Bid ‘ a and evolving conceptions of the shari ‘ a in Qing and Republican China

In this article I revise the conventional account of the contestation over Islamic reform in late Qing and Republican China. I argue that previous scholarship has overlooked important similarities between so-called “ reformists ” or “ Yihewani ” and “ traditionalists ” or “ Gedimu. ” Based on an analysis of several texts and their exposition of the concept of bid ‘ a , I show that scholars associated with opposite sides of this divide in the early twentieth century shared a legalistic understanding of the shari ‘ a as a system of categories for classifying human action; and that this classificatory conception of the shari ‘ a differed from the practice-centered approach reflected in earlier Chinese Islamic works

reality were porous, multivalent, and ephemeral networks of people and institutions (DeWeese 2016;Eden, Sartori, and DeWeese 2016;Ross 2020). A separate vein of scholarship offers a similar corrective to the study of Islamic reform in China. Although not explicitly revisionist, several studies have documented earlier transformations in and debates over Islamic education and practice in China (Benite 2005;Chen Chongkai 1993;Ma Chao 2017;Na Jufeng 2013;Nakanishi Tatsuya, Morimoto Kazuo, and Kuroiwa Takashi 2012;Weil 2016;Zhou Chuanbin 2009) and historicized the evolution of the traditionalist Gedimu (Li Xinghua 1985).
In this article I build on these previous studies while proposing a different revision of the conventional narrative. My purpose is not to downplay the dynamism of earlier periods of the history of Islam in China or the significance of the religious and educational reforms that shaped Islamic learning and practice in those periods. Rather, I aim to clarify important but underappreciated similarities between reformists and traditionalists, or Yihewani and Gedimu, in the early twentieth century as well as underappreciated differences between both of them on one hand and their late imperial predecessors on the other hand.
The common element in question is the conception of the shari'a among ahongs on both sides of what is now seen as the Yihewani/Gedimu divide. In the early twentieth century, disputes over Islamic ritual proliferated, and while observance and rejection of particular practices, such as accepting gifts in exchange for reciting the Qur'an, raising a finger at certain points during worship, and the collective performance of supererogatory worship, came to be associated with opposing factions. However, underlying the divergent rulings ahongs issued on such practices were mutually intelligible arguments rooted in a shared understanding of the shari'a as a system of categories for classifying practice rather than simply a set of commandments and prohibitions.
This article contributes to the study of Islam in China by tracing how conceptions of the shari'a developed over time even as Chinese Islamic scholars continually invoked it to delineate ethical and cultural boundaries (Erie 2016;Tontini 2016). The classificatory conception of the shari'a differed fundamentally from the practice-centered one that predominated in much of the late imperial period. In the practicecentered approach, ethical conduct consists of observing rules, which, in line with neo-Confucian ideas about ritual, were central both to cultivation of personal character and the operation of the natural order. By contrast, in the classificatory approach, ethical conduct is not simply a matter of observing the rules but also of doing so for the right reasons and with the proper mindset. It is particularly important that one recognize the status of a given act according to "rulings of the shari'a" (ah  kām al-sharī'a)whether it is obligatory, recommended, neutral, detested, or forbiddenand more generally that one understand where an act fits in the shari'a's classificatory scheme. 1 In this respect, the classificatory conception of the shari'a exemplifies what Dresch and Skoda call "legalism," the construction and negotiation of the moral order through appeals to abstract categories and rules (Dresch 2012;Skoda 2012). From the classificatory perspective, thinking in terms of the shari'a is part of observing it.
We can see the commonality of the classificatory conception of the shari'a by comparing two texts, each of which is conventionally associated with either the reformists or the traditionalists, the Yihewani or the Gedimu: Xing Mi Yao Lu (Registered Essentials for Awakening from Confusion), composed and printed in Xi'an, Shaanxi Province in 1916, and Ming Zhen Shi Yi (Elucidation of Truth and Resolution of Doubts), composed in 1919-1921 and printed in 1921 in Kaifeng, Henan Province. The latter was in part a response to the former and marked an eruption of intra-Islamic tensions in Kaifeng. According to contemporary but admittedly contested accounts, the distribution of Registered Essentials in Kaifeng disrupted the local religious order. In response to what were seen as the false teachings of this work and its distributor, the leadership of Kaifeng's largest mosque 1 Hanafi jurisprudence, predominant in China, actually classifies according to six judgments, since it distinguishes two varieties of obligation above the "recommended": the "necessary" (Ar. fard  , Ch. zhu ming, tian ming, or falize) and the "obligatory" (Ar. wājib, Ch. dangran or wazhibu), while the other schools include only "obligatory." The Hanafi classification is based on a distinction between certain (qat 'ī) and probable (z annī) scriptural evidence. See Hallaq 1990, pp. 5-7;Hallaq 1997, p. 40. Some Chinese Islamic scholars added yet another category, sunna (a prophetic practice, Ch. sheng xing or sunnati), in between "obligatory" (above it) and "recommended" (below it). composed their own work, Elucidation of Truth. Important studies of Islam in Xi'an, Kaifeng, and Republican China more generally have cast these texts and their authors (introduced below) as representatives of the Yihewani and Gedimu positions, respectively, and thus of divergent interpretations of Islamic orthopraxy (Bai Zongzheng 2014;Ma Bin 2015a;Ma Chao 2015a;Ma Jing 2014, pp. 128;143-4). However, by comparing the two texts' content and argumentation, I will show that they reflect a common understanding of the shari'a as a system of classification, which differs from the practicecentered understanding of the shari'a reflected in the late imperial writings.
