Ghosts and Religious Life in Early China by Mu-Chou Poo is a sweeping account of ghosts and the lifeworlds to which they were integral, in seven short and easy-to-read chapters. Although the book title has “Early China” in it, the time span of the book goes beyond its conventional end of the Han dynasty (202 BCE–220 CE) and covers from the earliest dynastic times of the Shang and Zhou in the late second millennium BCE to the Six Dynasties in the sixth century CE. Additionally, the book includes a brief comparison between Chinese ghosts and their counterparts in four other notable cultural traditions in the ancient world. Despite the ambitious timeline and broad comparative perspective, this book focuses on delivering one central argument: that ghosts are “the other side of humanity” (5). This argument is intended as a throughline to weave together disparate textual sources, which range from early classics such as The Zuo Tradition (Zuozhuan), to excavated Qin and Han hemerological texts (daybooks or rishu), to early medieval anomaly tales (zhiguai), to early Daoist and Buddhist scriptures. Characterizing ghosts as “the other side of humanity” also crystalizes the book's social-cultural approach to the subject of ghosts, which Poo argues “can be examined as a social imaginary or cultural construct that complements the world of the living” (5).

Chapter 1 “Ghosts: The Other Side of Humanity” serves as an introduction to the book. Poo starts with a list of idiomatic expressions in modern Chinese language—a device that Poo uses throughout the book, easily bringing the ostensibly archaic topic of ghosts to the present—that contain the terms gui (ghost) and shen (god; spirit). Poo defines ghosts primarily as the “postmortem existence” of humans, but at times he also includes spirits and forces of a non-human nature, demonstrating that the difference between ghosts and other supra-human forces is often functional instead of ontological. He notes that the negative connotation often associated with gui and the traditional emphasis on the “rational achievements” of Chinese culture have relegated ghosts to being “no more than the subject of curiosity that deserves no serious study.” This book is clearly intended to right this wrong, arguing for “a long tradition of ghost stories in Chinese literature,” and pointing out that “ghosts feature prominently in popular religion throughout Chinese history” (12). Even as Poo covers a wide range of periods and genres of text, he returns to pick up these two threads again and again. His argument about ghosts as humanity's “other side” justifies his approach: a social-cultural examination of the world in which ghosts reside as “conditioned by the social and cultural contexts of the living” (5). This chapter concludes with a convenient summary, laying out the gist of each chapter that follows.

Chapters 2–6 largely follow the chronological order from the early dynasties of Shang and Zhou to the time just before the Sui–Tang reunification, but each chapter also has a topical focus. Chapter 2, “The Emergence of Ghosts in Early China,” introduces early conceptualizations of gui in both transmitted and excavated pre-imperial sources, although the precise date of the sources is not always clear. Beginning by excerpting sources from Shang “oracle bone” inscriptions (ca. thirteenth century BCE) up to excavated Qin “daybooks” (ca. third century BCE) and many transmitted texts, Poo points out that early textual references in general show that gui was not limited to human ghosts, but rather was “a generic term referring to the spirits or souls of human beings, deities, or even animals” (25). In a similarly generalizing fashion, Poo further tries to argue that the frequent co-occurrence of gui with other concepts such as shen that connote supra-human forces “suggests the later distinction between gui as the spirit of the dead [human] and shen as the spirit of the gods had not yet been clearly made” (25). Besides examining the conceptual and semantic range of gui in arguably highly selective pre-imperial sources, Poo also provides textual accounts of an array of images of gui and various ritual prescriptions to deal with them in everyday life. He makes an interesting observation that a gradual change occurred in the image of gui and people's attitude toward them: from being horrifying and thus feared and avoided, to being humanized through attention to their “feelings and needs” (54). Chapter 2 concludes with a brief look at the social and religious background against which people created social imaginaries including gui, arguing that “this world of ghosts and spirits was actually interpenetrating with the human world, and local deities and local ghosts simply existed as part of the human social fabric” (53). This echoes the main argument of the book as proposed in the introductory chapter.

Chapter 3, “Imperial Order and Local Variations,” moves Poo's investigation of gui and associated practices into the early imperial era of the Qin and Han (ca. third century BCE to third century CE). As the political unification ushered in unprecedented efforts to centralize the control of the vast empire, including the unseen realm of gods, spirits, and ghosts, regional practices and local cults persisted as an integral part of people's daily life long before the establishment of the empire. Poo points to tension in beliefs and practices about spirits as a “tug-of-war between centralization and local variation” (60). This chapter hence oscillates between the imperial/official policies and literate discourses about spirits and ghosts, on the one hand, and on the other hand common beliefs and practices—in particular of a mortuary nature—to which many, especially the non-elite subjects of the early empires, subscribed. Poo argues that while the former shared “a common goal of establishing a rational management system that could sustain the operation of the ruling apparatus and the financial need of the government,” the latter revealed “an unruly reality” (78).

