A Poetics and Politics of Possession: Taiwanese Spirit-Medium Cults and Autonomous Popular Cultural Space Peter Nickerson

My intent in this essay is to discuss cults of spirit possession and healing in Taiwan, to consider the potential for this form of popular culture to create autonomous spaces, and to evaluate the larger social and political consequences of such autonomy. However, before turning to discussion of Taiwan, it would be useful to examine briey another recent study of possession cults in a perhaps not entirely dissimilar environment, namely, Michael Taussigs barely ctionalized Venezuela. (Taiwan is also a former colony with a well-developed modern economic sector.) In his Magic of the State, Taussig describes a kind of symbiosis between the state and popular religion, in which the same force that animates the bodies of the possessed is the history, and the historical gures and stereotypes, on which the state relies for its legitimation: a sixteenth century cacique, an early nineteenth century barefoot black cowboy freedom ghter, . . . the Liberator [Simon Bolvar] himself, coughing blood. And in thus physically manifesting, making
positions 9:1 2001 by Duke University Press

positions 9:1

Spring 2001

188

palpably real, the normally merely metaphorical gurations on which the nation-state depends for its own legitimating appropriation of history, popular religionthe bodies of the urban poorsuppl[ies] stately discourse with its concrete referents. At the same time, participants in the possession cults themselves become in some way empoweredable to make magic from the magic of the state.1 All of this might be applied to contemporary Taiwanese popular religion, except the magic involved is that of a state that went out of existence in 1911. Most discussions blithely skip over this issue, referring to the bureaucracy of the other world as an obviously more or less accurate reection of thisworldly administration, all the while failing to note that the Jade Emperors earthly counterpart can no longer be found.2 So that, perhaps, is one way of stating the question this essay seeks to address: What possibilities open up when the state chooses (as I shall discuss in more detail below) to rely on other, secular sources for its legitimacy? Taussigs analysis, which I by no means would wish to dispute, ts rather nicely with many currently accepted notions of what constitutes popular culture. As is well known, most debates on this topic have oscillated between two poles. On one end of the spectrum is the gloomy view of the Frankfurt school, according to which the public at large is subject to the hegemony of a culture industry. Ordinary people, while presented with the illusion of choice, consume culture the same way they consume any other industrial product, and thus are placed under the same restraints to which their bourgeois overlords subject them in the more purely economic spheres of labor and the consumption of material goods.3 On the other end of this spectrum, those who have sought elements of resistance within popular culture have done so by claiming that the dominated might make usein a subversive mannerof what is given to them by the dominant classes. Such resistant uses themselves then become a praxis that is not only consumption but also the production of new meanings or readings. One might mention Michel de Certeaus metaphor of the rented apartment, not owned but decorated to the renters taste.4 But, ultimately, only limited gains can be made. Consider John Fiskes comparison of popular uses of the products of the dominated to guerrilla warfare. The guerrillas may resist, and indeed may win concessions from the established authorities, but ultimately, like so many Viet

Nickerson A Poetics and Politics of Possession

189

Cong, they need to melt into the general village population again, once more indistinguishable from the rest of the compliant masses.5 Taussigs stance is quite consistent with this latter view: the people use what is given to them by the statethe history with which the State of the Whole constructs itselfwhile the state appropriates the magic through which those very symbols are made real. What appears to be a most outr form of religio-magical practice turns out to be a process of mutual appropriation between dominated and dominators, making the nation safe for state fetishism. In any event, the possibility of a genuinely autonomous culture, made not for but by the people, seems scarcely possible. This point has been clearly and authoritatively made by Pierre Bourdieu: Those who believe in the existence of a popular culture . . . must expect to nd . . . only the scattered remnants of an old erudite culture (such as folk medicine) . . . and not the counterculture they call for, a culture truly raised in opposition to the dominant culture and consciously claimed as a symbol of status or a declaration of separate existence.6 So now to Taiwans spirit-medium cults and an attemptwhich, in light of the above discussion must appear radical (and/or romanticist)to stake out a rather more extreme position. Some months of eld research on popular religion in the city of Tainan in southwestern Taiwan have convinced me that the question of popular cultural autonomy needs to be rethought. In some cases at least, popular culture can be theorized, not between the poles of hegemony and subversion, but outside them. We need to go back to Bakhtin, as it were, and to reconsider the possibilities for certain types of collective practices to create autonomous, popular cultural spaces, spaces that are more than simply poached upon by the people, but that genuinely belong to them. That is, neither the Frankfurt schools notion of popular culture as imposed from above and passively accepted by the dominated, nor the populist contention that such acceptance is far from passive and involves creative processes of rereading and subversion, allows for the possibility that the dominated might make something of their own. Owing to their roots in a particular discourse of possession trance, such autonomy (a reasonable word and, I think, at least more euphonious than auto-production or the like) is intrinsic to the popular religious groups I have studied.

positions 9:1

Spring 2001

190

Autonomy of this sort has its consequences. Precisely because the truths generated by the activities of the medium cults are empowered by cultural resources that lie outside the mainstreampossession trances that embody the magic of a no longer existent statethey cannot directly challenge those produced by the dominant culture. When the devotees of the spirit-altars confront mainstream discourses on religion, they can do so only by referring to a set of notions that is in many ways alien to the poetics of possession itself; these notions are instead largely consistent with the very conceptions of religion that marginalize the possession cults. Religion is unscientic; proper religion can be opposed, according to several criteria, to improper superstition. These ideas are broadly shared in Taiwanese society. As I shall try to show, by accepting these notions and employing them in their own explanations of and justications for their activities, the cults adherents, rather than subverting the dominant order, assist in perpetuating it. Paradoxically, whatever potential for resistance popular culture might possess according to de Certeau and others, in the Taiwanese case cultural autonomy blunts such potential rather than enhancing it. After rst describing the procedures generally followed in spirit-medium sances, I will turn to the poetics of spirit possession, the characteristics of trance behavior that make autonomy possible. Further sections, based largely on work at one temple and a case study of one womans experiences there, will then provide a basis for a discussion of the larger issue of popular religious discourses and the more authoritative ones with which they interact. The generation of authoritative discourses in response to popular religion is particularly visible in print media accounts written in the wake of religious scandals. I will examine one such case, the Song Qili affair of 1996. Finally, I will assess the discursive and social consequences ofand the limitations onthe kind of popular cultural autonomy I have found to characterize the medium cults.
Spirit-Medium Sances in Tainan