We can grasp the shift from the practice-centered to classificatory conception of the shari'a by examining how Islamic scholars in different periods have written about the idea of bid'a, or "innovation," which encompasses acts not prescribed in the sources of law and thus in some sense beyond the shari'a as a code of practice. In the next section I review some treatments of bid'a in several writings from the seventeenth-nineteenth centuries, including influential and widely available Chinese printed texts as well as Chinese and Arabic inscriptions in Kaifeng. I then turn to the confrontation in Kaifeng around 1919 and compare Registered Essentials and Elucidation of Truth and elaborate the classificatory conception of the shari'a that they have in common despite their divergent conclusions about normative Islamic practice.
The practice-centered approach to the Shari'a in late imperial China The predominant understanding of the shari'a among Islamic scholars in late imperial China was practical, in the sense of focusing on practice. Islamic scholars articulated the shari'a as a set of divine rules governing ritual. To observe the shari'a was to observe these rules. Scholars writing in Chinese expressed this understanding by rendering the Arabic shari'a as li, "ritual," or li fa, "ritual law" or "ritual method," in Chinese (Frankel 2011, pp. 93-114).
This is not to say that late imperial Islamic scholars did not theorize the shari'a; they did, but in cosmological rather than legalistic terms. Liu Zhi (1660Zhi ( -1730, the Jiangnan-based scholar who authored the 1706 Tianfang Dian Li (Rites and Norms of Islam), arguably the most influential Chinese work on the shari'a in the late imperial period, defined the shari'a as the "ritual vehicle" (li cheng). The shari'a was the first of three "vehicles" through which a person became one with the divine. On a more mundane level, it was through proper worship and other rites that a Muslim developed a proper disposition and right intentions. At the same time, the three vehicles together constituted fa, which Liu defined as the order or law of the "constancies of attainment and change of every motion and cessation." For Liu, the shari'a connected personal cultivation through ritual with the orderly activity of the universe (Frankel 2011, pp. 75-7;Liu Zhi 2008, p. 455).
We can also consider what elements are marginal to or absent from this practice-centered approach to the shari'a. A general lack of concern with classification according to the ah  kām and a lack of legalistic reasoning in shari'a categories characterize important Chinese Islamic writings from the late imperial period. We turn now to three Chinese-language works, including the aforementioned Rites and Norms of Islam by Liu Zhi, to see how these characteristics were reflected in the way some Islamic scholars from this period wrote about the concept of bid'a.

Some late imperial treatments of Bid'a
The definition of bid'a has been a subject of Islamic legal scholarship since at least the ninth century CE (Fierro 1992;Rispler 1991). A legal and ethical system that took the infallibility and perfection of revelation and the prophetic example as a fundamental tenet inevitably encountered the challenge of assessing acts not done in earlier times and/or about which an explicit ruling could not be found in the sources of law. Without abandoning the deference to revelation and the sunna (the acts and sayings attributed to Muhammad), jurists sought to classify bid'a into different categories. Some innovation (in the most literal sense) was unavoidable; the question was which sorts of innovation harmed religion and which sorts did not. Bid'a was subject to ethical-legal assessment, and a qualified jurist could International Journal of Asian Studies determine the h  ukm (ruling) for a particular innovation just as he could for any act mentioned in the sources of law. In theory, for some jurists at least, it was possible for a bid'a to be "recommended" or even "obligatory," while others were "forbidden," "detested," or "neutral." Despite this history in other parts of the Islamic world, bid'a was not a central concept in Chinese Islamic discourse for most of the late imperial period. The Jiangnan-based scholars who studied and translated Islamic and Arabo-Persian learning certainly believed that some practices were right and others were wrong, but they generally chose to make these judgments (in Chinese) using the generic and not specifically Islamic moralizing language of "deviant teaching" (xiejiao), "sinister way" (zuodao), "impropriety" ( fei li), or "heterodoxy/heresy" (yiduan). Three works composed in the late 1600s and early 1700s show that while these scholars understood bid'a in unambiguously negative terms, generally as something that contravened or was external to Islam, they did not elaborate it as a technical category of law or attempt to classify particular innovations according to the ah  kām.
The first work is a primer, Xiu Zhen Meng Yin (Guide to the Cultivation of Truth), completed in Nanjing in 1672 by Wu Zunqi (c.1598-1698). 2 This primer, mostly consisting of a translation of a Persian work, 3 refers to the Arabic term bid'a at the end of the third chapter, "judgments of the religious rules" ( jiao gui duan fa), which enumerates and provides examples of seven classifications: obligatory, necessary, normative practice of Muhammad (sunna), recommended, permitted, forbidden, and detested. The definition of bid'a concludes the chapter: When someone compels or views as permissible a matter that God has forbidden, or on the contrary rejects a matter that is permitted (lit: accords with principle)both [cases] fall under [the category of] unbelief (lit: a foreign way). Anyone who practices Islam, if he [acts] according his personal opinion and does not accord with sunna (lit: prophetic practice), it is bid'a (biede'ate).
