Meir Shahar presents us here with a very rich account of local religion in different parts of traditional China under the labels “King of Oxen” (niuwang 牛王) and “King of Horses” (mawang 馬王). He has united an abundance of sources, ranging from literature and anecdotes to ethnography and ritual manuals. He presents his findings in a very readable way, including analytical comments and appropriate methodological warnings when he is perhaps speculating a bit more. An important part of his argument is that the Indian taboo against eating the meat of milk-giving zebu (rather than all bovines) entered China through Buddhism and was a crucial factor in the taboo on eating beef (at the time meaning oxen in the north and buffalo in the south, neither of which were used for their milk). This taboo became increasingly important to literate groups (including, but not confined to the highest educated elite alone) from the tenth century onwards.

In ten chapters, with an introduction and an epilogue, Shahar takes us through the worship of horses and oxen in traditional China. In this review I provide the usual brief summary with some additions and comments, including possibilities of disagreement. Shahar’s work overlaps with and responds to the recent excellent English translation of Vincent Goossaert’s The Beef Taboo in China: Agriculture, Ethics, Sacrifice (Honolulu: University of Hawai‘i, 2025), by Barbara R. Ambros. Kings of Oxen and Horses is written in part as an alternative interpretation to the French original of Goossaert’s work from 2005 (240), since Shahar sees a much more important role for Buddhist beliefs than Goossaert, who connects the custom instead to ideological concerns of the intellectual elite of the pre-modern period. The custom spread with increasing literacy at lower social levels, which usually also entailed a certain adaptation to elite social values based on their interpretations of classical writings.Footnote ¹

The customs Shahar covers are a bit more disparate than the book title might suggest, since what applies in the southwest (where Shahar did most of his fieldwork) does not necessarily apply in the north (the principal region of documented temple worship of the two Kings of Horses and Oxen), or even in the southeast or the Yangzi River regions (which are not treated in any detail). Similarly, there were, and still are, considerable differences among different social layers in ritual practices. Thus, I would stress the connection between the beef taboo and literacy (with the accompanying inculcation of moral values more typical of the Classicist tradition). The historical process by which the taboo became accepted over the last ten centuries becomes much clearer from the account by Vincent Goossaert.

In the Introduction, Shahar explains why he selected oxen and horses, and what subgroups (such as donkeys and mules, where horses are mentioned) the larger labels refer to. He points out that they were all used as beasts of burden. He excludes the camel as animals from the northwest and therefore not part of what he seems to see as China proper (4). Now, much depends on our definition of the northwest, but surely Beijing and Song-dynasty Kaifeng would not count as outside the boundaries of “China Proper.” The famous Song painting with urban scenes of an idealized northern Song city (often interpreted as Kaifeng), entitled Qingming shanghe tu 清明上河圖, depicts the full richness of urban transport at the time, with a crucial role for equines, bovines, and camels alongside human labor.Footnote ² The scroll starts in the early morning with a convoy of donkeys led by one person into the city. As we move on, we see a man of some status riding a horse, accompanied by servants and a sedan chair (most likely hiding the man’s wife) moving towards the city, as well as two women riding donkeys in the other direction, with someone leading the first donkey and two persons on foot carrying some luggage. Closer to and within the town itself, we see equines in a variety of roles: waiting impatiently outside a shop, pulling carts (but never one of the big ships on the river, as used to be customary in my native Netherlands), or carrying people and heavy loads (several are shown crossing the famous bridge). Further on the picture, moving into the working day, we see several huge four-wheeled carts being pulled by two oxen. In front of the city gate, we pass the typical Chinese wheelbarrow with a single huge wheel, carrying a big load and pulled by a donkey with one person managing him and another behind the vehicle to keep the balance. A camel caravan is passing through the gate, probably destined for more distant regions. Camels remained a regular sight in Beijing city until 1976 when they were forbidden from entering (at least during the day—they still came in during the night in the early 1980s); I myself still saw a caravan crossing through the northern Chinese landscape in February 1980. While I agree that camels were relatively absent from religious worship, these mighty animals should not be relegated to oblivion as animals that are merely from the northwest.