The following describes events that took place on the evening of 13 July 1997:7

Nickerson A Poetics and Politics of Possession

191

The spirit-medium, a slight man in his late fties with a somewhat drawn face, walks to the altar and sits down silently, hands clasped and head slightly bowed. Soon (in about ten minutes) he will be the vehicle for the Living Buddha Jigong, a drunkard, vagabond monk who is said to have lived during the thirteenth century, and who later became the subject of both vernacular literature and religious worship (a not uncommon pattern). Devotees mill around, some chatting quietly with one another, others registering for consultations with the deity and making the appropriate preliminary offerings of incense before the temples images. The temple, or spirit-altar (shentan), is called San Qing Gong (Palace of the Three Pure Ones) and is located in Tainans eastern district. Public altars (gong tan), when the altar is opened (kai tan) for consultations with the god, are held here three times during every ten-day period. Anyone may come in and consult Jigong, and the sances are normally attended by around twenty or thirty people. An assistant (and longtime devotee) approaches the altar and repeatedly strikes a large wooden sh and a metal bell, signaling the incipient arrival of the god and the formal beginning of the sance. The mediums hands and head begin to tremble, and he belches intermittently. He gestures for the assistant to help him put on Jigongs patched robe. The medium himself places a rosary around his neck and, on his head, Jigongs usual hat, which is tall and rather comical, with the character fo (Buddha) written on it. He takes up his fan in his right hand and a container of rice wine in his left. His increasingly vigorous, rhythmic trembling and shaking motions cause him to tap more and more loudly on the altar with the fans handle. Then, after a smile and an exclamation of amusementha ha ha ha ha ha ha hahe begins singing, largely wordless tunes that, he had once explained, were from the opera of Nanjing (near the original Jigongs home region). The singing continues (oscillating oddly owing to the continued rhythmic trembling), interspersed with various ha-ha-has and heh-heh-hehs (in the same rhythm) as well as occasional swigs on the wine, until he is fully in possession trance. Jigong gets up from the altar, still singing, and staggers about for a bit, making dramatic gestures with his fan. The assistant gives him a long draught of smoke from an incense burner. Then Jigong processes to a chair, which has many nails protruding, points up, from seat and back, and sits for

positions 9:1

Spring 2001

192

a minute or two while still singing, fanning himself, and bouncing up and down slightly. He gets up (the nails make faint musical noises as, one by one, they gradually release his clothing), walks to the altar to set down his wine, and walks back to the center of the oor to speak with his rst questioners. A man and his wife want to ask about the husbands career and the health of their middle daughter; they both greet him with hands pressed together in a gesture of reverence. On any night of the week, all over Taiwan, scenes like the one just described take place. People go to shrines and temples to consult a variety of deities of the popular Chinese pantheon. The gods and goddesses are represented by their spirit-mediums (jitong, or tang-ki in Taiwanese), who enter possession trances and then answer questions on a variety of topics (both female and male tang-ki are common). Some smaller spirit-altars convene only when one or more people need to consult the deity, while the relatively larger ones (which may operate out of anything from a storefront shrine to a multistory public temple) usually have regular schedules that allow for public consultations once or twice a week. While I have attempted no overall numerical assessment of Tainans medium shrines, a recent survey of the cult to a single deity in Taiwan, the Reverent Lord of Broad Compassion (Guang ze zun wang), lists nine spirit-altars dedicated to that god in Tainan city alone, and another six in Tainan county. The majority hold at least weekly consultations with the Reverent Lord,8 and the Reverent Lord is of course only one of the many deities embodied by mediums in Tainan. These include popular deities associated with all three of Chinas textualized religious traditionsTaoism, Buddhism, and Confucianismsuch as Lord Guan, Guandi, the deied hero Guan Yu of the third-century Three Kingdoms era; Mazu, or Tian Hou, the Empress of Heaven; Guan Yin, Avalokiteshvara, the Bodhisattva Hearer of the Worlds Sounds; the various plague gods known collectively as Wangye or Lords; and even Bodhidharma, the rst Chinese patriarch of Chan/Zen Buddhism, to name just a few. The large number of active spiritaltars staffed by mediums indicates their continued centrality to Taiwanese popular religion and popular culture generally. As for the usual procedures at the spirit-altars, sances begin with the medium entering the trance-possession state. An assistant then reads the

Nickerson A Poetics and Politics of Possession

193

name of the questioner, usually giving other personal details, such as address and age. Then the questioner may present his or her inquiry. The majority of inquiries concern health, and people often come to seek the help of the goddesses and gods when standard medical treatment has failed to bring a cure or the usual diagnostic techniques fail to nd anything wrong. The other main category of questions concerns matters of luck and fate. People ask about the future in general, about business activities, or about transitional events such as marriage or moving a household and their timing. If reasons are attributed for medical or other misfortunes, they are usually of several often overlapping types: horoscopic (bad astrological conjunctions have caused an individual to collide with one or more malign astral deities), familial (a deceased family member is unhappy and causing the misfortune so that attention will be paid to his or her problem), or geomantic (i.e., involving bad fengshui, the arrangement of domestic or other spaces). The gods responses almost always include the writing of talismans, or fu, mysterious, often entirely illegible characters written with brush and ink. These represent a concretization of the healing power of the god, and the paper talismans most frequently are burnt and the ashes mixed with water and ingested.
A Poetics of Possession

An examination of the discourses generated by the medium cults must begin with the speech of the medium. Possessed speech is often unremarkable in terms of what it merely supercially says; what distinguishes it is, of course, that it emanates from a person whose body has been occupied by a deity. That process of enthusiasm, of engodment, is acted out publicly at the beginning of every sance. Mediums never arrive at the temples altar in a possessed state. They sit before or beside the altar and undergo, for all present to see, the process of possession. The witnessing of the mediums entrancement is an indispensable part of trance discourse. The procedure through which mediums become possessed, though inarticulate and supercially incoherent, is fraught with meaning. This meaning is best approached by means one might loosely call semiotic. Since, as I will claim, the process of entrancement, rather than the content of the utterances

positions 9:1

Spring 2001

194

of the possessed oracle, is most essential to trance discourse, it is insufcient merely to analyze the pronouncements of the deities once they have completely entered the bodies and begun speaking through their mediums. Rhythm is dominant in the rst stages of most Taiwanese trance performances, where it is almost always as much or more bodily than vocal. While mediums all enter trance differently, the majority during at least one stage tremble, twitch, shake, or rock, using hands, arms, legs, head, neck, or the entire body. The movements become more violent as the trance deepens, then they subside to some degree when identication with the deity is complete. All the bodily motions seem to be regulated by a master rhythm, usually the twitch of hand or arm, with which other motions or vocalizations are then harmonized. Rhythmic trembling and shaking is more often than not the principal marker of entry into trance, climaxed by larger bodily motions and postures and punctuated with (usually unintelligible) speech and other sounds. Glossolaliaspeaking in tongues, as it wereis also highly relevant. The vocal performances of one female medium in Tainan county were especially striking. (She is customarily possessed by a minor female deity called Immortal Maid Shi [Shi Xiangu] but sometimes also serves as the vehicle for the Bodhisattva Guan Yin.) As she nears the possession state, she alternates very loud, low-pitched, and rather masculine exclamations, somewhere between belches and yellsHWAAAAHP! or YAIYAIYAH!with rapid, higher-pitched noises, such as huaa-soho or yai-saha. Into this mix are then added gradually longer and longer, entirely unintelligible (both to myself and to other bystanders) declarations in a very high pitched, rapid, lilting voice. Her vocal range is truly striking (those who have listened to audiotapes have sometimes assumed that more than one person was speaking), and her unearthly utterances are haunting, even disturbing, despiteor perhaps in part because oftheir incomprehensibility. The necessity of entrancement behavior to the semiotic effectiveness of the sance is further conrmed by the mediums frequent reference back to that initial procedure. After settling into their trances to conduct divination or other rites, most Taiwanese mediums continually cite the entrancement process by means of small but continuous tremblings or occasional more noticeable shakes and twitches, belches or yawns, outbursts of gods

