Next Article in Journal
Scales of Justice by Rabbi Zvi Hirsch Kalischer: From Rabbinic Tradition to Public Participation
Next Article in Special Issue
Miyun Yuanwu 密雲圓悟 (1567–1642) and His Impact on 17th-Century Buddhism
Previous Article in Journal
The Evolution of the Spatial Distribution Pattern of Mosques in the Kashgar Region from 1955 to 2004
Previous Article in Special Issue
Lineages as Network: A Study of Chan Genealogy in the Zutang ji 祖堂集 Using Social Network Analysis
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:

The Social Networks of Gods in Late Imperial Spirit-Writing Altars

Vincent Goossaert
Ecole Pratique des Hautes Études (EPHE)—PSL, 75014 Paris, France
Religions 2023, 14(2), 217;
Submission received: 2 January 2023 / Revised: 30 January 2023 / Accepted: 1 February 2023 / Published: 6 February 2023
(This article belongs to the Special Issue Historical Network Analysis in the Study of Chinese Religion)


The late imperial-educated Chinese interacted with a very large array of gods through various means, especially spirit-writing, for which we have abundant detailed records. While a few prominent gods have been studied in this context, there are currently no comprehensive studies of the connections between humans and gods. Using the records of thirteen different spirit-writing altars in various parts of the Chinese world between the sixteenth and nineteenth centuries, this paper maps the 478 gods involved using standard social network analysis visualizations, and identifies the types of gods that played central roles (connecting many different gods and humans) and those that had fewer, more exclusive sociabilities.

