Shamanism, Eroticism, and Death: The Ritual Structures of the Nine Songs in Comparative Context

Abstract

Eroticism is an independent feature of human life that influences many areas of experience, including death and religion. While eroticism has received a good deal of scholarly attention in religious studies, the present study takes the Nine Songs as the starting point for a discussion of eroticism as a frequent element in the world of shamanism. These songs provide the earliest linguistic corpus in East Asia that allows us a glimpse into the world of the shaman, and they constitute one of the earliest sources of this type preserved anywhere in the world by giving depictions of eroticized gender relations between shamans and spirits. This study comparatively situates the ritual structures expressed in the Nine Songs to uncover deeper affinities between shamanism, eroticism, violence, and death.

1. Situating Agricultural Shamanism

This article attempts to contribute to the comparative study of shamanism, a phenomenon here defined as cultural representations1 of direct communication between a human and a spirit in a séance event for the benefit of the community. It recognizes several different forms of shamanism, but primarily explores agricultural shamanism by providing an outline of its fundamental ritual structures with reference to the Nine Songs (Jiu ge 九歌) from the Chu ci 楚辭.2

Because the erotic is only sometimes thematized in shamanism studies, one can approach it either by taking account of the records and evidences of its various cultural representations, or by assessing scholarly studies that gives it no recognition and declare them methodologically deficient. To embrace the latter approach is to misrepresent shamanic eroticism as a sui generis component of all shamanic institutions by confusing it with sexuality, as does, for example, Leo Sternberg (1925). However, although many shamanic institutions give little or no place to the erotic, this does not mean that existing scholarly work adequately manages to deal with it where it is functionally operative, as it is most conspicuously in agricultural shamanism.

Eroticism is one strategy deployed by shamans to attract spirits. Séance events in which displays of shamanic eroticism are minimal to non-existent primarily belong to pastoral shamanism (see Caroline Humphrey (1996) for the case of Mongolian pastoralism), as well as to bureaucratic shamanism (see Chongho Kim (2003) for the Korean case and Axel Morten Pedersen (2011) and Manduhai Buyandelger (2013) for the urban Mongolian one). Eroticism is centrally employed in hunting shamanism (see Roberte Hamayon (1990) for the hunter regions of Mongolia and Siberia), and in agricultural shamanism (see Nguyen Thi Hien (2015) for the Vietnamese case and Ellison Banks Findly (2016) for the Laos case). Despite surface similarities, the eroticism in these last two forms function according to very different central symbolic systems.

More essential than discerning the presence or absence of eroticism in séance events, and more important than the distinctions between shamanism and possession, is the legitimizing spiritual relationship between the shaman and the spirit: the shaman’s legitimacy as such always emerges from a particular type of relationship to a particular type of spirit in a particular type of economy. For the hunting shamanism of egalitarian subsistence economies, legitimacy emerges from the success of the shaman in forming a long-term marriage alliance with the spirit of game. For the pastoral shamanism of nomadic herding economies, it emerges in forming a filial relationship with the lineage ancestors. For the bureaucratic shamanism of centralized administrative economies, it emerges in forming a relationship with either the spirits of the dead with whom the shaman is not necessarily related, or nature spirits with whom there is no question of lineage relation. For these last two, it is as difficult to imagine eroticism employed to attract and seduce the spirits of one’s lineage as it is to imagine it employed for state rituals.3

Granted that the following claim rings as tautological, it is worth spelling out the sociological and economic correlates at issue: agricultural shamanism is most prominent, where it exists or once existed, only in agricultural economies.4 Its legitimacy emerges from the success of the shaman in forming a temporary love affair with nature spirits, but never with the spirits of game or ancestral spirits. Its measure of assessment concerns the shaman’s success either in crossing the boundaries separating humans and spirits from here to there in spirit-journeys, or in his/her success in enticing the spirit to cross the boundary from there to here in spirit-possessions. These relationships can be consummated or frustrated, but the success of the séance does not depend on it.5

This present work builds on previous studies of the Nine Songs that introduced the shamanic reading to a Western audience.6 This is a highly contentious reading, as demonstrated by the continued resistance against it from many sinologists who tend to assert exclusive ownership over the Nine Songs. Modern Western sinology has consistently exhibited a deep-seated resistance to recognizing early Chinese shamanism,7 and this situation has rendered it a virtual stranger to contemporary shamanism theory, thus making it more difficult to fruitfully situate the shamanism of the Nine Songs for either historical or comparative analysis. Pursuing a shamanic reading, therefore, requires some initial groundwork to allow it to take root against the allegorical reading well-staked out by the traditional Chinese exegesis inherited by modern sinology. The present study eschews the sinological in its adoption of a history of religions approach in its exploration into shamanic eroticism and shamanic death by calling upon the ritual structures evident the Nine Songs.8 By way of this comparative methodology, the arguments raised in this study also bear consideration as reflecting on one element of gender in the heritage of China.

If there is a viable shamanic reading to the Nine Songs, then it will be seen to adhere to the logic of eroticism. Gopal Sukhu (2012, pp. 77–78) provides a first step into the conjured world of its erotic relationships between shamans and spirits:

The spirits in the Nine Songs descend, like most spirits in the Chinese ritual context, to enjoy food. But even more than food, they seek amorous contact with the shamans, who adorn themselves in various ways to make themselves attractive to the spirits. The central event in several of the Nine Songs hymns is the love affair between the spirits and the shamans.

2. Two Readings of the Nine Songs: Shamanic and Sinological

Composed in strict metrical verse, the Nine Songs consist of eleven9 independent pieces that, taken altogether, are counted as one of seventeen sections (卷 juan or “poems”10) found in extant editions of the Chu ci edited by Wang Yi 王逸 (89–158 CE).11 The earlier version of the Chu ci that he received was edited by Liu Xiang 劉向 (77–6 BCE) and consisted of thirteen sections, but the earlier edition that he received had originally been compiled by Liu An 劉安 (179–122 BCE) and consisted of nine sections. Liu Xiang catalogued his edition in his “Bielu” 別錄,12 and although he did not list their specific titles, he ascribed authorship of twenty-five parts (篇 pian), a number that includes the eleven individual songs of the Nine Songs, to Qu Yuan 屈原 (343–278 BCE), a nobleman of the state of Chu who was slandered and banished to the remote south where he drowned himself in the Luo River in demonstration of his loyalty to king and country.

A memorial by Jia Yi 賈誼 (200–169 BCE) provides the earliest reference to Qu Yuan; it remarks on his banishment to, and suicide in, the south, and it also includes a quotation from Li sao (離騷).13 It was some decades later that Liu An composed a commentary to Li sao with direct reference to Qu Yuan’s authorship, which is lost save for sections preserved in later sources. Sima Qian’s 司馬遷 (135–86 BCE) Records of the Grand Historian (史記 Shi ji) provides the first “complete” biography of Qu Yuan and, while it makes ample reference to many of the poems attributed to him together with their relevance to his life career, it nowhere mentions a connection between the Nine Songs and Qu Yuan.

Still, the Chu ci stands as arguably the earliest collection of writings with a single author attribution (as opposed to writings entitled with the name of a master, products of group construction by disciples), marking what Mark Lewis (1999, p. 186) calls the “breakthrough of authorship”:

The appearance of the proto-Chu ci under the name of Qu Yuan was a crucial step in the invention of authorship in the late Warring States or early Han. A set of themes and images, probably defined by generic conventions, was redefined as the expression of an individual’s response to his experiences. The mutual echoes and resonances of the poems that appeared when they were read together were explained by reference to a single author, and ultimately each poem was linked to a specific stage in the writer’s life. However Qu Yuan, the figure of isolation, had no disciples and was thus credited with personally composing the poems.

Building on biographical materials about Qu Yuan including those brought together by Sima Qian, Wang Yi’s commentary systematically interpreted Qu Yuan’s writings through the lens of political allegory in which he was both external author and internal protagonist. Wang Yi did this based on what he believed to be his insight into Qu Yuan’s life and authorial intent.14 His commentary was the first of many composed by Confucian literati throughout traditional China,15 and it has exerted a dominant influence in Chu ci interpretation by predetermining the biographical structures through which traditional exegesis have been carried out and which are still often maintained today.