Nickerson A Poetics and Politics of Possession

195

language, or the like. It is as if the audience has to be periodically reminded of the source of the mediums utterances. There are certain convergences between Taiwanese trance performances and other forms of behavior that have been more thoroughly analyzed by European writers that help to shed some light on the former. Trance discourse closely resembles other texts and performances that have been variously labeled poetic, disruptive, spectacular, and (therefore) popular. Poetry and possession are not as distant as one might think; one need only consider the ancient Greek conception of poetic composition as possession by the Muse.9 Poetic and ecstatic forms of expression share similarities of both form and content. Poetry highlights aspects of words that are unconnected to denoted meaning but central to the sound of poetry, such as rhyme, alliteration, and above all, rhythm. Julia Kristeva has similarly highlighted other characteristics of poetic language that also are central to trance discourse, in particular glossolalia and related vocalizations, which she nds audible in both the rst echolalias of infants and the rhythms, intonations, [and] glossolalias in psychotic discourse. Indeed, poetic speech always borders on psychosis. Poetic language is thus the expressive aspect of the physicality of speech and is rooted in the body and in unconscious drives.10 I would not like to place too much weight on the links Kristeva makes between poetic and either infantile or psychotic speech. In Taiwanese society itself, communities and their religious leaders often require some time to decide whether an individuals trance behavior is evidence of a divine calling, demonic possession, or simply mental illness, as Marjorie Wolf has detailed.11 Part of what I have just called glossolalias (the rapid, high-pitched speech) were explained to me by the spirit-altars adherents as gods language (shen yu), intelligible to the divinities present, but not to humans. Whether such speech is culturalized as divine or psychotic is not the central issue. Instead, beyond the formal resemblances, I would like to point out the similarities of social effect that can be attributed to poetry and possession. As for poetic language, it is by nature subversive. According to Kristeva, Poetic language, in its most disruptive form (unreadable for meaning, dangerous for the subject) shows the constraints of a civilization dominated by transcendental rationality. Consequently it is a means of overriding this constraint. Thus there is an intrinsic connection between literature and the breaking

positions 9:1

Spring 2001

196

up of social concord, and between poetic language and the disruptive forces of Bakhtinian carnival.12 It is, of course, that same potential for resistance, for creating truths in opposition to those of the dominant culturefor making space for popular culture that does not ultimately have to be given back, guerrilla-like, to the power blocthat I would like to attribute to the spirit-altars of the mediums. Like Kristevas poetry, the mediums performance begins as unreadable for meaning, as well as dangerous for the subject in its continual suggestion of the threat of loss of mental stability (or of the mediums very personality, or soul). But the sance then continues as oracular pronouncement, as advice and instruction made clearly intelligible either by the medium alone or with the assistance of an interpreter. The poet, too, must compromise, of course, by enunciat[ing] rhythm, . . . socializ[ing] it, . . . channel[ing] it into linguistic structure, in order to create the writers universe.13 The sance establishes its universe through a true cosmogony: the creation of the ordered oracle from the chaos of entrancement. Also notable are the rather odd convergences between the Freudian and Lacanian Kristeva and the authors of Anti-Oedipus, Gilles Deleuze and Flix Guattari. The Taiwanese spirit-medium is rather like Deleuze and Guattaris schizo. The medium/schizo retreats into the body without organs, which in order to resist using words composed of articulated phonetic units . . . utters only gasps and cries that are sheer unarticulated blocks of sound (compare Kristevas glossolalias). The energy that passes through that organless body is divine, and the rst things to be distributed on [the schizos] body without organs are races, cultures, and their gods. For Deleuze and Guattari, and Kristeva as well, in psychosis/possession the unconsciousthe undifferentiated otherbreaks out to break down the established order. Finally, the schizophrenic, or the spirit-medium, I would claim, is the universal producer.14 Who produces the mediums trance? Not the culture industry. In this instance the means of production, as it were, belong to the medium cults themselves, establishing territory that can be held, not yielded to the dominant classes. The cultural space created at the spirit-altar via the mediums trance provides more than the metaphorical rented apartment; instead, it is a place of their own.

Nickerson A Poetics and Politics of Possession

197

Meiling

People in Tainan go to spirit-altars more often to make inquiries about illness and healing than for any other reason. The tensions between the ethos of the medium cults and modernist, scientistic notions is nowhere clearer than in the arena of medicine. A young womans case history will provide a basis for an account of how this conict is frequently understood and played out. In April 1997 a twenty-seven-year-old woman resident of Tainan city, whom I will call Chen Meiling, was due to have a minor eye operation. Prior to the operation, however, she had a ght with her husband, whom I will call Mr. Li. The husband took both his personal effects and their two children back to his parents home in Tainan county. Meiling later telephoned him, but he refused to speak to her. She then took eight tranquilizers, drank some alcohol, and telephoned her elder sister. (All of this information was reported by the sister.) Worried, the sister went to Meilings home in Tainan city. Meiling failed to answer the door, and the sister called the police and had them break in. Meiling was taken to a hospital emergency room, where the doctor claimed that she had no serious problem: Mei shenme. However, the sister felt that Meilings manner had changed, that she was in some way not her former self, and so brought her to consult Lord Guan (Guandijun) at the Prefecture of the Southern Heaven (Nantianfu), a large public temple in the eastern district of Tainan where I did my most intensive work on the medium sances. Meiling would attend sances about once a week from April until early July, when her treatment reached a climax. She began by participating in the regular Wednesday night sessions, for both diagnosis and treatment of her problem, which was described to me simply as a mental problem or nervous disorder ( jingshen wenti or shenjing wenti). Meiling frequently received treatment in the form of talismans written by the possessed medium. These were used for general purposes and at other times for more specic symptoms, such as nightmares. On other occasions Meiling would stand before the gods palanquin, or divination chair, while it shook up and down several times, then walk beneath it, a common apotropaic, healing and purifying procedure.15

positions 9:1

Spring 2001

198

Most of Lord Guans energies, however, were devoted to diagnosis specically, identifying the spirits responsible for Meilings maladyand to serving as an intermediary in order to reach an accommodation with those forces. Many of Meilings symptoms were acted out, and interpreted by the god and by Meilings family and the rest of the sance community, as spiritpossession. Both at home and at the temple she especially frequently spoke as a three-year-old child. Over the course of successive temple sances as well as two house callssances at Meilings home in the city and at the home of her husbands family in Tainan countyLord Guan developed an elaborate diagnosis involving the histories of both the agnates and the afnes. On the husbands side the problems were several. Meilings father-in-laws uncle (shu gong) had disrupted the usual pattern of agnatic succession (dao fang). By dying unmarried he lacked heirs who could make offerings to him as an ancestor. During a sance at the Li family house, Lord Guan declared that this initial problem, combined with certain geomantic issuesthe misplacement of an electric pole and the presence of a brick wall surrounding the familys front yardhad invited the entrance into the family compound of a host of Killers of the Yin-realm ( yin sha), or wandering, non-kin, malec spirits. In turn, this engendered the possession by one of these spirits, a threeyear-old child. During a sance at Meilings parents apartment, responses to Lord Guans inquiries determined that her mother had carried a female fetus that had died through miscarriage. As is usual, the girl continued to mature as a ghost in the other world. All of these spiritsthe heirless Li agnate, the outsider ghosts (especially the three-year-old), and the girl ghost in Meilings familywere unhappy, no doubt because of insufcient ritual attention from the living and consequent sufferings in the afterworld. Thus they had fastened on Meiling, causing her maladies in order to make people pay attention to their plights. The ultimate solution to these problems took the form not of violently exorcising the problematic ghosts but of helping and propitiating them. After a one-month cooling-off period prescribed by Lord Guan, during which Meiling continued to come on Wednesdays for checkups but little else was done, a Taoist priest was engaged. He performed a salvation ritual called an Attack on the Fortress (da cheng) in Tainans Palace of the Eastern Peak (Dong yue dian), a temple with a special relationship to the underworld

