Society and the Supernatural in Song China

1 Introduction

1 Introduction This book takes its title seriously. Its overwhelming concern is the relation of Chinese society with the supernatural and with the experience of the supernatural as an aspect of social relations. More specifically, it focuses on the experience of the supernatural in its most palpable and dramatic form—the descent of gods, ghosts, or ancestors , and their habitation within a human body. It focuses, in other words, on what we call “spirit-possession” and what Chinese writers of the Song period (960–1276) denoted by the term “pingfu.”1 I understand this experience both to be occasioned by a crisis in social relations and to be in itself an occasion for a transformation of these relations. Spirit-possession is necessarily a social experience. It has been succinctly defined as a “trance of identification,” and this identification— of the subject with a spirit—cannot be made except in the company of others.2 Moreover, far from being a sign of the private anguish of mental illness, spirit-possession was a means of avoiding it. Spirit-possession was both a role assumed in public and a shared and universally recognized idiom that allowed an individual person to convert emotion into culture, and symptoms into symbols.3 In an article written several years ago, Piet van der Loon asked us to examine and expose the “shamanic substrate” of all Chinese religions .4 Yet what is exposed by any such examination is the weakness of “shamanism” as a description of either China’s indigenous religion or the “substrate” of her religions. If there is a common basis of Chi1 nese religious practice, then that basis has more to do with forms of trance associated with spirit-possession than with shamanism. Moreover , if spirit-possession—and particularly the possession trance of the spirit-medium (wu)—has informed the religious practice of Daoist and Buddhist priests, then the textual and ritual traditions of China’s two organized religions have done just as much over the course of centuries to transform the practices of her spirit-mediums. By the Song dynasty, it becomes as difficult to talk about the “substrate” of Chinese religion as it is to trace the “origin” of a particular religious practice , belief, or divinity. Recent attempts to unearth the “shamanic substrate” of Chinese religion employ the term “shamanism” in such a loose way that it becomes almost meaningless, and therefore useless, as a category of historical and cultural analysis. In a chapter on the shamanic substrate of the Chinese Buddhist figure Mulian, Stephen Teiser relies on those authorities who identify shamanism and spirit-possession as more or less the same thing: Some scholars have tried to make historical and typological distinctions between two forms of mediumship, defining the first as mere “spiritpossession ” and the second as true “shamanism.” See, for example, Eliade , Shamanism, pp. 58, 499–507. But this distinction does not appear valid even for Central Asian spirit-mediumship, to which the distinction was first applied. As Ioan M. Lewis, Ecstatic Religion, pp. 55–56, writes, “the Tungus evidence makes nonsense of the assumption that shamanism and spirit-possession are totally different phenomena, belonging necessarily to two different cosmological systems and to separate stages of historical development.”5 Yet the Tungus evidence makes nonsense of no such thing. The analysis of the central funerary rituals of the Gold of Manchuria reveals that the Tungus shaman is master of the spirits in his journey to the underworld . He employs his auxiliary spirits as his guides, his servants, his means of transportation, his interlocutors. He converses with them and occasionally even imitates them, yet in all cases the subject of trance—the shaman—coexists with the guardian spirits.6 Possession, in contrast, is a trance of identification in which the persona of the divinity is substituted for, and does not coexist with, that of the subject. The possessed subject does not converse with or imitate the divinity; he (or she) is the divinity! Although it is true that 2 Society and the Supernatural shamanism and possession are both representations of trance behavior , these representations, as Gilbert Rouget has demonstrated, oppose and contradict each other at every point of comparison: The difference between shamanism and possession trance thus seems to rest on three factors: the former is a journey made by man to visit the spirits, the latter is a visit by a spirit...