Li sao recounts the spirit-journeys of a shaman in search of a mate among the ranks of spiritual beings. Jia Yi had already read its shamanic motifs as Qu Yuan’s emotional outpouring of his feelings of abandonment from his loved one, the king who banished him. This is the allegorical reading applied to each of Qu Yuan’s writings, with the exception of the Nine Songs, at least until Wang Yi’s commentary which stated that while Qu Yuan was sojourning among a southern barbarian tribe, he witnessed many of their religious performances and memorialized them in the Nine Songs.

Describing the experiences of shamans who induce spirits to share temporary love affairs with them, the Nine Songs are often interpreted as futile attempts that render the shamans reduced to tears at the failure of the spirits to consummate with them. Thus, the Nine Songs share with Li sao the motif of a shaman looking for his lover among the spirits. Wang Yi’s interpretation of the Nine Songs simply appropriated the mechanics of the allegorical interpretation that identifies the frustrated shaman with the banished minister and the numinous spirit with the benevolent king hoodwinked by the slanders of sycophantic ministers. Retelling Wang Yi’s account of the genesis of the Nine Songs, Zhu Xi (1987, p. 29) writes:

The Nine Songs are the work of Qu Yuan. Formerly, the custom of the people living between the Yuan and Xiang rivers in the district of Ying in the south of Chu was to believe in spirits and worship them. Their worship required the services of both male and female shamans who would make music and sing and dance to please the spirits. Among the coarse Jingzhou customs [of the shamans] were their crude lyrics. Placing themselves between yinyang and humans and spirits, they were confused and unable not to mix in disrespect and lewdness [toward the spirits]. Qu Yuan was banished there, and he witnessed their worship of the spirits and was moved. Thereupon he modified and fixed their lyrics by taking out their excesses, and he accorded with their sentiments in their service to the spirits to express his intention: “I am loyal to my lord and I love my country. I yearn for them and I will never forget them.”

This account provides a context for the genesis of the Nine Songs; however, alone among Qu Yuan’s other writings, Wang Yi does not interpret them as Qu Yuan’s personal creation: the “lyrics” (ci 詞) of the shaman songs that Qu Yuan “modified and fixed” (geng ding 更定) predated him and were originally at home in the shamans’ séance events performed in the rustic environs of country folk agricultural life.

Readers approach the Nine Songs differently. One way stems from Wang Yi who, despite recognizing for them an original environment of religious worship, subordinates that to their essence in political remonstrance. Another way is pursued by Zhu Xi who, although he admits the presence of remonstrance, subordinates that to their original uses in shamanic performance.16 Sukhu (2012, p. 13) writes:

Wang Yi thus incorporates the Nine Songs into the legend of Qu Yuan for the first time (they are not mentioned in the Sima Qian biography), interpreting some of them as literal accounts of Qu Yuan’s life in the barbarian south, others as his imaginary encounters with divinities, and all as figurative remonstrations addressed to his king.17 Zhu Xi also thought they were by Qu Yuan, but held that their remonstrative purpose was secondary to their function as shaman songs. And Zhu Xi, contrary to what one would expect, had a very deep knowledge and belief in shamanism. Unlike the interpretation of Wang Yi, Zhu Xi’s interpretation allows the hymns to be read primarily as religious artifacts.

These two approaches by Wang Yi and Zhu Xi are not entirely exclusive of each other because both recognize a religious genesis to the Nine Songs predating Qu Yuan, but they can nonetheless be taken as precursors for most modern readings. While I am tempted to oppose two of them, the sinological and the shamanic, this would be something of an oversimplification since “sinology” is an umbrella term for studies of Chinese literature, philosophy, history, and religion, among others, and should also include any shamanic readings of ancient Chinese texts. Two general types of readings that I will oppose, however, can be called the “sinological” and the “comparative,” where by “comparative” I specifically intend a “comparative shamanic” reading of the Nine Songs that I attempt to develop through the course of this study.18

The comparative shamanic reading that I pursue focuses on the ritual structures most evident in the Nine Songs that cohere around gendered erotic relations between shamans and spirits. Sinologists resist recognizing these structures as shamanism strictu sensu for a variety of reasons ranging from concerns of theoretical definition, historical interpretation, anthropological evidence, and phenomenological distinctions opposing shamanism and possession.19 Even as the Nine Songs are often heralded as providing the best available evidence for a tradition of early Chinese shamanism, scholars of early Chinese religion rarely engage them.20

Geoffrey Waters denies the shamanic hypothesis to the Nine Songs altogether, but he bases his judgment on Mircea Eliade’s (1964) influential work. Eliade made a strong separation between shamanism as ecstatic flight based on Siberian forms on one side, and possession as its degeneration on the other. Judiciously quoting from Eliade’s work, Waters (1985, pp. 74–75) writes, “What does a Chinese wu [shaman] have in common with a Siberian shaman? … Eliade dismisses this spirit possession,21 the descent of a ‘spirit,’ as an inferior form of shamanism… This is the criterion by which a shaman is to be tested. Spirit-possession is, in fact, an anti-criterion… If we accept this standard, we must insist that the wu may not be a shaman at all.” Waters (1985, p. 19) continues: “The shamans in the Chu ci are literary devices, and the Chu ci is only tangentially ethnographic.”

Waters harbored a deficient understanding of shamanism that he arrived at only through Eliade’s arm-chair conceptions of it.22 Still, most sinologists who interpret the Nine Songs accept the veneer of shamanism only to re-code it in terms of political allegory, and Gopal Sukhu toes this reading in both his monograph (Sukhu 2012) and translation (Sukhu 2017). His analyzes Li sao as strict political allegory even as he treats the Nine Songs with great trepidation, aware as he clearly is of the hazards involved with the shamanic reading; Sukhu (2012, p. 78) writes:

Clearly Qu Yuan had these hymns, and/or hymns very much like them, in mind when he wrote Li sao. The heretofore unnoticed feature that Li sao shares with several of the Nine Songs is the story frame of the descent and ascent of a spirit who has some sort of love affair with someone on earth, and this indicates a deeper kinship than could have been imagined before. Li sao now reveals itself to be a kind of Nine Songs hymn. The purpose of this new style hymn, however, is not sacred ritual but political complaint, which is to say the shamanistic surface hides (and reveals) another meaning.

23

Sukhu, however, refuses to explore the shamanism of the Nine Songs on its own terms, and the notes to his translation mainly discuss textual occurrences of the spirits mentioned therein found in other early Chinese sources; this too is a central feature of the sinological reading. Sukhu (2017, pp. 2–3) also suggests that Qu Yuan composed them for court performance, and that “once that thread [i.e., Wang Yi’s claim that they express the remonstrations of a slandered minister] is let go, one must admit that the elegant classical Chinese in which they are composed and the mention of objects such as bronze bells, associated exclusively with royalty, suggest that the hymns were meant for performance at a royal court.”24 But this begs the question as to why Qu Yuan, banished and on the verge of suicide, would compose them for this purpose in the first place.

Arthur Waley first introduced a shamanic reading of the Nine Songs to a Western audience. He attributed the strangle-hold that the allegorical reading has maintained over them to the long-standing literati habit of discounting expressions of popular religion. Affirming that the Nine Songs demonstrate an authentic tradition of early Chinese shamanism, Waley (1955, pp. 16–17) attempted to disentangle it from that allegorical reading through a history of religions approach:

As regards the time when these Songs were put into their present form, I should say that the traditional dating (fourth to third century B.C.) seems quite reasonable. But of course the prototypes on which they were founded may in some cases go back to a much earlier period. The Nine Songs owe their preservation to the fact that like other early Chinese songs they were interpreted allegorically. The shaman becomes a virtuous minister who after having for a time enjoyed the favour of his prince is discarded by him. The best-known similar case outside China is of course the Song of Songs, which would never have found its way into both the Jewish and the Christian Bibles if it had not been allegorized to meet the needs of later times. It was in this allegorical sense that the Nine Songs were understood till well into the twentieth century, although it was recognized from the second century A.D. onwards that the moral interpretation was only a sort of ultimate meaning, and that taken in their literal sense they were wu (shaman) songs.