Nickerson A Poetics and Politics of Possession

199

and the administration of the shades. The ritual was held for the benet of the heirless uncle and the three-year-old, on the husbands side, and the girl ghost in Meilings family. The priest liberated the troubled souls from the Fortress of Those Who Have Died Unjustly (wang si cheng) in purgatory and conducted them to the Western Paradise. Later on the same day, the medium, the divination chair crew, and other temple personnel returned to the husbands home in the countryside, where an altar and a long table with numerous food and wine offerings were set up and a massive pile of paper spirit money was burned to requite and send off the nonfamilial wandering ghosts. There are many observations that could be made about this case, and the psycho-physiological factors that both provide means for the aficted openly to express repressed thoughts and feelings (especially for women)16 and that often make the healing efforts of shamans and mediums effective17 have been discussed elsewhere. It is not difcult to see how the treatment provided by Nantianfu worked in analogous ways. Meiling got a way to articulate her needs and desires, through identication with possessing spirits whose problems were like her own (abandonment, outsiderhood), and she got her family circle to acknowledge those needs and take specic measures to satisfy them. Instead, I would like to focus on a different, and for my purposes more pertinent, question: What does cooperation in the treatment of Meiling and others like her mean for the larger temple community? Participation in medium sances not only can help assist the directly aficted; it also can support the creation, if not of on-the-ground autonomous communities, then of autonomous discoursesforms of knowledge that might oppose those of the dominant, secular, modernist culture and are available to all participants.
A Sociology of the Sance

Not only the poetics but also the sociology of the medium sance ensures the production of intelligible, and even didactic, discourse. The medium is joined by the spirit-altars core of regular participants, its inner circle, in the labor to create comprehensible statements that will solve the questioners

positions 9:1

Spring 2001

200

problems (or provide moral and spiritual guidance for the group at large). At many spirit-altars the work of the interpreter, or table-head (zhuotou), is essential, as the speech of the medium can be difcult or impossible for inexperienced bystanders to understand, or simply too laconic to apply concretely without the interpreters elaborations. Similarly, the Chinese characters traced on the table by the divination chair are entirely indecipherable for those who have not been specially trained and do not have unusual talent. The process of divination is often assisted by a rather large crew. In addition to the table-head, there is normally a separate individual designated as secretary, who records questioners names, personal particulars, inquiries, and the gods dicta. The most elaborate setup with which I am familiar was employed at Nantianfu. Frequently Nantianfu used two interpreters and always two secretaries, one who kept the aforementioned records and one who assisted with the paperwork connected with the creation of Lord Guans talismans. The use at each sance of the four-man divination chair, in addition to (or sometimes in substitution for) the medium, demanded the presence of numerous members of the palanquin crew, since it was necessary to alternate duty to prevent the labor from becoming too tiring for any one man (and to give more men the chance to serve and have contact with Guandi). Others in the inner circle, especially women, who could not carry the palanquin, helped to create the celestial prescriptions (tian fang) of incense ash and paper spirit money that were a specialty of the temple; still others rendered talismans into drinkable potions of ash and water. Additional members of the core group often simply clustered around the table at which the medium was stationed and were free to offer opinions and advice. (Outsiders or ordinary questioners were seated on the other side of the room and were often shooed away if a crowd of them inadvertently gathered.) Thus, any given sance at Nantianfu normally involved altogether a group of fteen to twenty or more individuals who assisted with the session. The speech of the medium begins from stationary silence, commences activity in rhythmic shaking, and climaxes in often cryptic oracular speech (and, one could add, in illegible writing in the form of talismans). The socialization of possessed speechits rendering into usable formis accomplished by the cooperation of the collective of temple regulars: divination by committee, as it were. Moreover, at public temples like Nantianfu, the management and

Nickerson A Poetics and Politics of Possession

201

operation of the temple is in the hands of the faithful themselves, who elect a management committee, its chairman, and other ofcers. At Nantianfu, at least, the crew that regularly assisted with sances was largely identical to the core management organization. Bourdieu notes the transformation that takes place when formerly suppressed individual experience is reected and thereby legitimated in the public discourse of the heretical group: Private experiences undergo nothing less than a change of state when they recognize themselves in the public objectivity of an already constituted discourse, the objective sign of recognition of their right to be spoken and to be spoken publicly. Objective crisis, of which the mediums trance is certainly a variety, is what brings the undiscussed into discussion, the unformulated into formulation, [and] in breaking the immediate t between the subjective structures and the objective structures, destroys self-evidence practically.18 Possession trance opens up a space for the generation of a new discursive eld that involves all three of the parties to the sance: the medium, the patient, and the spirit-altars community.
What Science Cant Explain

The most obvious and immediate use of the discourses generated by the medium sances concerns medicine. To be treated by a deity through his or her medium is to remain independent of the mainstream, Western-style, and state-supported medical establishment and the cultural nexus of power that legitimates and is legitimated by it. This is emphatically not to say, however, that those who seek out spiritual remedies to medical problems have entirely rejected Western-style medicine or the various scientic and other discourses with which it is connected. Nowadays, and judging from other studies,19 for at least a few decades, those who go to the gods for diagnosis and treatment normally have tried and failed with conventional therapies or are receiving both forms of treatment simultaneously. In fact, both for clarity of exposition and, perhaps, for the sake of drama (or antidrama), I have refrained from mentioning that while Meiling was regularly attending sances at Nantianfu, she was also regularly visiting

positions 9:1

Spring 2001

202

a psychiatrist; one diagnosis had been depression ( youyu zheng). Interestingly, Lord Guans talking cure seems more to resemble traditional psychoanalysis than what Meiling was getting from Tainans municipal hospital. Her sister described the treatment as eating medicine and getting injections (chi yao da zhen). Her doctor had also recommended electrotherapy and suggested that she be admitted to a hospital as an inpatient. I would speculate that Lord Guans ministrations at least provided a counterweight that enabled her family to resist long-term hospitalization. Meilings sister told me that she felt it was unwise to rely solely on Lord Guans help, that it was better to employ both religious and modern medical forms of healing. This is typical of a variety of formulations people use to explain how dealing with problems by consulting a medium or other religious practitioner might come into play after or along with the pursuit of secular methods. For example, it is quite common in Tainan for couples having difculty conceiving a child to go to a spirit-medium or diviner for a diagnosis of their problem. That specialist will then often make a further referral, sending the couple to a ritual-master (fashi) who can perform a ritual to correct the situation in the temple of the Lady by the Riverside (Lin shui furen miao), who deals particularly with the problems of women and children, as well as those connected with conception. This rite is only carried out, however (and several ritual-masters have conrmed this), if the couple has consulted a medical doctor to conrm that there are no organic disorders, of either partner, preventing conception. Only when the doctor is unable to detect anything is the problem then ritually treatable as one connected with the womans celestial ower bed, where a plant bearing white and red owers represents her fertility and the male and female children she will bear. This reects an essential distinction in the etiology of disease and other maladies, the difference between that which has form ( you xing) and the formless (wu xing). Doctors deal with the former, and when causes in the visible world cannot be identied or when treatment that assumes visible causes is not successful, only then does the search for origins in the invisible world begin. If, as in Meilings case, people are uncertain as to the nature of the cause, or if maladies are deemed to have both visible and

