the only haunted world.
Perhaps few contemporary disciplines are as haunted
as the field of anthropology. A haunting that has taken the
form of—among other things—treating our comparative
categories with deep suspicion. As our genealogies have
successfully demonstrated, many of our categories—including most of the ones mentioned above—emerged
out of a colonial past, and many are based implicitly on
Christian, and more specifically, Protestant, understandings (Asad 1993).
I certainly agree with these critiques, and will turn to
them below. But the question is where to turn from here.
Like many anthropologists, I would argue that we need
to take indigenous categories seriously in re-thinking
our theories. But taking indigenous categories seriously
in this sense does not simply mean that we must use indigenous understandings to explore practices within
that particular culture. To begin with, as we will quickly
see, it is not entirely clear what the “Chinese” understandings of any of these issues would be. We will, on
the contrary, see an intense debate over all of these practices. Moreover, the competing understandings of these
440
practices are extremely productive: no one set of indigenous concepts will explain them fully. But, perhaps
most importantly, we should still be committed to a
comparative project that would bring these understandings into our larger anthropological discussions. The
goal should be to work toward a more cosmopolitan body
of theory, in which indigenous theories are brought into
conversation with other indigenous theories, allowing us
to develop more robust anthropological understandings.
Our genealogies may have unleashed ghosts, but the goal
should be to continue working with these ghosts, not
abandon the project. This is exactly the approach advocated by da Col and Graeber (2011) in building what they
call “ethnographic theory.”
In turning to China, such a project may seem deceptively easy. The Protestant underpinnings of our categories could not be clearer than when dealing with China.
Nor, for the same reason, would the possible alternatives. During the development of social science theories
in the nineteenth and early twentieth centuries, China
became the Other for many of these categories. China
was repeatedly presented as a culture that emphasized
ritual over belief, harmony with the world rather than
tension with it, a worship of ancestors rather than a
following of one’s individual calling, a do ut des form
of sacrifice with the gods rather than a sincere communion with them. For generations of social scientists, from
Hegel to Weber, China became the embodiment of a traditional society. The original anti-Protestant culture.
Such an Other was relatively easy to construct when
working with Chinese materials. Statements that the
world is harmonious and that we should learn to accord with it, that we should follow the wishes of the
ancestors, and that the divine world is structured like
a bureaucracy, are extremely easy to find in Chinese
texts. And then asking what sets of assumptions would
have made such statements possible would seem like a
plausible mode of interpretation. Hence an entire anthropological literature devoted to reading so-called
traditional China in exactly these ways.
But discovering the assumptions behind such statements may be missing the point. To begin with, whether
statements such as these should be understood as ritual
statements or statements of belief was a topic of intense
debate in China. And, in either case, the work required
to make the claims is itself one of the most telling aspects of the practices in question.
The work of anthropology is, in this sense, much like
the work of everyday life. To which we now turn.
441
My goal in this paper will be to explore the indigenous understandings in the classical Chinese tradition
concerning sacrifice, ritual, and belief. In the first part
of the paper, I will explore the indigenous theories of
ritual and sacrifice that developed in the classical Chinese tradition—the understandings and practices of
sacrifice and ritual that were under attack in the second
century of the common era. As Veena Das has argued,
many of our anthropological theories of sacrifice are
based upon a restricted set of understandings. Das
(1983) expanded our anthropological understandings
by exploring classical Sanskrit theories of sacrifice, and
I hope to do the same here by exploring classical Chinese theories of ritual and sacrifice. I will focus on the
theories and practices that developed in the Warring
States and Han periods (ca. fourth to first centuries BCE),
arguing that they have much to offer our current understandings. I will then turn to the anti-sacrifice movements that emerged in the first few centuries of the
common era, movements that may in turn allow us to
expand our anthropological understandings of belief
and claims to coherence.
To begin with ritual and sacrifice.
Part I: Ritual and sacrifice in China
The body, energies, and death
A helpful way of analyzing ritual practices is in terms of
why they are being undertaken. In other words, what are
the perceived problems that need to be solved, and what
would things be like if the ritual were not undertaken?
A fundamental problem in dealing with humans is that
they are a mess of different energies (qi), souls (the hun
and po), and faculties. What we would call emotions were
in fact different energies that would be pulled out through
interactions with other people—themselves other messes
of energies. These responses tend to fall into certain patterns. What might elsewhere be called a “personality”
would in early China be seen more as the sets of patterned responses that these messes of energies have fallen
into.
Humans also have highly refined energies within them:
shen. We can provisionally translate this as “divinity” or as
“spirit.” These are the energies that give one consciousness, as well as the power to affect things—to move
things, to transform things. If ghostly energies are associated with the earth, divine energies are associated with
the heavens. The divinities above are pure shen—they
LIFE, DOMESTICATED
AND UNDOMESTICATED
move freely, can see clearly, and can transform things
flawlessly. They are not dominated by patterns of emotions—those seem to be associated more with the earthly
energies.
Living humans have a little shen. Not as much and
not as purely as the shen above, but enough to have some
degree of consciousness and some ability to transform
things around them.
This brief mapping of humans will give a sense of the
problems that humans confront. Minus any of the ritual
work humans can undertake, there is always a danger
that humans will fall into dangerous patterns of jealousies, angers, and resentments, at both a mundane, everyday level and a larger communal level. But the most
dangerous moment occurs at death. Released from the
confines of the human body, the demonic energies would
look back on the living and witness them continuing with
their lives; their worst angers, jealousies, and resentments
would be drawn out.
This is what would be called a gui—the term I have
been translating as “ghost,” although another equally
plausible translation would be “demon.” These demonic
presences pervade the earthly realm. When they are not
domesticated within something—a domesticated human body, a (as we will see) domesticated sacrificial
space—they are extraordinarily dangerous.
Over time after death, the most powerful element—
the shen—would tend to float up to the heavens. If the
danger of the demonic energies of the recently deceased
is that they become enmeshed in anger, jealousies, and
resentment against the living, the concern over time is
that the shen of those deceased long ago would float
into the heavens and become indifferent to human concerns—every bit as indifferent as the other spirits above.
The divinities above, including Heaven and the various spirits, are autonomous, self-productive, and selfdependent. They do not need humans and are often indifferent to human concerns. They can therefore also,
from the point of view of humans, appear highly capricious: since humans seem often not to be a part of their
concerns, the divinities will often—and perhaps unintentionally—undertake activities that are harmful for
humans. In short, if the danger of ghosts is that their
anger is often directed explicitly at the living, the danger
of spirits is that they can be indifferent and thus, from
the point of view of humans, capricious.
Given these sets of problems, the ritual work of dealing with the dead—and hence enhancing life—entailed
a process of both separating the various components of
Michael PUETT
what was once in a living human and transforming
them.
The body would be placed into a tomb, along with the
earthly energies and souls associated with the person
when alive. Various texts would call these souls the po,
or the hun and the po (Brashier 1996). The hope was that
these energies would remain with the body in the tomb.
Objects associated with the person in life—furniture,
food, clothing, texts—would be placed in the tomb. And
requests from the living would be placed at the tomb
calling for the dead souls to remain there, removed from
the living (Seidel 1987).
What would ultimately befall these energies was hotly
debated. Perhaps they simply dissipated in the tomb. Or
perhaps they went on to different spaces—western paradises, or subterranean hells. But, from the point of the
view of the living, one of the most important issues was
simply that they be kept away. Their presence among the
living would tend to bring out the most dangerous tendencies from both.
If the goal was to remove the earthly energies and
souls of the deceased from the living, the goal with the
shen was to transform it into an ancestor—into a supportive entity that would act toward the living as if they
were descendants.
After the corpse was buried, a tablet would be set up to
be the new form for the spirit on earth. The spirit would
thus hopefully be removed from the corpse, as well as
from the sets of patterned responses that had defined
the person while alive. The spirit would be given an ancestral title based upon the generation from the living
head of the family, and the tablet would be placed in
the ancestral hall based upon that rank. Now the ancestral sacrifices would begin.
This shift from the burial of the person to the rituals
involving the spirit as an ancestor would be called a
shift from “mourning” to “sacrifice.” As the Liji (Book
of Rites), one of the major compendia of ritual theory
from early China, puts it:
In sacrificing, one is called “filial son” and “filial grandson.” In mourning, one is called “grieving son” and
“grieving grandson.” (Liji, “Za ji,” 107/20.12/6)
This is the practice that has come to be known as “ancestor worship” in the anthropological literature. These
sacrifices to the ancestral spirits were also directly related
to the sacrifices that would be given to the other spirits
as well. The hope was to bring these energies of the shen
442
into networks of support that would enhance the life of
humanity.
Sacrifice and the world of forms: Corpses, names,
and tablets
The “Jifa” chapter of the Liji puts it succinctly:
Generally speaking, as for everything that is born between Heaven and Earth, all of these can be said to have
allotments. When the myriad things die, all are said to be
cut off; when humans die, they are named “ghosts.” The
kings of all under Heaven . . . established ancestral temples, ancestral halls, altars, and sacrificial areas, and they
offered sacrifices . . . Therefore the king erected seven
ancestral temples, with an altar and level area for each.
They were called: the temple for the father, the temple
for the grandfather, the temple for the great-grandfather,
the temple for the great-great grandfather, and the temple for the highest ancestor . . . When they removed each
tablet, they placed it at the altar; when they removed it
from the altar, they placed it at the level area . . . When
the tablet was removed from the level area, they were
called “ghosts.” (Liji, “Ji fa,” 122/24.4–5)
The designation “ghost” would be used before the creature became an object of sacrifice—an ancestor—and
then again after the sacrifices came to an end. During
the period when it was an object of sacrifice, it would
be given a temple name and a tablet, and also, as we will
see, a corpse impersonator to receive the offerings. The
name and body were used to give form to the spirit, so
that it could be domesticated and controlled, thus transforming the ghost into an entity supportive of humans.
Even writing was one of the inventions that allowed humans to begin domesticating and gaining control over
the ghosts and Heaven. As the Huainanzi put it: “In ancient times, when Cang Jie created writing, Heaven rained
grain and the ghosts cried all night” (Huainanzi, “Benjing,” ICS, 8/62/27–28).
But let us start with the impersonator.
After the corpse was placed in the tomb, it would be
given another ritual body. The ancestor would be called
down to inhabit what is often translated as “impersonator” (shi). But the term literally means “corpse.” (The
same word was used to describe the corpse now in the
tomb.) The ancestor was being given a new ritual body
to receive offerings. With its ritual body, the deceased
becomes (hopefully) dependent upon the living. The
living feed the dead just as, in reverse, the older fed the
younger while alive.