25

David Hawkes (1985, pp. 42–51) also comparatively contextualizes this shamanism, but his approach to the Nine Songs is more firmly grounded in literary history. Speaking of “the cannibalization of a new secular, literary tradition of an earlier religious, oral one,” Hawkes (1974, p. 44) writes:

I am referring, of course, to Wang Yi’s statement that the Nine Songs are a literary recasting of traditional religious material. He makes no such statement about any other work in the corpus, and what he says about the Nine Songs is notoriously suspect. Nevertheless, I believe that this statement was an inspired and fruitful guess which might well prove the key to a better understanding of the whole collection. If we could analyze the use which Chu poets made of an existing religious tradition, we should, I believe, be well on our way to understanding the nature of poetic inspiration and the workings of poetic imagination in that remote and formative era of Chinese literary art.

Hawkes (1974, p. 44) does not have in mind the “literary recasting of traditional religious material” fit for literati consumption, as Wang Yi and Sukhu assumed, but rather a literary effort to express that “religious material” in the literary language of poets, not historians (or anthropologists). Like Sukhu, he holds that the Nine Songs are performance pieces, but they are verses of performance, not verses for performance, despite the seeming contradiction in the first sentence of the following quotation where he (Hawkes 1985, pp. 95–96) writes:

The Nine Songs can best be described as religious drama; but though it is obvious from the most cursory reading of them that they were written for performance, the absence of stage directions indicating who at any given point was supposed to be singing, or what they were doing while they sang, makes it impossible to be sure how they were performed. In some cases we cannot even be sure whether what we are reading is monologue or dialogue or dialogue with choric interruptions. It appears that the actors or dancers were gorgeously dressed shamans; that musical accompaniment was provided by an orchestra of lithophones, musical bells, drums, and various kinds of wind and string instruments; that—to judge from one or two references to a ‘hall’—the performance took place indoors.

26

I support Hawkes’ claim that the Nine Songs are highly-polished poetic renderings of early Chinese séance events as they were performed in the early Chinese state of Chu. Walker (1982, p. 432) too appears to support this: “Written in the language of the Shijing but evincing a unique poetics, the Nine Songs poems appear to have been the products of an independent poetics: perhaps they were a translation into the major language and culture of pieces belonging to a local Chu tradition.” Without explicitly saying so, Hawkes also appears to deny Qu Yuan’s authorship to the Nine Songs, expressing as they do a certain joie de vivre presented in first person voices of ecstatic shamans and spirits that embody a religious sensibility more appropriate to a cultural consciousness than an individual poet.

3. A Brief Note on Codifying Shamanism

Waley and Hawkes apply a shamanic reading to the Nine Songs. Defending the sinological reading, Waters (1985, p. 75) writes, “As to assertions that these poems are shamanist hymns exclusively, with no allegorical or political importance, we leave it to the commentators and the simple facts of Chinese literary history to reject them.” True to the sinological project, his recourse is to the long history of traditional exegesis, which emerged subsequent to their being committed to writing. Ironically, Waters raises important considerations pertaining to the possible codification of shamanism: Can the ritual structures at play in the Nine Songs be taken as authentic demonstrations of actual séance events? Can they be comparatively situated to gain a deeper understanding of the logic of eroticism functional in agricultural shamanism?

At stake is the very amenability of shamanism to codification, an issue that can be considered in relation to two types of authority: shamanic and centralized.27 Hamayon (1996, p. 76) writes that when shamanism is present as “an all-embracing system,”28 it functions according to the dictates of shamanic authority in distinction to centralized authority, and Hamayon (1996, p. 88) continues:

Shamanic practice [is] an art to exercise, as opposed to a liturgy to apply… It appears as a deliberate refusal of dogmatism. Shamanic societies, and shamans in non-shamanic societies, reject the use of writing for strictly shamanic matters; this goes together with the absence of a church or a clergy. On the whole, shamanism seems to refuse its own codification, because this could hinder the play of partnership with super-nature—a partnership always available, so far as it consists in turning imaginary entities into partners with whom to exchange and negotiate.

The consequence of this “refusal” is that shamanism, in those societies where it is active as “an all-embracing system,” is hardly directly available to the modern researcher, simply because centralized authority has implanted itself throughout most known societies, and in the Chinese case since its first unification under the Qin Dynasty (221–206 BCE). And yet there remain the Nine Songs, considered by many (Waley, Hawkes et al.) to be authentic representations of early Chinese shamanism. However, they were not composed by shamans but by poets, therefore they can only be representations, not liturgy.

To uncover the shamanism of the Nine Songs, we can isolate two features in particular. The first concerns the temporal lapses that are apparent throughout the course of its depicted séances: in many, the shaman’s encounter with the spirit ends abruptly. Waley (1955, p. 14) notes: “Between these two parts may have come the shaman’s main ecstatic dance,” and Sukhu (2012, p. 77) supports this: “The sudden transition between the arrival and departure of the spirit may in fact hide a gap in the middle of the song where, in ritual performance, the love affair between the shamanka and the spirit was danced or acted out.” Because this is the core of the séance and most open to the shaman as “an art to exercise,” it is least amenable to both shamanic codification and poetic representation.

The second feature concerns the xi 兮 refrain found in the middle of each of its lines. Waley (1955, p. 14) comments, “The Nine Songs contain a number of meaningless cries or exclamations, and at the cesura of each line is the exclamation xi which may (but this again is only a very tentative suggestion) represent the panting of the shaman in trance, a sound very familiar to anyone who has attended mediumistic séances in Europe.” If Waley is correct, then this is likely a poetic strategy to represent the momentum of the séance and provide a phenomenological orientation to it one step removed.29

The remainder of this study takes the Nine Songs as valid representations of the shamanism in the early Chinese state of Chu. It does so by focusing on the ritual structures evident in each of its nine songs that cohere around the gendered relationships between shamans and spirits, as informed by agricultural shamanism’s logic of eroticism.

4. Ritual Structures of Séance Events in the Nine Songs

The Nine Songs are devilishly abstruse, and efforts to make even preliminary sense of them require a series of choices: who is speaking, what is their gender, what is happening, and where does it take place. In addition to these initial uncertainties lie the complicated shamanic structures of spirit-journeys and spirit-possessions. Waley (1955, p. 15) writes, “The main difficulty in interpreting the songs lies in the fact that the subject of the sentence is so often left unexpressed. Add to this the absence of number, gender and tense, and you will readily agree that there is bound to be room for differences of interpretation.” Hawkes (1985, p. 96) writes: “In some cases we cannot even be sure whether what we are reading is monologue or dialogue or dialogue with choric interruptions. Everyone who interprets the songs has to begin by making his own reconstruction; and because the uncertainties are so numerous, there are almost as many reconstructions as there are interpreters.” Sukhu (2012, p. 79) writes, “The central event in several of the Nine Songs hymns is the love affair between the spirit and the shaman. Even the traditional commentators acknowledge that some sort of love affair is hinted at, but the question they have debated for centuries is, who is the male and who is the female?”30

When Hawkes (1974, p. 47) sardonically writes, “It is impossible to be sure, and we are, it seems to me, entitled to choose whichever it amuses us most to believe,” it becomes tempting to give free rein to one’s interpretive imagination. However, more thoughtful strategies are required, and my own reconstructions are based on the application of agricultural shamanism’s logic of gendered eroticism, broadly illustrated by Zhang Jun (1994, p. 426):

Among the responsibilities of Chu shamans was calling the spirits; this involved the shamanic practice of seeking a love relationship with them, and it had a strongly local and popular quality… Young girls with beautiful voices and beautiful complexions were used to call male spirits, and pubescent boys of outstanding beauty were used to call female spirits.

In the Nine Songs, the logic of gendered eroticism works hand-in-hand with the much-related logic of yin (earth) and yang (heaven).31 This logic specifies the location of each of the nine spirits, six of whom are male (yang) and three of whom are female (yin). Five of the male spirits are at home in the sky (Donghuang Taiyi, Yunzhong Jun, Dong Jun, Da Siming, and Shao Siming), while the sixth is the single outlier at home in the Yellow River (He Bo). Two of the female spirits are at home in the Xiang River (Xiang Jun and Xiang Furen), and the third in a mountain forest (Shan Gui).