Nickerson A Poetics and Politics of Possession

203

invisible sources, then both types of therapies may be employed simultaneously. Sance participants and temple regulars often referred to the paranormal in their justications of their activities. There are things that science simply cant explain was a constantly recurring theme in such discussions. What science could not explain was the formless, the realm of gods, ghosts, and other supernatural forces, those things in which Chinese ritual practitioners have been specialists for millennia. Noting the limitations of science was inevitably a preface to one or more anecdotes of miraculous events and puzzling evidence. The stories of the unusual supernatural experiences of the individuals themselves, coreligionists, friends, and so forth were sometimes joined with accounts such as that of the successful predictions of Ronald Reagans astrologer, culled presumably from popular periodicals, or from the burgeoning genre of television programs devoted to such topics. It seems likely that this fascination with the paranormal, as well as its expression in the lowbrow popular media, represents a new transnational phenomenon of global religion and popular culture. Large-format magazines with names that include not only conventional evocations of the supernatural, like SpiritMarvels [Lingyi], but also imported notions, like Sixth Sense [Di-liu gan], recall less the traditional Chinese literature of the strange (e.g., zhi guai and chuan qi) than they do the National Enquirer, and they include both locally produced materials and reports clearly culled from their foreign counterparts.
A Politics of Popular Religion

I will return to the issue of science and popular religion below, but rst I would like to address the role of the Taiwanese state in the production of discourses concerning religion. However, my concern is not with the direct confrontation of the state with popular religion/civil society, but with the state as one actor among others, in particular by means of its judiciary. After the Nationalists retook Taiwan from the Japanese in the wake of World War II, occasional, feeble attempts were made to keep popular religion in check, to limit popular enthusiasm and in particular to limit the expenditure of wealth on popular religious activities. These measures were almost

positions 9:1

Spring 2001

204

completely ineffectual and were entirely abandoned by the 1980s.20 The Taiwanese government, by giving up on direct, overarching regulation of popular religionand by not attempting in a thoroughgoing manner to impose its own meanings on religious objects and practices, as had the imperial statehas allowed far greater latitude for other segments in society to appropriate religious institutions and practices for their own purposes. Nonetheless, the activities of the Taiwanese state continue to inuence popular religion signicantly, not through attempts at overt control over religious activities or interpretations but, rather, through the prosecution of religious institutions and gures. Allegations of nancial malfeasance allow the state to bring in judicial mechanisms and carry out such suppression on its own terms. The question ceases to be a religious one; it becomes a secular, legal issue of fraud. What I am interested in, though, is not the simple matter of state suppression of the relatively few popular religious practitioners who have been indicted. Court cases involving popular religion have an inuence far beyond the persons directly involved. Trials that attract public notice and media attention spawn the production of all kinds of discussions of religion in the media and among people at large. Thus, in order to convey a sense of the larger discursive eld within which the adherents of the spirit-medium cults struggle, I will examine below one of the most important cases of purported religious fraud yet to reach the courts.
Song Qili

The Song Qili scandal broke during the fall of 1996, when Song was fortyseven and some nine years after he had become active as a religious gure. Song himself claimed publicly to possess no supernatural powers but only to reveal the innate power of the original body (benti) every individual had. Realization of ones true nature did, however, allow one to manifest that body throughout the universe. Song produced numerous photographs (ultimately admitted by the photographer to have been doctored) that allegedly documented this ability to divide ones form (fen xing), often showing a larger-than-life image of Song hovering over his physical body. Daily viewing was claimed to bring enlightenment. According to one source, the photos were sold for about US$3,700 apiece. Audiences with Song are said to have

Nickerson A Poetics and Politics of Possession

205

required $74,000 to arrange, at which time one would be expected further to make donations of $370,000.21 Other members of the Song Qili Image Manifestation Association ascribed to Song abilities ranging from healing the sick to saving the dead. When Songs bubble ultimately burst, Xie Changting, a well-known politician, and other prominent political gures were seriously compromised and politically damaged by their intimate and supportive relationships with Song. Song himself was indicted for fraud, and others among his associates were jailed as well. On 30 October 1997 Song was sentenced to the maximum of seven years imprisonment, while his photographer received two years. The Song Qili affair was one of the biggest news items of the year and sparked a plethora of press accounts. The most interesting, for my purpose, are those marked out in some way as authoritativeas editorials that represent the viewpoints of major newspapers; as statements or opinion pieces by government ofcials or, especially, scholars; or as other representations of educated strata (authors may, for instance, exhibit their knowledge of Chinese and/or Western intellectual traditions). Such accounts more often than not take the case at hand as a springboard for broader commentary, in particular concerning the role of religion in society and the distinction between (good) religion and (bad) superstition. Consider, for instance, an article that appeared in the China Times (Zhongguo shibao) almost immediately after the Song Qili scandal broke. It reported the remarks of Qu Haiyuan, head of the Preparatory Ofce of the Social Sciences Research Institute of Academia Sinica. Qu remarked, Already in Taiwanese society there is an enormous number of adherents of popular beliefs, so that it is difcult for the government to regulate [popular religion] with laws. Recently the social and political situation has been unstable, and with the addition of the individual psychology of praying for blessings and seeking wealth, the trend toward popular belief has increasingly attracted a great number of the masses. . . . Most regrettable is that more and more of the media play up supernatural matters, giving assistance to this atmosphere of superstition in society. Moreover, that politicians take the lead is another important factor encouraging religious degeneration.22

positions 9:1

Spring 2001

206

Thus it is not criminality with which Qu is concerned but simply the wide diffusion of popular belief (minjian xinyang). Popular belief is clearly equated with superstition (mixin) and, if not with religion itself, with religious degeneration (zongjiao bianzhi). Another theme that runs throughout the press accounts of the Song Qili affair extends Qus dislike of mixing religion and politics to more general notions of what religion in a modern nation should concern. An editorial titled Seeking Personal Advantage and Superstitiously Believing in Miracles Disorders Ones Mind and Harms Oneself raised Xie Changtings unfortunate example. It then continued, But, nowadays, not a few people in politics and commerce enjoy the trend of seeking help from the gods and divination. This is enough to show that the Song Qili case is merely the tip of the iceberg as far as this problem is concerned. What deserves reection is that while everyoneincluding public guresof course has freedom of belief, and some can clearly draw the boundary between spiritual life and practical life, still the activities of many public gures . . . do not go beyond the level of personal benet: asking about their ofcial careers and trying to get promotions. . . . When people in the world seek the Way, they often do not stop at cultivating the mind but also look for miraculous conrmations. Going along in this way, bedazzlement by miracles nally surpasses the cultivation of the mind. Again expressing the idea that politicians should be exemplars in this regard, the editorial concludes with the same notion that it is the failure of so many political gures to separate religion from practical life that encourages the deluded fascination with the supernatural that made the Song Qili affair possible.23 Religion, this piece implies very clearly, is perfectly good if, rst, it concerns only individual moral self-improvement and, second, if it is kept well insulated from the world of daily affairs in general and politics in particular. The last source I would like to consider is a lengthy reection on the nature of religion published as one of several articles in a major newsmagazines report on Songs case. Its author, Nanfang Shuo, refers to a number of Western intellectuals, from Goethe to Max Weber to Walter Burkert. Nanfang