Beyond the very rich descriptive level of the book, which was a sheer pleasure to read, Shahar’s principal argument is that the inviolability of cattle (bovines or niu 牛) came to China through Buddhism. Of course, he does allow for local adaptations (11–15). His study of religious practices relating to horses forms the other part of this study, but does not have a similar core argument, although here, too, he devotes much attention to the Buddhist context. The role of Buddhism is an important area of disagreement with the work of Goossaert mentioned above—something that readers can appreciate only when they read the two works side by side. A basic problem with this argument is that the growing worship of bovines in particular aligns poorly with the historical development of Buddhism in Chinese culture. In particular, the beef taboo took off for real only in the tenth century; moreover it did so especially among those educated male groups who less often believed in Buddhism. Furthermore, the northern Chinese cults of the horse and the ox have very little to do with Buddhist religious practice. Shahar is of course aware of this, and tries to explain it away in various parts of the book as the result of historical change. But then: why? He does not say. Furthermore, Indian Buddhist influence really often means the influence of Central Asian forms of Buddhism. Why should one attribute to “Buddhist influence” a northern Indian taboo forgotten as Buddhism changed in Central Asia (where there were no bovines, but horses and camels) and then rekindled again in China? There must be a local Chinese reason why this mattered.

The first chapter, “Iconographic Prelude,” is devoted to the visual dimension of the two cults, although Shahar returns to this topic in later chapters as well.Footnote ³ Indeed, he makes excellent use of visual evidence throughout the book. Shahar points out that depictions of these deities do not look like the animals that they rule (21–34), although aspects of their cults are represented iconographically in other ways. This is important in choosing the right translation for the Chinese terms, Horse King or King of Horses and Ox King or King of Oxen. While the book title selects the second of these options, the book itself consistently uses the first. To my (admittedly non-native-English) ear, “Horse/Ox King” creates a potential misunderstanding about the appearance of these deities, while the second option does not. Luckily, Shahar clears away any misunderstandings early on.

Even with the decline of equine culture after the Tang dynasty, Chinese religious culture remained rich in horse representations. Examples are the White Horse (baima 白馬) (the transformation of a dragon) carrying the Tang monk in the Journey to the West (xiyouji 西游記) tradition, a theme that has been appearing in funerary contexts and religious (and eventually “literary”) texts for many centuries;Footnote ⁴ and even more so the Red Hare Horse (chituma 赤兔馬) of Lord Guan that can still be found in many of his larger temples as a statue or in murals.Footnote ⁵ At the same time, military-grade horses were always hard to come by and (a bit like the camel), required long-distance, often foreign, trade. We should often read “mule” or “donkey” where the King of Horses is mentioned. On the other hand, the Qingming shanghe tu mentioned above seems to depict both horses and donkeys (maybe more expert eyes than mine can discern whether some are mules).

Proving that Chinese religious culture had Indian origins must go beyond pointing out similarities. Glen Dudbridge has criticized the theory that the figure of Sun Wukong, the Monkey King from the Journey to the West tradition, was based on an Indian prototype in the form of Hanumat from the Rāmāyana tradition.Footnote ⁶ In my own backyard, authors continue to attribute the savior figure of the Luminous King (mingwang 明王) in the late Yuan rebellion that led to the Great Luminous Dynasty (damingchao 大明朝) or Ming dynasty to Manichaean influence, although that stems from religious stereotyping and there is a perfectly plausible indigenous messianic origin as well.Footnote ⁷ Victor Mair has proposed many Indian influences on China, beyond Buddhism.Footnote ⁸ Robert van Gulik once devoted many pages to the possible reverse influence of what he called “Chinese sexual mysticism” on Indian Tantric practices—something that is equally difficult to prove explicitly.Footnote ⁹ Shahar himself has written on these matters as well, among other places in his wonderful book on the Indian origins of the important religious figure Nezha.Footnote ¹⁰ Influences certainly flowed both ways, and degrees of skepticism and nationalism vary among scholars; but similarities in themselves prove nothing.

