Shamans, Mediums, and Chinese Buddhism: A Brief Reconnaissance

In traditional China, Buddhism was not a separate religious tradition or culture practiced in isolation from the rest of Chinese religious culture. This applied not only to people outside the monastic context, but also to people within that context. Even shamanic and medium practices could take place within a Buddhist context. Shamanic is here defined as spirit travel or communication whilst the practitioner stays himor herself, whereas a medium would be possessed and temporarily become the other spiritual being. Finally, future research should look at the way in which these practices may have been influenced and/or partially replaced by other forms of contact with the divine or supernatural world, such as dreams and visions. 203 SHAMANS, MEDIUMS, AND CHINESE BUDDHISM 1 Gao, Zhongguo wushu shi is the standard reference on wu from this broad perspective. Davis, Society and the Supernatural is the best Western book-length study. T China had a broad range of technologies for communicating with the divine or supernatural world. Here I will look specifically at two specific forms, which we usually characterize with the labels shamans and mediums, and the interaction between these two types of specialists and Chinese Buddhism. Since this analysis will depend very much on our definitions of these religious phenomena, some further delineation is in order. As the aim of this article is to take a broad view of the interactions between these religious fields or traditions, I will adopt an inclusive definition of Buddhist religious culture, rather than a strictly normative one. One might for instance define away any form of Buddhism that interacts positively with or even incorporates shamanic and medium culture as being not properly Buddhist. This would not be a very interesting approach, however. Therefore I will accept any self-identification as a Buddhist believer or practitioner as valid forms of a larger Buddhist religious culture. My primary interest is in showing the value of such an inclusive approach to the interactions between Buddhist religious culture, shamans and mediums. I hope to present sufficient examples to argue my basic point that these interactions helped to shape Buddhism and that Buddhist practices can be better understood within this larger religious culture. It is a common claim that the Song saw increased acceptance of local religious culture by Daoism and Buddhism, but the materials discussed show that this started already at a much earlier date. A full investigation cannot be attempted within the scope of one brief article and therefore I have termed this contribution a reconnaissance, in the hope that others will follow my lead. Problems of Terminology The terms that are conventionally interpreted as the Chinese equivalent of shamans and mediums are wu 巫 and xi 覡, supposedly referring to female and male ‘shamans’ respectively. The term xi is 204 BAREND J. TER HAAR 2 The following alternative form of writing xi suggests it originally was an alternative term in real usage. The Xunzi 荀子 (Institutions of the Kings 王制-chapter) has the following phrase: ‘the affairs of wu with humpbacks and xi with lame feet’ (傴巫跛擊之事). The character xi 擊 normally means ‘to attack’ and given the parallel with wu it is clearly used here for its sound value, rather than its meaning. This does not mean that xi was a shaman or medium as defined in this article. Indeed, the ‘xi with lame feet’ sounds more like a ritual specialist who walked the Step of Yu. See Harper, Early Chinese Medical Manuscripts, 167–69. 3 Early evidence is discussed in Boileau, ‘Wu and Shaman’. rarely used in primary sources and came to be used as an elite label in written discourse, rather than an oral term used in living religious culture.2 The term wu is used quite broadly from its first appearance in the earliest written sources onwards, for men as well as women, but it is unclear to me how it was used in living oral discourse.3 By the medieval period it already tended to be used to denote any kind of ritual expert that the author disapproved of. Nowadays it seems to mean little else than that part of traditional Chinese religious culture which falls outside institutionalized Buddhist or Daoist traditions. The appearance of the term wu in a primary source can therefore not be used as an indicator of a single underlying religious practice, unless sufficient additional information is provided in our sources. It is this additional information which counts, rather than the label. More general, we have to keep in mind that our primary sources are in fact rarely primary, but almost always composed in a stylized written language (‘Classical Chinese’) that is far removed from the Mandarin (guanhua 官話) spoken by the male elite in so far as they were involved in government, and even further from the many different vernaculars that most people spoke locally most of the time (including elites). Whether we call these vernaculars languages or dialects is a moot point, for they were certainly mutually unintelligible. In other words, the term wu is always a translation from a (usually lost) local term, in which written traditions would have strongly influenced the translation process. The terms shaman and medium in this investigation will be restricted to specialists of direct verbal and visual communication by and with the supernatural world. In terms of the culture itself, the 205 SHAMANS, MEDIUMS, AND CHINESE BUDDHISM 4 Lewis, Ecstatic Religion. 5 Eliade, Shamanism. medium is possessed by a being from this other world, on a regular and more or less controlled basis (which is then qualified as divine possession), rather than an incidental basis (which is then dismissed as demonic possession). It is the deity who speaks, usually in a form of messages or even through dialogic interaction, rather than lengthy narrative. A classical study in this field is by Ioan Lewis, which also provided inspiration for my own approach to the topic.4 Evidence of the activities of mediums in traditional China is very rich. The shaman remains in control of his or her soul and somehow travels to, or directly communicates with, this other world by means of trance. The shaman himor herself is the voice who speaks and narrates. Here there is a strong influence of ongoing work on religious practices among Siberian and Northern American peoples. A classical study, if now perhaps outdated, is that by Mircea Eliade.5 Evidence of this type of religious interaction does exist, but is much harder to find, although early Maoshan Daoism seems a good candidate. It is clear that from our own world view both types of specialist are actively engaged in the construction of their communicative realities, but the fact that they as well as their audience construct their role so differently suggests that this difference must be significant to them. I can easily imagine that a single specialist might play both roles, but I am presently unaware of explicit historical evidence to this effect. Here we have a common problem of sources in the study of Chinese religious culture, where we can easily build typologies of religious practices, but do not know what concrete set of practices was followed by any specific individual, whether as a customer or as a practitioner. Both types of specialist would often work together with other ritual specialists, who used them to gain information about the issues at hand (such as demonic possession or family misfortune) and then carried out rituals in order to remedy the problem. In our modern terms, we could say that they constructed an elaborate narrative to give meaning to the problem and its social context, in a similar way as a psychoanalyst does today. 206 BAREND J. TER HAAR 6 Pan, ‘A la recherche de l’unité’. 7 Gan, Soushen ji 5, 57. On this cult see Lin, ‘The Cult of Jiang Ziwen’. Shamans and Mediums When we no longer fixate on the term wu, we may obtain a more consistent analysis, but also face new problems in finding sources. In his recent dissertation François Pan (Pan Junliang) has circumvented this problem by focusing on the terminology that is used to describe possession in the so-called Period of Disunion, rather than the overall labels for the religious specialist in question. As it turns out, the sources use a rich vocabulary, such as the terms jiang 降 and xia 下 that both mean ‘to descend [in]’; terms such as ping 慿, zhuo 著, and fu 附 that refer to the demon or deity ‘attaching’ itself to someone; or chi 持 ‘to hold’, which expresses the same idea more strongly.6 Terms such as jiang also appear in sources from the Han and earlier periods. Take for instance the famous cult of Jiang Ziwen 蔣子文, who communicated through dreams and mediums. In the narrative that describes how his cult started from his death and subsequent appearance to the living, this is described as follows: [Jiang Zi]wen also descended (xia 下) in a medium (wuzhu 巫祝) [and said]: ‘I will provide assistance to the Sun-family on a great scale, you ought to erect a shrine for me. If not, I will cause insects to enter into human ears and cause disaster.’ Suddenly, [there were] small insects like deer flies (or stouts), when they entered [people’s] ears, [the victims] all died. The doctors could not heal it. The people were increasingly afraid. Lord Sun 孫 did not believe it. [Jiang Ziwen] descended (xia 下) again in a medium (wuzhu 巫祝) [and said]: ‘If you do not worship me, I will again cause disaster with a big fire.’ 7 The crucial terms here are ‘descend’ and ‘mediums’. We have no reliable evidence on the vernacular terms that were used at the time, although it is by no means impossible that the notion of descending was expressed in the form ‘coming down’ (xialai 下來) or a similar equivalent. 207 SHAMANS, MEDIUMS, AND CHINESE BUDDHISM 8 Gan, Soushen ji 5, 60. 9 There is the famous story of Ximen Bao 西門豹 from around 400 BCE who punished ritual specialists (wu 巫, but no mention of possession or soul travel) who sacrificed virgin women to the Lord of the River, see Sima Qian, Shiji 126: 3211–3212. In its present form this version of the narrative is suspect, since it introduces Han local functionaries, cash coins (wen 文, a Qin-Han innovation, rather than the then somewhat current bi 幣 in the shape of a variety of objects) and far too much oral dialogue to be a credible archival record. While this figure is documented in earlier sources, this particular anecdote is not. For a more traditional interpretations, see Lin, ‘The image and status of shamans’, 397–458, especially 417–19; and Lai, ‘Looking for Mr. Ho Po’. 10 Although I usually translate gui as demon(s), the term ‘ghost(s)’ reflects In another story featuring the same deity, we receive a very different indicator of the presence of the deity in the female medium, who has unknowingly encountered the deity on her way to his temple and received various favours from him: The fragrant smell of [Wu] Wangzi 吳望子 could be smelled several Chinese miles away. She had much divine efficacy and the entire county served her. After three years, Wangzi [who had started her career at sixteen Chinese years of age] suddenly developed an interest in another (lit. ‘the outside’) and the deity thereupon broke off his visits (wanglai 往來).8 Here the notion of possession appears to be expressed by the term ‘visits’. Clearly, once the young girl had become sexually interested, the deity could no longer monopolize her attention. The young age of mediums is probably also indicative of virginity in these materials.9 Such mediums have continued to function in different forms until the present day, although it is difficult at present to study them through fieldwork due to government repression and the need for building long term trust that is difficult for outsider observers. There was also a very different type of communication, which involved a very different world of beings. Here the specialist remained heror himself and became a ‘seer of ghosts’ (jiangui 見鬼).10 This 208 BAREND J. TER HAAR better the fact that we are dealing here with the ghosts of deceased relatives. An excellent study is Sun, ‘Youming zhi jian’. 11 Li, Taiping guangji, 121: 853–54. type of specialist is different from the medium and often denoted simply by his or her ability of ‘seeing ghosts’. The term wu may be used, as it is in the following example, but is much more frequently left out completely. In the youth of Su Ting 蘇頲 [formerly] Chancellor of the Tang someone read his appearance (xiang 相, the standard term for the technique of physiognomy): ‘You will reach the position of Chancellor and end in the second degree (pin). Later when he reached the position of Chancellor and the third degree, he became extremely ill. He called a shaman (wuxi 巫覡) to look at it (shi 視). The shaman (wu 巫) said: ‘The life of his lordship is about to end. You cannot rise again.’ For this reason Ting brought up the words of the physiognomist again and the shaman said: ‘Initially his lordship was [predicted to be] like this. When you were in charge of Guizhou examinations you had two people killed and now these two persons have accused you in the underworld. The person in charge has decreased your lifespan by two years. For this reason you will not get to the second degree.’ Ting had once been in charge of Guizhou and there had been two clerks who accused a county magistrate. Ting had had the clerks killed on behalf of the magistrate. Thereupon he sighed deeply and died.11 The use here of the unusual term wuxi suggests that the original oral account has been translated using assumptions from canonical sources. The term ‘seeing’ (shi) and the absence of any reference to possession indicate clearly that we are dealing with a very different kind of ritual specialist, who specializes in ‘seeing’ what happens after death. Typically, a medium does not visit the underworld or report on it, but allows the customer to communicate directly with the divine world. Thus, these are very distinct spheres of activity. In the following account from a Buddhist source, we again en209 SHAMANS, MEDIUMS, AND CHINESE BUDDHISM 12 I am still puzzled by the term yi 易 here. 13 T no. 2059, 3: 339. counter a ‘seer’, but this time he or she sees unidentified ghosts or demons, rather than the ghosts of deceased relatives. The monk Zhiyan 智嚴 was a man from Xiliang Prefecture. He left the family when he was young. ... There was a seer of ghosts (jianguizhe 見鬼者) who said that he had seen ghosts say to each other near the Altar of the Earth in the Western Prefecture that getting out of the way would be easiest12 once Lord Zhiyan arrives. This man (i.e. the seer) did not explain it [any further]. Suddenly, Zhiyan arrived. When they casually asked his various names, he was indeed called Zhiyan. They acknowledged him in silence and secretly paid him special respect. Ms Liu 劉, the wife of the Director Xiao Sihua 蕭思 話 of Lanling, was ill. She constantly saw ghosts come and implore [to her], which was very scary. They thereupon invited Zhiyan to explain the teachings [to them]. The moment that Zhiyan came into the outer hall, Ms Liu saw a host of ghosts scurry away. Once Zhiyan had come in to explain sutras to the lady of the house, her illness was healed as a result. Thereupon she received the Five Injunctions and the entire household took [the monk] as their master.13 Clearly, the monk himself had no dealings with the seer, but the household as a whole was convinced by his powers because of the initially obscure information that the seer had overheard from local demons. When the advent of the monks also definitively scared the demons away and Ms Liu was healed as a result, this then triggered the conversion of the household. The account stems from the Biographies of Eminent Monks (Gaoseng zhuan 高僧傳), showing how its compiler fully accepted the ability of a seer to observe ghosts (or demons). Without this prior event, in fact, the household of Ms Liu would never have been convinced of the monk’s charisma in the first place. The belief in seers did not disappear in subsequent centuries. Whereas in the above cases the sex of the seer is unclear, in later times 210 BAREND J. TER HAAR 14 Lewis, Ecstatic Religion. 15 Hong, Yijianzhi, ding 12: 639. the role seems to be played predominantly by older women, suggesting a lowly status.14 In the following example from the twelfth century, the second wife of an official, Ms. Zhang 張, has taken seriously ill and the doctors are unable to help her. The events take place in Changsha in modern Hubei. His (i.e. the official’s) mother was extremely worried about it and invited an old female shaman (wuao 巫媼) to inspect and see. She said that Ms. Wang 王 (the deceased first wife of the husband of the diseased woman) was standing in front of them and was haunting very seriously. She ordered to set up a [spirit-]tablet to say prayers and expel [Ms. Wang]. They promised [the dead woman] a jiao ritual as penance (jiaochan 醮懺), but she did not want to leave. The shaman (wu) said [her husband] should split up with her. He said: ‘You must write a letter of separation in imitation of the divorce between wife and husband among the living, and give it to her. Then you can get a separation from her.’ [Her husband] was unwilling to split up from her. When [Ms] Zhang became more seriously distressed by the day, [her husband] had no choice but to take up a writing brush in tears. He wrote the letter and handed it to the shaman (wu). [The latter] mixed it with paper money and burned it for [Ms. Wang]. The shaman (wu) said: ‘The woman was holding the letter, and after she had unfolded it and read it, she wept and left.’ [Ms.] Zhang indeed got better.15 Hong Mai writes without any pejorative judgment. He concludes by informing us that Ms. Zhang was also related to one of his own concubines (yushi 予室), indicating that at the time of writing seeing ghosts was still an acceptable practice in elite circles. On the other hand, all the principal actors were women, and the husband in question clearly was hesitant in complying with the required ritual divorce. The religious specialist is evidently a female seer and there is again no mention of possession. The ritual that is being promised could be either Daoist or Buddhist, given the combined 211 SHAMANS, MEDIUMS, AND CHINESE BUDDHISM 16 Jingyuan, Yuanrenlun faweilun (preface 1074). Maeda, Dai Nihon, X no. 1031, 58: 0725 (as part of a larger discussion of religious culture). 17 Ji, Yuewei caotang biji, 12: 15b–16a. reference to the jiao (which is Daoist) and the penance (which is usually Buddhist). Of course, the fact that the custom of a ghost seer was accepted did not mean that there was not also criticism of it. Thus a Buddhist monk from the northern Song lists the custom among a range of bad religious practices, ‘there are also shamans (wuxi 巫覡) ....who see (jian 見) people’s ancestors and parents, and can tell their forms and speech [as if] still alive.’16 Sadly, he is not more informative, but this would suggest that the custom was widespread, but also that Buddhist monks did not approve of it. We will return to the issue of Buddhist interaction with this custom further below. That the custom as such continued to exist whatever educated elites or Buddhist monks thought of it is abundantly clear, for instance from the following late eighteenth century account collected by Ji Yun: The elder sister of Ms. Tian 田 said that in one tenant household in Zhao 趙 Village husband and wife were very taken with each other. One morning the wife heard vaguely that her husband had an affair with someone else, but she was not certain about it. The wife was very gentle and sweet, and not very excitable either. She only said jokingly to her husband: ‘If you do not love me, but love her, I will hang myself.’ The next day, he was eating his meal in the fields when he encountered a shaman (wu 巫) who could see ghosts (neng shigui 能 視鬼). When [the shaman] saw him he (or she) was startled, and said: ‘Behind you are two ghosts who have hung themselves, why is that?’ Then he realized that those few words said [by his wife] in jest had already been heard by a ghost. Those who have died a violent death must always seek a replacement. ...17 Several beliefs come together here. It was thought that the ghost of someone who had killed herself (or himself) always tried to find some212 BAREND J. TER HAAR 18 For instance, see Elliot, Chinese Spirit-medium Cults in Singapore, and Jordan and Overmyer, The Flying Phoenix. 19 For two accessible studies, see Li, ‘On the Cult of the Four Sacred Animals’; Bunkenborg, ‘Popular Religion’. Also see the full version of his research, Bunkenborg, Porous Persons. 20 Potter, ‘Cantonese shamanism’. 21 De Groot, The Religious System of China, vol. 6: 1269–95 on mediums (and their larger ritual context) and 1332–41 on female ghost seers. one to replace her (or him). It is implied here that such a ghost had used the joke by the tenant’s wife and then somehow encouraged her to commit suicide prematurely. For us, the relevance is of course the ability of a ‘shaman’ to see ghosts and discover what had happened before he had had a chance to return home, where his wife had hanged herself. At present there is not much fieldwork on either shamans or mediums, as defined above, but they have not disappeared. Many who study Chinese religious culture will be familiar with the auto-mutilating mediums of Taiwan and the specific practice of spirit-writing,18 but there are also still animal mediums (especially the fox immortals) active in northern China (including the north-eastern provinces) and other types of mediums elsewhere.19 Elderly seer ladies are described briefly by Jack Potter on the basis of his 1960s fieldwork in the New Territories in Hong Kong, reflecting Cantonese practices.20 J. J. M. de Groot described both mediums, spirit writing and elderly seer ladies for late nineteenth century southern Fujian, on the basis of his experiences in the Amoy (Xiamen) region.21 Having established the ubiquitous nature of these two forms of contact with a supernatural world, we can return to the issue of the relationship between Buddhist practitioners and shamans or mediums. Seeing Ghosts and Speaking With Gods The first Buddhist monks were confronted with a sacrificial culture that conflicted directly with their own beliefs about rejecting killing 213 SHAMANS, MEDIUMS, AND CHINESE BUDDHISM 22 Campany, Strange Writing, 302 note 79, and 330 note 153. 23 For a strictly historical biography, see Forte, The Hostage An Shigao. Brief discussion of this purported miracle in op.cit., 86–87. The cult at Lake Gongting is discussed extensively by Miyakawa, ‘Local Cults around Mount Lu’. 24 Perhaps prognostication blocks. 25 Gaoseng zhuan, 1: 323. A different take on the story of the divine bamboo is given in the Mingbaoji 冥報記, see Gjertson, ‘A study and translation of the Ming-pao chi’, 264. and alcoholic drinks. This required them to take a stance at every step of the way. As history has shown, the sacrifice of freshly killed animals and liquor remained standard ritual practice throughout the imperial period and after. This is, however, not how they themselves represented their struggles about proper ritual. Numerous accounts preserved in the Biographies of Eminent Monks tell of the well-known battle between the Buddhist newcomers and established local cults, with a crucial role for shamans and mediums in telling the Buddhist side of the story.22 One of these accounts tells of a foreign monk with the Chinese personal name Gao who travelled past one of the best known cults of his time, located on an important transport node. Indeed, this was not just any monk, but the famous An Shigao 安士 高 (fl. 147–170).23 On his journey he reached the temple at Lake Gongting. This temple possessed charismatic power since ancient times. Merchants and travellers prayed here so that they would go with the wind upstream or downstream without any obstructions. Once someone requested the deity’s bamboo.24 When [the deity] did not allow it, he took it nonetheless. His boat then capsized and he drowned. The bamboo returned to its original location. From this moment onwards, the boatmen were full of respect and fear for it, and all were in awe of its shadow. Gao travelled together with over 30 boats, which offered life sacrifices and asked for good fortune. The deity thereupon descended in a medium (jiangzhu 降祝) and said: ‘On the boats there is a śramaṇa, you must call him to ascend [the shore]. The traders were really startled and frightened, and invited Gao to enter the temple.25 214 BAREND J. TER HAAR 26 For more examples where the medium or shaman assists Buddhist monks in the communication with the deities: see Gaoseng zhuan, 8: 380; Daoxuan, Xu Gaoseng zhuan (T no. 2060) 22: 623, 30: 702; Zanning, Song Gaoseng zhuan (T no. 2061), 19: 828. A lengthy conversation ensues, in which the deity explains in great detail that he had known the monk in a previous existence, but had been punished with this rebirth as a local deity because of his angry temperament. In order to make up for all of his misdeeds and in anticipation of his fall into hell, he wished to donate his divine wealth and this location to the monk. The monk then asked him to reveal his true form, which turned out to be a big snake. The monk built his monastery and eventually a young man was seen, who was the former deity released from his evil form. For our discussion, what matters here is that this entire conversation between the monk and the deity was carried out through a medium (who is labelled a zhu 祝), as is apparent from the term ‘to descend’ (jiang 降) and the two phrases ‘the deity told Gao ‘(shen gao gao yue 神告高曰) and ‘the deity said’ (shen yue 神曰). That the narrative was embellished over the years or maybe even largely invented does not alter the basic fact that the Buddhist conversion of the site was seen to depend on this form of direct divine communication. While the cult itself was deemed to be evil because of its sacrificial practices, though still very real, no disagreement existed on the validity of this form of communication through possession of a medium. Biographical sources on Buddhist monks contain a number of similar examples, in which deities communicated with them by possessing mediums.26 Some monks in this early period still operated on the border between their own expertise and that of a seer of demons or ghosts. In the following case a monk becomes a seer as well as an exorcist specialist himself: The monk Baotong 寶通, for whom we do not know his ancestry, was a man from Gaoqiu. He lacked in further knowledge or insight and only recited some spells from the seventh chapter of the Lotus 215 SHAMANS, MEDIUMS, AND CHINESE BUDDHISM 27 These lines and the following lines of dialogue are essentially in spoken


1
Gao, Zhongguo wushu shi is the standard reference on wu from this broad perspective. Davis, Society and the Supernatural is the best Western book-length study.