Nickerson A Poetics and Politics of Possession

207

Shuo also does not shy from broad generalizations. The piece begins, Humans are insecure animals who are tossed about between secularization and sacralization. Thus we have religion. The author then introduces Webers notion of the rationalization of social life and the ensuing disenchantment of the world, while at the same time noting that the appearance of new religious groups and the revival of existing ones seem to indicate a re-enchantment. Religion, the article holds, should partake of the rationalization that has dominated the political and socioeconomic realms of modern life: In the sphere of belief, religion, under pressure from secularization, cast off its original mystery and increasingly tended toward religious individualization. It became a kind of spiritual belief. By bucking this trend, Taiwanese religion remains mired in an atavism of the worst kind, votive religion (xuyuanshi zongjiao): Too many people see belief as a kind of medium of exchange, using offerings in exchange for protection. Thoughts of self-benet overcome those of beneting others. . . . The divine and the transcendent value represented by the divine lose their meaning. As one might by now expect, Nanfang Shuo then turns from religious degeneration to political decadence. Politicians similarly just see politics as a matter of winning elections and gaining power. Taiwanese politics lacks any transcendent public values or sense of responsibility, and Taiwans religious degradation is just like that of our government. . . . The Song Qili phenomenon is not a bit worth marveling at. He is nothing more than a skilled black magician putting on an extremely postmodern farce in our society, where the values system of votive religion ourishes.24 Thus the Song Qili affair gives this articles author a chance to indict what he takes to be the widely shared values of an entire society, as expressed particularly in religion. That Songs activities were perceived, in accounts like those cited above, as little different from those of other forms of popular Taiwanese religion is also plain from other sources. Just one example comes from the comments of a prosecutor made when one of Songs immediate disciples, while under examination, offered to demonstrate the art of making an object immovable by mental concentration alone. The prosecutor refused to allow him to do so, saying that that was merely the skill of spirit-mediums.

positions 9:1

Spring 2001

208

The photo of this man accompanying the same article identied him as Song Qilis table-head, the term commonly used for a spirit-mediums interpreter.25 Thus the Song Qili affair shows how the state can continue to play a role in the production of discourses on religion without actively attempting to regulate religious activity in general. Once a religious gure or organization can be prosecuted in the courts, ensuing media accounts ensure that religion will become a matter of public discussion. Moreover, as we have seen, such reports and accompanying opinion piecesmarked in various ways with the aura of authoratativenessgo well beyond reporting the facts of the case and devote much space to discussing what religion should and should not be. The most important observation that needs to be made about the notion of proper religion expressed in the above statements concerns the relationship between religion and daily life. Religion that involves praying for blessings and seeking wealth is superstition. Spiritual life and practical life should be clearly distinguished; the contamination of the spiritual by the material is reected in the pursuit of miracles, tangible conrmations of what is, purportedly, entirely intangiblea kind of spiritual belief. Chinese popular religion is perhaps above all based on the establishment of relationships of exchange with deities, using offerings of incense, food, and other material substances both to request and to requite supernatural help in connection with practical problems. (In the medium cults as well, one cannot simply ask ones question but must rst offer, at least, incense to the various deities represented in the shrine.) To deprecate the same as votive religion is not only to strike at the basis for popular practice; it is also to insist on a radical, and indeed modernist, distinction between otherworldly and this-worldly concerns.
Popular Responses

In the wake of the Song Qili scandal, spirit-mediums and other practitioners interviewed in my eld research were especially adamant that they not be tarred with the brush of superstition. Here we only have the gods true pneumas (zheng qi); people may bring their superstitions, but we dont

Nickerson A Poetics and Politics of Possession

209

encourage them, the medium at Nantianfu told me. Discussions of superstition at the spirit-altars clearly reected concern about the most basic form of the mixing of the spiritual and the material: monetary gains for the provision of religious services, which had provided the basis for Songs prosecution. The legitimacy of a particular spirit-medium altarits freedom from superstitious activitiesis often asserted in relation to money: accepting only voluntary donations (as at Nantianfu), charging only a relatively small amount for each consultation (e.g., as little as about US$4 or perhaps as much as about $11), or the absence or rarity of recommending further (and more expensive) curative rituals. The notion that bad religion (or superstition) is in part dened by its venality and exploitativeness is shared by both sides. More signicantly, in seeking to explain what they do by invoking the distinction between form and the formless, as discussed above, spirit-altar adherents replicate the same dualism expressed by critics of popular religion. The realm of religion is that of the intangible; practical affairs lie outside that realm. At least some devotees of the medium cults in fact superscribe to borrow Prasenjit Duaras termmeanings on their own activities that are highly consistent with the views of elites.26 Despite the fact that he frequently carried the divination chair at Nantianfu, one man stated that if one is healthy, one is numinous (ling), and if unhealthy, then not numinous; events depended on ones actions; only the ignorant believe in divination. Another man, who frequently acted as interpreter during Nantianfus sances, claimed that the use in sances of the palanquin for apotropaic purposes is merely a kind of skillful means (fangbian). Exorcism is not the issue; it is simply that pressure on the minds of questioners is reduced, and they can be encouraged to perform good acts. What happens to someone, he said, depends on his or her mind. Those who suggest interpretations like this reinscribe the same spirit/matter binarism that is inherent in the critiques of Song Qili I discussed above. The whole procedure of mediumistic divination and cure becomes nothing but an empty show for those of lesser understanding; religion again is made out to be something entirely spiritual. Acceptance by the medium cults adherents of the distinction between the scientic and the nonscientic leads to similar consequences. Following

positions 9:1

Spring 2001

210

the apparent assertion of a realm of knowledge independent of scientistic culture, sance participants often appeared to express a contrary desire for acceptance within that same culture. Things that science cant explain yielded to things that science will one day be able to explain, by means of ling xue, the study of the numinous (that is, scientic or other academic research into the supernatural). In order to be true, or at least to be more true, the supernatural ultimately must be validated by science. The formless exists, at best, in the epistemic penumbra of form. In their justications of their own practices, spirit-altar devotees accept the same basic distinctions as their critics: between the scientic and the nonscientic and between proper religion and superstition. The distinction between the formless and that-which-has-form entirely parallels more authoritative discourses; it involves a virtually identical dualism between the spiritual and the material and assigns the correct concerns of religion to the former. In accepting this binarism, devotees of the spirit-altars do more than share a certain notion of religion; they become complicit in a larger discourse about the legitimacy of educational capital as a determinant for social status. It is in this light that one may understand a thread that runs constantly through news accounts of Song Qilis downfall: surprise at the involvement of intellectuals and politicians in his association. Not only can Song Qili attract a large number of believers, but among them there is no lack of people of university lecturer rank or high-level public ofcials, remarked the China Times.27 A newsmagazine headed one section of an article The Highly Educated Also Believe in Song Qili.28 Another newspaper article, titled Religious Belief and Superstition, stated, Among Song Qilis followers there was no lack of those with higher educational degrees. . . . Although their beliefs are different, compared to the previous impression that only old ladies (lao ama) would burn incense and worship, certainly more and more so-called high class intellectuals are taking refuge in religion. Citing events ranging from imports of New Zealand beef to instability in the stock market, the article went on to attribute this change to an increased feeling of insecurity that had led to values systems starting to get mixed up.29 The involvement of intellectuals in popular religion is always portrayed as anomalous, deserving mention precisely because it is unexpected. Such a

Nickerson A Poetics and Politics of Possession

211

fact, if portrayed as an anomaly, actually reinforces the basic correlation between increased education and decreased interest in (especially popular) religionreligion in which the formless is always intimately engaged with that-which-has-form: the healing of bodily illness and the amelioration of all manner of everyday problems. In Taiwanese society today, religion functions much like art and other matters of taste did in Bourdieus France of the 1960s. Bourdieu writes that what he calls the aesthetic disposition is characterized by the suspension and removal of economic necessity and by objective and subjective distance from practical urgencies, which is the basis of objective and subjective distance from groups subjected to those determinisms.30 The clearest example of a disposition parallel to the Bourdieu aesthetic in Taiwanese religion was cited above in connection with the denunciation of so-called votive religion, the use of religion to gain help with practical necessities. Another writer warned against seeking personal advantage through religion. These texts continually deplore the mixing of the temporal and the spiritual. They criticize the same lack of distancing (characteristic of the lower classes) Bourdieu refers to in the case of aesthetics. It would be a mistake to correlate educational capital directly with other determinants of social level in Taiwan. The constituency for the forms of urban popular religion I studied consists not only of factory and other workers but also small business entrepreneurs and even a few factory owners and other bourgeois. However, my surveys of those who attend medium sances, as well as other eldwork data, indicate that even the best-off tend to be at most only moderately educated. While economic capital may differ, relatively low educational capital is a constant among those involved in the medium cults. Furthermore, the emergence and increasing predominance in Taiwan of a high-technological economy that rewards educational attainments, in combination with long-standing cultural values, can only mean that people like those who attend medium sances are hardly the beneciaries of an ideology that privileges formal education. By interpreting their own practices in light of elite notions of religion, they contribute to the marginalization of those practices and to their own sociopolitical marginalization.

positions 9:1

Spring 2001

212

Autonomy Reconsidered

We would appear to have come some distance from previous sections of this essay, which portrayed possessions poetics as empowering the medium cults to challenge the social regime and to establish truths in opposition to dominant ideologies. However, I would claim that this apparent contradiction between capacity for resistance in principle and complicity in factexists not in spite of but because of the particular form of autonomy that characterizes the medium cults. As I have noted, involvement in both the formless and the material realms, or the lack of a sharp division between them, is intrinsic to Taiwanese popular religion. That same lack is even more striking in the case of spirit-possession, the embodiment of a deity by a esh-and-blood human being. In possession, the presence of the divine is apparent not through subtle, spiritual signs but, rather, through the more than obvious spectacle of twitching limbs and unearthly speech. Such trance discourse is, as I have argued, the basis for the more elaborated truths produced by the spirit-altars communities: divinatory utterances, prescriptions for healing, remedies for all manner of lifes difcultiesthe solution of practical problems through means that are entirely distinct from prevailing medical or other modern technologies. Having examined the discourses of more dominant groups concerning religion, one would have to agree that these popular truths are enunciated through means that are independent of the dominant culture. The medium cults are not passive receptors for cultural materials produced elsewhere, nor do they take exogenously produced culture and turn it to their own uses. In this respect they are culturally autonomous. When in the face of opposition, why, then, do the devotees of the cults respond in ways that do not challenge that opposition but, instead, accede to it? This is due precisely to the very same autonomy. The discursive world of the medium cultsthe universe created through the poetic language of possessionstands apart from dominant understandings of religion. The spirit-altars cannot subvert the dominant order from within, because they operate outside it. They generate truths that cannot be recognized as such by the wider discursive community owing to their source; inherent in the fusion of form and the formless through spirit-possession is a cosmology entirely

Nickerson A Poetics and Politics of Possession

213

alien to the modernist version in which form and formless belong to distinct realms, with religion pertaining only to the latter. Thus, when attempting to respond to their critics, spirit-altar devotees use the language of their opponents and, in so doing, abandon the origins of the power of possession and add weight to the very discourses that would dismiss the medium cults as the products of mere delusion. In complete contrast to the populists metaphor of guerrilla warfare, the medium cults indeed establish territory that is their own. The enclaves that are the spirit-altars take nothing from the dominant order and thus, when under pressure, need not yield anything back. But of course they do seem to give something back: Is not the relegation of religion to the sphere of the purely spiritual and the acceptance of the explanatory superiority of science a complete capitulation? A distinction needs to be made between the medium sances, while in operation, and the subjectivities of those who attend them. The sance creates, as I have argued, a world of its own. And, pace Bourdieu, such a world represents far more than the remnants of a bygone erudite culture, retained now only among folk remnants. There is nothing more vital than a spirit-altar in session, nothing more riveting than the spectacle of possession. But that is not to say that those who attend the sance can remain in that world. In order, for instance, to participate in larger debates on religion, people cross over into the discursive sphere of the dominant, with the consequences I have noted. Again as opposed to guerrilla-held territory, the spaces created by the medium cults remain autonomous, but those who inhabit them must be migratory; they cannot remain there. That those who might oppose the social order often end up participating in their own subjection is of course far from new. What I have tried to show in this essay is that such subjection does not depend on the use by the dominated of the cultural products of the dominators. It can also come about when the dominated themselves participate in producing their own culture. Despite the fact that the carnival (or the sance) must end, time spent there might still be more protable than time spent watching television. At least I would like to think so; perhaps my conception of mass culture is still unreconstructably Adornist. In any event, I began work on this topic with directions that in many ways were set by the historical study of medium

positions 9:1

Spring 2001

214

cults in imperial China, in times and places where the cults were rmly based in on-the-ground local communities and were clearly used to buttress local forces against the outsideto serve the interests of local autonomy. Nowadays, worshipers in Tainan do not associate the spirit-altars with specic, bounded localities, such as urban neighborhoods. Spirit-altar attendees, when asked whether most who come to inquire of the deity are from the neighborhood, normally answer along these lines: People come here from all over. Its not a matter of where you live, but of having karmic links ( yuan) with the god. Whatever the actual composition of the worshiping group, people do not represent their group as a local one. Still, after attending many hours of sances, I could not help but be struck by how, even when the spirit-altars community was deterritorialized and despatialized, the sances created their own worlds. In these spaces exorcisms were carried out, healing took place, and people entrusted their life problems to a deity resident in a shaking, twitching human, a person whose presence dominated, even created, this space that was so different from the quotidian. Hence autonomy. Subsequent examination of both the devotees own understandings of the cults in which they participated and the religious discourses that disparage them necessitated reconsideration. The former and the latter are in many ways remarkably consistent. Thus this essay is rather more nuanced than initially planned. The medium cults are autonomous, but their adherents cannot escape dominant conceptions of religion; possession implies the convergence of the spiritual and the material, but sance participants restrict the mediums powers to the sphere of the formless. I am fully aware that it is now acceptable in academic writing to leave contradictions unresolved, at least sometimes. Still, it seems to me that further research should begin with the sance attendees but outside the space of the spirit-altar, when people have a choice about attending a sance or watching television. What do they retain when they leave the spirit-altar and return to the world outside it? How do they integrate these two very different modes of experience? As far as these questions are concerned, more remains to be done. If there is mediation between the medium cults and the dominant order, it is likely that much of that mediation is accomplished by the devotees of the spiritaltars themselves, those who are most directly and personally involved in migrating between one sphere and the other.

Nickerson A Poetics and Politics of Possession

215

Notes

1 2 3

4 5 6 7

9 10 11 12 13

I would like to thank the many friendly and helpful people at the various temples around Tainan where I carried out my work, in particular those at Nantianfu and San qing gong. Many thanks are due as well to the Chiang Ching-kuo Foundation for International Scholarly Exchange, which generously provided the funding for much of my research in Tainan, as well as to Wang Shan Shan, Chung Tin Yi, Huang Chien Ming, and Lu Pei Pei for their diligent and able assistance. Michael Taussig, The Magic of the State (New York: Routledge, 1997), esp. 184188. One notable exception is P. Steven Sangren, History and Magical Power in a Chinese Community (Stanford, Calif.: Stanford University Press, 1987), 130131 and chap. 8. Max Horkheimer and Theodor Adorno, The Culture Industry: Enlightenment As Mass Deception, in Dialectic of Enlightenment, trans. John Cumming (New York: Herder and Herder, 1972), 120167. Michel de Certeau, The Practice of Everyday Life, trans. Steven Rendall (Berkeley and Los Angeles: University of California Press, 1984), xxi. John Fiske, Understanding Popular Culture (London: Routledge, 1989). Pierre Bourdieu, Distinction: A Social Critique of the Judgment of Taste, trans. Richard Nice (Cambridge, Mass.: Harvard University Press, 1984), 395. This study is based on eldwork in Tainan carried out in the summer of 1996, the rst half of 1997, and again in the summers of 1998 and 1999. Altogether, in addition to the various written sources cited, the data for this study were gathered during attendance at over fty medium sances (as well as related interviews and more casual conversations), principally at six spirit-altars, during these periods. While I was generally able to converse with people in Mandarin Chinese, I frequently relied on assistants to make transcripts or summaries (in Chinese, either on the spot and/or from tape recordings) of events that were most often conducted in Taiwanese. Chen Meiqing, Taiwan de Guangze zunwang xinyang [Beliefs concerning the Reverent Lord of Broad Compassion in Taiwan], 2 vols. (unpublished manuscript in authors possession, 1997). See Gilbert Rouget, Music and Trance: A Theory of the Relations between Music and Possession, trans. Brunhilde Biebuyck (Chicago: University of Chicago Press, 1985), chap. 5. Julia Kristeva, Desire in Language: A Semiotic Approach to Literature and Art (New York: Columbia University Press, 1980), 2832, 125, 133134. Marjorie Wolf, A Thrice-Told Tale: Feminism, Postmodernism, and Ethnographic Responsibility (Stanford, Calif.: Stanford University Press, 1992). Kristeva, Desire in Language, 65, 137140. Ibid., 29, 134135.

positions 9:1

Spring 2001

216

14 Gilles Deleuze and Flix Guattari, Anti-Oedipus: Capitalism and Schizophrenia (Minneapolis: University of Minnesota Press, 1983), esp. 116, 85. 15 For a discussion of the divination chair, as David Jordan has called it, see his Gods, Ghosts, and Ancestors: Folk Religion in a Taiwanese Town (Berkeley and Los Angeles: University of Californa Press, 1972), 6467. 16 Carmen Blacker, The Catalpa Bow (London: Allen and Unwin, 1975); I. M. Lewis, Ecstatic Religion: A Study of Shamanism and Spirit-Possession, 2d ed. (London: Routledge, 1989); Ned Davis, Society and the Supernatural in Sung China (Honolulu: University of Hawaii Press, forthcoming). 17 Claude Lvi-Strauss, The Sorcerer and His Magic and The Effectiveness of Symbols, in Structural Anthropology, trans. Claire Jacobsen and Brooke Grundfest Schoepf (New York: Basic Books, 1963), 167185, 186205; Arthur Kleinman, Patients and Healers in the Context of Culture (Berkeley and Los Angeles: University of California Press, 1980); Zhang Xun, Jibing yu wenhua [Illness and culture] (Taibei: Daoxiang, 1989). 18 Pierre Bourdieu, Outline of a Theory of Practice, trans. Richard Nice (Cambridge: Cambridge University Press, 1977), 168170; emphases in original. 19 See, e.g., Jordan, Gods, Ghosts, and Ancestors, 139. 20 Robert P. Weller, Matricidal Magistrates and Gambling Gods: Weak States and Strong Spirits in China, in Unruly Gods: Divinity and Society in China, ed. Meir Shahar and Robert P. Weller (Honolulu: University of Hawaii Press, 1996), 250268, esp. 263. 21 Qiu Minghui, Song Qili miren de yuyan li you wuge moshu yaosu [In Song Qilis language of bewitchment there are ve essential black-magical elements], Xin xinwen [The journalist], 2026 October 1996, 4041. 22 Xuezhe yu zhengtan renshi wu daitou zhuzhang waifeng [Scholar calls for politicians not to take the lead in promoting unhealthy trends], Zhongguo shibao, 13 October 1996. 23 Zhuiqiu sili mixin shenji luanle benxin haile ziji [Seeking personal advantage and superstitiously believing in miracles disorders ones mind and harms oneself], Zhongguo shibao, 14 October 1996. 24 Nanfang Shuo, Gaoming de mofa shi wanle yi-chang houxiandai de huangdanxiu [A skilled black magician puts on a postmodern farce], Xin xinwen, 2026 October 1996, 5256. 25 Zuofa dingshen wan ru jitong: Song Qili da dizi Zheng Zhendong shouya jinjian [Performing the technique of immobilizing bodies like a spirit-medium, Song Qilis great disciple Zheng Zhendong is detained and interrogated], Zhongguo shibao, 14 October 1996. 26 Prasenjit Duara, Superscribing Symbols: The Myth of Guandi, Chinese God of War, Journal of Asian Studies 47, no. 4 (November 1988), 778795. 27 Benzun poxiang: jiekai Songqili shenmi miansha [The original venerable makes a fool of himself: Unveiling Song Qilis mystery], Zhongguo shibao, 13 October 1996. 28 Du Shengcong et al., Zhexie ren shuo: Weishenme yao yapo women de xinyang? [These people say: Why do you want to oppress our faith?], Xin xinwen, 2026 October 1996, 3839.

Nickerson A Poetics and Politics of Possession

217

29 Zongjiao xinyang yu mixin [Religious belief and superstition], Lianhe bao [United daily news], 4 January 1998. 30 Bourdieu, Distinction, 54.