Further confusion comes from translation practices.Footnote ¹¹ Thus, Shahar renders the perfectly Chinese name Shuicao ma mingwang 水草馬明王 as Water and Fodder Horse Vidyārāja (33–34), whereas undoubtedly everybody in traditional China would have read or heard this as “the Luminous Horse King of Water and Fodder,” with no Indian connotation.Footnote ¹² Translating an ordinary Chinese term with a presumed Sanskrit equivalent distorts or mis-construes how people actually believed in the particular deity. When a term is transcribed into Chinese characters it is of course perfectly fine to use Sanskrit transcriptions, but we should not unduly exoticize Chinese translations. Similarly, for us scholars today, “the Buddhist symbol of the lotus … attests [Guanyin’s] affiliation with the Indian-born faith” (29) as Shahar writes, but the connection was weaker for local believers. The lotus as plant and symbol in Chinese religious culture precedes the advent of Buddhism, and its popularity within Chinese Buddhism may have stemmed from its indigenous connections.Footnote ¹³ Likewise, Amitābha and his Western Paradise of salvation usurped a pre-existing slot of the Queen Mother of the West, who was widely depicted in Eastern Han graves.Footnote ¹⁴

In Chapter 2, Shahar argues that the King of Horses emerged from the Tantric Buddhist impact on Chinese religion during the Song. First, he discusses briefly the pre-Buddhist history of horses in China, which unsurprisingly included forms of horse worship and protective rituals for them, initially most clearly in a military context (38–45). Then he discusses the association of monkeys with horses (45–47), a horse-headed Avalokiteśvara (47–60), and the appearance of the Horse King as a local deity. Central to the argument for Buddhist influence is the Horse King’s third eye. But instead of having Tantric (i.e. Indian) origins, this could also reflect the indigenous ritual of opening a third eye (which I have undergone myself during fieldwork in Taiwan in 1992), and the depiction of this Heavenly Eye (tianyan 天眼) on the widely worshipped Second Lad (erlang 二郎), among other gods. Unless one wishes to explain all such third eyes as Tantric influence, of course.

In Chapter 2, Shahar adduces some fascinating evidence for Buddhist influence on the Horse King, using plays from the Southern Song and Yuan period, as well as discussing Daoist iconography and the appearance of the Indian (but surely not Buddhist) Śiva (60–65). But did people at the time connect all these pieces in a single body of belief? After all, Shahar also writes that Tantric forms of belief continued in Mongolia and Japan but not in China (70–71), whereas, as the cult spread in Ming and Qing, its Tantric origins were “forgotten” (65–70). On his evidence, China might never have experienced Tantric influence on the cult in the first place. The animal’s economic contributions to daily life could explain such a cult easily, with neither Buddhist nor Daoist influence.Footnote ¹⁵ In other words, the presence of such beliefs does not necessarily mean that they caused this cult.

Shahar calls the adoption of the horse-headed Avalokiteśvara (59) “testimony to the creativity of [the deity’s] native converts.” He does so, because the original Bodhisattva with this characteristic did not function as a tutelary deity of the horse. Seeing that as a Chinese “addition,” rather than an invention makes sense only if one assumes that characteristics shared by the Chinese and Indian horse-headed Avalokiteśvara/Guanyin resulted from influence. But the facial redness of the Horse King (53–54) already appears on other martial deities, most notably Lord Guan (the deified Guan Yu) early on. The connection with fire is likewise shared with Lord Guan, who is seen as an incarnation of the Fire Dragon Star in some local folklore.Footnote ¹⁶ Therefore, in Chinese thinking, the color and the fire element express fierceness befitting the martial nature of the Horse King. Even if one follows Shahar in linking the iconographic dimension of this Boddhisattva with the cult of the Horse King, Buddhism refined his imagery, but neither created nor changed the actual cult. More generally, the anthropomorphic cults for Kings of Horses and Oxen lack Buddhist paraphernalia, and most iconographic details, including their foreign faces (60) match the common depiction of deities as feudal (from marquis to king and emperor) or military figures (marshals, generals, and the like).Footnote ¹⁷ “Foreign” here also signifies martial and fierce.