T raditional China had a broad range of technologies for communicating with the divine or supernatural world. Here I will look specifically at two specific forms, which we usually characterize with the labels shamans and mediums, and the interaction between these two types of specialists and Chinese Buddhism. Since this analysis will depend very much on our definitions of these religious phenomena, some further delineation is in order. As the aim of this article is to take a broad view of the interactions between these religious fields or traditions, I will adopt an inclusive definition of Buddhist religious culture, rather than a strictly normative one. One might for instance define away any form of Buddhism that interacts positively with or even incorporates shamanic and medium culture as being not properly Buddhist. This would not be a very interesting approach, however. Therefore I will accept any self-identification as a Buddhist believer or practitioner as valid forms of a larger Buddhist religious culture.
My primary interest is in showing the value of such an inclusive approach to the interactions between Buddhist religious culture, shamans and mediums. I hope to present sufficient examples to argue my basic point that these interactions helped to shape Buddhism and that Buddhist practices can be better understood within this larger religious culture. It is a common claim that the Song saw increased acceptance of local religious culture by Daoism and Buddhism, but the materials discussed show that this started already at a much earlier date. A full investigation cannot be attempted within the scope of one brief article and therefore I have termed this contribution a reconnaissance, in the hope that others will follow my lead.

Problems of Terminology
The terms that are conventionally interpreted as the Chinese equivalent of shamans and mediums are wu 巫 and xi 覡, supposedly referring to female and male 'shamans' respectively. The term xi is 2 The following alternative form of writing xi suggests it originally was an alternative term in real usage. The Xunzi 荀子 (Institutions of the Kings 王制-chapter) has the following phrase: 'the affairs of wu with humpbacks and xi with lame feet' (傴巫跛擊之事). The character xi 擊 normally means 'to attack' and given the parallel with wu it is clearly used here for its sound value, rather than its meaning. This does not mean that xi was a shaman or medium as defined in this article. Indeed, the 'xi with lame feet' sounds more like a ritual specialist who walked the Step of Yu. See Harper,Early Chinese Medical Manuscripts, Early evidence is discussed in Boileau, 'Wu and Shaman'. rarely used in primary sources and came to be used as an elite label in written discourse, rather than an oral term used in living religious culture. 2 The term wu is used quite broadly from its first appearance in the earliest written sources onwards, for men as well as women, but it is unclear to me how it was used in living oral discourse. 3 By the medieval period it already tended to be used to denote any kind of ritual expert that the author disapproved of. Nowadays it seems to mean little else than that part of traditional Chinese religious culture which falls outside institutionalized Buddhist or Daoist traditions. The appearance of the term wu in a primary source can therefore not be used as an indicator of a single underlying religious practice, unless sufficient additional information is provided in our sources. It is this additional information which counts, rather than the label. More general, we have to keep in mind that our primary sources are in fact rarely primary, but almost always composed in a stylized written language ('Classical Chinese') that is far removed from the Mandarin (guanhua 官話) spoken by the male elite in so far as they were involved in government, and even further from the many different vernaculars that most people spoke locally most of the time (including elites). Whether we call these vernaculars languages or dialects is a moot point, for they were certainly mutually unintelligible. In other words, the term wu is always a translation from a (usually lost) local term, in which written traditions would have strongly influenced the translation process.
The terms shaman and medium in this investigation will be restricted to specialists of direct verbal and visual communication by and with the supernatural world. In terms of the culture itself, the 4 Lewis, Ecstatic Religion. 5 Eliade, Shamanism. medium is possessed by a being from this other world, on a regular and more or less controlled basis (which is then qualified as divine possession), rather than an incidental basis (which is then dismissed as demonic possession). It is the deity who speaks, usually in a form of messages or even through dialogic interaction, rather than lengthy narrative. A classical study in this field is by Ioan Lewis, which also provided inspiration for my own approach to the topic. 4 Evidence of the activities of mediums in traditional China is very rich.
The shaman remains in control of his or her soul and somehow travels to, or directly communicates with, this other world by means of trance. The shaman him-or herself is the voice who speaks and narrates. Here there is a strong influence of ongoing work on religious practices among Siberian and Northern American peoples. A classical study, if now perhaps outdated, is that by Mircea Eliade. 5 Evidence of this type of religious interaction does exist, but is much harder to find, although early Maoshan Daoism seems a good candidate.
It is clear that from our own world view both types of specialist are actively engaged in the construction of their communicative realities, but the fact that they as well as their audience construct their role so differently suggests that this difference must be significant to them. I can easily imagine that a single specialist might play both roles, but I am presently unaware of explicit historical evidence to this effect. Here we have a common problem of sources in the study of Chinese religious culture, where we can easily build typologies of religious practices, but do not know what concrete set of practices was followed by any specific individual, whether as a customer or as a practitioner. Both types of specialist would often work together with other ritual specialists, who used them to gain information about the issues at hand (such as demonic possession or family misfortune) and then carried out rituals in order to remedy the problem. In our modern terms, we could say that they constructed an elaborate narrative to give meaning to the problem and its social context, in a similar way as a psychoanalyst does today. 6 Pan, 'A la recherche de l'unité'. 7 Gan, Soushen ji 5, 57. On this cult see Lin, 'The Cult of Jiang Ziwen'.

Shamans and Mediums
When we no longer fixate on the term wu, we may obtain a more consistent analysis, but also face new problems in finding sources. In his recent dissertation François Pan (Pan Junliang) has circumvented this problem by focusing on the terminology that is used to describe possession in the so-called Period of Disunion, rather than the overall labels for the religious specialist in question. As it turns out, the sources use a rich vocabulary, such as the terms jiang 降 and xia 下 that both mean 'to descend [in]'; terms such as ping 慿, zhuo 著, and fu 附 that refer to the demon or deity 'attaching' itself to someone; or chi 持 'to hold', which expresses the same idea more strongly. 6 Terms such as jiang also appear in sources from the Han and earlier periods.