1. Introduction

Chinese social and cultural life in modern times involves intense interactions with a seemingly boundless cohort of spirits, gods, and other divine entities—in this article I use “god” as an encompassing term for all of these entities, despite their various ontological statuses. Scholarship on Chinese gods tends to favor the study of their cult, that is, the various ways in which different groups invent or shape a divine persona according to their own needs, values, and preferences. Divine entities, however, are not entirely malleable, such as floating signifiers waiting for humans to fill them with any substance, but come with inherent characteristics and constraints (themselves shaped by history, but that cannot be ignored once created). Source material on gods affords them considerable agency as well as singular personalities, and treating them as mere stand-ins for human voices is not the only way to make historical and anthropological sense of this material. For this reason, a fruitful way to study gods and their interaction with humans is to treat them as persons, that is, entities with a unique identity, history, and value, with a capacity to act (that is, to be perceived and understood by humans to act) as a subject, out of their own volition and choices. Similarly to humans, divine personas develop and subjectify themselves as they relate to other gods and humans, and are informed by such relationships.
The SNA work theories readily treat non-human as well as human actants; thus, it is quite natural to treat Chinese gods as such, even though this has rarely been carried out. My project here, however, is markedly different from research inspired by actor-network theory, which aims at bringing attention to actants (institutions, objects) that are not persons.1 By contrast, I am especially interested in the processes through which divine entities subjectify themselves through their involvement of social networks (this is a reason as to why I do not call them actants). In fact, in the case we will explore here, the boundary between human and non-human entities is quite moot: the gods’ late imperial literati dealt with are mostly (although not exclusively) dead humans who maintain their own identity in their divine existence, and the living members of the networks have in some cases known them (in the case of recently deceased and apotheosized masters and fellows), and in others have been exposed to their presence in multiple ways. Furthermore, spirit-writing adepts are engaged in self-divinization processes, and often consider themselves as future gods. To them, in emic terms, gods are not a different category of beings from themselves; their main difference is temporality, as gods remain present over many centuries. As a result, we can use SNA (Social Network Analysis) to explore how humans and gods build social networks that allow them to subjectify themselves (affirm their values and identities, spread their writings and ideas), with the caveat that, different from the usual SNA dealing only with historical living humans (as found in the other essays in the present special issue), chronological analysis is warped.
This avenue of enquiry can be pursued with all kinds of historical and ethnographical material; I wish to explore it with one specific sort of document: the records of late imperial spirit-writing altars—the most common generic terms are jitan 乩壇 and luantang 鸞堂. At these altars, gods were invited to possess the hand of one or two altar members, or a writing implement that these members wielded; altar members asked questions to the gods, and the latter also spontaneously offered instructions, wrote poems, tracts, essays, and sometimes full scriptures. Many of these were published.2
Among the vast literature published by late imperial spirit-writing altars, some books consist of the pronouncements of one given god, but others compile the revelations of a number of gods who all made themselves present at the same altar. It is uncertain whether this phenomenon—one altar receiving the revelations of many gods—was the norm in late imperial spirit-writing practice (it is common in contemporary times), but it certainly was not rare.3 It is this type of texts that I use for mapping divine social networks, because they allow us to see which gods associated with each other at different altars. The conversations between gods and humans found in such sources, on which I focus exclusively here, provide ideal material for SNA; however, I would like to briefly mention in passing that, in the future, this would ideally be combined with other lists of gods, notably those found in the litanies of repentance, chanfa 懺法—many of them produced and performed by spirit-writing groups. Another separate type of sources is the anthologies of revelations from one or a set of gods; they are not considered in this article because they are edited together, but were originally revealed by different groups.
I hasten to add that the late imperial spirit-writing records provide a view of divine-human social networks in very specific contexts, and that in other contexts (communal rituals, hagiographies), we would see other gods, absent or marginal here, take a central place, and other types of connections appear; even gods who were crucial in early modern spirit-writing activities (such as Zhenwu 真武) are almost absent from my corpus. Therefore, this article by no means offers a comprehensive view of the society of Chinese gods, but merely a glimpse of the dynamics of their social networks building.
This whole exercise is supported by my conviction that the altar members—by definition, people with some education, mostly men, but in some cases women—did have some knowledge about the gods, acquired through reading, participating in rituals, watching opera, or reading stories, and therefore participated in a shared discourse of popular theology. The spirit-written literature shows that they did not simply make up gods’ names and make them say whatever they wanted to hear; by and large, the spirit-written discourse of any given god—and the gods frequently talk about themselves—is broadly coherent, not only within a given collection, but also between different altars. In other words, divine personas are quite stable, and do not entirely depend on the humans who make them speak.
In the context of spirit-writing altars, relations between humans and gods are often personal and intense; gods accept living humans as disciples, and instruct, encourage, and chastise them individually, sometimes with emotionally charged language. Furthermore, many of the gods active at spirit-writing altars express a unique personality; their writing style, the type of topics they discuss, the ways they interact with humans are all specific. Understanding the processes of creation and maintenance of such unique divine personalities would require in-depth case studies—as I have attempted to do in a separate piece with Divine official Wang, Wang lingguan 王靈官, the fierce thunder god enforcing moral laws (Goossaert 2022b). The point of the present article is that each god’s divine persona is linked to their place in various networks connecting them to other gods. During spirit-writing sessions, gods present themselves in turn, sometimes discussing their relations to other gods in the process and referring to each other.
Gods do have an intense social life, and their interventions at spirit-writing altars show this abundantly. The types of relationships vary, including hierarchy—many gods are presented as subordinates to higher-ranking ones, and doing things on the orders of their higher-ups—but there is also bonding between equals (there are many instances of joint revelations by all members of a group wherein gods act together, such as the Eight immortals, or various groups of divine generals). Bonding is nurtured by exchanges of words, but also other means; the role of wine-sharing was important at a good number of altars, for expressing bonds between gods, and also with humans.