Chronologically, the anthropomorphic cults to the Kings appear most prominently in Ming and Qing, much later than their presumed Buddhist and Daoist predecessors; but Shahar rightly asks whether the cults might not have existed long before (60–61), with the lag an artifact of available sources. Instead, I would tentatively argue for a thirteenth-century date for the temple-based worship of protectors for horses and oxen. Northern China has many early local inscriptions from the eleventh to fourteenth centuries, but few evince temples for the King of Horses or his divine colleague. Unlike for some other cults, Ming gazetteers do not suggest earlier dates. And the early thirteenth-century collection of northern anecdotes by Yuan Haowen, the Xu yijian zhi (Sequel to The Record of a Listener), includes a story on the beef taboo, but nothing on the worship of horses, oxen, or their protectors.Footnote ¹⁸ I suggest that the Ming–Qing rise of these two northern Chinese temple cults was inspired by earlier northern anthropomorphic cults (e.g., Lord Guan, King Tang 湯王) dating to the late-tenth and eleventh centuries.

Given the absence of references to a Horse King cult before the Ming period in northern China, we need to reassess the evidence that Shahar found in three southern Song or Yuan plays.Footnote ¹⁹ First, an early fourteenth-century play refers to the title of a lost comedy. The title refers to the deity, but not necessarily to a temple cult: Ma mingwang cunli hui jiaqi 馬明王村裏會佳期, which Shahar translates “A Happy Betrothal at the Horse-King-of-Spells Village.” Wilt Idema and Steven West note an alternative title, Minghuang cun yuanhui jiaqi 明皇村院會佳期, or “An Auspicious Meeting in a Temple in the Village of the Luminous Emperor.”Footnote ²⁰ The name of the Horse-King-of-Spells Village sounds very similar to Village of the Bright Emperor King (Emperor Xuanzong of the Tang). The ease of the change suggests that an actual cult with the specific indigenous name “Luminous Horse King”—a much more literal translation than Meir Shahar’s interpretation—played no important role in the comedy’s narrative. The play quoting the comedy’s title is preserved in the early fifteenth-century Yongle dadian 永樂大典, so it probably dates back to the fourteenth century.

Shahar’s second piece of dramatic evidence merely refers to the Horse King deity and not the temple cult. The third play is the famous Baitu ji 白兔記 or “White Rabbit Play” (61), about the tenth-century warlord Liu Zhiyuan 劉知遠 (895–948). For Shahar, the appearance of the Ma Mingwang Temple 馬鳴王廟 as a locale for crucial events evinces a Yuan date for the cult of the King of Horses. But the version Shahar uses is not a Yuan-dynasty play, but rather a seventeenth-century Ming production intended for an elite readership.Footnote ²¹ A version of the story dating to before 1372 does not include the Temple of the Singing Horse King, as most people at the time would have understood it in a non-Tantric way.Footnote ²² A printed play from Hangzhou, found in a late fifteenth-century grave, in the same narrative tradition, mentions the Ma Mingwang Temple, but mentions no elaborate festivities.Footnote ²³ Thus, the only thing we can prove is that there is a fifteenth-century reference to a Ma Mingwang Temple, which is entirely in line with the evidence from local gazetteers about increasing Ming-period foundations.