The combination of these two logics allows clear designation for who is speaking at any one point (the shaman, the spirit, or the chorus), the speaker’s gender, and what is happening (spirit-journey or spirit-possession), and it also reveals three separate ritual structures. In the first, a male shaman spirit-journeys in quest of a female spirit, as in “Xiang Jun” 湘君 (The Queen of the Xiang), “Xiang Furen” 湘夫人 (The Lady of the Xiang), and “Shan Gui” 山鬼 (The Mountain Spirit). In the second, a male spirit descends to possess a female shaman, as in “Donghuang Taiyi” 東皇太一 (The Eastern August Supreme One), “Yunzhong Jun” 雲中君 (The Lord Amidst the Clouds), and “Dong Jun” 東君 (The Lord of the East). In the third, a male spirit descends to possess a female shaman before escorting her on a spirit-journey, as in “Da Siming” 大司命 (The Greater Master of Fate), “Shao Siming” 少司命 (The Lesser Master of Fate), and “He Bo” 河伯 (The Earl of the River). The combination of the logic of eroticism and the logic of yinyang further decrees that males participate in yang by being active and mobile: male shamans spirit-journey to the homes of female spirits and male spirits descend to female shamans in the séance arena, while females participate in yin by being passive and sedentary: female shamans receive male spirits in the séance arena, while female spirits receive male shamans in their homes in rivers or mountains.

In the Nine Songs, the logic of eroticism is often displayed in the idiom of consummation or frustration. Male spirits are more susceptible to, even nonchalantly desirous of, the seduction strategies employed by female shamans than are female spirits to those employed by male shamans, as seen for example in the words of the male spirit in “Dong Jun” as he drives his chariot in the skies:

• The sounds of their song and their beautiful faces delight me,
• And I, the spectator enchanted, forget to go home.
• Yes! Strum loudly the zither to the beat of the drums!
• Strike the bronze bells on their white jade frames!
• Sound out the flutes and sing forth the pipes!
• I pine for the shaman, worthy and ravaging,
• Whirling through the air like a king-fisher in flight.
• Performing her song in time with the dance,
• The resonating pipes accord with her steps.
• The descent of my spirits darkens the sun!

The three songs depicting the spirit-journeys of male shamans (“Xiang Jun,” “Xiang Furen,” and “Shan Gui”) describe the female spirits as wary of the male shamans as if they were intruders, and they remain compunctious and mindful of themselves as females (even if they are spirits) in relation to males, coy and possibly distracted by thoughts of other male spirits to whom they may or may not be otherwise predisposed. Still, their hesitancy to give themselves to the male shamans is not entirely insurmountable. In “Xiang Jun,” the male shaman sings: “My Goddess does not come, she is too hesitant. Oh, who amuses her on the island?… Playing her bamboo flute, who lingers in her thoughts?” The male shaman then finally recognizes that he has lost her as he sings: “Unfaithful intercourse breeds bitter resentment. She does not honor our rendez-vous, saying, ‘I have not the time’.”

In “Shan Gui,” after the male shaman has spirit-journeyed to the mountain domain of the female spirit in his attempt to seduce her, she turns the tables on his erotic strategies by tantalizing and titillating him, as he sings:

• I see a figure appearing on the flank of the mountain,
• Her body covered in fig leaves, girdled with rabbit silk.
• With a seductive glance and a disarming smile,
• She says, “You desire me, for I am beautiful and gracious.”

Ultimately realizing that his rendez-vous will bear no fruit, he sings: “Bitterly sorrowful at her failure to come, in my frustration I forget to go back. My Goddess has me in mind but says, ‘I no longer have time for you’.”

These eroticized gender relations share certain structural affinities with those of hunting shamanism, but each conforms to a different logic: eroticism for the former and alliance for the latter. For hunting shamanism, Hamayon notes that the shaman’s eroticism is expressed in his declarations of love for the female spirit who he will, if everything goes well, go on to marry (and who elected him to become a shaman in the first place); Hamayon (1990, pp. 469–70) writes,

The single element of a conception of alliance which is given at the outset is the representation that the shaman marries his elector, the daughter of the spirit of the forest, for purposes which are above all sexual; the love that the shaman owes to his supernatural spouse is plain to see but not spoken about out loud… However, existing everywhere, at least latently, is the conception of sexual or romantic relations between the shaman and his electing spirit… The principal spirit is always female for a [male] shaman and male for a female shaman: this love ought to motivate the election.

Granting that these kinds of shamanic ritual structures could bolster arguments that, cross-culturally, spirit-journeying is a special prerogative of male shamans while spirit-possession is primarily associated with female shamans, with I. M. Lewis (1989) and Laurel Kendall (1985) providing the classic arguments, I would like to avoid such an approach because the Nine Songs also give a prominent position to the female shaman’s spirit-journeys.

5. Shamanic Eroticism

Georges Bataille famously uncovered three forms of eroticism: physical, emotional, and religious; he (Bataille 1986, pp. 15–16) writes,

My aim is to show that with all of them the concern is to substitute for the individual isolated discontinuity a feeling of profound continuity. It is easy to see what is meant by physical or emotional eroticism, but religious eroticism is a less familiar notion. The term is ambiguous anyway in that all eroticism has a sacramental character, but the physical and the emotional are to be met with outside the religious sphere proper, while the quest for the continuity of experience systematically pursued beyond the immediate world signifies an essentially religious intention. In its familiar Western form religious eroticism is bound up with seeking after God’s love, but the East, intent on a similar quest, is not necessarily committed to the idea of a personal God.

Religious eroticism differs from the physical and the emotional as a predominantly mystical phenomenon: eminently personal, it is exclusively shared between two beings, one human and the other spiritual. Bataille (1986, p. 21) further specifies its nature: “Through the beloved appears full and limitless being unconditioned within the trammels of separate personalities, continuity of being, glimpsed as a deliverance through the person of the beloved.” Next to Bataille’s three recognized forms of eroticism, I suggest an additional fourth: shamanic eroticism. While the practical thrust of both religious and shamanic eroticism appears similar on the surface (“the experience of profound continuity”), they are quite different with respect to their separate institutions: private mysticism versus public shamanism. And while both concern the erotic relationship between a human and a spirit, their fundamental goals differ: personal fulfillment versus the community benefit that accrues from bringing together the human and the divine in the public performance of the séance event.

The erotic shamanism of the Nine Songs is powerfully depicted in the spirit-journey of the female shaman in “Da Siming”:

• Our towering flight, soaring serenely,
• Harnessing the pure vapors, driving onward yinyang,
• Together with my Lord in purified velocity.
• Leading the Emperor through the Nine Mountains,
• His sacred robes in billowing folds,
• His jade pendants dangle and dazzle.
• Now in yin, now in yang,
• No one knows what it is that we do.

This spirit-journey depicts the female shaman escorted through the skies by the male spirit in starkly sexualized images: flying “in purified velocity” 齊速 (zhai su) while “driving onward yinyang” (yu yinyang御陽陰).32 This language does more than border on the orgasmic; it is in fact the purest depiction of “profound continuity” (between the shaman and the spirit as much as between the human and natural worlds) and demonstrates the consummation of shamanic eroticism. The Nine Songs never explicitly articulate erotic consummation; it lingers, rather, in the hazy and mysterious center of the logic of eroticism.

This eroticism, initially exercised by the shaman, is activated in the direct encounter of the shaman and the spirit either in the spirit-journey or the spirit-possession. Within the context of public performance, it is channeled toward producing a sense of the continuity of being with the divine experienced by and for a public community, in distinction to the religious eroticism typically performed in personal isolation that only concerns the individual’s sense of a continuity of being with the divine. For this reason first of all, the religious eroticism of the mystical experience cannot be identified with the shamanic eroticism of the séance. Rita Perintfalvi (2014, pp. 237–39) bears this out in her study of the mystical experiences of Mechtild of Magdeburg:

Eroticism and mysticism have structural similarities since both involve the experience of transgression. The mystic Mechthild was downright driven by an insatiable desire to overcome the isolation and the finiteness of her person through a union with God. The mystical experience is a radical confrontation with one’s self here where one recognizes one’s real boundaries and is at the same time invoked to transgress them. Mysticism is also an intense devotion to life but in a way that the touch of death, a kind of an annihilation of one’s previous Ego that leads to a profound transubstantiation of the person, cannot be avoided: [quoting Mechthild] “I delight in loving him who loves me, and I long to love him to the death, boundlessly, and without ceasing.”