2. Research Questions

The list of gods present at each spirit-writing altar forms what might be called a pantheon for that particular group, if we keep in mind that pantheon here does not mean a fixed, authoritative structure but a network where gods with different personas and roles cooperate to instruct living disciples, and engage them in the divine work of saving humanity; for altars that have a long, continuously documented history, these pantheons constantly develop and grow. The research question then is, how do we understand the logics and grammar that inform the formation and development of these altar-specific pantheons? More specifically, what types of divine personas play central roles in these processes?
I will unpack this question by proposing two typologies of the gods. The first, which I call “theological”, is essentially independent of the spirit-writing context, and concerns the ritual contexts where these gods are deployed, as known from other (often earlier) sources. The second, which I call “relational”, is the result of visualizing the dataset and observing the position occupied by the gods in the altar pantheons. I detail these two typologies below. Matching the gods’ status in the two typologies will help us better understand the roles of the various types of gods in the operation of the society they form with humans. In a further stage of research—which this article merely points to—revisiting the actual contents of their discourse in the light of such SNA analysis will allow us to better grasp how and why specific gods play specific roles, such as provider of personal self-cultivation instructions, mentor in ritual practice, introducer of higher gods, teller of stories of retribution and divinization, or enforcer of moral codes.
A second research question, once we have developed an operating typology to understand the various positions gods can occupy within divine networks as created by spirit-writing, will be whether there are substantial differences between the pantheons of the various altars. Some (not all) of the altars in our corpus can be assigned to a dominant Daoist, Buddhist, or Sectarian identity; to what extent does that affect the types and roles played therein by the gods? For instance, an important turning point in the history of spirit-writing was its adoption during the mid-nineteenth century by sectarian traditions, notably the Xiantiandao 先天道 (Goossaert 2022a, chap. 8). In the process, the Xiantiandao adopted gods already active in earlier altars, but also introduced its own: what was the impact on the divine networks and the repartition of roles between gods? Moreover, can we see differences in these pantheons according to region or period? This article provides a preliminary foray into such questions, and will not be able to answer them all, but it will hopefully lay the foundations for future investigations based on larger source bases.

3. The Corpus

The corpus on which the dataset for this article was built is composed of the published revelations at 13 different spirit-writing altars.
  • Huangjing jizhu 皇經集註 (Collected commentaries on the (Jade) Emperor’s scripture). This collection of commentaries on the Jade Emperor scripture, a major Daoist text itself spirit-written in the thirteenth century, is part of the 1607 supplement of the Daoist Canon (Xu Daozang). It was compiled ca. 1585, and the spirit-written comments by 49 gods must have been revealed shortly before that date; the location of the altar is unknown, but most likely in Shandong or Beijing.
  • Taiyi Jinhua zongzhi 太乙金華宗旨 (Essential meaning of the Great One golden flower). This important inner alchemical text was revealed at an altar in Changzhou 常州 (Jiangsu) in 1668, and again in 1692; seven gods were involved (Lai 2016).
  • Yuquan 玉詮 (Instructions from the Jade [Altar]). This is a collection of revelations by 35 gods at the Jade altar, Yutan 玉壇, in Suzhou between 1668 and 1722.
  • Zhishenlu 質神錄 (Record of soliciting confirmation from the gods). This is a collection of revelations between 1678 and 1720 by 19 gods at altars at which Peng Dingqiu 彭定求 (1645–1719) was present as a mortal, in Suzhou and Beijing, with the final inclusion being a spirit-altar descent by the recently deceased Dingqiu himself (Burton-Rose, forthcoming). It was edited by Dingqiu’s great grandson, Peng Shaosheng彭紹升 (1740–1796).
  • Xu Zhishenlu 續質神錄 (Continued record of soliciting confirmation from the gods). This is presumably a sequel to the former one, but details are lacking. It involves 18 gods.
  • The various altars managed by Jin Bencun 金本存 in Sichuan and Guizhou during the 1730s and 1740s, the revelations of which were included in several Wenchang 文昌 canons; they concern 11 gods.4
  • Jueshi zhengzong 覺世正宗 (True transmission to enlighten the world) is a ten-juan collection of morality tracts revealed by spirit-writing at an altar in Jinzhou 金州, present-day Dalian 大連, between 1855 and 1858. The main deity of this altar was Lüzu 呂祖 (the immortal Lü Dongbin 呂洞賓), but a total of 52 gods contributed.
  • Gaochun Pujitang zhi 高淳普濟堂志 (Gazetteer of the Hall of universal succor, Gaochun county) is the record of a spirit-writing temple and charitable hall in Gaochun, southwest of Nanjing. It was published in 1900, and contains the revelations of 133 gods between 1892 and 1900.5
  • Xixin yulu 洗心語錄 (Recorded sayings to cleanse minds) is a collection of revelations from 16 gods, at the Qiyunge 齊雲閣 in Hankou (modern Wuhan, Hubei province) between 1793 and 1805, with a total of 16 gods.
  • Jiushengchuan 救生船 (A boat to save lives). Revelations from 117 gods between 1860 and 1863 by the Qunyingtan 群英壇 in Sichuan.
  • Zhilu baofa 指路寶筏 (Precious raft to show the way), one of the earliest spirit-writing collections emanating from the Xiantiandao, likely produced during the 1850s or 1860s in southwestern China; the preface is dated 1868. It features 56 gods.
  • Quanshi guizhen 勸世歸真 (Exhorting the world to return to transcendence) is a collection of revelations from 75 gods to an altar in a township some 40 km south of Beijing. The prefaces are all dated 1886, and the revelations must date from this or preceding years.
  • Xiuxin zhuanxing jiujie huisheng 脩心轉性救劫回生 (Saving [humans] from the apocalypse and returning them to life, through cultivating the mind and reorienting one’s nature). A collection of spirit-written instructions revealed at two related altars in Hunan in 1870–1872; 75 gods.
This corpus has a good spread in space, with northern China, Jiangnan, Hubei, Hunan, and the southwest all represented. It also runs from the late sixteenth century to 1900, even though the second half of the nineteenth century is dominant. The size of the revealed corpus and, most importantly, the number of gods involved, vary widely, from 7 (altar #2), to 133 (altar #8), with the average being 51.