The seventeenth-century version of the play used by Shahar, with its detailed description of local temple festivities, corresponds nicely with the topic of Chapter 3: the actual cult activities of the King of Horses. Shahar starts with the wonderful find on the old book market of a manuscript called “Records of the Association of the King of Horses (mawangshe ji 馬王社記).”Footnote ²⁴ From this source, he describes local cultic organization. He then discusses the cult in Beijing itself (72–96).Footnote ²⁵ The guild of horse merchants in Beijing naturally provides especially good documentation (87–91). Shahar points out, as others have also shown, that among the ritual and protective activities of local temples, shadow puppet theatre is especially important (Chapter 2, 78; Chapter 4, 114–17; Chapter 6, 160).Footnote ²⁶

After treating northern temple cults to the Horse King, Chapter 3 moves to ritual scriptures for protecting horses from the southwestern provinces, discovered and collected by Professor Hou Chong of Shanghai Normal University (96–107). Such manuscripts from the second-hand book market generally lack context, but are still useful. No matter how fascinating, however, material from the southwest cannot explain much about the northern cults, and a key difference is that the southwestern rituals were not connected with cultic practice in a temple but were performed for the protection of horses by ritual experts. They show a connection with tutelary deities of the silkworm (99–104). Chapter 4 deals with the scanty legends of an equine god, primarily the northern Horse King (108–17).

The second half of the book moves to the topic of oxen. Chapter 5 deals with the important socio-religious phenomena of Ox worship and the beef taboo, which Shahar sees as historically connected. As I have already noted, there is a debate in the background with the work of Goossaert. First, Shahar demonstrates successfully the early existence of a temple cult to the King of Oxen (122–24), using a 1333 inscription for a Shanxi temple stage. David Johnson lists two additional temple stages for this deity, from 1283 and 1342.Footnote ²⁷ Thus the evidence is much clearer than for the King of Horses. Shahar also uses theatrical references, but again, late-Ming editions do not evince Yuan practice.Footnote ²⁸

The next step in the argument is a detailed description of the rise of the taboo on eating beef in the Song period (and a little before) (124–34). What Shahar cannot demonstrate is that northern Chinese worshippers were actually inspired by this taboo in their worship of the King of Oxen. First, most of his and Goossaert’s evidence of the taboo comes from the south and not from the north. Absence of information, in this case, cannot be entirely meaningless, and the geographic provenance of a source matters. For instance, when Shahar discusses the taboo based on a lengthy poem by Yao Shouzhong 姚守中 (1290), he does not inform readers that this author came from Luoyang (modern Henan). That is exactly the part of China where a cult for the King of Oxen would eventually thrive, yet the author makes no reference whatsoever to the cult and refers only to the bitter fate of the ox in local agriculture (131–34). Moreover, the primary concern of the author is not Buddhist, despite the appearance of the underworld deity King Yama, but rather the ideological point about the importance of oxen in agriculture.

Shahar begins Chapter 5 by noting that worshippers of the King of Oxen might well have eaten beef (135–38) but then claims that there is an undeniable, intimate connection of the bovine cult to the beef taboo (138). I do not find this convincing: the evidence is ritual scriptures from the late imperial and Republican periods (138–43), and from the southwest. What do they tell us about an earlier, northern cult? Further, the scriptures describe protection rites for bovines with no necessary link to a temple cult. Chapter 9 (see below) shows that temple cults for bovines did exist in the southwest, but a communal cult (temple) and ritual specialists (manuals) are different socio-religious phenomena. A ritual specialist is not going to transfer his rituals to the local community just like that, since they are his source of income.

While keepers of the beef taboo did at times draw on Buddhist principles, this is not the same as historical causation; just as in Goossaert’s materials, educated authors drew extensively on what they thought of as Classicist texts, yet that does not show causation, especially when, as in both cases, the texts were around for centuries before the beef taboo. Something important must have happened to make some people of the tenth century change their minds about eating beef. Likewise, Goossaert argues that the beef taboo arose from both agricultural practice and the ideological views of educated elite about its importance. Since buffaloes/oxen have a long history of use in agriculture, that use cannot in itself explain the dating of the taboo to the tenth century or later. Well-attested changes in educated elite ideology right about that time, during the Song period (if anything at a remove from Buddhism!) mean that Goossaert’s choice of ideological explanation makes more sense. If the critical ideas were Buddhist, why the time-lag after its general adoption? And why taboo beef alone? It seems to me that the taboo on eating beef was in fact the non-Buddhist version of caring for (some) animals, as expressed in the popular late-Ming Buddhist custom of Setting Free Life (fangsheng).Footnote ²⁹