Take for instance the famous cult of Jiang Ziwen 蔣子文, who communicated through dreams and mediums. In the narrative that describes how his cult started from his death and subsequent appearance to the living, this is described as follows: [Jiang Zi]wen also descended (xia 下) in a medium (wuzhu 巫祝) [and said]: 'I will provide assistance to the Sun-family on a great scale, you ought to erect a shrine for me. If not, I will cause insects to enter into human ears and cause disaster.' Suddenly, [there were] small insects like deer flies (or stouts), when they entered [people's] ears, [the victims] all died. The doctors could not heal it. The people were increasingly afraid. Lord Sun 孫 did not believe it. [Jiang Ziwen] descended (xia 下) again in a medium (wuzhu 巫祝) [and said]: 'If you do not worship me, I will again cause disaster with a big fire.' 7 The crucial terms here are 'descend' and 'mediums'. We have no reliable evidence on the vernacular terms that were used at the time, although it is by no means impossible that the notion of descending was expressed in the form 'coming down' (xialai 下來) or a similar equivalent. 8 Gan, Soushen ji 5, 60. 9 There is the famous story of Ximen Bao 西門豹 from around 400 BCE who punished ritual specialists (wu 巫, but no mention of possession or soul travel) who sacrificed virgin women to the Lord of the River, see Sima Qian, Shiji 126: 3211-3212. In its present form this version of the narrative is suspect, since it introduces Han local functionaries, cash coins (wen 文, a Qin-Han innovation, rather than the then somewhat current bi 幣 in the shape of a variety of objects) and far too much oral dialogue to be a credible archival record. While this figure is documented in earlier sources, this particular anecdote is not. For a more traditional interpretations, see Lin, 'The image and status of shamans', 397-458, especially 417-19; and Lai, 'Looking for Mr. Ho Po'. 10 Although I usually translate gui as demon(s), the term 'ghost(s) ' reflects In another story featuring the same deity, we receive a very different indicator of the presence of the deity in the female medium, who has unknowingly encountered the deity on her way to his temple and received various favours from him: The fragrant smell of [Wu] Wangzi 吳望子 could be smelled several Chinese miles away. She had much divine efficacy and the entire county served her. After three years, Wangzi [who had started her career at sixteen Chinese years of age] suddenly developed an interest in another (lit. 'the outside') and the deity thereupon broke off his visits (wanglai 往來). 8 Here the notion of possession appears to be expressed by the term 'visits'. Clearly, once the young girl had become sexually interested, the deity could no longer monopolize her attention. The young age of mediums is probably also indicative of virginity in these materials. 9 Such mediums have continued to function in different forms until the present day, although it is difficult at present to study them through fieldwork due to government repression and the need for building long term trust that is difficult for outsider observers.
There was also a very different type of communication, which involved a very different world of beings. Here the specialist remained her-or himself and became a 'seer of ghosts' (jiangui 見鬼). 10 This better the fact that we are dealing here with the ghosts of deceased relatives. An excellent study is Sun, 'Youming zhi jian '. 11 Li,Taiping guangji, type of specialist is different from the medium and often denoted simply by his or her ability of 'seeing ghosts'. The term wu may be used, as it is in the following example, but is much more frequently left out completely.
In the youth of Su Ting 蘇頲 [formerly] Chancellor of the Tang someone read his appearance (xiang 相, the standard term for the technique of physiognomy): 'You will reach the position of Chancellor and end in the second degree (pin). Later when he reached the position of Chancellor and the third degree, he became extremely ill. He called a shaman (wuxi 巫覡) to look at it (shi 視). The shaman (wu 巫) said: 'The life of his lordship is about to end. You cannot rise again.' For this reason Ting brought up the words of the physiognomist again and the shaman said: 'Initially his lordship was [predicted to be] like this. When you were in charge of Guizhou examinations you had two people killed and now these two persons have accused you in the underworld. The person in charge has decreased your lifespan by two years. For this reason you will not get to the second degree.' Ting had once been in charge of Guizhou and there had been two clerks who accused a county magistrate. Ting had had the clerks killed on behalf of the magistrate. Thereupon he sighed deeply and died. 11 The use here of the unusual term wuxi suggests that the original oral account has been translated using assumptions from canonical sources. The term 'seeing' (shi) and the absence of any reference to possession indicate clearly that we are dealing with a very different kind of ritual specialist, who specializes in 'seeing' what happens after death. Typically, a medium does not visit the underworld or report on it, but allows the customer to communicate directly with the divine world. Thus, these are very distinct spheres of activity. In the following account from a Buddhist source, we again en-12 I am still puzzled by the term yi 易 here. 13 T no. 2059, 3: 339. counter a 'seer', but this time he or she sees unidentified ghosts or demons, rather than the ghosts of deceased relatives. Clearly, the monk himself had no dealings with the seer, but the household as a whole was convinced by his powers because of the initially obscure information that the seer had overheard from local demons. When the advent of the monks also definitively scared the demons away and Ms Liu was healed as a result, this then triggered the conversion of the household. The account stems from the Biographies of Eminent Monks (Gaoseng zhuan 高僧傳), showing how its compiler fully accepted the ability of a seer to observe ghosts (or demons). Without this prior event, in fact, the household of Ms Liu would never have been convinced of the monk's charisma in the first place.
The belief in seers did not disappear in subsequent centuries. Whereas in the above cases the sex of the seer is unclear, in later times 14 Lewis,Ecstatic Religion. 15 Hong, Yijianzhi, ding 12: 639. the role seems to be played predominantly by older women, suggesting a lowly status. 14 In the following example from the twelfth century, the second wife of an official, Ms. Zhang 張, has taken seriously ill and the doctors are unable to help her. The events take place in Changsha in modern Hubei.
His (i.e. the official's) mother was extremely worried about it and invited an old female shaman (wuao 巫媼) to inspect and see. She said that Ms. Wang 王 (the deceased first wife of the husband of the diseased woman) was standing in front of them and was haunting very seriously. She ordered to set up a [spirit-]tablet to say prayers and expel [Ms. Wang]. They promised [the dead woman] a jiao ritual as penance (jiaochan 醮懺), but she did not want to leave. The shaman (wu) said [her husband] should split up with her. He said: 'You must write a letter of separation in imitation of the divorce between wife and husband among the living, and give it to her. Then you can get a separation from her.' [Her husband] was unwilling to split up from her. When [Ms] Zhang became more seriously distressed by the day, [her husband] had no choice but to take up a writing brush in tears. He wrote the letter and handed it to the shaman (wu). [The latter] mixed it with paper money and burned it for [Ms. Wang]. The shaman (wu) said: 'The woman was holding the letter, and after she had unfolded it and read it, she wept and left.' [Ms.] Zhang indeed got better. 15 Hong Mai writes without any pejorative judgment. He concludes by informing us that Ms. Zhang was also related to one of his own concubines (yushi 予室), indicating that at the time of writing seeing ghosts was still an acceptable practice in elite circles. On the other hand, all the principal actors were women, and the husband in question clearly was hesitant in complying with the required ritual divorce. The religious specialist is evidently a female seer and there is again no mention of possession. The ritual that is being promised could be either Daoist or Buddhist, given the combined reference to the jiao (which is Daoist) and the penance (which is usually Buddhist).
Of course, the fact that the custom of a ghost seer was accepted did not mean that there was not also criticism of it. Thus a Buddhist monk from the northern Song lists the custom among a range of bad religious practices, 'there are also shamans (wuxi 巫覡) ….who see (jian 見) people's ancestors and parents, and can tell their forms and speech [as if] still alive.' 16 Sadly, he is not more informative, but this would suggest that the custom was widespread, but also that Buddhist monks did not approve of it. We will return to the issue of Buddhist interaction with this custom further below.
That the custom as such continued to exist whatever educated elites or Buddhist monks thought of it is abundantly clear, for instance from the following late eighteenth century account collected by Ji Yun: The elder sister of Ms. Tian 田 said that in one tenant household in Zhao 趙 Village husband and wife were very taken with each other. One morning the wife heard vaguely that her husband had an affair with someone else, but she was not certain about it. The wife was very gentle and sweet, and not very excitable either. She only said jokingly to her husband: 'If you do not love me, but love her, I will hang myself.' The next day, he was eating his meal in the fields when he encountered a shaman (wu 巫) who could see ghosts (neng shigui 能 視鬼). When [the shaman] saw him he (or she) was startled, and said: 'Behind you are two ghosts who have hung themselves, why is that?' Then he realized that those few words said [by his wife] in jest had already been heard by a ghost. Those who have died a violent death must always seek a replacement. … 17 Several beliefs come together here. It was thought that the ghost of someone who had killed herself (or himself) always tried to find some- one to replace her (or him). It is implied here that such a ghost had used the joke by the tenant's wife and then somehow encouraged her to commit suicide prematurely. For us, the relevance is of course the ability of a 'shaman' to see ghosts and discover what had happened before he had had a chance to return home, where his wife had hanged herself.