443
A similar logic is used in sacrifices to other spirits as
well. The goal is to make the ghosts and spirits dependent
on humans by involving them in webs of exchange—
webs of dependent relationships.
And what is the nature of these relationships?
In one of these impersonation rituals, the grandson
of the deceased would be called upon to serve as the
corpse of his deceased grandfather. The Liji theorizes
this impersonation as follows (Puett 2014):
Now, according to the way of sacrificing, the grandson
acted as the impersonator of the king’s father. He who
was made to act as the impersonator was the son of he
who made the sacrifice. The father faced north and
served him. By means of this, he made clear the way
of a son serving his father. This is the relation of father
and son. (Liji, “Ji tong,” ICS, 131/26/14)
In this example, the father is the ruler, and his father is
the deceased. The grandson plays the corpse of his
grandfather, and the father plays the role of son to his
own son. Through this role reversal, a normative father–
son relationship is built up across the three generations.
Within the ritual space, the ghost becomes a supportive ancestor and the living become filial descendants,
with perfect father–son dyads extended across the three
generations.
The father, therefore, is the sacrificer. It is he who offers sacrifices to his father, via his son as corpse. And it
is thus also he who becomes the center of the web of relationships created by the sacrifice. The deceased (ideally)
becomes dependent on the father, just as the son serving
as the corpse of the grandfather is dependent on the father outside the ritual.
At the royal level, the dyad is extended. The sacrifices
are also given to Heaven, and the leftovers are spread
to the various families of the realm. The ruler thus becomes, ritually speaking, a Son of Heaven, and also, going in the other direction, the father and mother of the
myriad people. In the world created within the ritual,
the disparate lineages are connected by their ritual kinship relations with the ruler, who in turn is connected
in ritual kin relations with Heaven (Puett 2005).
The goal, in short, is to create an ordered pantheon
of spirits and ancestors that would operate as much as
possible on behalf of humanity. These relations, moreover, ideally place the ghosts and even higher spirits
into ritual kinship relations with humans. Within the
LIFE, DOMESTICATED
AND UNDOMESTICATED
sacrificial space, the cosmos would become, ritually speaking, a single lineage. And the ruler would be at the center
of this web of relationships, making, to quote again from
the Liji, “all under Heaven as one family” (Liji, “Li yun,”
9.22/62/5).
So how are we to understand this?
Ritualization and domestication
The “Li yun,” one of the chapters of the Liji, compares
the constructions of reality created by ritual to the domestication of the landscape through agriculture (Puett
2010a). The world that would have pre-existed agriculture was one in which a basic level of harmony existed.
But it was also a world that was inherently dangerous
for humans. It would at seemingly random times be
too cold or too hot, and humans would die from exposure; rains and dry spells would occur, leaving floods or
droughts; animals would eat humans, and humans would
unwittingly kill themselves by eating poisonous plants.
The development of agriculture involved finding general
patterns to the cold and heat (which would then be
termed “seasons”), clearing forests, domesticating crops
and animals, and killing those animals that could not be
domesticated. The entire world was thus organized into
a new type of ordered harmony—but one now organized
by and for humans. As a result, what had before been a
series of forces, powers, and elements that—from the
point of view of humans—interacted in dangerous ways
were domesticated and transformed into a new web of
relationships in which humans, now at the center, would
flourish. The abundant fertility of life, although never
completely controlled, was organized into forms beneficial to humanity.
Ritual, according to the chapter, works the same way.
Just as agriculture involved domesticating discrete and
dangerous entities and organizing them into new relationships with humanity at the center, so does ritual involve the same. If we think of life as flows of qi, including
the highly refined qi of shen (spirits), then the problem
in dealing with humans is much the same as the problem
in dealing with plants and animals: the issue is one of organizing the qi into workable patterns (Farquhar and
Zhang 2012). By so domesticating the world, what was
before a dangerous (for humans) world of highly capricious spirits and potentially malevolent ghosts is transformed into a world of benevolent spirits and ancestors
acting in support of humanity (Puett 2013a). The more
removed from the center one is, the less domesticated
one is. At the furthest remove, one is simply a ghost.
Michael PUETT
Ritual theory
It is precisely this endless work of domesticating ghosts
into ancestors and spirits that becomes the focus of classical Chinese ritual theory. Unlike a theorization that
proceeds from a view of ritual as a single transformative
moment—a Christian baptism, for example, or an initiation, or a wedding ceremony—the theorization in
this case is of a ritual based upon the ongoing work of
domestication, a domestication that is never complete.
For all of these participants—the living, the remains
of dead humans, powers in the larger cosmos—the normative dispositions being hopefully inculcated through
the sacrifices are an ideal world which only in part alters the actual behavior of the participants after they
leave the ritual: sons will often not be filial, the populace
will usually not view the ruler as a father, the ghosts will
often continue to curse the living, Heaven will continue
to be capricious. The transformations, in other words,
are temporary and inadequate. Ghosts and spirits are always more powerful than the ritual attempts to domesticate them. They always exceed the rituals. And when
the rituals fail, the ghosts return. Underlying the ritual
world of supportive gods and benevolent ancestors are
ghosts, just as a capricious world of disasters underlies
and frequently breaks through the domesticated world
of agriculture. Despite human attempts at domestication, the world continues to be disparate and dominated
by capricious and highly dangerous powers.
And, to a large extent, the failure of these rituals is
why in a larger sense they work (Seligman et al. 2008:
30; Puett 2010b, 2013b). As Seligman, Weller, Simon,
and I have argued, building upon Chinese ritual theory:
These arguments imply that ritual always operates in a
world that is fragmented and fractured. Moreover, the
subjunctive world created by ritual is always doomed ultimately to fail—the ordered world of flawless repetition
can never fully replace the broken world of experience.
This is why the tension between the two is inherent
and, ultimately, unbridgeable. Indeed, this tension is the
driving force behind the performance of ritual: the endless
work of ritual is necessary precisely because the ordered
world of ritual is inevitably only temporary. . . . If the
world is always fractured, and if ritual always operates
in tension with such a world, then we need to think of
ritual in terms of such an endlessly doomed dynamic.
Ritual should be seen as operating in, to again quote
Robert Orsi, “the register of the tragic.” Although the
claims of ritual may be of an ordered, flawless system,
the workings of ritual are always in the realm of the limited and the ultimately doomed. (Seligman et al. 2008: 30)
444
The vitality here, the energies of qi, is not something
that is produced by or within these domesticated ritual
spaces. The domesticated spaces serve rather to shift this
energy into workable—but ultimately doomed—worlds.
The world that exists within the ritual space is one of
pure patriarchal hierarchy. A domesticated space, but
one lacking in women, fertility, life. Ritual creates an ordered world, but it is also a world in which life—or at
least the processes that would allow for a continuation
of life—is erased. One is reminded of the opening lines
of Sylvia Plath’s “Munich Mannequins”: “Perfection is
terrible, it cannot have children./ Cold as snow breath,
it tamps the womb” (Plath 1981). Ritual domestication
temporarily orders a capricious world, but such a world
of ritual perfection is a world without life.
The world outside the ritual space has abundant fertility, but also death; a world of reproduction, but a world
as well of ghosts. In contrast, the ritual world has neither
life nor death. A flourishing life, then, actually resides in
between the two—in between the ritual and non-ritual
worlds, in the disjunction between the two.
Part II: Toward a more global anthropology
Multiple worlds, multiple selves
Within the ritual space, we find a world that perfectly
matches the list of characteristics commonly attributed
to China—reverence for ancestors, a harmonious cosmos, a do ut des form of sacrifice. But all of these are ritual statements, not statements of belief. Far from being
assumptions, these notions are on the contrary highly
counter-intuitive, made in explicit disjunction with what
exists outside the ritual space. They refer to ritual attempts
to create such a world—with, again, the full (and necessary) understanding that such attempts will ultimately
fail. And that is also why they can be so effective.
Disjunctions have certainly been encountered before
in anthropological theory. Take, for example, Maurice
Bloch’s famous critique of Geertz’s (1973) interpretation of Balinese conceptions of time as cyclical (1977).
As Bloch argued, Geertz’s method consisted of taking ritual statements concerning time cycles and then reading
these as indicative of a larger worldview. But isn’t it possible that these were indeed just ritual statements, and
not at all indicative of what conceptions existed outside
the ritual space?
But Bloch’s further argument (1989) is that ritual—
and the world of religion in general—should be understood as naturalizing a set of hierarchies that are in fact,
outside the ritual space, potentially contingent. Such a
445
reading is based in part upon Bloch’s understanding
(1989) of Malagasy ancestor worship, a ritual that grants
authority to the elders of society by conferring sacred
legitimacy to them. In other words, the problem that ritual is trying to solve is that the world of everyday life is
contingent, based on linear time and thus ever-changing.
The world of ritual, on the contrary, is based on the timeless authority of the ancestors.
This framework also informs Bloch’s argument concerning rebounding violence, which Bloch defines as a
set of rituals directed at solving the problem of “how human beings can be the constituent elements of permanent institutional structures” (1992: 19). As Bloch argues:
To achieve this they must appear, in a certain light at
least, to be immortal and unchanging, and therefore
other than human; at the same time, they must also
be truly alive, in a human body which cannot but be
perceived as transformative and mortal. The construction of the ritual drama of rebounding violence is an
attempt to avoid the force of this contradiction. (p. 19)
To resolve this contradiction, the world of ritual posits a
division between a transcendental order of the divine
that is eternal and unchanging, and a vitality associated
with the world of plants and animals. Rebounding violence in ritual then consists of a two-step process. In
the first stage of a ritual (be that initiation, possession,
or sacrifice), one is taken over by the transcendental order. In initiation, for example, one is taken from a childlike state, in which one is closer to animals, and then
symbolically killed and brought closer to the divine powers (be they ancestors or gods). In sacrifice, the vitality of
the sacrificer is identified with the animal that is then
killed and taken over by the transcendent powers. In the
second stage of the ritual process, there occurs a reversal,
in which the participant, now associated with the transcendent powers, returns and attacks the vital, animal
world. In an initiation, this is when the initiated children
(now adults) return to attack and appropriate sources of
fertility—i.e., they are now warriors and hunters. In sacrifice, this is when the sacrificed animal is consumed. At
this point, after the victim has been killed, it no longer
represents the sacrificer in his vital aspect; it is simply
meat that is then consumed by the participants. This restores vitality to those who had just lost it symbolically
because of the killing of their vital aspects during the
sacrifice.