Much like the spiritual ravishments of religious eroticism,33 shamanic eroticism reveals the primal yearning for the renewal of human life in conjunction, alliance, or consummation with a spiritual being. Not reducible to religious eroticism, shamanic eroticism is only one possible feature of the shamanic institution. Still, the eroticism at play in hunting shamanism fulfills a purpose (the successful hunting of game) very different from that of agricultural shamanism, given that the eroticism of both is expressed in deeply personal terms as a love affair between the shaman and the spirit. This is borne out in “Da Siming,” as the female shaman anguishes: “My constant thoughts of him torment my heart, what can I do with this torment? I wish only that our tryst of this day could last forever. Oh! Obstinate destiny, unchangingly fixed, who is able to alter its dictates?”

Recognizing that both hunting and agricultural shamanism place eroticism front and center allows a particular access into their distinctive ritual structures. Some might counter that the shamanic institution is incompatible with the love relation between the shaman and the spirit. Catholicism, for instance, often presents the relation between Christ with his Church as wife, which emerged as a structural element of the covenantal relation in medieval Christianity, which is also seen, for example, in the love relation Teresa of Avila (1988) imagines she has with Christ. More important than this, however, is that religious and shamanic eroticism are both grounded in distinct institutions, in the ritual and symbolic sense: if the eroticism of the shaman is primarily institutional in that the shaman is playing a part in representing his or her community to the spirits, then so too is the eroticism of Teresa of Avila, Mechtild of Magdeburg, and Hildegard von Bingen.

As long as we are clear that we are not attempting to break into the psychic state of the shaman (or the medieval Christian mystic for that matter), then we must continue to speak of both religious and shamanic eroticism as discrete institutionalized relations (again in the symbolic, not the sociological sense) between a human and a spirit in terms of their specific representations. This is so whether such representations are ritual, symbolic, literary, musical, mythic, political, or otherwise; psychological explanations, on the other hand, will inevitably lose track of this institutional nature because of their singular focus on the inner experience of the shaman or mystic.34

Representations of shamanic eroticism are legion in the literature, even as there has yet to appear a systematically comparative or comprehensively theoretical study of it. Hamayon laid the groundwork for such a project, but she does not thematize a strictly shamanic eroticism; her focus is on “the religious alliance” between the shaman and the spirit of game in hunting shamanism. While I would prefer to designate this as “the shamanic alliance” to clearly differentiate it from other, non-shamanic forms of spiritual marriage such as those sometimes found in mysticism, she in fact notes that its particular shamanic meaning emerges from an anthropological understanding directed to communities in a structural and systematic relation to game rather than to individual mysticism; Hamayon (1998, p. 25) writes that “alliance is distinguished as a pronounced term from a series of modes of religious relations, ‘pact,’ ‘union,’ etc., which also assume the alterity of partners and their mutual choice for each other, but which do not form a system.”

Hamayon (1998, p. 26), finally, intends “to sketch an inventory of the modes of religious ‘alliance’ by proposing to consider together a relation between the phenomenon of ‘marriage relations’ [allié] (the beloved, the loved) and the desire to attract ‘luck’ (fortune, providence).” Hunting shamanism’s logic of alliance, activated for “luck” in hunting game, is thus not identical with agricultural shamanism’s logic of eroticism, and the difference lies in the fact that when the whole community is the ritual agent as it is in the exchange of human life for animal life, the relation with the spirit is clearly institutional as the covenant is in medieval Christianity, read “the King’s two bodies” (Kantorowicz 1997). Still, the logic of alliance is shamanic insofar as the shaman publicly shamanizes for the community together with his spirit wife, whereas the logic of eroticism is activated in the relationship between the shaman and the spirit that centers on temporary love affairs.

6. Shamanic Gender, Shamanic Sex, and Shamanic Eroticism

Exploring the ritual structures of eroticized relationships between shamans and spirits requires us to recognize that sex and gender necessarily impinge on any attempt to thematize the topic. As early as 1908, Roland Dixon (1908, p. 2) noted:

One of the broadest distinctions which may be made, in the connection with the making of shamans, is that of sex—whether the practice of shamanism is open freely to both sexes, or is more or less restricted to one or the other… The element of sex appears again in another way, as among the tribes of Patagonia, where there was a curious custom which prescribed the wearing of female clothing by male shamans… It appears that in America women are widely permitted to become shamans, male shamans even in some cases having to assume women’s dress.

In her discussion of the Japanese shamanic rite of initiation called kamsauzu uki,35 Manabu Waida (1994, p. 463) substantiates the fact that different shamanic traditions, especially those in which shamans marry spirits, often are restricted to one or the other gender: “It is only women who are entitled to the ceremony… Essentially, it is a wedding ceremony for the future shamaness and her guardian god… From now on, she will serve the god as his spiritual wife.” However, even when societies outwardly only recognize one gender as shamans, members of the opposite gender in some cases are provided the possibility to serve as shamans if they adopt the opposite gender through transvestism, male to female or female to male. Instances of shamans cross-dressing are quite prevalent, and Wolfram Eberhard (1968, p. 46) even discusses early Chinese records of certain ceremonies in which male shamans cross-dressed. In her study of male maibas and female maibis shamans of Manipur, Otojit Kshetrimayum (2009, pp. 21–22) writes, “The maibis dress themselves very distinctly. They dress in white. Even a man can also become a maibi. They usually wear phanek (dress for women in Manipur) of white color.”

Still other traditions recognize members of either gender as shamans. In his formidable 1909 study of the Chukchee, arguably the most ethnographically shocking examination of the many variations of “Sexual Perversions and Transformed Shamans” ever to appear in print, Waldemar Bogoras (1909, p. 415) writes, “Shamanism is not restricted to either sex. The gift of inspiration is thought to be bestowed more frequently upon women, but it is reputed to be of a rather inferior kind, and the higher grades belong rather to men. The reason for this is, that the bearing of children is generally adverse to shamanistic inspiration.”

Options between male and female shamans, as well as the gendered relations between shamans and spirits, are not infinitely variable; they remain constrained by gender, albeit with wide variation. They are often central to shamanic ritual structures where they supply the building-block mechanisms that propel movements of the séance, and communal assessments of the legitimacy of the shaman are often based upon them.

While shamanic eroticism is culturally represented in only a limited number of ways, from flirtations, dalliances, and affairs to marriages or alliances of both the temporary and permanent sort, shamanic sex is less containable in its transgressiveness because it rarely complies with the dictates of established ritual structures. Bataille (1986, p. 11) marks the tenuous distinction between eroticism and sex: “Sexual reproductive activity is common to sexual animals and men, but only men seem to have turned their sexual activity into erotic activity. Eroticism, unlike simple sexual activity, is a psychological quest independent of the natural goal: reproduction and the desire for children.”

If the natural goal of sex is reproduction in normal sexual relations between males and females, then this marks a decisive difference from shamanic sex. For what possible fruit could emerge from the sex of a shaman and a spirit, and how could it be implicated with shamanic invigorations of the sexual reproduction of game or the spiritual well-being of the community? Michel Perrin (1987, p. 572) provides one possible answer in his study of Guajiro shamanism:

Tobacco is an object which is ingested by shamans to give concrete form to their power… The action of the tobacco juice on the body is indirectly associated with fertilization through the mouth, comparable to the impregnation of a woman during the sexual act: the tobacco juice, ingested by the novice, literally causes “shamanery” to “ripen” or “grow” in the belly of the shaman to be.

Perrin (1987, p. 574) otherwise notes that “the shamanistic function is considered to be incompatible with normal sexuality…” This idea is echoed by Verrier Elwin (1955, pp. 147–48) who, in his study of the Saora of India, writes, “After the marriage, the shaman’s spirit-husband visits her regularly and lies with her till dawn. He may even take her away into the jungle for days at a time, feeding her there on palm wine. In due course a child is born and the ghostly father brings it every night to be nursed by the girl.”