Chapter 4, “Stories that Reveal the Dark Corner,” picks up a recurrent theme, namely, the image of gui. This chapter focuses on the shift from the horrifying, impersonal, and occasionally malicious, image of gui in the earlier sources to the emergence of colorful, and almost ironically lively, images of ghosts in the post-Han era of the Six Dynasties. Unlike the previous chapters, which rely on a variety of sources, this chapter primarily utilizes one genre of literary texts, anomaly stories (zhigui). Acknowledging that the anomaly stories were usually written or collected by the literate elite, Poo argues that they were still “based on the concept of ghosts and spirits circulating in society, and served to articulate or even to refashion the image of the ghost” (123). He tries to approach this body of literature by addressing two aspects of the anomaly stories: “the social need and the mentality of the audience” that kept the stories circulating (92) and the “intentionality of the ghost-story writers” (101). The chapter is hence divided between a somewhat random typology of ghost stories—“ghosts who speak their mind,” “vulnerable ghosts,” “female ghosts,” “vengeful ghosts,” “benevolent ghosts,” “ghosts in need of help,” and “self-asserting ghosts,” nearly all of which are selected and translated from one study by the famous modern writer Lu Xun (Guxiaoshuo gouchen, 1986)—and a short list of intentions of mostly anonymous story writers: “justice,” “morality,” “humor and skepticism,” and “proselytizing.” Poo emphasizes the role that the anomaly stories, which he hails as “the most important field for the growth and formation of the literary image of ghosts” in the Six Dynasties period, played in the “development of religious beliefs hereafter” (121). This observation transitions to his investigation of ghosts in Daoist and Buddhist religious traditions in the next two chapters.

Chapter 5 “Ghosts in Early Daoist Culture” and Chapter 6 “The Taming of Ghosts in Early Chinese Buddhism” can be seen as a paired examination of how ghosts were conceived (origin and image) and dealt with (rituals) in the two rival religious traditions. The chapters also address the differences in these two religious traditions when it came to gui, as well as their relationship with the long existing common concepts and practices about gui in pre-Buddhist China as already examined in previous chapters. Without differentiating among various Daoist traditions, Poo concludes from surveying a selection of early Daoist texts such as Taipingjing, Baopuzi (by Ge Hong, ca. 283–343) and writings by Lu Xiujing (406–477) that the Daoists saw gui as part of the spirit world and the “cosmic ether” (131). However, gui were not viewed as individuals but as a malicious “category” that should be controlled through exorcistic rituals prescribed and provided by Daoist religious specialists (135). Poo argues that exorcism was a main function of many texts in the Daoist Canon that “was unseen in the previous eras” (148). In this negative conception of gui as “the enemies of human beings” (145) and in their own active functions, the Daoist traditions differed from and competed with Buddhism, a foreign religion that eventually became “a natural component of Chinese society” (150).

Largely based on a 2017 chapter in a volume that Poo also edited (150n2), Chapter 6 begins with the connotation of gui in Chinese Buddhist sutras as denoting various Sanskrit terms of “dead humans and various kinds of demons” (154). Poo uses an effective example of hungry ghost to elucidate a key difference between the Buddhist conception and a pre-Buddhist one seen in the daybooks of the third century BCE examined previously. Despite the Chinese Buddhist adoption of the Chinese term egui “hungry ghost” to render the Sanskrit preta, Poo argues that the Buddhist hungry ghost—a pathetic being that eternally suffers from hunger—“falls into that condition because of crimes (being greedy, gluttonous, rapacious, etc.)” while the daybooks hungry ghost originated as a person who “was hungry and died of hunger,” but committed no crimes and suffered no moral deficiencies (157). The rest of Chapter 6 is devoted to the Buddhist ritualistic means to “tame” ghosts, and how those rituals interacted with both existing indigenous traditions and the Daoist exorcistic rituals through “mutual influencing and borrowing, and even competition or confrontation” (160). The morally charged nature of Buddhist concepts of ghosts, Poo argues, also contributed to their different proselytizing strategy. Buddhists encouraged people to follow the Buddhist path to live a moral life, in contrast to the Daoist practice of exorcism that aimed at persuading people of the power and authority of the Daoist priests (168–69).

Chapter 7 “Chinese Ghosts in Comparative Perspective” compares the conceptions, beliefs, and practices about gui in the Chinese context with their equivalents in four other cultural traditions in the ancient world (Mesopotamia, Egypt, Greece, and Rome). Given the limited space and the need to stay true to the main argument of the book that ghosts are “the other side of humanity” as a social imaginary and a cultural construct, Poo focuses on teasing out the “reciprocal relationship between the ideas of ghosts and society, and the cultural constructs that developed around this relationship” (174) in a cursory examination of each tradition in this short chapter. Poo argues that cross-cultural perspectives on ghosts—considered a psychological-cognitive universal—matter because they can shed light on the “cultural psychology” of each tradition and deepen our understanding of “a common humanity” (171).