The passage juxtaposes bid'a with sunna and equates deviating from the sunna with committing bid'a. This contrastive formulation may resonate with modern Salafi and revivalist calls to do away with the corrupt practices that emerged in period between the time of Muhammad and the present and to emulate Muhammad and his companions as closely as possible. These calls convey a clear temporal or sequential aspect of bid'a; bid'a is innovation, that which appeared after the perfect model provided by Muhammad and his companions. However, no such temporal aspect is present in Wu's brief treatment of bid'a.
Indeed, there is no justification based solely on Wu's text to translate bid'a as "innovation" rather than "deviance," "heterodoxy," or "wrong practice." The short gloss translated above indicates that Wu did not conceive of bid'a (biede'ate) as an important legal concept. The primer mostly addresses the fundamentals of Islamic belief (Ch. yimani, Ar. īmān) and the essential Islamic rituals and their varied ranks (in terms of obligation), conditions, and disqualifications. Notably, bid'a does not appear in the final chapters of the primer, which cover death ritual, an aspect of Islamic practice often associated with contestation and accusations of innovation (Becker 2013;Bowen 1993;Halevi 2007).
This lack of elaboration of innovation as a legal concept becomes clearer if we consider another work, the roughly contemporary Tianfang Wei Zhen Yao Lüe (Summarized Essentials for Preserving Truth in Islam). Composed by Ma Junshi (c.1600-c.1680), a Confucian-educated Islamic scholar from and active in Nanjing, Summarized Essentials comprises a set of over sixty hypothetical scenarios, each of which was ultimately deemed an act of unbelief or "a foreign teaching" (wai jiao) (Ma Jing 2011). Such a polemic is precisely the sort of text in which one would expect to find invective against rampant bid'a. The term yiduan (meaning heresy, heterodoxy, or literally "different end"), a common Chinese rendering of bid'a, did appear in the opening line of the preface to later edition of the book printed in 1892. But the neither yiduan nor any transliteration of the Arabic bid'a appears in original text of Summarized Essentials (Ma Junshi 2008, passim). Their absence in a work dedicated to identifying religious offenses suggests that "innovation" (in the sense of a novel as well as deviant practice that emerged after the time of Muhammad) was not a central concept in seventeenth-century Chinese Islamic thought.
The term yiduan, which on a literal level does not connote the aforementioned temporal or sequential sense of "innovation," did acquire such a sense in Liu Zhi's 1706 Norms and Rites of Islam. Notably, Norms and Rites does not include a section devoted to the ah  kām. For Liu Zhi, the question of heterodoxy ( yiduan) also seems unrelated to shari'a classification. He defines yiduan briefly in the following passage, near the end of the work: Do not commit heterodoxy. Do not follow deviant teachings. Do not give credence to any sorcerers or witches.
What in ancient times did not exist but later people have created is called "heterodoxy" (yiduan). Contriving something that has no basis in reason is called a "deviant teaching" (xie shuo). One who spreads occult doctrines and falsehoods about evil forces to deceive and confuse people is called a "sorcerer or witch" (men who do so are sorcerers; women who do so are witches). None [of the above] may be done or believed. If you even slightly concern yourself with something doubtful, it is near to apostasy. (Liu Zhi 2008, p. 586) Here yiduan has a clear temporal aspect: it refers to those practices and beliefs that arose after an idealized ancient time and thus should be understood as a deviant or heterodox innovation. But in the context of "deviant teachings" and "sorcerers and witches," and in the absence of any discussion of the ah  kām, this conception of bid'a as yiduan does not amount to a technical category of the shari'a. As in Wu Zunqi's primer, yiduan here is an unambiguously bad but also nebulous category, not a precise legal term referring to a class of actions and beliefs with a particular relation to other classes within a larger taxonomy.

A Mosque inscription in Kaifeng
So far we have examined the treatment of bid'a in three important Chinese-language expositions on Islam from the seventeenth and eighteenth centuries. To what extent did their practice-centric approach to the shari'a, with its lack of concern with classification, influence or resemble local understanding in Henan, and within it Kaifeng? These three works originated in the Jiangnan region (around Nanjing), but as Benite and others have shown, they were composed by members of scholarly network that included numerous schools in Henan (Benite 2005;Na Jufeng 2013). Ma Chao has documented ongoing intellectual ties between mosques in Henan and in other provinces, both in terms of traveling scholars originally from Henan and other scholars who studied and taught in Henan (Ma Chao 2018). Liu Zhi, author of the Rites and Norms of Islam, visited Zhuxianzhen (about 15 miles southwest of Kaifeng) and consulted Islamic texts held in private library there (Weil 2016, pp. 117-8). To get a clearer sense of local understandings of the shari'a, we can examine a set of inscriptions held in Kaifeng's Great North Mosque, meticulously transcribed, analyzed, and partially translated in previous studies (Ma Chao 2016;Nakanishi Tatsuya, Morimoto Kazuo, and Kuroiwa Takashi 2012;Weil 2016).