Saladin d’Anglure and Philibert (1993, p. 85) expose a different side of shamanic sexuality: “The shaman was always rewarded when he offered his services privately and it was not unusual for him to ask for payment in the form of sexual services from one of the women of the family which had called him in, whether the wife or the daughter of the man of the house.” There are many such cases as well of such sexual hospitality for the shamans of Inner Asia.

Shamanic sex is not to be confused with shamanic eroticism, and Eliade’s imputation of “ecstasy” as the goal of the shamanic institution entirely misses the point of eroticism. He nevertheless differentiates it from the sex between a human and a spirit based on a specific feature of the former: it entails a more enduring component of tutelage and mentorship, over and beyond the reproduction involved with sexuality; Eliade (1964, p. 73) writes that “anyone can have sexual relations with female spirits, yet without thereby acquiring the magico-religious powers of shamans.” When a sexual relationship develops between a non-shaman human and a spirit, Eliade (1964, p. 79) calls it a “deviation,”36 but his assessment changes when it pertains to shamans:

Celestial wives… help him either in his instruction or in his ecstatic experience. It is natural that the “celestial wife’s” intervention in the shaman’s mystical experience should be accompanied by sexual emotion; every ecstatic experience is subject to such deviations, and the close relations between mystical and carnal love are too well known for the mechanism of this shift in plane to be misunderstood. Then too, it must not be overlooked that the erotic elements present in shamanic rites exceed the mere relationship of the shaman with his “celestial wife.”

While Eliade’s insight is important, there are many exceptions to what he wants to normalize in the relations between shamans and spirits. One such is evidenced in the findings of Carmen Blacker (1983, p. 140) concerning the Japanese itako: “The deity to whom they are formally ‘wedded’ at their ceremony of initiation seldom comes to their aid thereafter with the supernatural gifts.”

Shamanic eroticism is most tangible where gendered relations between shamans and spirits are deemed primary to the ritual structures of the séance, and this mitigates claims that would systematically explain shamanism as a function of aberrant sexuality. Although he does not adequately distinguish the erotic from the sexual, Eliade (1964, p. 74) concurs with this: “It does not seem that sexual relations with spirits constitute the essential and determining element in the shamanic vocation.”

Exploring this very issue in Korean shamanism, Laurel Kendall (1998, p. 145) poses the question, “Is possession erotic? The idea that the shamanic experience implies a sexual union between the female shaman and her guardian spirit appears in fact vaguely indicated in the recesses of the Korean popular imagination.” It can be noted that in several languages, as for instance Korean, shamanic possession is assimilated with sexual penetration in that the possessed one (whether man or woman) is conceived as a female having intercourse with a male spirit, yet Kendall nevertheless resists a reduction of the shaman’s experience to a sexuality that could be interpreted in terms of psychological repression, oppression, or substitution.

Kendall’s conclusions are persuasive because shamanic eroticism is not about any individual person or class of people who become shamans. And unless the goal of shamanic eroticism is the satisfaction or consummation of the erotic desire of the shaman, then the difference between sexuality as a goal-directed activity (reproduction) in distinction to eroticism must be maintained. If, on the other hand, shamanic eroticism devolves into the personal satisfaction of the shaman, then it would be no more than an aberrant off-shoot of religious eroticism. In the end, shamanic eroticism is not about the shaman but the community, even as the female shaman of Korea still performs within the midst of a gendered difference, sexualized as it might be, in her relationships with the spirits, and Eliade (1964, pp. 461–62) writes, “In Korea, where shamanism is documented as early as the Han period, male shamans wear women’s dress, and are far out-numbered by shamanesses.”

Bataille (1986, p. 18) recognized an element of violence inherent in all varieties of eroticism, which helps to explain shamanism’s gendered eroticism, where it exists:

When I come to religious eroticism which is concerned with the fusion of beings with a world beyond everyday reality I shall return to the significance of sacrifice. Here and now, however, I must emphasize that the female partner in eroticism was seen as the victim, the male as the sacrificer, both during the consummation losing themselves in the continuity established by the first destructive act.

Sacrifice37 is far less pronounced in hunting shamanism: if the spirits allow you to capture their wild animals, what logic is there to then go on to use them as sacrifice to the spirits who just gave them to you? Still, other kinds of offerings to the spirits remain requisite, and the violences identified with hunting shamanism remain as part and parcel of its own logic of alliance, as Hamayon (1990, p. 29) notes: “This conception assigns to the holder of the shamanic function, whether in terms of the individual or the group, a male position: to be a taker of game is to be a taker of the female. Reciprocally, it assigns a feminine position to the supernatural partner. This is how shamanic alliance is oriented.”38

7. Shamanic Eroticism, Consummation, and Death

Shamanic eroticism, enforced by notions of shamanic gender and shamanic sexuality, plays a constitutive factor in the gendered relations between shamans and spirits. The erotic desire of the shamans driving them to consummate their love with the spirits is the structural principle that endows thematic unity to each of the séance events depicted in the Nine Songs. While some of them depict consummation and others depict frustration, the success of any one séance does not depend on either outcome but rather on other factors. Consummated desire is indirectly represented in “He Bo”:

• I roam together with you along the Nine Rivers,
• As a mighty wind arises, unleashing the waves.
• Mounted on a water chariot, with its lotus dais,
• Pulled by two dragons and flanked with serpents on all sides.
• We ascend Mount Kunlun to survey the four quarters.
• My heart is dizzy and flutters, overcome with giddiness.
• Enraptured by you under the setting sun, I forget to go back.

It is also represented in “Xiang Furen”:

• At dawn my steeds run along the banks,
• At dusk I pass the western shore.
• I hear my beloved beckoning to me,
• We mount the chariot and race off together,
• We build a house in the water,
• The walls of lotus, and so too the roof…
• The spirits from Mount Nine Doubts welcome us,
• Their arrival approaches like gathering clouds.

Mount Nine Doubts (Jiuyi 九嶷) is the legendary burial place of the god Shun, to whom Xiang Furen was supposedly married, and when the spirits “approach like gathering clouds” (ling zhi lai xi ru yun 靈之來兮如雲), it is not unreasonable to suspect that they are literally giving cover to the intimacies of the male shaman and the female spirit alone in their temporary love-nest constructed entirely of flowers.

Frustrated desire is given more space in the Nine Songs, but some of these cases are complicated because in them the shamans reveal that they had previously experienced consummation with the spirit, as with the male shaman in “Xiang Furen”: “The time that we shared never again shall we have, and so for now I wander,” and also as with the female shaman in “Shao Siming”:

• With you I wandered the course of the Nine Rivers,
• The whirlwind that followed us made waves on the water.
• With you I bathed in the waters of the Xian Pond,
• And dried your hair on the banks of the Yang Valley.
• I yearn for the Fair One but he does not come.
• My face to the wind, desperate, I sing out my song.

Representations of erotic consummations and frustrations in the Nine Songs do not significantly differ from those of the mystical experiences of religious eroticism undergone by, for example, Teresa of Avila and Mechtild of Magdeburg. While Bataille understood eroticism as a charged response to an impulsive force capable of crossing boundaries, he never accounted for a specifically shamanic eroticism. Shamanism, however, concerns nothing if not the crossing of boundaries, and this is nowhere clearer than with spirit-journeys, although boundary crossings are not always easy to distinguish from boundary dissolutions identified with spirit-possessions. Boundary crossings are a primary feature of many religious activities (see Van Gennep 1961; Smith 1993; Douglas 2002), but what differentiates those of the shamanic sort are that they result in direct communication between humans and spirits in a séance event. No other boundary crossings can be considered shamanic in this strict sense.

Communities depend on the shaman’s ability to cross or dissolve boundaries, and it would not be unreasonable to construct a definition of shamanism that separately highlights this as its central function and feature. If the shaman is not able to accomplish boundary crossings and boundary dissolutions to the satisfaction of the community, he or she will be stripped of the position; Hamayon (1990, p. 437) writes, “The idea is generally clear that the period of shamanic practice is that of sexual activity, and the spirits suffer from the aging of the shaman and will demand another who is younger.” This is also a concern central to the male shaman in “Shan Gui” as he stays with the female spirit on her mountain fairyland:

• The darkness deepens and spreads, enveloping the day in its shadow.
• The east wind gusts as the goddess releases her magical rain.
• Content, I remain with my goddess, with no thought of return,
• But my years grow old; who besides her offers such youthful glory?

Shamanic boundary crossings are typically represented in the idiom of spatial movements between the profane and the sacred, but they can also be translated into cognitive crossings. In his study of the shamanic elements in early Finnish religion, Veikko Anttonen (1994, p. 104) identifies the Finnish term phyä with the sacred:

What I mean by sacred are the acts of “boundary violation” performed by shamans in the context of rituals conducted at times of transformation involving important cultural values… Interdictions, prohibitions and rules of avoidance, i.e., taboos, are cognitive derivatives of the notions of boundaries. The concepts of impurity, danger and power can be interpreted as symbolic markers and representations of cognitive boundaries by which places, objects and phenomena categorized as sacred and held as anomalous were made binding.

Boundaries in shamanic séances often function to separate the sacred and the profane, whether we understand this in the terms given by Durkheim (2001), Eliade (1987), or even Bataille himself. Boundary crossings, of both the shamanic and the mystical sort, are inherently dangerous and transgressive. Crossing them always involves a leap that requires a willingness to lose everything one has, including one’s own sense of identity and presence, but also one’s very hold on life itself, and eroticism is inextricably bound with death; Bataille (1986, p. 11) writes, “Indeed, although erotic activity is in the first place an exuberance of life, the object of this psychological quest, independent as I say of any concern to reproduce life, is not alien to death.” Bataille (1986, p. 24) continues: “Eroticism opens the way to death. Death opens the way to the denial of our individual lives.”

Eroticism, distinguished from love, lust, and sex, maintains a particularly acute complicity with death, as Bataille (1986, p. 11) also recognizes: “Eroticism, it may be said, is assenting to life up to the point of death,” and this includes shamanic eroticism. Even Eliade (1964, p. 509), who did not particularly attend to the complex of eroticism, shamanism, and death, writes, “The shaman has been able to contribute decisively to the knowledge of death.” Bogoras (1909, p. 417) as well recognized the shaman’s continuous proximity to death: “The slightest lack of harmony between the acts of the shamans and the mysterious call of their ‘spirits’ brings their life to an end. This is expressed by the Chukchee when they say that ‘spirits’ are very bad-tempered, and punish with immediate death the slightest disobedience of the shamans.”

A similar complex of eroticism, death, and mysticism runs through the writings of Christian female mystics, as evidenced in the mystic vision of Angela of Foligno (Lachance 1993, p. 182): “In a state of ecstasy, she found herself in the sepulcher with Christ. She said she had first of all kissed Christ’s breast—and saw that he lay dead, with his eyes closed—then she kissed his mouth… Afterwards she placed her cheek on Christ’s own and he, in turn, placed his hand on her other cheek, pressing her closely to him. At that moment, Christ’s faithful one heard him telling her: ‘Before I was laid in the sepulcher, I held you this tightly to me’.”

The Nine Songs are surrounded by death on many sides. Ancient legends remark on the human and animal sacrifices that were offered to some of the spirits named in them (Hawkes 1985, pp. 113–15; Sukhu 2012, pp. 206–7). The tenth song, “Guo Shang,” represents a ceremony conducted for the spirits of the dead who gave their lives to defend Chu. Qu Yuan, supposed author of the Nine Songs, was believed to have drowned himself immediately after composing them.

Shamanic death is not exactly physical death. Next to the initiatory death experiences of the “shamanic sickness,”39 shamanic death (call it boundary-death), a necessary feature of the shamanic institution, has two modes: in the first death, the shaman “dies” either when launching the spirit-journey or when the spirit takes possession of the shaman’s body, and in the second death the shaman “dies” when returning from the spirit-journey or when the spirit departs from the shaman’s body.

Shamanic death is synonymous with boundary crossings as the shaman utterly surrenders the self in its individual “discontinuity” against a radically different realm of “continuity,” and it is in some ways comparable to the brief weakening of consciousness in the petite mort of sexual orgasm. In crossing boundaries, shamans typically enact the moment of the first shamanic death by physical movements ranging from minor gestures likened to nervous fits to all-out fainting spells, and they enact the second shamanic death by being rendered either comatose or in a state of complete and anguished exhaustion.

Still, it is necessary to be precise concerning how we deal with shamanic death as it relates to the eroticized relations between shamans and spirits, and Hamayon (1990, p. 525) provides the model for this: “The shaman’s attack of furie is ordinarily contained in a rather abrupt manner, which gives place to a surrender: he collapses, inert, stretched on his back, for a time more or less long, in the course of which he (his ‘soul’) is supposedly in the supernatural, in direct contact with the spirits; his state is assimilated to one who is dead.” Hamayon (1990, p. 534) continues:

In this way can be defined the sequence of physical behaviors to which the shaman is given, in two moments of contrasting styles which are necessarily linked: the succession of leaps and cries, tremblings and jolts, which translate the sexual excitement appropriate to the process of ensauvagement, followed by the plunge into the inertia that leads the proceedings to its conclusion, the death on earth for the life in the supernatural.

Shamanic death is well-represented in the Nine Songs.40 We see the second death immediately after the spirit departs from the body of the female shaman in “Yunzhong Jun” as she sings:

• The deity descended in a shower of radiance, he remained within me,
• When of a sudden he flew off into the distance returning to the clouds…
• I remember my Lord and heave a great sigh,
• In utter exhaustion, my heart is pure anguish.

“Shao Siming” also gives representation of the second shamanic death:

• Without a word you came to me, without a word you left,
• Riding your wind-vortex with cloud-banners flying.
• Oh! There’s no affliction more heart-wrenching than separation,
• Oh! No happiness more poignant than our first knowing each other.

The first shamanic death of the initial boundary crossing involves more phrenetic movements and gestures than the second, which signify the shaman’s furie that is the sine qua non of the séance. This French term is far more apt for characterizing the shamanic death than Eliade’s preferred term, “ecstasy,” particularly when it is implicated with shamanic eroticism. While furie is close in meaning to the English term “passion,” it has altogether different connotations that bring out the deeper implications at play with the shamanic death; Hamayon (1990, p. 524) writes:

The furie of the shaman in the séance, with its animal coloration, is entirely and without equivocation positive, simultaneously a necessity and a gauge of efficacy… Artificially induced or not, represented with more or less naturalness, the burst of furie is the single ritual action in which the shaman cannot be replaced. To say that a ritual requires the presence of the shaman comes to mean that it requires this episode. This is a reserved element in an open order of behavior, like the marriage in the order of the love relations with the supernatural, but it is still tempting to bring them together. An entire range of arguments converge in supporting this identification and showing that the ritual episode constituted by the burst of furie and the state of death which immediately follows, the two moments of ensauvagement, are the physical expression of this coupling that leads to the conclusion [of the séance].

The shaman’s furie in the Nine Songs is several times depicted in the shamanic dance, which remains de rigueur the main source of the female shamans’ erotic appeal to male spirits, deployed to entice them to descend and take possession.41 It appears as a strategy exclusive to the female shamans in keeping with the gendered eroticism at play in the Nine Songs and, in its charged eroticism and provocativeness, might in some ways be compared to the Persian belly dance or the classical Hindu Bharatanatyam dance.

Shamanic dance embodies a feminine force capable of evoking the deepest of erotic male desires, and it remains a primary expression of agricultural shamanism’s logic of eroticism. It excites a basic instinct to take possession of the other and to be taken possession of by the other, totally and completely. Shamanic dance radically transgresses standards of propriety and moral decorum; besides, it is notably dangerous, ever on the verge of unlocking the Pandora’s box of primal sexuality with the violence of the libido. The raw primality, the furie, of the shamanic dance situates it at the opposite side of the spectrum from courtly ballroom dancing where single drops of sweat are considered an unforgiveable faux-pas. Shamanic dance can never be entirely choreographed, and its efficacy depends on the erotic furie into which the shaman is believed to tap and surrender in her performance.