As a whole, this book's ambitious scope and accessible style clearly attest to Poo's broad knowledge of the subject and his deep familiarity with the relevant sources. It is a confidently written book with many insights, smoothly walking the reader through an authoritative account of ghosts in ancient Chinese society. It will certainly appeal to general readers who prefer an expert overview with little of the interruption and distraction that a specialized monograph usually contains in the form of extensive scholarship review and meticulous footnotes presenting and explaining different viewpoints and sources. That being said, this book may more substantively benefit readers already familiar with the sources and the secondary literature on ghosts in ancient China. Poo has very good reason to keep the footnotes sparse and references minimal and expect more of his readers. He has published substantially in both English and Chinese on the topics of death and ghost, so it is not surprising that many of the references in the book direct the reader to his past scholarship. The desire to reach a broader audience without repeating his scholarly arguments may explain the absence of fuller contextualization of both primary sources that are excerpted—at times, seemingly randomly—throughout the book, and the limited secondary literature on ghosts. A fuller context would have not only better informed readers but also provided curious readers convenient guidance to explore the topic beyond this book.

To illustrate the importance of clarifying sources and broadening scholarship, I will discuss the section (2.4) on funerary rituals (40–42). Poo refers to “an inscription from a bronze vessel dated to the ‘Western Zhou’” as the “earliest textual evidence” for voluntary or compelled human-sacrifice of a subject himself to follow the death of his lord, “the Zhou King,” to the afterlife (41). Poo cites Zhang Zhenglang's 1981 annotation of the said bronze inscription as the source and as the basis for his claim that in the Shang and Zhou there existed “a belief in the netherworld in which the ghosts of the sacrificial victims were expected to serve the dead rulers” (41). As broadly reasonable as this interpretation of human sacrifices may be, one may find Poo's dating of this bronze vessel—known as the Ai Cheng Shu ding-tripod—to the Western Zhou perplexing. Since the said bronze inscription is not specified in the main text, if one is curious enough to check Zhang Zhenglang's original Chinese article and subsequent scholarship including a brief archaeological report of the vessel,Footnote ¹ one would not only find the vessel has been securely dated to around the fourth century BCE (i.e., the Warring States period of the Eastern Zhou in the traditional chronology), but also that Zhang's interpretation, later adopted by Hayashi,Footnote ² suggests that this inscription may show a belief that deceased—not sacrificed—subjects were expected to continue to serve their lord (in this case, a certain Lord Kang, not a Zhou King) in the afterworld. There is little indication that human sacrifices or the ghosts of human sacrificial victims are present in the inscription, as Poo suggests.

Another example is found in the section (3.4) on the bureaucratization of the netherworld in the early imperial time. Poo uses a document found in Fenghuangshan tomb M10 in which the deceased, named Zhang Yan, a self-claimed wudafu, reported his relocation (i.e., death) to the “Underworld Lord” (dixiazhu). Again, Poo cites a single source by Qiu Xigui—a preliminary reading, albeit by a prominent scholar of Chinese paleography, not too long after the poorly preserved tomb was excavated in 1973, and when this unknown genre of funerary documents had just become available to scholars—as the basis of his translation of the entombed document, despite the existence of later better-edited texts and studies. Notwithstanding, Qiu Xigui did correctly date the tomb to the fourth year of Emperor Jing of Han which would be 153 BCE, not 176 BCE as Poo writes (80).Footnote ³ Additionally, the self-claimed wudafu (not wufu as Poo reads because of the ligature marker after the graph fu that commonly renders it as dafu, which Qiu Xigui also made clear in his transcription,Footnote ⁴) is more likely a rank (the ninth order) in the Han twenty-rank system, not “a low-level official in charge of the affairs of the hamlet (li) in the Han local administration” as Poo claims (80). The number of unearthed funerary documents has dramatically increased since the 1970s, and this particular genre has been fruitfully studied in both Chinese and English.Footnote ⁵ Poo cites one recent work in English in this section but it is not for this particular document.Footnote ⁶ It may be prudent to cross-check primary sources and provide more secondary scholarship.

Other minor issues are mostly formal and editorial. Some Chinese characters are provided while others are not. Conventions are inconsistent; for example, a chapter from Yili is italicized as Shisangli on p. 41 but put into quotation marks as “Shisangli” on p. 43. Footnotes are sometimes insufficient; for example, for the anecdote of Emperor Wu of Han employing a recipe master to conjure the image of a beloved concubine after her untimely death on p. 63, Poo cites two sources for the anecdote as it is preserved in both Shiji and Hanshu but fails to note that these two records differ in the identification of the deceased lady, i.e., the Shiji account has Lady Wang while the Hanshu account has Lady Li. There are a number of typos; to just give one example, Zhuangzi is misspelled as “Zuangzi” and “Zhangzi” on p. 50.

These issues and examples do not render Poo's main argument unsound; nor do they diminish the value of the book. Yet a wider range and more up-to-date references matter for scholarly works, even for an otherwise excellent synthesis of the decades of extensive research by an established authority like Poo.