The Inscriptions in question include Arabic and Chinese writing and date to 1840. The Arabic text was copied from an earlier inscription from 1743-44, also held in the Great North Mosque. The inscription lists 13 "well-known practices" that became the source of disagreement among local Muslims. The inscription opens with a verse from the Quran (4:59) 4 and then locates itself in both 4 "You who believe, obey God and the Messenger, and those in authority among you. If you are in dispute over any matter, refer it to God and the Messenger, if you truly believe in God and the Last Day: that is better and fairer in the end." (Abdel Haleem 2004, p. 56) space and the Islamic legal tradition, positioning itself against bid'a: "This stone tablet is the record of those who have preserved the Book and the prophetic tradition from the Prophet. What it consists of is disseminated among the community of followers of Abu Hanifa, God's mercy upon him. The way of the North Mosque of Kaifeng was entirely traditional, Sunni, and Hanafi, keeping afar from innovation (bid'a) and heresy and meticulously embracing evidential scrutiny." 5 The inscription lists the 13 practices, which have to do with worship, recitation, and death rites, as well as 27 Arabic and Persian texts as evidence for their scriptural basis. The 1840 inscription in Kaifeng reproduces the original Arabic and provides a summary of the thirteen practices (but not the rest of the Arabic inscription) in Chinese.
The enumerated practices are ritual instructions. In general they are not explained according to the ah  kām scheme of moral classification (whether something is "obligatory," "recommended," "licit," "detested," "forbidden"). Only the tenth rule, concerning the full prostration for the nighttime witr prayers ("The tenth is the two prostrations that are deemed recommended after the witr"), deploys a h  ukm category ("that are deemed recommended," allatān tustah  babān, referring to the judgment "recommended"); and only the fifth rule, concerning wearing shoes for the janāza funeral prayer ("the fifth is the performance of the funeral wearing shoes, because it is not [done as] worship"), is a complex logical claim with a "because" construction. The other acts are listed in straightforward, unjustified terms; for example, "the eleventh [act] is the pointing with the index finger during the shahāda" (the testament of faith).
To the extent that the inscribed rules involve classification, it is represented as esoteric knowledge available to the few people in the community who could read Arabic. The shari'a reasoning limited to the fifth and tenth Arabic rules was not conveyed in the 1840 Chinese summary. Both Chinese renderings are compressed and logically simplified. The fifth rule becomes "Whenever there is a funeral and burial and the janāza is held for the deceased, shoes must be worn and [people] may not go barefoot" (the "because it is not [done as] worship" is omitted); the tenth rule becomes "The final prophet, after the witr (weiteilie) prayers, kowtowed twice" (the "deemed recommended" is omitted). For the majority of congregants not literate in Arabic, recognizing and affirming this classification was not enjoined as part of performing the ritual.
Taken together, the seventeenth-and eighteenth-century Chinese expositions on Islam and eighteenth-and nineteenth-century inscriptions in Kaifeng exhibit a practice-centered approach to the shari'a. These sources attest to the importance of the shari'a in scholastic and communal articulations of Islam, but also to the particular way in which the shari'a was, and was not, understood. Of particular significance for our purposes is the lack of classification according to the ah  kām or concern with bid'a as a technical category of legalistic reasoning and argumentation. As we will see now see, these features would become central to local scholastic engagement with the shari'a in early twentiethcentury Xi'an and Kaifeng.

A confrontation in Kaifeng
We now turn to the confrontation between the two texts I argue share the same classificatory conception of the shari'a in contrast to the practice-centered approach exhibited in the Qing-era writings reviewed in the previous section. I will first provide some background on the authorship of the two texts and then describe their confrontation in Kaifeng.
Registered Essentials for Awakening from Confusion, today associated with the Yihewani reformist position, was composed in Xi'an in 1916 by the Tianjin-born ahong Xiao Dezhen (1884Dezhen ( -1947  ( Ma Jing 2014, pp. 143-4). Xiao was a cousin of the Tianjin-born Wang Jingzhai (1879-1949, one of the celebrated "four great imams" of modern China. He was a disciple and son-in-law of Liu Yuzhen (1861-1943), a Xi'an native who made the hajj pilgrimage twice and who took up a post as cleric at the Sajinqiao Mosque in his home city in 1912. According to Ma Bin, in 1914, Liu issued a series of twelve rulings "helpful to the living and faithful to the dead," i.e. concerning matters of worship and social life as well as death ritual. Some supported Liu Yuzhen's reforms, including Xiao Dezhen, who decided to demonstrate the correctness of many of his teacher's rulings, as well as a few others, in Registered Essentials. But others opposed these reforms and were angered by the distribution of Xiao's work among the city's seven main mosques. Amid increasingly bitter relations between Liu and Xiao's supporters and opponents, the head of the city's oldest mosque sought official intervention. A public "scripture debate" ( jiang jing) was organized, but it failed to ease tensions. Xiao and Liu were expelled from Xi'an, though both would find their way back to the city in later years (Ma Bin 2015a; Ma Bin 2015b).