At present there is not much fieldwork on either shamans or mediums, as defined above, but they have not disappeared. Many who study Chinese religious culture will be familiar with the auto-mutilating mediums of Taiwan and the specific practice of spirit-writing, 18 but there are also still animal mediums (especially the fox immortals) active in northern China (including the north-eastern provinces) and other types of mediums elsewhere. 19 Elderly seer ladies are described briefly by Jack Potter on the basis of his 1960s fieldwork in the New Territories in Hong Kong, reflecting Cantonese practices. 20 J. J. M. de Groot described both mediums, spirit writing and elderly seer ladies for late nineteenth century southern Fujian, on the basis of his experiences in the Amoy (Xiamen) region. 21 Having established the ubiquitous nature of these two forms of contact with a supernatural world, we can return to the issue of the relationship between Buddhist practitioners and shamans or mediums.

Seeing Ghosts and Speaking With Gods
The first Buddhist monks were confronted with a sacrificial culture that conflicted directly with their own beliefs about rejecting killing 22 Campany, Strange Writing, 302 note 79, and 330 note 153. 23 For a strictly historical biography, see Forte, The Hostage An Shigao. Brief discussion of this purported miracle in op.cit., 86-87. The cult at Lake Gongting is discussed extensively by Miyakawa, 'Local Cults around Mount Lu'. 24 Perhaps prognostication blocks. 25 Gaoseng zhuan, 1: 323. A different take on the story of the divine bamboo is given in the Mingbaoji 冥報記, see Gjertson, 'A study and translation of the Ming-pao chi', 264. and alcoholic drinks. This required them to take a stance at every step of the way. As history has shown, the sacrifice of freshly killed animals and liquor remained standard ritual practice throughout the imperial period and after. This is, however, not how they themselves represented their struggles about proper ritual. Numerous accounts preserved in the Biographies of Eminent Monks tell of the well-known battle between the Buddhist newcomers and established local cults, with a crucial role for shamans and mediums in telling the Buddhist side of the story. 22 One of these accounts tells of a foreign monk with the Chinese personal name Gao who travelled past one of the best known cults of his time, located on an important transport node. Indeed, this was not just any monk, but the famous An Shigao 安士 高 (fl. 147-170). 23 On his journey he reached the temple at Lake Gongting. This temple possessed charismatic power since ancient times. Merchants and travellers prayed here so that they would go with the wind upstream or downstream without any obstructions. Once someone requested the deity's bamboo. 24 When [the deity] did not allow it, he took it nonetheless. His boat then capsized and he drowned. The bamboo returned to its original location. From this moment onwards, the boatmen were full of respect and fear for it, and all were in awe of its shadow. Gao travelled together with over 30 boats, which offered life sacrifices and asked for good fortune. The deity thereupon descended in a medium (jiangzhu 降祝) and said: 'On the boats there is a śramaṇa, you must call him to ascend [the shore]. The traders were really startled and frightened, and invited Gao to enter the temple. A lengthy conversation ensues, in which the deity explains in great detail that he had known the monk in a previous existence, but had been punished with this rebirth as a local deity because of his angry temperament. In order to make up for all of his misdeeds and in anticipation of his fall into hell, he wished to donate his divine wealth and this location to the monk. The monk then asked him to reveal his true form, which turned out to be a big snake. The monk built his monastery and eventually a young man was seen, who was the former deity released from his evil form.
For our discussion, what matters here is that this entire conversation between the monk and the deity was carried out through a medium (who is labelled a zhu 祝), as is apparent from the term 'to descend' (jiang 降) and the two phrases 'the deity told Gao '(shen gao gao yue 神告高曰) and 'the deity said' (shen yue 神曰). That the narrative was embellished over the years or maybe even largely invented does not alter the basic fact that the Buddhist conversion of the site was seen to depend on this form of direct divine communication. While the cult itself was deemed to be evil because of its sacrificial practices, though still very real, no disagreement existed on the validity of this form of communication through possession of a medium. Biographical sources on Buddhist monks contain a number of similar examples, in which deities communicated with them by possessing mediums. 26 Some monks in this early period still operated on the border between their own expertise and that of a seer of demons or ghosts. In the following case a monk becomes a seer as well as an exorcist specialist himself: The monk Baotong 寶通, for whom we do not know his ancestry, was a man from Gaoqiu. He lacked in further knowledge or insight and only recited some spells from the seventh chapter of the Lotus 27 These lines and the following lines of dialogue are essentially in spoken Chinese. 28  [Baotong] also asked the deity: 'Is that deity of the Bailu 白露 village also involved in this?' The deity said: 'He is also involved in this.' Baotong said: 'In that village is a woman of the Hao 郝 family. She is possessed by the deity of that village. Can we make her get better or not?' The deity said: 'This can be obtained.' Then [Baotong] sent someone to summon that deity of the Bailu Village and also punished him with a hundred beatings and the woman of the Hao family thereupon got better. Later the old illness of the wife of the Zhao family flared up again. She would sing the entire day long. Her husband told Baotong and he went to them. He then saw (jian 見) the deity who had been punished previously standing next to the ill person. Baotong spoke to her: 'Previously I said you had to leave for ever. How could you return today? I will now recite a spell.' Then the deity kowtowed, begging that he would be allowed to return and not suffer a spell. The deity thereupon left to a faraway place and thereupon she got better forever. 28 This account is fascinating on several levels. The liveliness of the dialogues does not necessarily suggest that this is how it was transmitted, 29 The similarities with later Daoist vernacular ritual are striking and should be investigated more systematically.
for the final editor of our story most likely improved the narrative by adding dialogue. Despite this likely editing, this will still be a plausible story in terms of the religious practices that are described showing that at the time a Buddhist monk could also operate as a seer himself. In addition, he functioned more or less as a ritual specialist, akin to vernacular (Daoist) specialist of later items, down to the use of subordinate officials to summon the deities. His ritual functioned therapeutically, but there is no mention of a conversion. Instead the deities are punished and dismissed. 29 It was not always clear how to distinguish an unwanted demonic possession from regular Buddhist visions. The following example from the Period of Disunion shows how the local 'shamans' initially interpreted the young girl's 'meditation' as possession. Indeed, if she had not eventually become a Buddhist nun, she might have been qualified as suffering from demonic possession and that would have been the end of it.
The Song nun Tanhui 曇輝 came from Chengdu in Sichuan. Her family name was Qingyang 青陽 and her personal name was Baiyu 白玉. When she was seven Chinese years of age, she enjoyed sitting in meditation. Each time she would obtain a vision of a realm. She did not quite understand it and said that it was just a dream.
This account was of course written retrospectively and the language of this initial event already betrays a Buddhist interpretation. This is true both for the term 'sitting in meditation' and the curious term 'realm' (jingjie 境界), most likely here an abbreviated reference to the Pure Land or some other ideal world.
Once she slept together with her older sister, when she entered into meditation in the night. Her older sister found her at the corner of the chamber screen, her body stiff as wood or stone and no longer breathing. Her sister was frightened a lot and rallied the rest of the family. Together they held her in their arms, but in the morning she still had not woken up. They rushed off to ask the shamans (wuxi 巫 覡), who all said that she was possessed by gods or demons. When she became eleven, the foreign meditation master Kālayaśas 畺良耶 舍 (383-442) came to Sichuan. Tanhui inquired about her visions. Kālayaśas considered that the meditational practice of Tanhui was already predestined, and wished to exhort and transform her, and to make her leave her family. 30 I have left out the next stage in the narrative, which is described in similar detail. The family attempted to marry her out, but she eloped to a monastery with the help of a local monk. While her family did not approve, with the support of the local magistrate she managed to stay in the Buddhist monastery. In this story the events are told from a Buddhist perspective, but even then the 'shamans' are merely deemed to be misinformed and there is no fundamental theological doubt about their abilities. The vision of another realm could be interpreted as a shamanic or mediumistic one, as well as a Buddhist one, showing how the basic experience was the same, but the ways of labelling them were different.