From this point in Chapter 5, the author moves far back in time to discuss the ox in early Chinese ritual culture (143–49). Ox sacrifice was an elite practice—not surprising given the cost—and the imperial state always worshipped the ox, giving rise to decrees against its slaughter (147). Here Shahar rightly stresses a crucial difference with the beef taboo, for sacrificed oxen were eaten as part of the ritual. Worship and sacrifice therefore did not create a beef taboo, until Buddhism provided the animal with a voice and agency of its own (149).Footnote ³⁰ Shahar also argues that the possibility that oxen/buffaloes could be incarnations of one’s ancestors would have inhibited consumption (140–41, 161–62). But this would apply to all animals: Buddhist teachings hold that all animals are part of the same cycle of transmigration as humans, and concerns about cruelty to animals embrace dogs and even crabs (but rarely pigs).Footnote ³¹ Only beef-eating became taboo. People went right on eating pigs, not to mention drinking alcohol—another Buddhist taboo. A possible reason is that refraining from beef was easy for most non-Muslims. Ordinary people could not afford it anyway, and it did not matter in their rituals.Footnote ³² Further, the immensely popular cult to the Lower Yangzi deity of Shrine Mountain (cishan 祠山), sacrificed large numbers of oxen, but Goossaert suggests that this practice declined during the Yuan dynasty, most likely due to incessant elite attacks for Classicist, rather than Buddhist reasons.Footnote ³³ The state and local communities did not care one way or the other, but ideologically-minded intellectual elites did.

Chapter 6 makes extensive use of religious texts bought online. As in the case of the King of Horses, these sources on the King of Oxen come from southwest China and were ritual texts to heal and prevent animal illnesses. Whether more Buddhist or Daoist in background, they share many details (150–52). To make up for lack of a fieldwork context for these texts, Shahar followed them back to the countryside together with his colleague Hou Chong (see above) (153–56) and found additional materials, although present-day politics hampered proper fieldwork. Shahar then discusses how farmers long into the twentieth century used rituals based on such texts, with the addition of shadow puppet theater, to atone for their moral guilt in causing rinderpest (157–60 and 160–67). As Shahar points out numerous times, these rituals concern the southern buffalo (i.e., the draught animal with the long horns), which traditionally stands in for humans (166–67). He links this to an interesting observation, namely that southwestern ritual manuscripts partially reflect pre-Han Chinese local cultures. However, that would be yet another reason not to extrapolate from southwestern practices to northern ones with very different precursor cultures.

Shahar next discusses the famous Chan Buddhist metaphor in which herding the ox (or buffalo, more precisely, see esp. 180–83) refers to harnessing one’s inner nature (167–73). For him this is more evidence of the special place of the bovine in Buddhist culture (173–80). However, surely it was the importance of the buffalo in farming that led to the rituals and to the buffalo’s representation in painting, and thence to the Chan metaphor. Thus this metaphor stems from an indigenous ideological background, rather than a Buddhist one. Either way, I fail to see how the metaphor is connected to these southwestern rituals.Footnote ³⁴

Chapter 7 starts off with a discussion of the myth of the Herd Boy and the Weaving Girl. Although the myth goes back to ancient times, the ox started to play an active role in the story only in the later imperial period when the animal became divinized (184–88).Footnote ³⁵ The folktales with which Shahar begins this chapter, however, are hard to date precisely, and much of the chapter is an attempt to argue for the divine nature of the ox in earlier narrative traditions. Shahar traces in seventeenth-century Precious Scrolls and in the same late-nineteenth- and twentieth-century southwestern ritual texts a narrative in which the ox generally is a bodhisattva who has sacrificed himself to become the draft animal of humankind (188–96).