Elucidation of Truth and Resolution of Doubts, associated with the Gedimu position, was composed between 1919 and 1921 in Kaifeng by leaders of the Great East Mosque, the largest congregation in the city (Ma Chao 2015a; Ma Jing 2014, p. 128). The more than two-dozen rulings that constitute the main text were originally written in Arabic in 1919 by the mosque's head cleric Hong Baoquan (1860s-c.1936), who titled the work Munīr al-Dīn (Illuminator of Religion). A group of five other mosque leaders and congregants then then summarized and elaborated (they are not always strict translations) the rulings in Chinese and added a bibliography and Chinese title, Elucidation of Truth and Resolution of Doubts. Three other prominent congregants, including one of the mosque headmen, are credited with editing the work, and one of them contributed a preface, dated October 1921 (Hong Baoquan et al. 2008, pp. 337-40). 6 Earlier studies have demonstrated that one of the developments that prompted the writing of Elucidation of Truth was the religious activism of Ma Guangqing (1880Guangqing ( -1951. Born just outside the city, Ma headed west as a young man in pursuit of religious learning and became a disciple of Hajji Guoyuan, probably when the latter was serving as ahong in Ankang, Shaanxi Province. Ma Guangqing returned to Kaifeng in 1917 and began preaching in favor of religious reform by one of the city's numerous Muslim-owned bathhouses close to both the Great East Mosque and the Wenshu Temple Street Mosque (henceforth "Wenshu Mosque") (Ma Chao 2015a; Ma Chao 2015b; Guo Qingxin and Guo Chengmei 2015). It is unclear if at this point Ma Guangqing already had a program of reform comparable to Liu Yuzhen's rulings, but one rather partisan (in Ma's favor) report by Wang Jingzhai, who visited Kaifeng in 1919, says that Ma refused to accept gifts in exchange for reciting the Quran (Ma Zuowu (Ma Guangqing) and A Xiao (Wang Jingzhai) 1921). His opposition to this practice probably won him enemies among the city's numerous "floating ahongs" (san ban ahong), who lacked formal employment and remuneration from a mosque and therefore depended on such gifts for their livelihood. But Ma evidently managed to win over some influential supporters in the Wenshu Mosque congregation, which hired him as cleric ((Ma Chao 2015b, pp. 197-8;Wang Jingzhai 2006, p. 11,217)). In 1919, Hong Baoquan was hired as cleric at the Great East Mosque (Hu Yunsheng 2007, p. 153).
It was during Ramadan of that year, around June, that the two ahongs first clashed. The precipitating issue was whether the scent of youxiang, fragrant oil-fried cakes traditionally eaten at the end of Ramadan, compromised the fast (Ma Chao 2015b, p. 198). This led to more disputes concerning gift-giving and recitation, worship, veiling, tajwīd, donning mourning robes, and other matters. As Hong Baoquan set to work refuting Ma Guangqing in writing, Shang Qingxuan (dates unknown), the imam of the Great East Mosque, took legal action. Shang petitioned the provincial government to intervene, alleging that Ma Guangqing belonged to a group recently banned by the governor of Xinjiang and was "coercing people to join his association" (qiang ren ru hui) (Da gong bao 1919). 6 The copy of Elucidation of Truth included in the Huizu Diancang Quanshu anthology is blurry or otherwise unclear at points. Where illegible, I have consulted a separate copy of the same edition provided to me by a Chinese scholar.
Shang's initial petition failed, but a subsequent appeal to the Ministry of the Interior ultimately succeeded in getting Ma Guangqing expelled from Kaifeng in May of 1920though he would manage to return again two years later (Ma Chao 2015b, p. 198; Ma Zuowu (Ma Guangqing) and A Xiao (Wang Jingzhai) 1921; Wang Jingzhai 2006, p. 11,217).
Thus far the story behind the Elucidation of Truth in Kaifeng seems similar but not directly connected to that of Registered Essentials in Xi'an. Evidence for a connection between the two can be found in a report on the dispute from May 1920, after Ma's expulsion. The report quotes part of imam Shang Qingxuan's complaint to the Ministry of the Interior, which accuses Ma of following a "New New Teaching" and distributing "numerous copies of the Xing Mi Lu, composed by Bao Dezhen…" (Shen bao 1920). Here, either Shang or someone in the chain of bureaucratic transmission made a few errors; the book in question was Xing Mi Yao Lu ("Registered Essentials for Awakening from Confusion"), not Xing Mi Lu ("Registry" or "Record of Awakening from Confusion"), and the author's surname was Xiao, not Bao, and the characters for Dezhen were homophones of Xiao's actual given name. An irony almost certainly lost at least on the officials involved was that the erroneous Xing Mi Lu was actually the title of a seventeenth-century translation by She Qiyun (1630-1703), an Islamic scholar who taught in central Henan (Wu Jianwei and Zhang Jinhai 2010, p. 23). In a final plea defending himself against Shang's accusations, Ma Guangqing claimed ignorance of the specified text: "What teaching is the New New Teaching? What book is the Xing Mi Lu? I do not know. At present although the books distributed by each teaching are numerous, I, [Ma] Guangqing, based on my affiliation with Islam, only know to act according to the classical texts of Islam. I do not concern myself much with anything else" (Da gong bao 1920).