The furie of the dance is expected, and the shaman, if successful, will fulfill the expectations that the community brings to the séance in ways not very different from our own judgments of theatrical or musical performance. If the shaman lacks efficacy, he or she will be effectively fired. This standard and measure of the shaman’s efficacy relies on the spirits’ assessment of it as borne out by the success of the séance itself, but that again is judged by whether the shaman has accomplished the socially expected results. As Hamayon (1993, p. 15) writes, “Usually a successful practice is the condition for a shaman to be recognized as such. He may be deprived of his social role for lack of efficacy.” Or, as I. M. Lewis (1989, p. 136) states: “But success here is often ephemeral. A shaman’s position depends upon public recognition, and reputations can be destroyed as easily as they can be built up.”

The shamanic dances of the Nine Songs demonstrate shamanic eroticism unmasked and unleashed, as depicted for example in “Donghuang Taiyi.” After a series of preliminary rituals that likely included a purifying orchid-water bath as mentioned in a later song as well as her raising up of enticing offerings of food and drink for the male spirit, we witness her dance from the beginning as its crescendo builds in accord with the music that in the finale eventuates in the male spirit taking possession of her.

• Raise the mallets to beat the drums.
• Gradual pulses open the slow tempo of the song.
• Enter the ranks of the zithers and flutes to the raucous rhythm!
• Descending in sinuous movements, the spirit possesses her.
• The wafting hazes of fragrant incenses engulf the temple with a splendid fragrance.
• The frenzy of the five notes comes to accord in dazzling array.
• The Lord savors his pleasure in ease.

8. Concluding Thoughts

Although the shamanism displayed in the Nine Songs has drawn much attention from scholars of Chinese mythology, religion, and literature, few of them deal with it from a cross-cultural comparative perspective. The present study has attempted to approach and interpret the Nine Songs as a textual gateway leading to an engagement with a fundamental practice embedded in the shamanic institution: the courting of a celestial partner in order to gain insight, experience, and power not obtainable on the earthly plane. Still, the Nine Songs do not quite explain what they are, and perhaps what was gained by performing them could only be known to the shamans themselves, and the secrecy contributed to their authority and status.

In depicting the intimate encounters between shamans and spirits, the Nine Songs offer deeper possibilities for cross-cultural engagements with representations of the erotic, including its tangible sensualities as well as its hardly-disguised violences. But let this not fool us into imagining certain types of shamanism as anything other than a unique religious phenomenon fundamentally grounded in eroticism, sacrifice and offerings, and violence and death, for which the Nine Songs provide a surprisingly early model.

If there is a truth to shamanic eroticism, it is not, to quote Foucault, “the truth of sex,” and the ars erotica has something more akin to the religious eroticism of Teresa of Avila than to the shamanic eroticism of the Nine Songs; Foucault (1978, p. 57) writes:

In the erotic art, truth is drawn from pleasure itself, understood as a practice and accumulated as experience; pleasure is not considered in relation to an absolute law of the permitted and the forbidden, nor by reference to a criterion of utility, but first and foremost in relation to itself; it is experienced as pleasure, evaluated in terms of its intensity, its specific quality, its duration, its reverberations in the body and the soul. Moreover, this knowledge must be deflected back into the sexual practice itself, in order to shape it as though from within and amplify its effects. In this way, there is formed a knowledge that must remain secret, not because of an element of infamy that might attach to its object, but because of the need to hold it in the greatest reserve, since, according to tradition, it would lose its effectiveness and its virtue by being divulged.

Significantly different from the ars erotica, shamanic eroticism, systematically “divulged” to the audience in every lyric, gesture, expression, and convulsion, there for all to bear witness, is put to work for the community. Nor does it approach the metaphysical subtleties of the Buddhist yab-yum embrace that divulges the truth of the unity of compassion and wisdom in supremely sexualized form.42 If there is a truth to shamanic eroticism, as the personalized and ritualized seduction of the spiritual other involving both body and mind in the pursuit of spiritual benefits, it is the truth of the continuously flowing and synergetic relationship holding life, sacrifice, and death together for an entity bigger than the self, namely the constantly reproducing community of human beings jointly locked in life, labor, loss, and celebration.

The present study has attempted to open up various pathways for considering shamanic eroticism, and it has done so through situating the shamanism of the Nine Songs in a comparative framework that seems most appropriate, but not exclusively so, to agricultural shamanism. One way concerned the recognition that while many anthropologists have long noted the presence of sexualized relationships between shamans and spirits in both ritual and discourse, they have neglected the erotic aspects in particular by confusing them with sexuality. This distinction between sexuality and eroticism might contribute to account for the importance attached to the shaman’s individual qualities, for the strongly personalized character of their practices, hence for the permanent rivalry between shamans within the same society. In various types of shamanism such as hunting and agricultural, a shaman must enter a love affair with a spirit to become legitimate as a shaman, and eroticism could be a good way to be better than other shamans at seducing the spirit.

Another way that this study has attempted to open up pathways for engaging shamanic eroticism concerns the evidence of love affairs between shamans and spirits that speaks to a cosmic perspective on the relationship between nature and culture or, to use more traditional jargon, between the sacred and the profane. Many are the ways to understand nature: as totally, transcendentally other, or as chaos, or as danger; or nature can be seen as nurturing, or as supportive, or even loving. A foundational component of most every shamanic worldview rests on the nearness and approachability of the spirits, and there are those humans capable of having congress with them. The spirits inhabit liminal locales within this very world: on mountains, above the clouds, in forests, or under water. In the Nine Songs, the spirits of nature are radically humanized in ways that ancestors never can be: they are erotic beings most susceptible to those most human of emotions: passionate joys and lustful delights together with rabid possessiveness and petty jealousies, but they feel such emotions all the way. The world is alive with their presence and shamans tap into it, they bring the spirits home to the community and openly share them. Shamanic eroticism, with its succulent offerings of delicacies and liquors, its lush atmospheres of music and incense, its floral decors on head-dress and garments—the celebration of life richly spiced with the flavors of death.

The agricultural shamanism of the early Chinese state of Chu provides many of the most powerful indicators that have led people to posit “a southern hypothesis”43 beginning from the earliest times; Hawkes (1985, p. 19) writes, “As for the men of Chu’s superstitious beliefs in spirits, their addiction to shamanism and fondness for ‘lewd rites,’ these were already notorious when the Chu ci poems were being written and remained so for centuries afterwards.” The shamanism of the Nine Songs is itself also subject to history,44 which the present study has attempted to situate in relation to its possible sociopolitical and economic correlates, although much work remains to be done here. In terms of its own particular Chinese history, this shamanism was targeted by the increasing centralization of state power following the establishment of the Qin Dynasty directly leading into the Han (206 BCE–220 AD). On one level, these developments can be approached with the awareness that the forms of any sexual practice, even if they are naïve misrepresentations of a fundamental eroticism, relate to biopower, which, as Foucault (1995) explains, is subject to the panopticon of state power. Thus, sexual congress with otherworldly beings, who cannot be seen on the mundane level, could be perceived by those in power as a threat posed by the shamans to the centralized state of imperial rule. From a different vantage point, the Confucians also attempted to tame the shamanic eroticism so vividly displayed in the Nine Songs by turning it into political allegory, which was taken as a sort of archetypal model for the frustrated minister but was, in the end, more easily containable, despite the costs to personal well-being.45

These movements eventually and inexorably led to the marginalization and eventual disappearance of the shamanic institution revealed in the Nine Songs of Chu as a complete symbolic system, although it would persist for many more centuries and, as Hawkes (1985, p. 19) comments, is visible even today:

Suppression of shamans and destruction of their holy places were part of the ‘civilizing’ policy vigorously prosecuted by Confucian administrators in this area in the early years of the Tang dynasty. To judge from the numerous poetical accounts of shaman shrines and shaman ceremonies dating from the eighth and ninth centuries, it does not look as if they were altogether successful. Relics of old shamanism are to be found even today among the Chinese settler communities of South-East Asia.

46

China has much to offer and much to teach, and far too many of its jewels such as the Nine Songs have for too long been hostage to the iron grip of sinological captivity. In the end, it is my hope that the present study can make a contribution, not necessarily to the insulated institution of Chu ci studies per se (but who knows) or even to Western sinology more generally, but rather to wider awareness and recognition of the role played by shamanism within Chinese history, and that this role can find a deserved place within the comparative study of shamanism as one important component within the total panoply of human experience.