4. The Dataset

For the purpose of the present article, I am interested in how a god is associated with other gods as well as with the living members of a given altar, and therefore I make no assumption about the relative importance of gods and their pronouncements; obviously, in some cases, some gods only produced one short poem, and others dozens of pages. I have not attempted to weight this in the dataset; I have, however, made use of my qualitative reading of the material to analyze the data and their visualizations. I do not consider chronology either; even though at some altars, revelations spanned many years and are all dated, for the sake of simplicity, I have simply considered the list of all active gods without noting when exactly they intervened; neither have I considered the number of times they intervened, even though this varied widely. In future developments of this project, weighting the gods (by number of appearances) could further refine our assessment of their roles; in the current dataset, however, my focus is on which god is related to other gods, and this is not fundamentally affected by weighting.
Identifying these gods was in many cases an issue, and I had to make many choices and educated guesses. In some cases, I identified as one single god different names and titles used at different altars (or even the same) when I was quite confident the same divine persona was being referred to, but I left as different entries in the gods’ list names who might in fact have referred to the same god, but for which I judged the evidence to be inconclusive.
As a result of the above choices, the dataset was thus formed by two types of nodes: the 13 altars and the 478 gods. The 669 edges link one god to each of the altars at which they are present. I have visualized them with Gephi. In the graphs, I have made the altar-nodes black, that is, invisible, so that one can see the links between gods and altars, but does not visually confuse altar-nodes and gods-nodes.

The Types of Gods (“Theological” Typology)

Since the list of gods in the dataset is long and largely composed of rather obscure figures, I have tagged some of them—about half, the others not being easily assignable—in categories that, even though largely etic and external to the data, can serve our analysis. I fully admit to the ad hoc nature of this classification, and I recognize that other classifications would help us see other patterns in the data.
  • (code: lei): thunder gods (leishen 雷神), comprising the fierce generals central to the early modern Daoist thunder rites and the staff of the Thunder ministry (18 gods);
  • (code: bax): the eight immortals, Baxian 八仙 (8 gods);
  • (code: con): Confucian saints and major philosophers, from Confucius and his disciples to Zhu Xi 朱熹 (1130–1200) and later luminaries (14 gods);
  • (code: poe): famous poets and literati, such as Li Bai 李白 (701–762) (11 gods);
  • (code: bud): Buddhist figures, including Buddhas, bodhisattvas, and famous monks (8 gods);
  • (code: qzh): patriarchs of the Daoist Quanzhen 全真 monastic order (4 gods);
  • (code: loc): local gods, including city gods (Chenghuang 城隍) and earth gods (tudigong 土地公) (16 gods);
  • (code: imm): immortals known from ancient lore (such as figures in the Liexianzhuan 列仙傳, and other classical hagiographies) (13 gods);
  • (code: cle): clerks of the divine administration, referred to by their title, with or without a name (gongcao 功曹, etc.) (24 gods);
  • (code: top): high-ranking gods with titles such as Sovereign, di 帝, or Heavenly worthy, tianzun 天尊; some of these gods have their own ritual traditions and hosts of subordinate gods under their orders: prominent examples include Wenchang 文昌帝君, Lüzu, and Guandi 關帝 (15 gods);
  • (code: guf): “old Buddhas”, gufo 古佛, who are sent by the Unborn Mother 無生老母 to save humans in the Sectarian traditions (12 gods);
  • (code: dao): famous historical (or considered to be so) Daoists, other than Quanzhen patriarchs and legendary immortals (8 gods);
  • (code: gen): historical generals and military heroes (13 gods);
  • (code: diz): deceased altar members or their kin, and other local persons known by altar members (27 gods).