It is possible that Ma Guangqing never distributed Xiao's Registered Essentials. It is also possible that he was coyly taking advantage of the erroneous title in the accusation, since he had never distributed a book by that title. We should note that to the south in Changde, Hunan, a similar petition was brought against Li Renshan (1881Renshan ( -1931, who had studied together with Ma Guangqing under Hajji Guoyuan, and who in 1919 was also accused of stirring up trouble using the (properly titled) Registered Essentials (Hunan sheng difang zhi bianzuan weiyuanhui 1999, pp. 27: 308-12; Wang Jingzhai 2006, p. 11,216). The Changde petition failed, and we cannot rule out the possibility that these were copycat framings, perhaps inspired by knowledge of Liu and Xiao's recent expulsion from Xi'an. But even if Ma and Li were not directly involved in its distribution, Registered Essentials circulated widely in this period (Wang Jingzhai 1939, p. 21). Close ties between mosque congregations in Xi'an and Kaifeng make it likely that Registered Essentials would have found its way to the latter quickly. 7 Moreover, as we will see below, the content and structure of Registered Essentials and Elucidation of Truth, suggest that the latter, written by Hong Baoquan et. al in Kaifeng, was a response to the former, written by Xiao Dezhen, with the alleged intermediary of Ma Guangqing.
The classificatory conception of the Shari'a in the early twentieth century I turn now to a comparison of Xiao's Registered Essentials and Hong's Elucidation of Truth, ostensibly representative of the Yihewani and Gedimu positions, respectively. I first examine the two texts' treatment of bid'a and then broaden the comparison to other aspects. Although the two texts reach divergent conclusions on substantive questions of normative practice, their argumentation is similar and reflects a shared understanding of the shari'a as a system of classification.
Bid'a as a technical category of the Shari'a I showed earlier how in some of the major Chinese-language expositions of Islam in the seventeenth and eighteenth centuries, bid'a held an unambiguously negative but also somewhat vague meaning, roughly interchangeable with "heresy" or "deviance." By contrast, for both Xiao Dezhen and Hong Baozhen, bid'a is a technical category of the shari'a referring to innovation and contrasted against sunna, the normative example of Muhammad. According to Xiao, For those of our religion, upholding the prophetic practice (i.e. the sunna) is a matter of the utmost importance. If one contravenes [the sunna], then it is heterodoxy ( yiduan), called bid'a (bide'erti) in the Western language (i.e. Arabic). Heterodoxy is those matters innovated in religion after the death of the Noble Prophet and all the Great Worthies (i.e. the salaf, the first three generations of Muslims) (Xiao Dezhen 2008, p. 170).
In its dichotomization against sunna, Xiao's definition of bid'a resembles that of Wu Zunqi examined above. However, for Xiao the concept of bid'a also has a clear temporal aspect ("after the death of the Noble Prophet…"). Moreover, whereas Wu dealt with the concept only in passing, Xiao addresses it throughout his work.
Hong Baoquan likewise introduces the sunna/bid'a dichotomy in the second chapter of Elucidation of Truth, "Explanation of the Differences Between Sunna and Bid'a ( yiduan)," but takes it in a different direction. Hong reasons that if one is uncertain about whether something is bid'a or sunna, it is preferable to avoid that thing, since it is worse to commit a bid'a than to neglect something that is sunna. Thus, unless one is certain, one should not automatically follow what one thinks to be sunna. Hong employs this rationale later on, in the chapter concerning raising a finger during worship, in which he first presents the divergent rulings on the shari'a status of the act (some say it is sunna, some say it is detested or even forbidden) and then concludes that it is preferable to follow the ruling that it is not permitted. Xiao, by contrast, argues that raising the finger during the shahāda is sunna and must be done (Hong Baoquan et al. 2008, pp. 356, 416-25;Xiao Dezhen 2008, pp. 187-91).
Having repudiated the idea that one should always do what one thinks is sunna, Hong then turns to counter the idea that bid'a is always bad. In his third chapter, "Explanation of the Five Types of Bid'a ( yiduan)," Hong applies the ah  kām to the category of bid'a, opening the possibility for licit, recommended, or even obligatory bid'a in addition to detested and forbidden bid'a: … What is more, bid'a is divided into five types. The first is harām. The second is "obligatory" (wājib). For example, establishing proofs to repudiate all those who go astray, and studying grammar to understand sunna and the Quran. The third is mandūb. For example, all those innovations of such kind as repairing schools, halls, shops, and homes, all of which are good and which were not done or found in the time of the Noble Prophet. The fourth is makrūh. For example, adorning mosques. The fifth is mubāh. For example, often preparing a tasty meal or good-quality clothing (Hong Baoquan et al. 2008, pp. 357-8).
Hong's application of the ah  kām to bid'ato acts that were by definition beyond the explicit religious commandments and prohibitions found in the sources of lawreflected an understanding of the shari'a as a set of ethical-legal categories rather than simply a ritual code. Xiao Dezhen's engagement with the concepts of bid'a and sunna reveals a similar understanding of the shari'a as system of categories for classifying human action. In contrast to Hong, Xiao rejects the possibility of good bid'a, a notion he addresses at least twice in the text (Xiao Dezhen 2008, pp. 171-2, 183). This categorical judgment follows from Xiao's strict definition of bid'a as matters "within religion" ( jiao zhong). Whereas Hong defined bid'a broadly as anything "not done or found in the time of the Noble Prophet," and thus as a category applicable both good and bad acts, Xiao narrows bid'a to matters of religion, in which there can be no legitimate innovation. At the same time, Xiao leaves room for classification according to the shari'a of acts that are not explicitly enjoined by the Quran. Xiao opens the chapter on bid'a and sunna with a call to scrutinize Muhammad's behavior in its entirety: "Evidently, the Final Prophet received the True Scripture (the Quran), which fully established the way of [our] religion; none of it was [his] personal whim. What [he] did was all in accordance with what the True Scripture instructs, and if [there was something he did that] the True Scripture had not [explicitly] instructed, then there must be a concealed instruction (an yu) [for that thing]" (Xiao Dezhen 2008, p. 170). In other words, according to Xiao, one must take the "prophetic practice," i.e. the sunna, as the standard for assessing all actions, not just matters of religion.