The existence of visions during meditation practice is well-attested, but I would argue that we also need to understand them in terms of the widespread belief in seeing ghosts and demons, whether a sign of demonic illness or prophetic warning, or alternatively an extraordinary ability that could be drawn upon to help others. This sentiment was then transferred to Buddhist religious culture. The question that remains, but need not be answered here, is whether we can recover elements of this culture of seeing ghosts and demons in actual Buddhist visions in more detail as they were shaped and reshaped in Chinese traditional culture.

Enlisting Seers Within Buddhist Culture
In the previous examples, the use of seers and mediums served to further the Buddhist teachings-or were post facto constructed as such, whether directly by facilitating the conversion or reworking of local cults, and/or indirectly by demonstrating the superiority of Buddhist beliefs and practices. Here I want to look at some examples in which Buddhist believers made use of these specialists for their own religious purposes, rather than as a means of convincing others from a different religious paradigm. In the following instance from the early seventh century, a man who had returned from the dead has witnessed several local monks in the underworld, who discussed how two other monks were about to die. This actually happened in that very year, causing some commotion.
Later on the monks of the monastery made a shaman (wuzhe 巫 者) go to the old rooms of Hongliang 弘亮 c.s. to summon the two monks and question them. The comment ascribed to Hongliang is interesting in itself, as a reflection on the difficulties of sociability in a monastic community. The person labelled as a 'shaman' is not possessed, but operates as a seer, by means of his enactment of dialogues with the dead. He seems to be well-informed, which allows him to play a crucial function in the communal life of the monastery. By this time, Buddhism had 32 Davis, Society and the Supernatural, 126-51 (including a discussion and translation of the present example) and Xie, 'Mifa, daoshu yu tongzi'. long become the dominant religious culture of its time and this coexistence of practices can no longer be explained away from the need of Buddhism (or rather Buddhist practitioners) to adapt itself to Chinese religious practices. For this particular monastic community, seers of demons and ghosts continued to be a normal part of their religious environment. People might witness events in the underworld because they nearly died and then came back to life, as was the case in the above account, but specialists were still needed to ask more specific follow-up questions to the dead.
In the following account from the twelfth century, the ritual of the Ucchuṣma vajra/Huiji jin'gang fa 穢跡金剛法 takes central place. The ritual practice is well-documented in the secondary literature, and here we will only give one example which is dated to 1151. 32 According to Hong Mai, the ritual was widely practiced in what is now southern Fujian, using mediums to communicate with the gods.
The monk Ruochong 若衝 was living in the Guangfu 廣福 Hall in the Western Hills of Quanzhou. In the night there was a monk who requested to see him. Ruochong was surprised about the untimeliness of his [request]. The monk said: 'I am very poor. Apart from my clothes and begging bowl, I only have several ounces of silver. They were stolen by someone. I happened to invite a Person of Way to carry out the ritual [of the Ucchuṣma Vajra] and the deity said: "The abbot must come and then I will say it." I hope that you will briefly come over.' Ruochong went with him to his room.
Ruochong must have been the abbot of the monastery and therefore had the ultimate responsibility for all goings-on. The boy was of course possessed by the deity who now started to speak. The Person of the Way could of course be a Daoist priest, but also a lay Buddhist or another type of ritual specialist. I have therefore left my translation here somewhat vague. The text now continues as follows: A village boy leaning on a sword was standing on a chair. When he [the village boy] saw Ruochong he greeted him, and said: 'Do sit down. Deep in the night, I should not have bothered you.' Ruochong said: 'I did not know that the venerable deity would come down here. I was remiss in burning incense. May I ask why it is that you wish to see Ruochong.' He said: 'I am a high deity from Heaven. Because things were lost in the monastery, I need you to provide testimony. This affair is very easy to know, but I am afraid of causing legal disputes, which would go against my intentions. If you will indeed not put an accusation before the magistrate, I will search it for him.' Ruochong apologized himself two, three times: 'I will respectfully follow your injunctions.' The deity said: 'I will start performing the ritual.' He rested on the sword and went out. Now a more theatrical part of the ritual follows, in which the boy who is possessed by the deity enacts a successful search for the lost silver.
He jumped and he walked, suddenly he threw himself in a deep well. After a long period he jumped out again. He hurried straight to a heap of cow dung outside the monastery gate, and jumped and hurled around it. He stuck his sword in it three times. Suddenly he passed out on the ground. After a while the boy came to. When they asked him about it, he remembered nothing. Thereupon they lifted the cow dung up and saw that one brick stuck out irregularly. When they lifted it, the silver was underneath. It was said that it had been hidden by the thief. 33 What matters here is not so much the activities of the possessed medium, but the totally matter of fact way in which the monks interacted with the deity. They fully accepted the reality of the divine presence within their monastery and even showed him their respect, though the ritual event as such did not really serve any Buddhist doctrinal purpose. We usually see this level of acceptance as a Song development, but the material discussed before suggests that this type of reconciliation had always been there. 34 A reference to the Analects (lunyu 論語) ascribed to Confucius, see Confucius,The Analects, Zhuhong 袾宏, Zhuchuang suibi 竹窗隨筆, in: Zhuhong, Yunqi fahui, 12: 0025b29.
The prominent late Ming monk Zhuhong 祩宏 (1535-1615) has preserved another example for us, in which a deceased lay practitioner in Hangzhou possesses a medium (fu yi tongzi 附一童子) to tell his next of kin that he had practiced his fasts and injunctions well enough to be comfortably off in the underworld, but had not been reborn in the Pure Land.
One day he berated his wife and son, saying: 'Why did you kill a rooster and prepare some millet 34 in the rituals for my grave? Now I have a clerk following me and it no longer quite resembles my previous freedom [in the underworld] anymore.' The wife of his son was pregnant and they asked him about it. Then he said: 'You will give birth to one son without harm. If you exceed this and give birth to a[nother] son, mother and son will both die.' I (Zhuhong) made a respectful note of this, to wait whether this would come true or not. Soon after this she gave birth to a son. Then she was pregnant again, and again gave birth to a son. The boy died subsequently and the mother died as well. Then we knew that none of [the medium's] words had been wrong. Given this, how could a filial son kill living beings on behalf of his parents? 35 The fact that Zhuhong made a mental note to check whether the medium was right or not suggests some skepticism, but ultimately he accepted the statements, not in the least because they fitted his Buddhist worldview. He used the incident to advocate consistently vegetarian sacrificial practice for the deceased as much as for the living. To us as outside observers, the story also tells us how strong ritual requirements were. Not sacrificing meat was a difficult road to follow as next of kin, who not only had to deal with the wishes of the deceased, but also the social pressures of the people living around them. 36 Gjertson, 'A study and translation of the Ming-pao chi'. 37 Wang, Mike Amituo jing zhonglun, (X no. 0401, 22: 017a20-0172b14). For more details and background, see my discussion in ter Haar,Guan Yu,

Alternative Forms of Communication
A lot of religious experiences within Buddhist religious culture are quite akin to the shamanic journey, such as visions of the Pure Land or visits to the underworld. While these visions would count as 'Buddhist' and the visits perhaps not, it seems to me that both visions of a better realm and visits to the underworld served the same purpose as shamanic travel and mediumnistic conversations in providing assistance in dealing with the past and making decisions about the future. Indeed, I suspect that in narrative terms the visions and visits would be heavily indebted to these earlier types of communication, but proving this would exceed the boundaries of this essayistic contribution.