5. The Findings

5.1. Types of Gods (“Relational” Typology)

Figure 1 below provides a view of the whole corpus. It clearly shows that the majority of gods are clustered around one altar with which they have one unique connection, while a much smaller number with two or more connections are found in the middle.
The gods active at the altars, based on their connectivity with other gods, can be roughly organized into three relational types. First, central gods who are linked to most other gods (through the many altars where they are co-present) and played a crucial role in the operation of the altars. When we visualize the corpus by filtering out gods who have four or fewer connections (i.e., are active at fewer than five out of thirteen altars), we are left with 15 gods (that is, 3% of the total). When we increase the threshold to a minimum of seven connections, we are left with just six gods. Coming as no surprise, this list includes the three most prominent spirit-writing gods: Lüzu (present at 12 altars), Wenchang (also 12 altars), and Guandi (7 altars); more interestingly, the remaining three are Zhongli Quan 鍾離權 (9 altars), Wang lingguan (9 altars), and Han Xiangzi 韓湘子 (7 altars). This should draw our attention to the importance of these gods and their role in divine-human sociality; they have been studied for the literary aspects of their lore, but our data show that their divine persona and connectedness made them present on a regular basis in the meetings of humans and gods at many different altars.6 Zhongli Quan, for instance, is the master of Lüzu and one of the Eight immortals; his occasional presence as such could be expected, but his very high connectedness strongly suggests he has developed a strong divine persona and a unique role that deserve specific attention.
The second relational type, at the other extreme of the arc, is gods unique to one altar. This type, which accounts for the largest number of gods, comprises two main theological types—many more are simply not well-identified. First, local gods, with their own temples, naturally intervene at nearby altars, but are not known elsewhere. These gods are part of the altar members’ everyday lives, and it is natural that they should be part of the human-divine sociality performed at spirit-writing altars. In some cases, they play an important role, such as in the Xiuxin zhuanxing jiujie huisheng where the earth god (tudigong) serves as a master of ceremonies, announcing all other gods, but does not produce a substantial discourse of his own. The second type is the deceased altars’ members and their kin. It is a well-known fact that one of the main aims of spirit-writing altars was self-divinization, and that deceased members were worshipped by their surviving fellows, and wrote to them. We have several examples of that phenomenon in our corpus; for instance, in Xiuxin zhuanxing jiujie huisheng, a member who was instructed by a god when alive intervenes later, forty days after his death.7 The altars were also venues for the divinization of members’ parents. The Quanshi guizhen contains remarkable material on the mothers of five of the altar leaders, who produced a series of revelations after their death. They were apparently not altar members when alive, but now in the afterlife, they studied literary Chinese with famous philosophers and thus, became able to produce well-written revelations.8 Figure 2 shows the local gods and deceased disciples (theological types G and N) in blue and red, respectively, almost all in the clusters around one altar, and with uneven repartition, some altars giving them more place than others.
The third relational type concerns gods of middling connectivity who belong to categories that are frequently tapped by spirit-writing altars for building their divine networks, but among which no individual god takes a strong, dominant position. Figure 3 shows the place in the dataset of three types of historical figures: M (generals) in red, D (poets) in green, and C (Confucians) in blue: none of them is highly connected, but most are located toward the center, with at least two connections. It seems as if these types work to some extent like generic categories, where individual heroes perform similar roles, and the choice of one or another is linked to the specific circumstances of the human-divine society at hand. To take famous Confucian authors, for instance, Zhu Xi and Han Yu 韓愈 (768–824) both appear at three altars, but they only have one altar in common.
Based on this simple typology distinguishing broad levels of connectedness (high, isolated, and intermediary), it is interesting to explore how specific theological types map onto it. Consider for instance the question of the place of Buddhist gods: among the eight Buddhist gods (not including the sectarian Old Buddhas), only two are present at three or more altars: Guanyin 觀音 (five altars) and Jigong 濟公 (three altars); even though most of these Buddhist gods are universally known divine personas in late imperial Chinese culture, their connectedness in the context of spirit-writing sociality is quite weak. By contrast, Daoist gods tend to have a higher connectedness; to take one important category, the thunder gods are not only quite numerous (18), but are also quite well-connected, with three present at three or more altars (see Figure 4).
Another way to look at the connectedness of categories, rather than individual gods, is to chart the presence of my “theological” categories among the various altars, as shown in Table 1. Besides J (“top” gods), the only category to be present at all altars, the best represented are the thunder gods (A) and the Eight immortals (B), both present at 12 out of 13 altars; all other categories are significantly less equally spread. This suggests that these two categories play an essential role in divine sociality. The key role of thunder gods can be related to the rituals practiced at sprit-writing altars, which often are closely related to thunder rituals—as are the spirit-writing procedures themselves. Thunder gods, notably Wang lingguan, serve as ritual instructors to human disciples. As for the Eight Immortals, their importance is linked to the figure of Lüzu, who quite often is accompanied by some or all of the other seven members of this group.
In addition to theological types, our dataset also allows us to look at gender. Among the 478 gods, I have only been able to tag 16 (3.2%) as female, even though this is without a doubt an underestimate, as some of the unidentified gods, many of them are just referred to by a family name and a gender-neutral epithet such as “immortal” 仙, may be female.9 Whatever the case, Figure 5 maps these female gods, and although not very numerous, they are not peripheral; while female deceased altar members are obviously of the isolated relational type, other female deities are quite well-connected. Two are present at five or more altars (they are the three red nodes in the center): Guanyin (five altars) and He xiangu 何仙姑 (five altars). I leave aside the case of the immortal Lan Caihe 藍采和 (six altars), an androgynous figure.