Views on the possibility of good bid'a did not align neatly with the Yihewani/Gedimu or reformist/ traditionalist divide. The renowned Guangzhou-based ahong Ma Ruitu, descended from an Islamic scholarly lineage in Yunnan and a prominent critic of what he considered unlawful practices among Chinese Muslims, employed the five-fold classification of bid'a in a 1934 work (Halevi 2019;Ma Ruitu 2005, p. 603). Wang Jingzhai later listed Ma Ruitu as a member of the same reformist, "scripture-observing" (zunjing) milieu as Xiao Dezhen and against Hong Baoquan, whom Wang criticized as a member of the "custom-following sect" (congsu pai) (Wang Jingzhai 1939, pp. 22-3;Wang Jingzhai 2006, p. 11,218;Wang Jingzhai 2012). Ma Yizhi, another ahong and Ma Ruitu's uncle, also affirmed the five-fold classification of bid'a, yet Wang Jingzhai labeled him as one of the traditionalist obstacles to the scripture-observers (Ma Yizhi and Ma Jinxiang 2015; Wang Jingzhai 2006, p. 11,218).
What all of these scholars have in common is consideration of bid'a as a category of shari'a classification. They all engage in a sustained form of "legalism" or legalistic thinking: essentially, a schematization of the moral order in terms of explicit, abstract categories understood as independent of the actions and relationships they classify (Dresch 2012;Skoda 2012). This classificatory approach to the shari'a was not limited to Xiao's and Hong's treatment of bid'a. We may now extend the comparison to look for other commonalities across these ostensibly divergent texts. Their structure, positioning within the Hanafi jurisprudential tradition, and use of sources constitute additional commonalities between them.

Organization, jurisprudence, sources
The two texts are organized in similar ways and belong to the same genre: the houkun maisailai (or maisailie, from the Arabic h  ukm and mas'ala, "question" or "matter"), a collection of rulings or judgments on legal questions (Wang 1999, p. 51). Both begin with introductory chapters outlining the methodology according to which particular houkun are derived in later chapters: Chapters One, Two, and Three of Xiao's Registered Essentials introduce the ranks of belief, the ranks of legal canons, and the dichotomy of sunna and bid'a, while Chapters One, Two, and Three of Hong's Elucidation of Truth review the ranks of the maisailai, the typology of bid'a, and the priority of avoiding bid'a in cases where sunna is in doubt. Table 1 presents the contents of both works.
The authors are also similar in their self-positioning within the Hanafi school of law (madhhab). Both ahongs were what we can think of as critical Hanafis; they sought truth through the Hanafi tradition while recognizing that the tradition was ultimately a scaffold erected to aid in that search rather than the truth itself. In other words, they sought answers to specific questions about the shari'a without either abandoning the Hanafi tradition or blindly following the authority of an allegedly Hanafi text.
In their respective prefaces, each claimed to be expounding authentic Islam against corruption and ignorance. After bemoaning the historical influence of false doctrines and heretical innovations (yiduan) among Muslims in "the central lands" (zhong tu, referring to China), Xiao explains that he has studied and synthesized true texts now available thanks to the opening of travel to "awaken" (hence the title, Registered Essentials for Awakening from Confusion) fellow believers to truth. Hong, writing in Arabic, presents his work as an "illumination of the path of the religion of the pious forebears and an elimination of the creed of the innovators" (hence the Arabic title Munīr al-Dīn, "Illuminator of Religion"). Xiao also stresses the need to follow the example of the pious forebears (Ch. xian xian), refers to Abu Hanifa respectfully as the "Great Imam" (Ch. yimamu ai'erzan, from the Arabic al-imām al-a'z am) describes how the latter, together with his students, wrote down Muhammad's words and actions as a "raft to practice the Way"though the reminder that Abu Hanifa lived after the death of Muhammad could also be interpreted as a gentle indication that even the "Great Imam's" perspective was limited (Hong Baoquan et al. 2008, pp. 341-7;Xiao Dezhen 2008, pp. 161-3). In subsequent chapters, Xiao and Hong lay out their basic methodology for studying Islamic jurisprudence. In his second chapter, "Explanation of the Conditions for Observing the Scriptural Canons," Xiao cites the late Hanafi scholar 'Abd al-Hayy al-Laknawi's' (1848-1886) commentary 'Umda al-Ri'āya 'ala Sharh  al-Wiqāya' to introduce a five-fold hierarchy of legal texts within the Hanafi school. Again drawing attention to the limits of the Hanafi tradition, Xiao relates two quotations attributed to Abu Hanifa that say, in effect, that the Quran and sunna are higher sources of law than his own teachings, and that if there is a contradiction, one should observe the Quran and sunna. We should note, however, that these quotations are drawn from the Hanafi Laknawi's 'Umda and so in some sense are still given from within the Hanafi tradition. Moreover, Xiao opens the chapter with a decidedly Hanafi statement: "The compositions of the former and later scholars may all be followed because they were all composed according to the Great Imam (Abu Hanifa)." Xiao continues, "When analyzing and selecting a ruling, review those compiled by later [scholars]; if they accord with the texts of the former scholars, they should be followed, and if they contravene them, they should be abandoned" (Xiao Dezhen 2008, pp. 168-70).