Shahar then takes a crucial step in his argument that the beef taboo descends from the sacred cow in Indian tradition and was transmitted to China by Buddhist missionaries, by linking the relatively late tradition of the ox as a bodhisattva to earlier Indian beliefs about the purity of the cow (the milk-giving zebu) (197–98). This is a rather big jump to take in terms of time, since he does not really present evidence from Tang or earlier of the ox as a bodhisattva. To strengthen the connection of the beef taboo with imported Buddhism, Shahar therefore proceeds to analyze traditions about Guanyin or the historical Buddha Shakyamuni riding an ox (198–208); the ox as a star in the Herd Boy and Weaving Girl legend (208–9); and Daoist and folkloric versions of the self-sacrifice of the ox to humankind (209–15). Again, most evidence is quite late and surely cannot prove much about origins. Instead, Buddhist and Daoist stories may have been primarily intended to rationalize the demanding usage of the buffalo or ox in hard labor over many years, despite their status as a family friend (one of many good observations by Shahar, see 182). In other words, these stories derive from the ideological concern about the ox in agriculture that became current in the eleventh century, rather than having been imported with Buddhism many centuries earlier.

In Chapter 8, Shahar develops his introductory reflection on the different meanings of niu 牛 with a comparison between the Chinese ox/buffalo and the Indian zebu as carriers of meaning (rather than pure biology). Although ox and buffalo already mattered in ritual practices before Buddhism entered China,Footnote ³⁶ the taboo on eating the milk-giving zebu of Indian and Buddhist narratives (217–24); the ritual significance of cow dung in Indian narrative (217–30); and the value of ox bezoars (230–38) all appear in Shahar’s argument for Indian influence on the special importance of bovines in Chinese Buddhism (217–19, 238–45). But the change from valuing the zebu for its milk to valuing the ox and buffalo as draught animals (esp. 245) is hardly a minor one. Since the ability of an ox to pull did not contribute to the zebu taboo in India, why would the presence of the ox-drawn cart in the very popular Burning House parable in the Lotus Sutra (221–22) contribute to a beef taboo in China?Footnote ³⁷ And one more time: why did the beef taboo develop so long after Buddhism entered China?

In Chapter 9, Shahar turns to the actual worship of the King of Oxen. Once again, we accompany him on his search for the traces of this worship, through texts and actual field visits. Local gazetteers confirm the wide distribution of a temple cult in the southwest, and Shahar was able to visit several surviving exemplars (246–49); but he cannot link the scriptures with the temple cults, even though he can describe southern temple festivals in wonderful detail thanks to a description by an early Ming prince, local gazetteers, and surviving folklore (249–54). The interesting Thai parallels are limited with respect to the beef taboo, since Thais and other non-Han southwesterners sacrifice buffalo—the latter because they are not Buddhists, says Shahar. Moreover, what happened in the southwest has limited or no relevance to understanding the northern cult.

Chapter 10 and the Epilogue are devoted to Daoist scriptures for protecting livestock (259–64) and the Christian saints in medieval Europe with a similar function (264–71). This last intercultural comparison makes a final, important, point: whereas European peasants prayed for the health of their livestock, their animals did not possess a soul and could not be saved in a following incarnation, as Chinese animals could (269–71). For Shahar this further proves that the transfer of a taboo on eating beef came from India to China with the Buddhism.

With these two shorter chapters, a fascinating forage through an extremely broad range of primary sources and secondary research comes to an end. The book is extremely well-written and full of insight and information, although it jumps rather easily between periods and genres. The author puts forward his case with great conviction and analytical care. He also shows how a broad approach to historical sources, requiring considerable expertise in different genres, brings a much better understanding of traditional Chinese society. Despite my analytical quibbles, there is much to enjoy in and learn from this book.