5.2. Systems

Wang Chien-chuan has argued for the existence of different systems of spirit-writing altars with a strong regional character: a system centered on Guandi in the southwest, another on Lüzu in Jiangnan (Wang 2016). Indeed, some of our 13 altars are centered around one god, and each of the central gods has a retinue of subordinate gods around him that often reveal instructions on their boss’s order: Liu Shouyuan 柳守元 for Lüzu, Zhou Cang 周倉 and Guan Ping 關平 for Guandi, Kuixing 魁星 and other stellar gods for Wenchang. These collaborators rarely intervene at altars where their boss is not active—but then, Wenchang and Lüzu are present at all altars but one, and Guandi in the majority of them. Another argument in favor of distinguishing various systems among the spirit-writing altars is that the central gods promote the dead (including altar members) into their own administration: since early-Qing revelations, devotees of Wenchang are promoted in the Jade Bureau, Yuju 玉局; in certain texts, devotees of Lüzu are promoted in the Central Palace, Zhonggong 中宮.
However, in our corpus, these systems are not exclusive. For instance, in the Quanshi guishen, deceased altar members are enrolled in the administration of several different gods (including Lüzu and Bixia yuanjun 碧霞元君). None of the altars in my corpus is exclusively associated with one system. This may be a bias; there certainly were altars affiliated with one system only, that produced single-god revelations (with a few additions by the subordinates); however, our multi-god altars show that they were largely understood as compatible within an extended divine society; an exception is the sectarian gods. The “Old Buddhas” (13 of them) all appear in one single altar (n°11, Zhilu baofa) and, therefore, none has more than one connection. Including more twentieth-century altars in the dataset would very likely change the picture, with many sectarian gods enlarging their sociality.