Xiao then cites the 'Umda to define his categories: "former scholars" are those who met the three imams, i.e. the Great Imam (Abu Hanifa) and his two principal students, Abu Yusuf and Muhammad al-Shaybani. The "later scholars" are those who did not meet these three imams. The five-fold hierarchy crosses this former/later boundary and ranks Abu Hanifa's students and their followers, with Abu Yusuf and Muhammad al-Shaybani at the top (rank one), i.e. the most authoritative beneath Abu Hanifa. Xiao moves down through the various ranks, giving examples of authors or in some cases just the texts that correspond to the given rank. The lowest rank, the fifth, are "the most detailed of the works of the later scholars; they generally are not arbitrary personal opinions (si zhuan) and do not relate dubious statements or embellished or deficient remarks." Xiao concludes by mentioning an even lower category of "ruling explanations" (tiao jie), including "compositions of we ordinary people," which do not qualify as evidence (bu wei zhi ju, i.e. they are merely explanatory).
Hong includes a parallel discussion in the first chapter of Elucidation of Truth, titled in Chinese, "Explanation of the Three Ranks of the Maisailie." Hong's taxonomy is three-fold instead of five-fold and pertains to the collections of rulings on legal questions (responsa) and not the rank of the jurists (Hallaq 2001, pp. 47-8). The three types are: (1) those that are "fundamental" (Ar. us ūlī), i.e. of "manifest transmission" (Ar. z āhir al-riwāya) from the highest authorities of the Hanafi school through trustworthy subsequent jurists; (2) those that are "rare" or "singular" rulings (Ar. nawādir, sing. nādira) attributed to the highest authorities of the school but lacking highly qualified or numerous transmissions; and (3) those that are "occurrences" (nawāzil, sing. nāzila, or wāqi'āt, sing. wāqi'a), judgments issued by later jurists within the school of law for issues not addressed by the more authoritative predecessors.
Finally, we may briefly consider the question of sources. New or previously unavailable texts brought back to China by pilgrims have been seen as key ingredients in religious reform in modern Chinese Islam. On the other hand, scholars as well as Gedimu and Yihewani partisans have cast opposition to reform in terms of a disagreement over the authority of certain texts (Qi Mingde 2004, pp. 3-9). The particular texts associated with Hajji Guoyuan's return in the early 1890s and  Weilin, Yu Zhengui, and Na Guochang 1997, pp. 355-6;Pang Shiqian 1937;Qi Mingde 2004, pp. 3-9). Of these, Ma Bin lists (2), (4), and (5) among the works Xiao Dezhen consulted when writing Registered Essentials (Ma Bin 2015a, p. 179). Indeed, he cites those works throughout the text, in addition to (3) as well as the aforementioned (6) 'Umda al-Ri'āya, another work associated with the Yihewani. Remarkably, (1), (3), (5), and (6) are all listed in the bibliography in the frontmatter of Elucidation of Truth, and (4) is cited in the various rulings themselves (though it does not appear in the bibliography) (Hong Baoquan et al. 2008, p. 339). Evidently, Hong Baoquan studied and cited the very texts thought to distinguish the Yihewani from the Gedimu.

Conclusion
The foregoing has shown that a pair of authors and texts that ostensibly represent two opposing groups within Chinese Islam share much more in common than is generally recognized. While Xiao Dezhen's Registered Essentials and Hong Baoquan's Elucidation of Truth stake out different positions on numerous questions of ritual, they are similarly organized, similarly position themselves within the Hanafi tradition, and cite many of the same texts. Underlying these commonalities, I argue, is a shared legalistic understanding of the shari'a as a system of classification, reflected in their engagement with bid'a and other Islamic legal categories. Moreover, this classificatory conception of the shari'a differs fundamentally from the more practice-centered approach examined earlier and exhibited in late imperial expositions of Islam and treatments of bid'a.
In Muslim Chinese: Ethnic Nationalism in the People's Republic, the anthropologist Dru Gladney outlined a typology of Chinese Islamic groups arranged based on their relative distinctiveness in relation to Chinese culture. The "early Yihewani" and Gedimu fall on opposite ends of this spectrum, characterized as "fundamentalist" and "synthesist," respectively (Gladney 1991, p. 61). This axis of accommodation is certainly a useful heuristic in the study of Islam in China. But it is not the only one available. When we attend to evolving conceptions of the shari'a in the early twentieth century, we see that champions of these seemingly opposite groups in fact had a great deal in common, and that the shari'a, sometimes construed as a boundary between "China" and "Islam," afforded Muslims a range of options for accentuating, negotiating, and even transcending difference.