There is an abundance of stories of visits to the underworld by devout Buddhist believers or people who become believers afterwards. This material goes back to the early days of Buddhism in China. 36 The difference with the mediation by shamans and mediums is, of course, that in these stories the principal agent in the vision or visit is also the one who interprets, possibly together with family or friends, and the one who draws conclusions for his or her own life. I end this contribution with an example from the cult of Lord Guan, but again with a strong Buddhist context. The story is told by Wang Dingzhu 王定柱 (1761-1830), the great-grandfather of a late Qing and early Republican lay Buddhist, Wang Gengxin 王耕心. Wang Gengxin and his brother were prominent lay believers in Taizhou in Jiangsu in this period. We have a long account by Wang Dingzhu himself, which tells us how his family had become followers of the Buddhist way in the second half of the eighteenth century. I quote extensively from it elsewhere and here I will only discuss the crucial moment in which he has a vision of the deity and then reconverts to Buddhism. 37 Dingzhu came from a classicist (ru 儒) household, which traditionally also worshipped Lord Guan (Guan Gong 關 公), but with considerable room for Buddhist beliefs and practices such as the worship of Guanyin. Especially on the female side of the family, what we might call Buddhist devotion was kept alive. Wen Dingzhu became older and started to study seriously for the civil service examinations he lost his faith. In 1786 a devastating epidemic occurred in his county, and he himself became dangerously ill as well.
Corpses on the road were as numerous as sesame seeds, all of them dying after seven days [with high fever] without sweating. By the 21st day [of my disease] I had suffered three periods of shaking and not sweating. Suddenly I blew out my last breath. My soul (shen 神) passed through the rooms and doors out into the street. I was on my way to the Shrine of the Marchmount in the eastern suburbs (i.e. the Temple of the Eastern Marchmount, associated with sentencing and punishing the dead), when I saw a cloud of spears and harnesses approach from the east, the men and horses were all roughly one fathom large. (… ) Suddenly I saw the divine martial one (i.e. Lord Guan) with his bow and arrows riding a red horse, holding his dragon halberd. His divine countenance was shining brightly. On the street I paid my respects in welcome, and pressed my forehead in the dust. When I glanced up, his official robes were as good as complete. The divine martial one was in his court robe and wore his court hat, with his gui-tablet in his hands. Looking sideways he had a spare moment, and in tears I set out before him what I suffered from. I begged for his protection. The divine martial one sharply scolded me: 'Do you know your sin? Because of the accumulated virtue of previous generations it had been recorded in the registers that four years after succeeding for the prefectural exams you would ascend to the degree of Presented Scholar in 1781. Your good fortune and emoluments would not have stopped there. But who could have told that after obtaining your degree you would be without discernment in your youthful blindness, forgetting the merit of previous generations and bragging of your individual talent. (…) In addition you have not paid attention to the writings of [Buddhism and Daoism]. … The travelling envoy made countless daily reports, as a result of which the Jade Emperor shook with rage. He first erased your name from the exams and now erased you from the records of long life. You will fall into an evil rebirth, so how could you become the recipient of protection by me?' Lord Guan is referring to the merit accumulated by generations of family worship of Guanyin; he is blaming the young man for leaving the ritual practices of his ancestors behind, by blindly following the words of the classicists. In his early youth our protagonist had been thoroughly imbued with Buddhist worship by his mother and this faith now came back to him at a moment of crisis. He returns to his old faith and is healed.
Similar to visions and underworld visits, dreams could also fulfill the function of communication with supernatural world. In the following account the dream functions remarkably like a conversation through a medium, providing confirmation about someone's next level of existence. From a Buddhist perspective it provided reassurances to other believers in the Pure Land, and from the perspective of the grandchildren in this particular story it provided emotional certainty that their grandmother was now okay.
Ms. Jin 金 [the wife of] Zhang Taiyi 張太宜 was a person from Jinzhou. … Her family had been prominent for generations, but she was very simple in her personal life. She lost her partner when she was middle aged. In teaching her sons she had moral standards. Her sons Zhengdao 正道 and Zhengxue 正學 both became successful in the examinations.
[She] received texts on the Pure Land in her late years and read them. Thereupon she fixed her mind on the [Land of] Great Joy. In the morning and evening she paid obeisance and recited. One evening she warned her grandchildren. 'You must read the letters of your grandfather well. I will soon be gone.' She called a servant girl to light incense, sat down in meditation posture and passed away. Several days later, she showed a dream to her grandchildren, saying: 'I have just come from the Western Paradise.' Only then did they know that [she] had really been reborn in the Pure Land. 38 38 Peng,Jingtu shengxianlu,9: 29b (349), quoting from the late Ming collected works of the prominent literatus and lay Buddhist Yuan Zhongdao 袁宗道. This and other Buddhist sources are full of similar stories.
The difference with shamanic journeys and medium conversations is that visions and dreams did not offer quite the same possibilities of interaction, but on an individual level they provided the same kind of insight in the other dimensions of existence that a Buddhist believer (and most others as well) wanted to get assurances from and about.

Concluding Comments
In this brief investigation I have only touched the surface of a rich culture of communication in which Buddhism was as involved as were other doctrinal and ritual approaches. Studying Buddhism, however we define it, in traditional China and indeed to this very day without this larger social and religious context would be to reduce it considerably in its religious import. While early Buddhist monks attempted to fight local cults, they did so with the assistance-or at least the claim of assistance-of local mediums. In order to know about the fate of the dead, they consulted another kind of specialists, whom we might call shamans. I see at least two useful ways of looking at the evidence. On the one hand, we have long since accepted that Buddhism is not really a single religious culture separate from everything else, but in actual practice we still tend to study it in isolation independent of the rest of China's rich religious culture. This has to do partly with our way of specialisation, which focuses on either Buddhism or Daoism or local religious culture. Or, as in my own case, social history. The study of local religious culture tends to be practiced outside the religious studies departments, since it lacks the kind of theology that appeals to colleagues in those departments. In reality, everybody, monk, nun or lay believer, wanted to communicate with the supernatural world of the divine, the demonic and the dead. Or maybe we should say, they had to communicate, since they may not have felt that there was a choice. They did so through people whom we might call mediums and shamans or seers, or alternatively they accomplished it through dreams and visions of their own.
In preparing these comments, two more historically minded questions arose that I can only ask, but not answer at this point. One field of inquiry might be to compare the ways in which shamans, mediums and their ways of communicating with other dimensions of being may have influenced specific Buddhist practices, especially in the field of dreams and visions. In the same vein, one can imagine that dreams and visions in some ways replaced these older forms of communication and certainly competed with them. After all, in order to prove their case, Buddhist monks had to collect witness accounts and what better than statements by the established specialists in communication with the supernatural world.
Another field of inquiry would be to study the interaction between different types of religious specialist more carefully than has been done in this investigation. In another article published some years ago I already suggest the notion of Buddhist practices as an option and this concept could be elaborated to see how shamans and mediums fitted in this larger picture. 39 A preliminary hypothesis based on the sources I have seen until now would be that interaction between Buddhist specialists, mediums and shamans decreased over time. Part of the reason might be that there is a general downward social mobility of these alternative specialists, but another hypothesis would be that all types of specialists eventually acquired their own niches in local society-albeit in different niches in different regions.