6. Conclusions

I hope to have shown that SNA provides us with an effective tool to take a comprehensive look at all gods involved in the discussions between humans and gods at spirit-writing altars, rather than focusing (more or less arbitrarily) on a few of them and without drowning in the endless lists of the hundreds of gods involved, even in a very limited dataset such as here. It does justice to certain individual gods and certain groups and categories that have not produced long, original revealed texts and, therefore, have not caught the attention of the few scholars who have read these texts, yet they play a crucial role in making all these discussions and revelations possible.
This opens up a whole field of research that is germane, not only to understanding spirit-writing, but more generally to Chinese religion and culture, as connections with gods have long been and remain an important element in Chinese ways of developing one’s identity. This field could develop in many directions; obviously, by expanding the dataset, it can refine my current hypotheses and identify more chronological, geographical, and other variations. It also offers the tantalizing perspective of offering an open-access collaborative and authoritative list of all Chinese gods, which would be very useful for many different purposes. Finally, on a more theoretical level, it may contribute meaningfully to discussions on the relations between subjectivity and social connectivity, i.e., between personal and social identities in the Chinese cultural context, a topic which had caused much ink to flow, but which may be fruitfully revisited from this rather unexpected angle.


This research received no external funding.

Data Availability Statement

The dataset used in this article is published online as part of all datasets of this special issue: see Marcus Bingenheimer’s introduction to this special issue.

Conflicts of Interest

The author declares no conflict of interest.


See for instance, as applied to Chinese religion, (Chau 2012). Faure (2015) uses a rather similar approach to Japanese gods.
On the history of spirit-writing, see (Goossaert 2022a).
The literature is documented in the CRTA open-access database (, accessed on 4 February 2023), which also allows tracing the productions of specific gods.
On this corpus, see (Hu 2017).
On this altar and its publication, see (Goossaert 2019).
On the lore around Han Xiangzi, see (Clart 2011).
Xiuxin zhuanxing jiujie huisheng, 315.
Quanshi guizhen, 3.19a passim.
(Naquin 2022) has shown that at the same time as they grow more present (at least in northern Chinese culture) in late imperial times, goddesses also tend to become more generic and less personalized.


  1. Primary

    Gaochun Pujitang zhi 高淳普濟堂志. Dai Fengyun 戴鳳筠 et al., comp., 1900. in Jiangsu jinxiandai shehui jiuji yu cishan wenxian congkan 江蘇近現代社會救濟與慈善文獻叢刊. Nanjing: Fenghuang chubanshe, 2015.
    Huangjing jizhu 皇經集註. DZ 1440.
    Jiushengchuan 救生船. 1905 edition, in Ming Qing minjian zongjiao jingjuan wenxian xubian, vol. 9.
    Jueshi zhengzong 覺世正宗. 1875 edition in Ming Qing minjian zongjiao jingjuan wenxian 明清民間宗敎經卷文獻. Compiled by Wang Chien-Chuan 王見川 and Lin Wanchuan 林萬傳, 12 vols. Taipei: Xinwenfeng, 1999, vol. 11.
    Quanshi guizhen 勸世歸真. 1889 edition, available at the ANU Library website.
    Taiyi Jinhua zongzhi 太乙金華宗旨. In Lüzu quanshu 呂祖全書 (1775), j. 49. Zhonghua xu Daozang, vol. 20.
    Xiuxin zhuanxing jiujie huisheng 脩心轉性救劫回生. Undated reprint of the 1872 edition in Shanshu baojuan yanjiu congshu 善書寶卷研究叢書. Li Zhengzhong 李正中, comp. Taipei: Lantai chubanshe 台北蘭臺出版社, 2014, vol. 5.
    Xixin yulu 洗心語錄. 1824 edition kept at the University of Chicago Library.
    Yuquan 玉詮. In Daozang jiyao 道藏輯要.
    Zhilu baofa 指路寶筏. Undated edition in vol. 11 of Ming Qing minjian zongjiao jingjuan wenxian.
    Zhishenlu 質神錄. In Wendi quanshu 文帝全書, j. 46–47. 1775 edition at Waseda University Library. Available at: Accessed 4 February 2023.
    Xu Zhishenlu 續質神錄. Appended to Zhishenlu, 1842 edition at Gest Library, Princeton University.
  2. Secondary