Such an argument comes out of a very long tradition
in anthropology, in which religion is read as a reified
LIFE, DOMESTICATED
AND UNDOMESTICATED
version of social hierarchies: society creates gods in its
image, but then forgets that society is in fact the creator.
It takes a modern social scientist to unmask the sacred
and show that it is really a reified form of what are in
fact contingent social practices.
The dispute between Geertz, arguing that culture is
first, and Bloch, arguing that ritual, and religion in general, are a projection of otherwise contingent hierarchies
in society, has played out in analyses of China as well. Is
Chinese ancestor worship, as well as its bureaucratic
pantheon of gods, a simple projection of a patriarchal
family structure and bureaucratic government respectively (Wolf 1974), or were these religious assumptions
first, and the respective social structures emerged from
them (Keightley 1978)?
But the arguments in the Liji give another possibility, and it is one that opens up some interesting issues
for anthropological theory. The ritual theory in the sacrifice chapters of the Liji is quite clear that what appears in the ritual is not based upon an assumption
of a harmonious world, benevolent ancestors, and supportive gods. But the rituals are equally not trying to instill such a belief either. On the contrary, rituals are presented as constructs, radically disjunct from the world
outside the ritual. And the nature of the substitutions
and role reversals works to emphasize these disjunctions. There is nothing to unmask here. No one is being
asked to believe that the ancestors are really benevolent,
or that the ruler is really the Son of Heaven, or that he is
the people’s father and mother. The supportive gods
and benevolent ancestors are clearly a product of ritual
action. And such relations are not even being presented
as an ideal that we could or even should try to live within.
The interest in the theory is rather in the disjunctions
themselves, in the constant work of going back and forth
between the spaces.
This is a very different model than one that emphasizes coherence—whether that coherence is to be located
in the coherence of a belief system or in the (partial) coherence provided by a structured, enduring, transcendental religious sphere.
To return to Bloch’s theory of rebounding violence.
The problem underlying these rituals in China is not
one of reconciling the needs of permanent social and
political institutions with the fact of the mortality of individual humans (a tension represented symbolically as
a contrast between a transcendental world above and
a vital world on earth). The problem is one of working with a fundamentally discontinuous world of potentially dangerous interactions, and the work of ritual
Michael PUETT
involves attempts to domesticate both the natural and
divine worlds—with, again, the constant realization that
such a domestication will never (and ultimately, for life
to continue, should never) be complete.
The implications of this can be rather far-reaching.
But let’s begin with some of the most immediate consequences. Think of the entire body of scholarship devoted to proclaiming that in so-called traditional China,
people thought themselves to be, to quote one of the
most famous studies, “under the ancestors’ shadow”
(Hsu 1967). According to such a view, the living would
normatively follow the paths laid out by their ancestors,
and ancestor worship was the ritual means of instantiating such values. Otherwise brilliant ethnographies of
the self in contemporary China will occasionally fall into
this paradigm as well when they contrast a modern self
with a traditional vision of the self, in which humans
saw themselves as residing under the ancestors’ shadow
(Liu 2000, 2002). But, as we have seen, the rituals were
not aimed at instilling a belief that one should follow the
ancestors; the rituals were on the contrary seen as (endlessly failing) attempts to domesticate ghosts into ancestors—an ancestral world actively being constructed by
and for the living.
A similar point can be made in terms of cosmology.
In opposition to the tendency to read cosmology in
China as based on an assumption of a harmonious cosmos, Philippe Descola (2013) has argued that cosmology in China should be thought of as an endless attempt to connect disparate things—what Descola terms
“analogism.” Analogism in this sense operates at a cosmological level as sacrifice operates at a ritual level. In
both cases, the goal is to work with a fragmented, discontinuous world and to construct from this a world of continuity—a world of relationships in which humans can
flourish. From the point of view of discontinuity, sacrifice and cosmology are both acts of domestication. To
read statements concerning the harmony of the cosmos
as statements of belief is to miss the extraordinary ritual
work underlying the utterance.
None of this is unique to China. Let’s return to ancestral rituals in Madagascar. David Graeber (1995) has
challenged Bloch’s understanding, arguing that Malagasy
ancestral rituals do not involve socializing the participants to submit themselves to the will of the ancestors.
On the contrary, the rituals involve acts of violence against
the deceased, and result in an empowerment of the living vis-à-vis the dead. Putting it in terms of Chinese ritual theory: the rituals work because they play on the (ultimately unresolvable) disjunctions between the power
446
that the deceased once had while alive and the hopes
for the living vis-à-vis the dead.
And one wonders too about Bali. It is possible that
there was an overall assumption of time as cyclical. It
is also possible that there was a split between such a notion within ritual space and a more universal experience of time as linear outside of it. But one wonders
if it might not also be possible that the really interesting
questions would come in with the constant interplay
and disjunctions between several different worlds. What
was happening when people moved back and forth
across these different worlds? This is, of course, precisely
the re-reading of the Balinese material that Howe (2000)
provides. Labeling any of these worlds as universal, or as
based on the need for a permanent set of institutional
structures (represented as a transcendental order), is to
domesticate this complexity every bit as much as one
does by denying the complexity and reading the ritual
statements as indicative of a coherent worldview.
Thought of in this way, sociality consists of these
endless orderings and re-orderings of the world, from
mundane daily life to the larger cosmos. Multiple selves,
multiple worlds. If we take ritual theory from China seriously, we have to give up our models based upon assumptions of coherence (however defined). To refer
back to Orsi: the work of ritual always operates in the
register of the tragic.
Continuity, discontinuity, and the play of substitutions
and transformations in sacrifice
What would this mean for our theories of sacrifice? As
we have seen, sacrifices in China do not, contrary to the
way they have so often been portrayed, involve a do ut
des logic—such a logic that would only be operative
within the ritual space, whereas the real work of the ritual involves the disjunction between the ritual space
and what is outside it. And they are equally not based
upon the interplay of a transcendental order and a vital
order as theorized by Bloch. So how then should we
understand sacrifice?
Sacrifice is one of the many comparative terms in
anthropology that is being treated with suspicion—a
concept overly dependent on Christian understandings
(see, for example, Detienne 1989: 20). And the classic
work on the subject by Hubert and Mauss (1981) is often taken as an example of this. The theory in Hubert
and Mauss does indeed consist of several different and
potentially contradictory strands organized together
under a vaguely evolutionary framework culminating
in the Christian eucharist.
447
But I would argue that sacrifice is still an extremely
helpful comparative concept. And, if our goal should be
not to reject our comparative terms but rather to rethink them through indigenous understandings, then
we should also recognize that Hubert and Mauss were
attempting to do precisely this. One of the key reasons
the work has proven to be so productive is that it was
developed primarily not from Christian understandings
but rather from Sanskrit theories, applied to Catholic
evidence as well as a smattering of ethnographic material from around the world. Even if the result was at
times confused, it has nonetheless succeeded in opening up a huge number of provocative questions. The
goal in its aftermath should be to continue the work of
bringing in more indigenous understandings to enrich
our understanding of the phenomena.
One very successful example of this approach has been
the work that figures like Lévi-Strauss, Sahlins, Valeri,
and Bloch have done in developing Hubert and Mauss’s
arguments. Hubert and Mauss discuss the sacralization of
the sacrifier, and elsewhere discuss the victim as a substitute for a god. Taken out of the general evolutionary theory Hubert and Mauss employ, one finds a fascinating
insight into the interplay of sacralization and desacralization in sacrifice.
We have already seen the importance for such an
idea for Bloch: in his theory of rebounding violence,
the notion of the victim as a substitute for the sacrifier
allowed Bloch to argue that this represented the first
stage in the ritual complex of rebounding violence, with
the transcendental overcoming the vital aspects of the
sacrifier. But let us expand our comparative focus to
look at other ways this substitution plays out. If we take
ritual theory from early China seriously, we should be
looking at the way that sacrifice creates disjunctions that
operate in productive tension with the world outside the
ritual space. Exploring the different ways that this occurs
will help to place the Chinese material into a larger comparative perspective, while also allowing the Chinese material to frame some of the terms of the comparison.
For both Sahlins and Valeri, working with Polynesian materials, the focus has been on the notion of the
victim as a substitute for the sacrifier, representing the
sacrifier in his disordered state. As Sahlins has argued,
“It could even be argued that cannibalism exists in nuce
in most sacrifice, inasmuch as the victim must be identified with the sacrifier and is often consumed: either
by the congregation as communion or by the sacrifier
(priest) as representative of the god” (Sahlins 1983: 82).
The result of these substitution processes would be, rit-
LIFE, DOMESTICATED
AND UNDOMESTICATED
ually speaking, the sacralization of the sacrifier. As Valeri has put it:
In a sacrifice, the offering—which is a substitute of the
sacrifier—is eaten by the god and thus feeds him. But
it also becomes part of him and thus participates in his
powers. Insofar as part of the offering so transformed
returns to the sacrifier to feed him, he acquires part of
the divine powers. The sacrifier may be viewed as undergoing, through his substitute, symbolic cannibalization and resuscitation: he is transformed by being eaten,
incorporated by the god. (Valeri 1989: 224)
The human sacrifier is ostensibly submitting himself to
the divine, but in fact the ritual act involves the sacrifier
acquiring—one could say stealing—divine powers.
The substitution processes of sacralization and desacralization can operate in other ways as well. As Valeri
argues with hunting rituals of the Huaulu:
All wild animals, and particularly pig, deer, and cassowary, are imbued with sacredness for the Huaulu.
What this means is that they are intangible, forbidden,
and that there are nonhuman, occult forces that represent that intangibility and actively enforce it. To appropriate the animals, then, human must desacralize them,
which means persuading these occult forces to let them
go. Or perhaps is it more correct to say that every time a
Huaulu kills an animal, he commits a sacrilegious act
and thus becomes vulnerable to retaliation from the occult forces that protect his victim. These forces must
therefore be neutralized and propitiated through a variety of ritual means, some of which are sacrificial by
any definition of the term “sacrifice.” The main one is
a metonymic preservation of the sacredness (i.e., intangibility) of the animal through the “offering” of its
prime part—the head—to the occult powers. (Valeri
1994: 116–17)
The goal is to remove material from the world controlled
by divine powers in order to allow human appropriation.
The overt act of the sacrifice thus involves an offering to
the divine powers, but the result is that humans are able
to appropriate the resources they want.
Despite the many differences, the practice of sacrifice
in all of these Austronesian examples operates in terms
of introducing discontinuity. A world without sacrifice is
a world of continuity, in which all beings are connected
in a single order controlled by and often descended from
the divine. Human appropriation is thus an act of introducing discontinuity either by appropriating animals
from this divine world or by appropriating divine power.
Michael PUETT
The Vedic material from which Hubert and Mauss were
working operated in very similar ways.
As Veena Das (1983) has argued, however, an entire
body of Sanskrit theory concerning sacrifice does not focus on the act of killing, or on the notion that the victim
represents the sacrifier in his disordered state. And much
the same holds true for the classical Chinese theories as
well. There is little concern with the act of killing in classical Chinese theories of sacrifice, and the play of substitutions is far more generalized than simply a focus on the
victim and the sacrifier. So how might we understand
this from a larger theoretical perspective?
Building upon Lévi-Strauss (1966: 224–5) and Marshall Sahlins (1985), Michael Scott (2007) has argued
that a key question for any practice is whether the problem underlying and motivating the practice is seen as being one of too much continuity (in which case the goal is
to create discontinuity) or too much discontinuity (in
which case the goal is to create continuity).
What is intriguing about Chinese sacrificial practice
from this perspective is that, as we have seen, it is on the
contrary based on a concern with discontinuity. The
point of the sacrifice was to create continuity. To move
from a world of disparate things, dangerous ghosts, and
capricious spirits—a world of discontinuity—to one of
hierarchical continuity, in which the sacrificer connects
these entities into a continuous hierarchy with himself
at the center of the web of relationships. The goal was
not to make (ostensibly, anyway) amends for removing
elements from a divine world but rather to domesticate
the divine world by building sets of relationships with
humans at the center.
Similarly, the play of substitutions in sacrifice involves more than simply the identification of the offering with the sacrifier. The play also occurs among the
participants (with the grandson, for example, acting as
the corpse of the deceased grandfather, but doing so in
the ritual role of the deceased as an ancestor), and the
role of the offerings is to define these relationships. The
result is an endless interplay of substitutions and transformations: the son becomes the grandfather, Heaven
becomes the father, the ruler becomes the son, the ruler
becomes the father and mother, etc. The concern is
thus less to sacralize the human than to humanize the
divine, domesticating the divine into a web of human
relationships.
To return to Bloch. In early China, the divine is not a
transcendental realm in Bloch’s sense of the term. The
goal in sacrifice is to domesticate the world—most cer-
448
tainly including the divine. And the play of substitutions
involves a series of substitutions in the various subject
positions of the (human and divine) participants, rather
than identifications with the victim. Fitting this into
Bloch’s model risks losing what seems to be the real concern of the ritual: the focus is not the endless interplay of
vitality and transcendence but rather to empower the
sacrificer by domesticating both the natural and divine
worlds—with, again, the constant realization that such
a domestication will never (and ultimately, for life to
continue, should never) be complete.
Read this way, Bloch has powerfully articulated the
workings of a ritual complex based upon a concern with
an essentially dualistic cosmology. There may well be
many societies that fit this model well (although Graeber
has raised questions about the degree to which the Malagasy material fully fits). But I would prefer to see this as
one permutation of a much larger set of possible concerns. What if the problem is not seen to be one of reconciling the needs of permanent social and political institutions with the fact of the mortality of individual
humans (a tension represented symbolically as a contrast between a transcendental world above and a vital
world on earth)? What if the problem, as is the case
here, is one of working with a fundamentally discontinuous world of potentially dangerous interactions? The
work of ritual thus involves a different set of endless acts
of domestication and substitutions.
So, if sacrifice can work in terms of either the (limited) establishment of either continuity or discontinuity,
and in terms of either the sacralization of the human or
the domestication of the divine, then what is the larger
problematic that leads to these permutations? If we again
take ritual theory from China seriously, one possibility
might be that sacrifice serves to establish a limited empowerment of life for humanity against a larger order—
an empowerment that is limited because that larger order is inevitably more powerful than human attempts to
work with it. This can work from the point of view of either continuity or discontinuity—either trying to break
from a continuous divine order or, as in China, trying
to build continuity out of a discontinuous order. Either
way, the work of sacrifice involves the productive disjunction between the created ritual order and the world
outside.
The sacrificial ruse
As Valeri has argued, there is a bit of a ruse in any sacrifice (1994: 104). And, as is now clear with the inclusion
449
of the material from China, this ruse can operate either
way—in terms of either continuity or discontinuity.
In terms of the failed construction of discontinuity,
one of the most famous examples is Hesiod’s myth of
Prometheus (Vernant 1989). Prometheus’s sacrifice to
Zeus is ostensibly a submission to the divine, but in fact
Prometheus gives Zeus the inedible bones, wrapped enticingly in fat, while Prometheus himself gets to eat the
meat. This is, of course, the same division that was made
in most sacrifices in ancient Greece: the goal of the sacrifice was to create discontinuity, to gain the support of
the gods while winning autonomy for humanity. Ostensibly a submission, but in fact a ruse.
And that is most certainly the case with China as well.
But the concern was not to create discontinuity but rather
to create continuity. Yes, humanity gives food from itself
(domesticated by human labor) to the divine, but it does
so with the goal of—to whatever extent possible—domesticating the divine. The sacrifice does not in any
way involve a submission of the living to the ancestors.
On the contrary. It is the sacrificer, not the ancestor,
who is empowered by the sacrifice. Indeed, one of the
results of the sacrifice is that everything, including the
ancestor, becomes dependent on the chief sacrificer.
But not fully. The key with sacrifice is that the ruse is
never complete. Going back to Greek sacrificial practice.
The sacrifice was a ruse to gain autonomy from the gods,
but such autonomy was short-lived: humans were dependent on the gods for their livelihood. And the ruse
in part only underlined that dependence. Yes, humans
get to eat the meat in the sacrifice, while the gods get only
the inedible bones. But humans need to eat the meat to
survive, while gods, being immortal do not. The gods
have life; humans do only briefly. The ruse only underscores the dependence that humans have on the gods.
If Greek sacrificial ritual involved an ultimately
doomed attempt to gain autonomy from the divine,
Chinese sacrificial ritual involved an ultimately doomed
attempt to domesticate the divine. In both cases, the sacrificial ruse was based upon an ultimately failing creation of a disjunctive ritual world.
Perhaps this explains part of the motivation for sacrifice in general. There have been attempts to link the
practice of sacrifice to the emergence of domestication
(Smith 1987; Descola 2013: 228). But, as Valeri (1994)
has pointed out with the Huaulu, sacrifice can also occur with hunting rituals, using non-domestic animals.
But if sacrifice is thought of, as hypothesized above,
as a limited empowerment of life for humanity, then
LIFE, DOMESTICATED
AND UNDOMESTICATED
it becomes clear that domestication is only one of the
many modes in which sacrifice can operate. In the form
we see in China, this empowerment of life for humanity
certainly does take the form of domestication, and very
much operates, as Descola (2013) has correctly argued,
in terms of attempts to build a continuous world out of
what is perceived to be radically discontinuous. When
the empowerment takes the form of appropriating animals and materials from what is seen as a continuous
order—an order in which everything is inherently related—it involves an introduction of discontinuity. In
this sense, Huaulu and ancient Greek sacrifice, for all
of their many differences, share a similar concern. Even
though Huaulu sacrifice is focused on non-domestic
animals, and Greek sacrifice is focused on domesticated
ones, the problematic is similar: appropriation in both
cases involves creating a rupture from a divine, continuous order, and the sacrificial ruse, so to speak, involves
stealing some aspect of this divinity for the sake of human flourishing.
In both cases, too, it is important to note not only that
the ritual is less powerful than the divine figures and will
thus gain only a momentary support for humanity, but
the ritual practice assumes this to be the case. The source
of life is beyond the rituals, and that source would be destroyed if the rituals were to ever be successful in colonizing the world outside it.
Sacrifice, then, plays on the tensions produced through
the inevitably doomed human attempts to control, appropriate, and domesticate life. The precise permutations
of sacrificial practice depend on the ways in which these
tensions are articulated. But the exploration of these permutations should focus on the disjunctions, the plays of
multiple worlds, and the nature of the ruses required to
create doomed ritual worlds in which humanity can,
briefly, capture life. Pre-defining any of these worlds as
a coherent worldview, or pre-defining the tension as being one of a transcendental order defined by the need for
permanent structures and the vitality of life, limits our
ability to see the complexity of these disjunctions.
Sacrifice and rejections of sacrifice
The texts I have been working from here—the sacrifice
chapters of the Liji—are attempts to theorize sacrificial
practice. The chapters were written between the fourth
to second centuries BCE, but the particular focus was
to theorize sacrificial practice from the Bronze Age—
practices that the authors felt were being increasingly
rejected. To explicate this further, a brief history of
Michael PUETT
these sacrificial practices will be helpful. This will allow
us to see the agon of these practices, the conflicting attempts at encompassment, as competing figures tried
to order different worlds.
Through paleographic evidence, scholars have been
able to reconstruct a fair amount of sacrificial practice from the Bronze Age. Our first written evidence
from China is from the late Shang period (ca. 1200–
1050 BCE). The writings are divination records made
to the ancestors of the Shang kings. Many scholars have
tried to read the inscriptions as evidence of a do ut des
vision of sacrifice in which the kings would give offerings to their ancestors and in return would receive blessings from above, and many have further argued that
such sacrificial practices reveal an early emphasis on a
belief in an ordered, bureaucratic pantheon (Keightley
1978).
But the indigenous theories we have been exploring
may fit the evidence much better. Putting it in that terminology, the rituals are attempts to deal with a highly dangerous set of recently deceased figures as well as highly
capricious deities—most importantly the high god Di.
The recently deceased are constantly cursing the living.
Thus, when the living get sick, the divinations seek to
determine if the sickness is a result of a curse by one of
the recently deceased figures, and, if so, what sacrifices
might be used to alter the situation. If the sacrifices succeed, the result would be that, using the later terminology, the ghost would be transformed into a supportive
ancestor, acting on behalf of the living. And, of course,
often this would not occur. Even after the sacrifices, we
are often told in the divination records that the sickness—the curse from the recently deceased—continued.
The deceased are more powerful than the rituals. In the
later terminology, the ancestors often revert to being
ghosts. As I have argued elsewhere:
There is, thus, in the late Shang, a constant agon between humans and spirits, with spirits controlling natural phenomena and humans attempting to appropriate aspects of the natural world for their own benefit.
This results in seemingly endless attempts by humans
to placate, coax, and influence the spirits through sacrifice and divination. And the attempt seems often to fail:
the spirits are capricious and far more powerful than
the rituals humans use to control them. (2002: 44)
Tellingly, such divinations focused on ending the
sicknesses of the living would only be aimed at the re-
450
cently deceased, as they were the ones who would curse
the living. Those who had been dead longer seemed to
be more powerful, but also less directly focused on the
living. The most powerful—and most distant—entity
was Di, the high god. Di controlled the winds and rains,
but was relatively unresponsive to human manipulation.
Thus, human attempts to affect more powerful entities
would begin with the recently deceased, who were at least
potentially more pliable by human activity. The recently
deceased would be sacrificed to, and they in turn would
be called upon to “host” (bin) the ancestors in the next
generation. This would continue up the entire pantheon, ultimately to the point where the highest ancestor would be called upon to host Di (Puett 2002: 47–50).
Returning to later ritual theory: the goal of the sacrifices
was to domesticate a series of highly dangerous ghosts
and capricious spirits and to form them into a coherent,
hierarchical pantheon of ancestors and spirits operating
on behalf of the living.
Well, on behalf of a certain group of the living. We also
have divination records from the Zhou, a group to the
west who were also making sacrifices to the Shang ancestors (Shaughnessy 1985–87). The Zhou later defeated
the Shang, and instituted a system of sacrifices linking
its own ancestors to the high god, now called Heaven.
Far more records survive from the ensuing Western
Zhou dynasty (ca. 1050–771 BCE). Perhaps most importantly, we have in the bronze sacrificial vessels used
to make offerings inscriptions addressed at least in part
to the ancestors. Here we find many of the ritual statements that would become so important for the Liji theories: within the ritual space, Heaven is a benevolent
deity, the ruler is a Son of Heaven, the ancestors of
the royal family are arrayed in ascending generational
order between the living and Heaven, with the dynastic
founders, Wen and Wu, serving Heaven (Puett 2002:
54–68). These are, of course, ritual statements, aimed
at constructing such a world.
It was this Western Zhou version of these rituals that
the Liji was most concerned with reviving. And what
were they being revived against?
By the fourth through second centuries BCE, these
rituals, and the social worlds being created and worked
against through these rituals, were being destroyed. The
centralized forms of statecraft that emerged during this
period were designed to gain as much direct control over
territory as possible. With such a goal, at least the highest
level of state rituals shifted dramatically as well. Although
ancestral rituals continued, the ritual title of “Son of
451
Heaven” came to be rejected, as did state rituals aimed
at bringing a capricious Heaven into a relationship with
the ruler. The imperial rituals, developed by the First
Emperor during the Qin empire (221–206 BCE) and
consolidated by Emperor Wu (153–89 BCE) during
the subsequent Han dynasty, instead involved the sacralization of the ruler. The ruler would engage in sacrifices, but the goal was not to become a Son of Heaven,
centering the cosmos as if it were a lineage. The goal,
on the contrary, was for the ruler to become divinized
and ultimately connected to the Great One—a high god
more primordial and more powerful than Heaven (Puett
2008).
Along with these changes at the state level, attempts
began developing at the non-state level as well to transcend the entire process of sacrifice and the inevitable
result of becoming a (partially) dependent ancestor.
Sets of techniques that I have termed “self-divinization”
movements began arising in the fourth through second
centuries BCE (Puett 2002). These tended to be antisacrificial movements, oriented toward cultivating the
energies and spirits within the body. The ultimate goal
was to directly become a spirit and ascend into the heavens without dying. The result would be life as a spirit—
autonomous, living as long as Heaven and Earth, and
completely removed from the webs of dependent relationships that sacrifice would have wrapped them within.
Tellingly, part of the regimen for such practitioners
involved not eating grains—grains being the perfect
embodiment of human domestication (Campany 2009:
62–87). The goal of the practices was to transcend the
world of human domestication—including both agriculture and sacrifice—altogether.
Self-divinization is, in a sense, the under-current of
sacrificial practice. The goal here is not to domesticate
the divine but rather to divinize the human. One of the
key aspects of this is to reject or transcend the domesticated world of humanity. And the goal is also to transcend the capricious spirits above and move as close as
possible to the Great One, a divine power more primordial even than Heaven and one that could not be
domesticated to even the limited degrees that Heaven
might hopefully be domesticated.
Part III: Sincerity movements
Moral ghosts
This brings us back to the world of ghosts. Exploring
indigenous understandings of sacrificial practice can be
LIFE, DOMESTICATED
AND UNDOMESTICATED
very revealing. Equally revealing are the attempts to transcend or avoid such rituals, as in the self-divinization
practices. And revealing as well are the critiques that were
made of the practices, and the attempts to create alternate communities altogether.
In the fifth century BCE, a figure named Mozi founded
a new community based upon the claims that Heaven
was not capricious at all but rather a perfectly moral deity who had created the cosmos precisely so that humans could thrive. The emphasis was not on the human
domestication of an otherwise capricious and dangerous world but rather on the harmonious cosmos created
by a benevolent deity:
Moreover, there are ways that I (Mozi) know Heaven
loves the people deeply. It shaped and made the sun,
moon, stars, and constellations so as to illuminate and
guide them (i.e., the people). It formed and made the
four seasons, spring, autumn, winter, and summer, so
as to weave them into order. It sent down thunder,
snow, frost, rain, and dew so as to make the five grains,
hemp, and silk grow and prosper, and sent the people
to obtain materials and benefit from them. It arranged
and made mountains, streams, gorges, and valleys, and
distributed and bestowed the hundred affairs so as to
oversee and supervise the goodness and badness of the
people. It made kings, dukes, and lords and charged
them with, first, rewarding the worthy and punishing
the wicked, and, second, plundering the metals, wood,
birds, and beasts and working the five grains, hemp,
and silk so as to make the materials for people’s clothing and food. (Mozi, “Tianzhi, zhong,” 7.6b–7a)
This same emphasis on a properly ordered cosmos
can be seen in the divine sphere in general. The ghosts
were not highly dangerous entities but rather figures in
a perfectly ordered pantheon, ruled by the benevolent
deity Heaven, who rewarded the good and punished
the bad so as to give proper guidance to humanity:
Therefore, in ancient times the sage kings made manifest and understood what Heaven and the ghosts bless
and avoided what Heaven and the ghosts detest so as
to increase the benefits of all under Heaven and eradicate the harms of all under Heaven. This is why Heaven
made coldness and heat, placed the four seasons in
rhythm, and modulated the yin and yang, the rain and
dew. At the proper time the five grains ripened and
the six animals prospered. Diseases, disasters, sorrows,
plagues, inauspiciousness, and hunger did not arrive.
(Mozi, “Tianzhi, zhong,” 7.6a–6b)
Michael PUETT
The harmonious world posited in the Liji as a product
of human domestication is here seen as a creation of a
moral high god. The entire cosmos is already ordered
and perfectly moral.
And what about sacrifices? They were to be continued,
but, since the ghosts were already benevolent and already
acting on behalf of humans—indeed, providing guidance to humans—the entire emphasis on transforming
ghosts into gods, goddesses, and ancestors was rejected
altogether. The work of domestication and the play of
substitutions have no role. Sacrifices were simply instituted by the sage kings to inculcate within the practitioner a proper reverence for the ghosts:
Therefore, if it were like this, then Heaven would send
down cold and heat without moderation, snow, frost,
rain, and dew at the improper time, the five grains would
not grow, and the six animals would not prosper . . .
Therefore, in ancient times, the sage kings clarified what
Heaven and the ghosts desire and avoided what Heaven
and the ghosts detest. They thereby sought to increase
the benefits of all of Heaven and push away the problems
of all under Heaven. They thereby led the myriad peoples under Heaven to purify themselves, bathe, and
make libations and offerings to sacrifice to Heaven and
the ghosts. (Mozi, “Shangtong, zhong,” 3.5a–5b)
The key, as several chapters of the Mozi argue, is that
one must simply have faith in the ghosts.
In short, human behavior was not aimed at domesticating a capricious world so that human life could
flourish. The concern was rather to live properly within
the world created by Heaven and ruled over by Heaven
and the ghosts. Sacrifices were to continue, but only
to inculcate within humans a proper reverence for the
moral ghosts. Other rituals that did not serve a comparable purpose were to be rejected.
This emphasis on belief in a benevolent creator-deity,
on a coherent, harmonious world that we must learn to
live within properly, and on a suspicion of rituals, is, of
course, quite comparable to generally Protestant practices. Those anthropological approaches devoted to reconstructing a coherent worldview that is purportedly
believed in by the figures involved, with rituals serving
essentially to inculcate those beliefs within the practitioners, would fit the Mohist community very well. Here,
one could indeed take a ritual statement as (ideally) a
statement of belief. But, it is important to note, this is
an oppositional movement, explicitly devoted to denying the disjunctions and transformations that underlay
452
the ritual practices of the day and instead calling for a belief in coherence. As Luhrmann (2012) has argued, belief in a counter-intuitive way of thinking requires incredibly difficult work.
The Mohist communities were relatively small, and
they seemed to have died out not long after the consolidation of the empire. But later communities with a comparable focus on moral gods and ghosts were to emerge.
Millenarian movements
In 142, Laozi appeared to give revelations to a certain
Zhang Daoling. Although Laozi had earlier been understood to be a human sage who had written the Laozi,
Laozi was in these revelations the cosmos itself, now
taking human form in order to offer revelations to humanity. The Laozi text was an earlier revelation from
the same god. But that earlier revelation had been so
misinterpreted that a further revelation was necessary.
Indeed, the misinterpretations were so severe that human activity had been progressively destroying the cosmos. Laozi was thus returning to offer further revelations
to a community of followers who would hopefully survive the coming apocalypse and seed a new group of humans for the cosmos to come. The resulting community
of followers, called the Celestial Masters, created a separate state in southwestern China, preparing for what was
to come.
At roughly the same time, another salvationist movement, called the Taiping (Great Peace), emerged in the
eastern portion of the north China plain. Similarly apocalyptic, the Taiping movement held that the cosmos
had become overridden by ghosts—a problem created
directly by human sacrifice. The cosmos was in danger
of coming to an end, and a higher—and purely good—
deity was offering revelations to save humanity (Puett
2015).
The Taiping Jing
The early portions of the Taiping Jing, mentioned at the
start of this paper, are part of this context (Hendrischke
2007). The concern is that the world has been overrun
with ghosts. And this is entirely a consequence of how
humans lived while alive:
Those who exhaust themselves by bringing distress and
bitterness to themselves become distressed and embittered ghosts. Those who exhaust themselves in vileness
become vile ghosts. This is something that can clearly
453
be seen. All humans are able to understand this, and yet
none are willing to become good and make their hunsoul and spirit joyous. This is a truly severe transgression. (Taiping jing hejiao, 52.73)
The reason that ghosts should not be a problem is
that humans in fact are perfectly ordered in relationship to the cosmos.
When humans are born, they receive correct qi from
Heaven and Earth, and the four seasons and five phases
come to join as [the qi] becomes human beings. This was
the ordered form of the former humans. Their bodily
forms resided [properly] within Heaven and Earth, the
four seasons, and the five phases. The bodies of the former humans were always joyous, good, and without anxieties, turning back to transmit more life. (Taiping jing
hejiao, 52.73)
Humans in antiquity lived properly. They were thus
joyous at death, just as they had been joyous in life:
As such, hold fast to the good and study. Those who
roam joyously to the utmost become ghosts who roam
joyously. (Taiping jing hejiao, 52.73)
As such, humans—both during life and afterward—
were able to undertake their cosmic duty of helping
Heaven to generate and nourish the myriad things:
In between Heaven and Earth, all of the spirits and essences must together help Heaven generate, nourish,
and grow the twelve thousand things. Thus all of the
spirits and essences fully obtain ranks and sustenance.
This is like the myriad ministers and worthies who all
help the emperor and kings nourish the people and myriad things; they all receive ranks and sustenance. Thus
they follow Heaven as their model, always with the fifteenth day of the month a small report is sent up; at
the beginning of the next month a medium report is sent
up; and each year a large report. Therefore those with
great merit will receive promotion and those without
merit will be sent away or punished. (Taiping jing
hejiao, 151.407–408)
The wording here is particularly telling. The way that
the spirits and essence can become helpful for the nurturance of life is that Heaven itself functions as a perfect
bureaucracy, rewarding those who act properly and punishing those who do not. Far from being a cosmos filled
with dangerous ghosts, the cosmos on the contrary looks
LIFE, DOMESTICATED
AND UNDOMESTICATED
very much like the one described by the Mohists: it is a
cosmos governed by a benevolent Heaven, and overseen
by a proper bureaucracy (Puett 2009).
So why have things fallen apart? It is entirely the fault
of humans. Heaven has been sending down sages to
provide proper guidance to humanity. But despite the
clear teachings, humans have misunderstood them. These
misunderstandings have accumulated, and as a result
later generations have increasingly failed to live properly
within the cosmos and failed to undertake their proper
cosmic duties. This has led to an increasing level of resentment on the part of humans, and this in turn has resulted in an increasing number of dangerous ghosts:
Later generations were unworthy. They have on the
contrary long embittered the bodies of Heaven and
Earth, the four seasons, and the five phases. This has
caused them to be all the more resentful when they
die, distressing their hun and po souls. (Taiping jing
hejiao, 52.73)
The result is the situation mentioned at the beginning
of the essay: the ghosts have become dangerous and allpervasive, feeding off the living and thus throwing off
the balance between Heaven and Earth.
The world of dangerous ghosts, in other words, is
not an inherent part of the cosmos that humans are trying to alter through their work of domestication. The
cosmos on the contrary is inherently harmonious, and
proper human behavior plays a crucial role in maintaining that harmony. The dominance of a world of
dangerous ghosts is purely a result of improper human
activity. And among the most important of these improprieties is sacrifice.
Done properly, sacrifice for the Taiping Jing is simply a way of paying proper obeisance to higher powers
and of therefore inculcating within the living a proper
sense of reverence for the ghosts and Heaven. What the
Taiping Jing is rejecting is precisely the play of substitutions in the sacrificial ritual. There is no transformational play at all: sacrifice is simply an act of obeisance
to the ghosts. A view of sacrifice, in other words, directly reminiscent of the Mohists.
But in later generations—what the text calls “late antiquity”—sacrifice on the contrary started being used
incorrectly. As the ghosts became progressively more
dangerous and progressively more numerous, humans
have responded by increasing the sacrifices. And, by
foolishly over-indulging in sacrifice, humans have only
Michael PUETT
made the problem worse. The result is a cosmos overrun by ghosts.
How do we solve the problem? If Heaven sends down
yet more sages, they will simply give words that will yet
again be interpreted incorrectly and misunderstood:
If they [Heaven and Earth] were to wish again to give
birth to a sage, it would just be the same yet again.
Heaven has been troubled for a long time. For this reason it sent me down to give its words as announcements
to you, the Perfected. (Taiping jing hejiao, 91.350)
Accordingly, there can be no more human sages: “Therefore, Heaven does not again make a sage speak, as he
would be unable to fully eradicate all of the problems”
(Taiping jing hejiao, 91.356). And there can be no more
interpretation of any kind.
So what will the Celestial Master ask the Perfected to
do? The Perfected will instruct humans to take all of the
earlier writings and, without any interpretation, simply
put them together. Since the teachings of Heaven are
one, simply organizing by topic everything that remains
from different sages of the past together will result in a
single, clear sagely statement:
If one has these follow one other by category and
thereby supplement each other, then together they will
form one good sagely statement. (Taiping jing hejiao,
132.352)
The words should be taken literally, with no interpretation.
No interpretation, no sacrificial substitutions, no
transformations. The world is coherent and harmonious, and guided by a moral deity. The key is simply
to have faith in that coherence and in the precepts offered from above, and to accept the precepts literally.
The Celestial Masters
At roughly the same time, another group, the Celestial
Masters, formed in the southwest (Kleeman 2016). The
Celestial Masters were followers of the revelations that
the high god Laozi made to Zhang Daoling in 142. The
leader was Zhang Daoling’s grandson, Zhang Lu, who
also purportedly authored the Xiang’er commentary to
the Laozi.
The Xiang’er commentary reads the Laozi not as a
text written by a human but rather as a revelation of
the high god Laozi (Bokenkamp 1997). According to
454
the Xiang’er, Laozi is in fact the Way, and has taken human form in order to offer revelations to humanity.
Even the cosmos itself is a form that the Way has, reluctantly, taken:
“I” refers to the Way. It desires to be without a body. It
simply desires to nourish spirits; that is all. And it desires to compel humans to model themselves on this.
(Xiang’er, 154–5)
The Way would prefer not to have a body. But it needs
to nourish spirits, and it needs humans to use their
bodies to do the same (Puett 2004).
Humans, therefore, must simply follow the admonitions of the Way to use their bodies to accumulate
more essence and thus nourish spirits for the Way:
Humans should only preserve their bodies; they should
not love their bodies. What does this mean? If you
maintain the admonitions of the Way, you accumulate
goodness and complete accomplishments; accumulate
essences and complete spirits. (Xiang’er, 161–2)
If they do so, they can become transcendents and live
long:
When spirits are completed, the transcendents live
long. Consider them [the spirits] the treasures of the
body. (Xiang’er, 163)
Life, then, consists of following the precepts of the
Way and using one’s body to nourish spirits on behalf of the Way. Death means failing to follow these
precepts:
The Way is life; aberrance is death. The dead belong to
the earth; the living belong to Heaven. (Xiang’er, 295)
Life is of Heaven, while death is of the Earth.
Like the Mohists, the high god—in this case the Way—
created the cosmos with a clear system of rewards and
punishments to ensure proper human behavior. In this
case, death itself was created as punishment. Life is the
reward given to the good:
The Way established life in order to reward the good,
and established death in order to punish the bad. As for
death, this is what all men fear. The transcendent rulers
and nobles, like the common people, know to fear death
and enjoy life; it is what they practice that is different . . .
455
Although the common people fear death, they do not try
to trust in the Way, and they enjoy committing bad acts.
Is it surprising that they are not yet trying to escape from
death? The transcendent nobles fear death, trust in the
Way, and hold fast to the precepts. Therefore they join
with life. (Xiang’er, 299–303)
Insofar as the transcendent nobles follow the precepts
of the Way and have faith in the Way, they will live.
They are freed from death. Those who do not are but
corpses:
The bodies of those who do not understand the Way of
long life are all just corpses that move. It is not the
Way that moves them; theirs is entirely the motion
of corpses. The reason that people of the Way are able
to obtain the long life of transcendents is that theirs is
not the movement of corpses. They are different from
the vulgar. Thus, they are able to fulfill the potential of
their corpse and command themselves to become transcendent nobles. (Xiang’er, 72–4)
The proper use of the body, therefore, is to focus
simply on the cultivation of essences and spirits:
The body should constantly generate itself and take
calming the essences and spirits as its basis. (Xiang’er,
456)
What about other uses of the essences and of the
body—such as copulation for reproduction? The Way
decrees that this should only be done to the minimum
degree necessary to generate more bodies:
Now this [copulating to reproduce] produces great calamities. Why did the Way create (zao) it? The Way values ancestral sacrifices and values that the species does
not end. It desires that humans join their essences and
generate life; therefore [the Way] teaches it. (Xiang’er,
57–63)
The Way would prefer that there be no bodies, and it
would also prefer that there be no waste of essence
for the purposes of copulation. But it needs bodies to
nourish spirits, and it therefore needs the human species to continue. Accordingly, it has created copulation.
And, for those who do not go on to use their bodies
properly, it has created death and sacrifice.
But those who do follow the precepts of the Way cease
copulation at an early age and thereafter use their bod-
LIFE, DOMESTICATED
AND UNDOMESTICATED
ies only to complete spirits. They become transcendents,
and are freed from the world of sacrifice:
However, humans with utmost power . . . are able to not
unite and produce life. From a young age they stop this
[i.e., the losing of their essences through copulation] and
they are able to complete good spirits earlier. These are
called the essences of the Way. Thus, Heaven and Earth
have no sacrifices, dragons no offspring, transcendents
no wives, the Jade Maiden no husband. (Xiang’er, 57–
63)
Sacrifice is associated with death, not life. Indeed, it is
forbidden to those who follow the precepts of the Way:
Those who practice the Way live; those who lose the
Way die. The correct method of Heaven does not reside in sacrificing, praying, and offering. The Way
therefore forbade sacrifices, prayers, and offerings.
(Xiang’er, 374–5)
Freed from the world of reproduction, death, and sacrifice, they become self-generating and achieve long life:
They are able to model themselves on the Way, and are
therefore able to generate themselves and endure and
live long. (Xiang’er, 65–6)
This results in a division of humanity. The transcendents are able to follow the Way, nourish spirits, and
obtain long life. They do not copulate and do not engage
in sacrifices. Those who do not follow the precepts, on
the contrary, at least serve the minimal function of reproducing more bodies. But theirs is the way of death.
They die, and they become ghosts to whom must be
given sacrifices.
So why does the Way need human bodies to be used
to nourish spirits?
The body is the vehicle of the essence. Since the essence
can leave you, you should carry and orient it. When spirits are completed and the qi comes, they carry and orient
the body. If you wish to bring this task to completion, do
not depart from the One. The One is the Way. Where
does it reside in a person’s body? How does one hold fast
to it? The One does not reside in the human body . . .
It exists outside Heaven and Earth. When it enters between Heaven and Earth, it comes and goes in the human body. It moves everywhere within your skin; it does
not rest in one place. The One disperses its form as qi
and collects its form as the Taishang Laojun [i.e., Laozi],
Michael PUETT
who rules Kun Lun. It is sometimes called emptiness
and nothingness; it is sometimes called spontaneity; it
is sometimes called the nameless; all are the same.
(Xiang’er, 103–10)
The Way created bodies in order to accumulate essence
and complete spirits. The Way disperses qi throughout
the cosmos, and humans are called upon to use their
bodies in order to accumulate essence and complete spirits. Those who do so can become transcendents and live
long. But those who do not follow these precepts will
die and become ghosts, and sacrifices will have to be
given to them.
The entire cosmos thus functions normatively like a
giant factory to generate spirit. When humans fail to
follow their proper role in doing so, the entire cosmos
is thrown into danger.
The problem in what the Taiping Jing would call “late
antiquity” is that fewer and fewer humans have been
following the precepts. As a consequence, ghosts and
sacrifices are becoming ever more prevalent, and the
cosmos is not getting the support it needs for the cultivation of spirits. The result is that the cosmos is in danger. This is why Laozi took human form yet again to
offer further revelations. Those who would follow the
precepts will survive the coming cataclysm and become
the seed people for the new cosmos to come. As the Da
Dao jia ling jie puts it:
You will see Great Peace. You will pass through the catastrophes unscathed and become the seed people of
the later age. Although there will be disasters of war, illness, and flood, you will confront them without injury.
(Translation by Bokenkamp 1997: 173)
The result of these practices will be the creation of the
“seed people,” who will help to generate humanity in the
cosmos to come. Proper seeds come not from human
domestication but from transcending such domestication and simply completing the spirits within the body.
The cosmology of the Xiang’er commentary operates
as an inversion of that seen in sacrificial practice. If the
goal of sacrificial practice was to use forms to domesticate the spirits, the Laozi, according to the Xiang’er, is
written from the point of view of the spirits and, behind
them, the Way. The world of forms is simply being used
by the Way in order to generate more spirits. The body
is being used not to domesticate and control spirits but
simply to nourish and cultivate them (Puett 2010c).
456
And how does the commentary get such a reading
from the Laozi? The commentary presents itself as simply a literal reading of the text (Puett 2016). The reason
the Laozi has, from the point of view of the commentary,
been misread as being a highly mysterious text filled with
complex metaphors is because humans have failed to
simply read the text as it is: a clear set of revelations that
should be understood at face value. This, of course, is
the same argument we saw in the Taiping Jing: the revelations are clear and offer the reader direct, unmediated access to the words of the purely good deity. Just
as there are no substitutions in the sacrificial process,
so are there no interpretations necessary for understanding. The world is as it is, and the precepts for understanding it and living properly within it are as well.
All one needs is faith.
Coherence and disjunctions
We have seen an opposition. The dominant sacrificial
practices were devoted to domesticating dangerous
ghosts and capricious divinities. The world was discontinuous and fragmented, and life consisted of the endless attempts to order, encompass, and link these disparate things.
This was opposed by a series of movements—many
of them millenarian movements—that put an emphasis on the importance of belief in a coherent and harmonious cosmos, created by a benevolent deity. In the
form this takes in Chinese late antiquity, the cosmos
was seen as being thrown into chaos by human interventions that had resulted in the ghosts becoming empowered and dangerous. Among these human interventions, indeed, were precisely the sacrificial practices
devoted to domesticating the natural and divine worlds.
For the Taiping Jing, the ghosts were inherently good,
but human sacrifices had thrown the balance of life and
death out of whack, thus leading to an endangerment
of the cosmos. The Celestial Masters took this further
and rejected sacrifice altogether for the transcendents.
In both cases, humans were being called upon to cease
sacrifices and undertake self-divinization techniques—
only now in order to help the cosmos. These millenarian movements were accordingly building upon selfdivinization techniques while holding a generally Mohist
cosmology.
In short, if the goal of sacrificial action was to domesticate the dangerous ghosts and capricious divinities
and transform them, to whatever extent possible, into
a pantheon of supportive ancestors and deities, the
457
critique of such action was to argue that this domestication is on the contrary empowering the ghosts and
throwing the cosmos out of balance. Within the world
being created through sacrifice, the flourishing of human life is at least in part a product of the domestication of the world. For the millenarian movements, it is
precisely the world created through sacrifice that is leading to an empowerment of the ghosts and an increasing
push toward death—the death of humans, and ultimately the death of the entire cosmos. In opposition to
the ritual and non-ritual interplay of the world of sacrifice, then, these millenarian movements called for a coherent world that we must have faith in, with revelations
that must be read directly, without any mediation and
without any human interpretation.
To explore the further implications of the point, it
may be helpful to return to Bloch. For Bloch, millenarian movements can be characterized as an attempt to
turn fully to the transcendental, with a rejection (or partial rejection, depending on the extremity of the movement) of the vital elements. In terms of the ritual complex of rebounding violence, millenarian movements
undertake the first part of the ritual move, in which
the vital elements are overtaken by the transcendental,
but not the second, in which the vital elements are then
reincorporated.
Bloch is certainly in part correct to say that the millenarian movements involve an attempt to create a transcendental order. But what such a framework misses is
the distinction between the work of ritual and the work
of belief. Millenarian movements are often what I would
like to call “sincerity movements.” Sincerity movements
are based upon calls for a belief in a counter-intuitive
set of transcendental claims—claims that the world is
coherent, created by a moral deity, and that such a belief
requires a rejection of the sacrificial and ritual traditions
that otherwise guide life. Indeed, in China, it is precisely
in the sincerity movements that a transcendental order
along the lines articulated by Bloch is posited as such.
And the point can almost assuredly be generalized. Millenarian movements are not a rejection of vitality but
a (deeply counter-intuitive) claim that the vital world
is part of a larger coherent order created by a benevolent
deity.
In terms of his overall model of rebounding violence, it is as if Bloch has taken the visions of transcendence more associated with sincerity movements and
then read those into a ritual world. But it is precisely
the interplay of ritual work and sincerity movements
LIFE, DOMESTICATED
AND UNDOMESTICATED
that is of interest in exploring the dynamics of life in
human communities. The former rely upon an endless
play of disjunctions and substitutions in the (ultimately
doomed) attempts to gain, appropriate, and control life.
The latter are based upon a (highly counter-intuitive)
faith in coherence and the literal in a call to see life as
(a lesser) part of a larger comprehensive system.
Conclusion
A pervasive danger in anthropology is to take a statement, such as “the world is harmonious,” or “ancestral
spirits are benevolent,” and ask what set of assumptions or worldview would make it possible for someone
to believe such a statement. But that may be precisely
the wrong question to ask. Whether such statements
should be taken literally, and treated as beliefs, is a very
complicated question. In many cases, such statements
are ritual statements, clearly counter-intuitive to the
world outside the ritual space. Or, when they are made
as statements that are supposed to be believed, and believed quite literally, we must recognize again that such
statements are highly counter-intuitive, and the work
of believing them is every bit as fraught and difficult as
the work of ritual world-making. This work, and the agony of this work, is a key issue that anthropology must
address.
One of the goals of this paper has been to take seriously indigenous ritual theories from China as well as
the theories from some of the movements that arose
against the sacrificial practices in question. Doing so
highlights that agony.
If we use these indigenous theories to think comparatively, we come up with some intriguing results. Let’s
begin with the disjunctions of ritual. The practices involved—here we have focused on sacrifices—in generating these disjunctions can be focused on building
continuity or creating discontinuity. But in either case
the sacrificial practices assume a disjunction between
the ritual world being constructed and the world outside—a world larger than, and governed by powers stronger than, human sacrificial work. The practices can therefore only be understood not as an ideal world that would
hopefully be created more generally but rather as a contingent attempt to empower humanity—whether that
be through gaining some degree of autonomy from the
divine (as, for example, in ancient Greek sacrificial practice) or through domesticating the divine (as in Chinese
Michael PUETT
sacrificial practice). The practices work through an endless play of substitutions and transformations, and they
assume multiple worlds; reading them as a coherent
worldview—taking ritual statements as indicative of a
larger set of cultural assumptions—by definition misses
the work of ritual. Again, the rituals only work because,
ultimately, they fail. The sources of life lie beyond the
rituals. A successful ritualization of the world would also
mean a loss of life.
In contrast, sincerity movements, based upon a rejection of ritual more generally and sacrifice more particularly, call on practitioners to believe in a coherent world,
usually created by and certainly governed by a benevolent deity. The structure of that world, and the precepts given by the deity governing that world, must
be taken literally, directly, and without any mediation.
Human acts of appropriation, of interpretation, destroy
this coherence, and only sincere faith can bring redemption. Such claims of coherence are every bit as counterintuitive as the ritual worlds created in sacrifice.
Taking these indigenous debates in China seriously
certainly helps to underscore the degree to which many
of our theoretical models may indeed be overly indebted to a generally Protestant vision. But we would
therefore also have to expand the category of “Protestant” to say “sincerity” claims more generally (Seligman
et al. 2008). Claims for the existence of a coherent world
that one must believe sincerely, for the importance of
following a truth (located in a transcendent god, within
the practitioner, or both) that one must follow literally,
for a degenerative/progressive vision of history culminating in a radical break—be that modernity or an
apocalypse—are recurrent in world history. Protestantism is simply a recent instantiation of a recurrent way of
acting in the world. But, ironically, the complexity of
that acting is what has been lost when we read statements along these lines as simply being the result of a
pre-given set of cultural assumptions or ontology. We
are in essence applying a Mohist/Taiping Jing/Xiang’er
hermeneutic of reading texts and the world literally,
without the play of substitutions, and yet we have ignored the incredible difficulty that such a literal reading
entails. These beliefs, in other words, operated much like
rituals, and were just as counter-intuitive (Severi 2015).
For all of our concern with questioning the Protestant assumptions underlying our theories, we may have missed
some of the most important ones.
Putting the distinction between ritual and sincerity
in terms of the agony of dealing with life, the questions
458
would be the following: Does life come from following the precepts of a divine order, or through the ultimately doomed ruses that appropriate that life for human growth? If the former, what is the work required
to cohere with these precepts of life? And, if the latter,
does that appropriation take the form of creating discontinuity from the continuous word of the divine, or
does it take the form of creating continuity against a
fundamentally discontinuous world?
The powers of life lie beyond humanity, and attempts
to appropriate, domesticate, or accede to these powers
are always limited and ultimately doomed. Whether
one is undertaking analyses through a sincerity or ritual
lens, these attempts—the work of culture—should become one of the concerns of our analysis. Ironically, our
concepts have been so successful at domesticating this
work that our analyses of Chinese practices often risk
replicating the lifeless world of the ritual space. We analyze a domesticated space as if it were a coherent world,
ignoring the ghosts underlying that space. Lost in such
analyses is the work of rituals and the work of belief, the
always doomed efforts to corral vitality, the wrestling
with ghosts.
References
Asad, Talal. 1993. Genealogies of religion: Discipline and reasons of power in Christianity and Islam. Baltimore, MD:
Johns Hopkins University Press.
Bloch, Maurice. 1977. “The past and the present in the present.” Man 12: 278–92.
———. 1989. “Symbol, song, dance and features of articulation: Is religion an extreme form of traditional authority?”
In Ritual, history and power: Selected papers in anthropology, 44–5. London: The Athlone Press.
———. 1992. Prey into hunter: The politics of religious experience. Cambridge: Cambridge University Press.
Bokenkamp, Stephen. 1997. Early Daoist scriptures. Berkeley:
University of California Press.
Brashier, K. E. 1996. “Han thanatology and the division of
‘souls.’” Early China 21: 125–58.
Campany, Robert Ford. 2009. Making transcendents: Ascetics
and social memory in early medieval China. Honolulu:
University of Hawaii Press.
da Col, Giovanni, and David Graeber. 2011. “The return of
ethnographic theory.” HAU: Journal of Ethnographic Theory 1 (1): vi–xxxv.
459
LIFE, DOMESTICATED
AND UNDOMESTICATED
Das, Veena. 1983. “Language of sacrifice.” Man 18 (3): 445–
62.
Plath, Sylvia. 1981. The collected poems, edited by Ted
Hughes. New York: Harper and Row.
Descola, Philippe. 2013. Beyond nature and culture. Chicago:
University of Chicago Press.
Puett, Michael. 2002. To become a god: Cosmology, sacrifice,
and self-divinization in early China. Cambridge, MA:
Harvard University Asia Center.
Detienne, Marcel. 1989. “Introduction.” In The cuisine of
sacrifice among the Greeks, edited by Marcel Detienne
and Jean-Pierre Vernant, 1–21. Chicago: University of Chicago Press.
Farquhar, Judith, and Qicheng Zhang. 2012. Ten thousand
things: Nurturing life in contemporary Beijing. Brooklyn,
NY: Zone Books.
Geertz, Clifford. 1973. “Person, time, and conduct in Bali.” In
The interpretation of cultures, 360–411. New York: Basic
Books.
Graeber, David. 1995. “Dancing with corpses reconsidered:
An interpretation of ‘famadihana’ (in Arivonimamo, Madagascar).” American Ethnologist 22 (2): 258–78.
Hendrischke, Barbara. 2007. The scripture on great peace:
The Taiping Jing and the beginnings of Daoism. Berkeley:
University of California Press.
Howe, Leo. 2000. “Risk, ritual and performance.” Journal of
the Royal Anthropological Institute 6 (1): 63–79.
Hsu, Francis L. K. 1967. Under the ancestors’ shadow: Kinship, personality, and social mobility in China. Stanford,
CA: Stanford University Press.
Hubert, Henri, and Marcel Mauss. 1981. Sacrifice: Its nature
and functions. Translated by W. D. Halls. Chicago: University of Chicago Press.
———. 2004. “Forming spirits for the way: The cosmology
of the Xiang’er commentary to the Laozi.” Journal of Chinese Religions 32: 1–27.
———. 2005. “The offering of food and the creation of order: The practice of sacrifice in early China.” In Of tripod and palate: Food, politics, and religion in traditional
China, edited by Roel Sterckx, 75–95. New York: Palgrave
Macmillan.
———. 2008. “Human and divine kingship in early China:
Comparative reflections.” In Religion and power: Divine
kingship in the ancient world and beyond, edited by Nicole
Brisch, 199–212. Chicago: Oriental Institute of the University of Chicago.
———. 2009. “Sages, gods, and history: Commentarial strategies in Chinese late antiquity.” Antiquorum Philosophia
3: 71–87.
———. 2010a. “Ritualization as domestication: Ritual theory from classical China.” In Ritual dynamics and the science of ritual, Volume I: Grammars and morphologies of
ritual practices in Asia, edited by Axel Michaels, Anand
Mishra, Lucia Dolce, Gil Raz, and Katja Triplett, 365–
76. Wiesbaden: Harrassowitz Verlag.
Keightley, David. 1978. “The religious commitment: Shang
theology and the genesis of Chinese political culture.” History of Religions 17: 211–25.
———. 2010b. “The haunted world of humanity: Ritual theory from early China.” In Rethinking the human, edited
by J. Michelle Molina and Donald K. Swearer, with Susan
Lloyd McGarry, 95–111. Cambridge, MA: Center for the
Study of World Religions.
Kleeman, Terry F. 2016. Celestial Masters: History and ritual
in early Daoist communities. Honolulu: Harvard University Asia Center.
———. 2010c. “Becoming Laozi: Cultivating and visualizing
spirits in Early Medieval China.” Asia Major, Third series,
23 (1): 223–52.
Lévi-Strauss, Claude. 1966. The savage mind. Chicago: University of Chicago Press.
———. 2013a. “Economies of ghosts, gods, and goods: The
history and anthropology of Chinese temple networks.”
In Radical egalitarianism: Local realities, global relations,
edited by Michael M. J. Fischer, Felicity Aulino, Miriam
Goheen, and Stanley J. Tambiah, 91–100. New York: Fordham University Press.
Liu Xin. 2000. In one’s own shadow: An ethnographic account
of the condition of post-reform rural China. Berkeley: University of California Press.
———. 2002. The otherness of self: A genealogy of the self in
contemporary China. Ann Arbor: University of Michigan
Press.
Luhrmann, Tanya. 2012. When God talks back: Understanding the American evangelical relationship with God. New
York: Random House.
———. 2013b. “Critical approaches to religion in China.”
Critical Research on Religion 1: 95–101.
———. 2014. “Ritual disjunctions: Ghosts, philosophy, and
anthropology.” In The ground between: Anthropologists
engage philosophy, edited by Veena Das, Michael Jackson,
Michael PUETT
Arthur Kleinman, and Bhrigupati Singh, 218–33. Durham, NC: Duke University Press.
———. 2015. “Ghosts, gods, and the coming apocalypse:
Empire and religion in early China and Ancient Rome.”
In State power in Ancient China and Rome, edited by Walter Scheidel, 230–59. Oxford: Oxford University Press.
———. 2016. “Manifesting sagely knowledge: Commentarial strategies in Chinese late antiquity.” In The rhetoric of
hiddenness in traditional Chinese culture, edited by Paula
M. Varsano. Albany, NY: SUNY Press.
Sahlins, Marshall. 1983. “Raw women, cooked men, and
other ‘great things’ of the Fiji Islands.” In The ethnography of cannibalism, edited by Paula Brown and Donald
Tuzin, 72–93. Special Publication of the Society for Psychological Anthropology, Washington, DC.
———. 1985. Islands of history. Chicago: University of Chicago Press.
460
Severi, Carlo. 2015. The chimera principle: An anthropology
of memory and imagination. Translated by Janet Lloyd.
Chicago: HAU Books.
Shaughnessy, Edward L. 1985–87. “Extra-lineage cult in the
Shang Dynasty: A surrejoinder.” Early China 11–12: 182–
94.
Smith, Jonathan Z. 1987. “The domestication of sacrifice.” In
Violent origins: Walter Burkert, René Girard, and Jonathan Z. Smith on ritual killing and cultural formation, edited by Robert G. Hamerton-Kelly, 191–235. Stanford,
CA: Stanford University Press.
Valeri, Valerio. 1989. “Death in heaven: Myths and rites of
kinship in Tongan kingship.” History and Anthropology
4: 209–47.
———. 1994. “Wild victims: Hunting as sacrifice and sacrifice
as hunting in Huaulu.” History of Religions 34 (2): 101–31.
Scott, Michael W. 2007. Severed snake: Matrilineages, making place, and a Melanesian Christianity in the Southeast
Solomons. Durham, NC: Carolina Academic Press.
Vernant, Jean-Pierre. 1989. “At man’s table: Hesiod’s foundation myth of sacrifice.” In The cuisine of sacrifice among
the Greeks, edited by Marcel Detienne and Jean-Pierre
Vernant, 21–86. Chicago: University of Chicago Press.
Seidel, Anna. 1987. “Traces of Han religion in funeral texts
found in tombs.” In Dôkyô to shûkyô bunka, edited by
Akitsuki Kan’ei, pp. 21–57. Tokyo: Hirakawa shuppansha.
Wang Ming 王明. 1992. Taiping jing hejiao 太平經合校 (Collated edition of the Taiping jing). Beijing: Zhonghua shuju.
Seligman, Adam, Robert Weller, Michael Puett, and Bennett
Simon. 2008. Ritual and its consequences: An essay on the
limits of sincerity. Oxford: Oxford University Press.
Wolf, Arthur P. 1974. “Gods, ghosts, and ancestors.” In Religion and ritual in Chinese society, edited by Arthur P.
Wolf, 131–82. Stanford, CA: Stanford University Press.
Michael PUETT is the Walter C. Klein Professor of Chinese Anthropology and History at Harvard University. He is
the author of The ambivalence of creation: Debates concerning innovation and artifice in early China and To become
a god: Cosmology, sacrifice, and self-divinization in early China, as well as the co-author, with Adam Seligman, Robert Weller, and Bennett Simon, of Ritual and its consequences: An essay on the limits of sincerity.
Michael Puett
Department of Anthropology
Harvard University
21 Divinity Avenue
Cambridge, MA 02138
USA
puett@fas.harvard.edu