  3. Burton-Rose, Daniel. Forthcoming. Instantiating the Genealogy of the Way: Spirit-Writing in the Construction of Peng Dingqiu’s Confucian Pantheon. In Spirit-Writing in Chinese History. Edited by Matthias Schumann and Elena Valussi. Leiden: Brill.
  4. Chau, Adam Yuet. 2012. Actants amassing (AA). In Sociality. New Directions. Edited by Nicholas J. Long and Henrietta L. Moore. Oxford: Berghahn Books, pp. 133–55. [Google Scholar]
  5. Clart, Philip. 2011. The Story of Han Xiangzi: The Alchemical Adventures of a Daoist Immortal. Translated by Erzeng Yang. Seattle: University of Washington Press. [Google Scholar]
  6. Faure, Bernard. 2015. The Fluid Pantheon: Gods of Medieval Japan, Volume 1. Honolulu: University of Hawai’i Press. [Google Scholar]
  7. Goossaert, Vincent. 2019. Guérison, révélation et société locale: Gaochun en 1900. Etudes Chinoises 38: 37–64. [Google Scholar]
  8. Goossaert, Vincent. 2022a. Making the Gods Speak. The Ritual Production of Revelation in Chinese History. Cambridge: Harvard University Asia Center. [Google Scholar]
  9. Goossaert, Vincent. 2022b. Ritual Techniques for Creating a Divine Persona in Late Imperial China: The Case of Daoist Law Enforcer Lord Wang. Journal of Chinese Religions 50: 45–76. [Google Scholar] [CrossRef]
  10. Hu, Jiechen 胡劼辰. 2017. Wendi quanshu yanjiu: Qingdai Wenchang dijun xinyang de wenxianshi 《文帝全書》研究:清代文昌帝君信仰的文獻史. Ph.D. dissertation, Chinese University of Hong Kong, Hong Kong, China. [Google Scholar]
  11. Lai, Chi Tim 黎志添. 2016. Revisiting the Jingming Origin of the Taiyi jinhua zongzhi: Pan Jingguan and his Spirit-Writing Community in Changzhou during the Early Qing. Cahiers d’Extrême-Asie 25: 45–75. [Google Scholar] [CrossRef]
  12. Naquin, Susan. 2022. Gods of Mount Tai: Familiarity and the Material Culture of North China, 1000–2000. Leiden: Brill. [Google Scholar]
  13. Wang, Chien-ch’uan 王見川. 2016. Spirit Writing Groups in Modern China (1840–1937): Textual Production, Public Teachings, and Charity. In Modern Chinese Religion: 1850 to the Present. Edited by Vincent Goossaert, Jan Kiely and John Lagerwey. Translated by Vincent Goossaert. Leiden: Brill, pp. 651–84. [Google Scholar]
Figure 1. General overview. Altar-nodes in green, god-nodes in black, all node sizes proportional to connectivity.
Figure 1. General overview. Altar-nodes in green, god-nodes in black, all node sizes proportional to connectivity.
Religions 14 00217 g001
Figure 2. Disciples (red) and local gods (blue).
Figure 2. Disciples (red) and local gods (blue).
Religions 14 00217 g002
Figure 3. Historical figures (generals in red, poets in green, and Confucians in blue).
Figure 3. Historical figures (generals in red, poets in green, and Confucians in blue).
Religions 14 00217 g003
Figure 4. All thunder gods (in red).
Figure 4. All thunder gods (in red).
Religions 14 00217 g004
Figure 5. All female gods (in red).
Figure 5. All female gods (in red).
Religions 14 00217 g005
Table 1. Categories and altars.
Table 1. Categories and altars.
Type/AltarsA leiB baxC conD poeE budF qzhG locH immI cleJ topK gufL daoM genN dizTotal / 14
1xxx xx x x x 8
2xx x x x 5
3xxxx x x x 7
4xxxx x x x 7
5xxx x x x 6
6x x 2
7xxxx x x x 7
8xxxxx xxxx xx 11
9 x x xx 4
10xxxxxxxxxx x 11
11xx xx xxx x 8
12xxxxx xxxx xxx12
13xxx x x xx xx9
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Goossaert, V. The Social Networks of Gods in Late Imperial Spirit-Writing Altars. Religions 2023, 14, 217.

AMA Style

Goossaert V. The Social Networks of Gods in Late Imperial Spirit-Writing Altars. Religions. 2023; 14(2):217.

Chicago/Turabian Style

Goossaert, Vincent. 2023. "The Social Networks of Gods in Late Imperial Spirit-Writing Altars" Religions 14, no. 2: 217.

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop