WHAT IS ISLAMIC HISTORY?

Efforts to define the concepts “Islam” and “history” have separately engendered rich debates with long intellectual genealogies. Both debates serve as a foundation for this essay’s attempt to delimit the subject of “Islamic history.” However, the essay also argues that a close examination of the interaction between the two categories offers its own insights. Chief among these is the argument that a reliance on subjects’ self-ascription as “Muslims” for definitions of “Muslim” and “Islamic” is far more than the empty or “nominal” approach that some critics have described. Rather, Islamic self-ascription is historically entangled, both an artifact of historical processes and an evocation of them, even an integral element of the phenomenon it seeks to define. The essay begins with an evaluation of the “islams not Islam” approach to defining the Islamic, rooting the argument not only in self-ascription—a common social-science tool for category definition and boundary making—but also in Islamic historical traditions themselves. It then demonstrates this historical rootedness through an unusually difficult test case: Chinese-language Islams that eschewed the words “Islam” and “Muslim.” After proposing a definition of “Islamic history,” one that is particularly open and expansive, the article outlines some common characteristics of Islamic history across its many forms, asking what makes it distinct and where it can contribute to a global comparative historiography. Finally, it argues that when we generalize about these traditions, describing the features most widely shared among them, we find an Islamic history that reflects and substantiates the centrality of self-ascription in delineating the scope of Islam.


INTRODUCTION
Any effort to assess the Islamic past ultimately rests on some delineation of the phenomena we are assessing, including some means of determining what past is Islamic and what past is not. This implicates us in the rich debate over the nature and definition of Islam, but by no means does it require that we construct a new 9 the linkage of nominalness and meaninglessness is not as obvious or unproblematic as it might appear at first sight, and the conceptual baggage of such a move, particularly in the Islamic context, will be examined later in this essay. Nonetheless, such critics worry that by accepting everything, Islam becomes nothing. The door to an Islam without Muhammad, or without the Quran, is seemingly left wide open. A secondary complaint (and the root of Asad's brief critique) is that the self-ascription approach in fact dismisses certain Muslim claims about Islam. Many, perhaps most, Muslims define Islam in ways that exclude some other self-identified Muslims' beliefs from the category of Islam. These particular claims about the boundaries of Islam are openly rejected by the self-ascription approach as analytical definitions of Islam (though they are not rejected as forms of Islam). Thus the self-ascription argument has been said to contradict Muslims' beliefs even as it claims to emerge from them. 5 This essay addresses questions about self-ascription that have received little attention from either side. Most important, why do people self-identify as Muslims in the first place? And what exactly does self-ascription comprise? What effects follow from the act of self-ascription? Finally, if we survey scholarship on Islam, how does the term "Islam" actually get used? In addressing these questions, I aim not only to mount a defense of the self-ascription approach, but to understand more fully the implications of self-ascription in Islamic contexts across time: the inducements to self-ascribe, the ramifications, and the consequent entanglements with historical representation. I argue that self-ascription is not merely an act of naming, but rather an engagement with a long, manytendrilled historical process, and that this process, the reproduction and spread of ascription, is also a uniting focus of diverse Islamic historiographical traditions. The self-ascription approach leads to an expansive, flexible definition of Islamic history, embracing a wide variety of traditions. And when we generalize about these traditions, that is to say, when we describe the features most widely shared among them, we find an Islamic history that reflects and substantiates the centrality of self-ascription in delineating the scope of Islam.
"NOMINAL" AND "UNDERDETERMINED" Explicit critiques of the self-ascriptive approach are prominent but not all that numerous. Of the two main critics already mentioned, Talal Asad devotes only a few sentences in his influential article "The Idea of an Anthropology of Islam." For a sustained and direct critique, we must turn to Shahab Ahmed, who detailed a long series of objections over the course of twenty-eight pages of his What is Islam? 6 Ahmed's objections are too numerous to address individually in the space I have here, but there is a connecting thread that makes his work an effective starting point for my argument. That thread is the notion that what he called the "islams-not-Islam" argument and particularly its "whatever-Muslims-say-it-is" subset are "underdetermined" and "nominal." 5. Ironically, rejections of self-ascription themselves not only contradict some Muslims' beliefs about the limits of Islam but also implicitly reject some self-identified Muslims' claims to Muslimness.
For Ahmed, mistaken conceptions of Islam are placed along a spectrum of overdetermined to underdetermined (in his metaphor, they turn in opposite directions along an axis of determinedness). This is one of two main problems he finds in all such failed conceptions of Islam. The implication is that this is a matter of proper calibration, that the proper understanding of Islam is not overor underdetermined but rather just-right-determined, a Goldilocks Islam that turns neither clockwise nor counterclockwise around the axis of determinedness. Ahmed objects to definitions that have given Islam too strong an essence, but he does not want Islam to have no essence. He advocates for essentialism "lite." In this view, islams-not-Islam falls on the underdetermined side, or, rather, spins in the underdetermined direction.
Ahmed argues it is not enough to say that there are Islams, in the plural, without defining what Islam, the singular, would then represent. Thus, Ahmed asks, "is there a single Islam of which these plurals are somehow expressions?" Those in the islams-not-Islam camp would simply answer that there is not. But this hints at a practical question: of all the varied local cultural and religious phenomena, Islamic and not Islamic, how do we determine which are Islams and which are not? It is much later in the book that Ahmed confronts the simplest solution to this problem, one that has been employed not just in the study of religion, but also, and more notably, in the study of ethnicity: self-ascription. 7 Under this view, an Islam is any phenomenon that is claimed to be Islam by someone who considers her-or himself a devotee of Islam, that is, a Muslim. To return to Ahmed's question, although any individual Muslim's Islam can exist in the singular, it is no more an expression of a single Islam than individual birds are expressions of a single ur-bird.
Ahmed's central criticism of self-identification is that it is "nominal" (all emphases original): [Hourani's] analogy suggests that, merely because Muslims identify with Islam . . . , we should not assume that there is in fact a relationship between the act of identification and the putative object of identity. Rather, the relationship is with a "symbol" that can be invoked in diverse circumstances with no necessary continuity-that is to say, no necessary coherencebetween the various instances of invocation. This would imply that symbol is only nominally the same in each instance, but is substantively different. 8 [I]f Muslims are not doing what they are supposed to be doing in order to be Muslimsif they are doing "much that would be non-Islamic or even anti-Islamic" according to "Islamic revelation vouchsafed by God to the Prophet Muhammad and embodied in the Holy Book of the Qur'ān" and according to "the whole imposing corpus of Islamic law, theology, tradition, and practice"-then what is the connection between what they were supposed or required to think and do and what they actually thought and did? If there is no connection, then how is it in any way meaningful to name what they are doing "Islam"? In the absence of a meaningful connection, Braude and Lewis' use of the word "Islam" is entirely nominal-they have named the civilization Islamic, but This idea that the "whatever-Muslims-say-it-is" approach is nominal, meaningless, and unsubstantive shares much with Ahmed's essentializing (though he objects to this description) critique of the islams-not-Islam approach as "underdetermined." Ahmed worries that we have opened the door so wide that it is no longer a door at all, or that we have reduced Islam to nothing more than a name. Is this not an abandonment of the entire Islamic tradition? Where is the Prophet Muhammad? Where is the Qur'ān? EXTREMES Rather than seek a moderate approach that softens such objections, let us explore them at the extremes. Let us see how the self-ascription model operates in its most uncompromising form and on the most troublesome cases, where the objections outlined above are most attractive. A particularly pure reliance on self-ascription for the formation of subjective categories can be found in Fredrik Barth's revolutionary essay on ethnicity in the 1969 volume, Ethnic Groups and Boundaries: The Social Organization of Culture Difference. To this day Barth's work remains the most influential formulation of ethnicity for anthropologists, and it certainly lies at the root of the "whatever-Muslims-say-it-is" position, if not also the "islams-not-Islam" argument. Indeed, anthropologists have been central to the formulation of these positions. 10 And Barth's theory leads directly to an extreme real-world example. Following Barth's lead, anthropologist Dru Gladney described as Muslims a group of people who would seem to confirm Ahmed's fears about the "nominal": the Chendai Hui of Fujian, China, who called themselves Huijiao ren (回教人) and "do not follow Islamic practices." 11 Barth's concerns about ethnic categorization share much with Ahmed's goals in delineating Islam, even if his solutions are different. Just as Ahmed repeatedly frames his study as a search for "coherence" in the face of diverse and contradictory understandings of Islam, Barth asks what coherence it is that makes us regard an ethnic group as identical across long stretches of time, even when the cultural traits of that group change diachronically. Barth was reacting to the then-widely accepted claim that an ethnic group: 4. has a membership which identifies itself, and is identified by others, as constituting a category distinguishable from other categories of the same order. 12 His answer was to discard all but the fourth element of the definition, to regard "ethnic groups . . . as a form of social organization" expressed in ascription, rather than an assemblage of particular biological, cultural, or social traits. 13 Barth would later be accused of disregarding cultural content entirely, and he would rebut that charge in useful ways I will return to below. 14 Nonetheless, Barth's understanding of ethnicity gained a wide following among anthropologists, including those who applied it to the case of Islam and Muslims. So it was that Dru Gladney surveyed "Muslim Chinese" across the People's Republic of China by studying any groups that called themselves "Hui," a Chinese word roughly corresponding to "Muslim" (though also indicating an ethnic status). These groups included the Ding lineage of Chendai, Fujian, who, as of 1940, referred to themselves as people of the Hui teaching (Huijiao ren). When Gladney met them in the 1980s, the Dings were fighting to be recognized officially by the state as Hui. According to Gladney, the Dings openly professed not to participate in Islamic practices, so we might say, for example, that their consumption of pork is not regarded by them as violating a tenet of Islam. 15 In a strict understanding of self-ascription, the practices and beliefs of the Dings would not be considered Islam, because the Dings count themselves as nonbelievers and nonpractitioners of Islam, even though they are "of" Islam (this by virtue of their descent from Muslims). Without the original Chinese-language transcripts of the Ding interviews, it is difficult to analyze the Ding case in much more detail, but their example raises more general questions about self-ascription and the relationship of Chinese Muslims to the category of Islam.
When confronted with purported Muslims calling themselves Huijiao ren, we are immediately forced to ask what it means to self-identify as a Muslim. An extraordinary historical phenomenon, almost entirely unremarked upon in the scholarly literature, becomes suddenly apparent: with only a few exceptions, every group that has been considered Muslim, and every tradition (or religion or civilization or discourse) that has been considered a manifestation of Islam, has employed words whose etymological linkage to "Islām" and "Muslim" is not only close but conspicuous. Thus we have people who call themselves musulmān rather than muslimūn, but only rare cases of etymologically divorced terms like Huijiao ren for Muslim and Qingzhen for Islam.
The case of Chinese Muslims who call themselves by some form of the word Hui may be one of very few exceptions to the rule, but it is a significant one. In the People's Republic of China, the term Hui is entangled with ethnicity, as the Hui are one of fifty-six official minzu (nationalities). Before this codification, however, Hui and Huihui usually denoted either adherents of Islam in general, or, later, Chinese-speaking Muslims. The etymological roots of the term are not fully traceable, but it is widely understood to be a derivation of Huihu. Huihu, in turn, was the Sinicized form of Uighur, the name of a kingdom and tribal confederation to the west of China. Today, approximately ten million people in China (half of the country's Muslims), call themselves Hui. Among them, the most common term for Islam is Qingzhen, a combination of Chinese words for pure and true.
A key moment in the rise of these etymologically distant Islamic terms was the first publication of Islamic religious texts in the Chinese language. This happened only at the end of the sixteenth century, more than eight centuries after the first Muslims arrived in China. The earliest work to have lasting influence was Wang Daiyu's Great Learning of the Pure and Real (清真大學) of 1642, in which Wang attempted to explain his belief system, one derived from Sufism in the vein of Ibn 'Arabi, 16 to Confucian, Buddhist, and Daoist literati, expressing concern that "the books of Islam are seldom seen by Confucians." 17 Six decades later, Liu Zhi composed Sufi treatises that also appear to be aimed at highly educated non-Muslims, and attempted to gain respect for Muslims in the Confucian-dominated world of Qing scholar-officials. Both authors based their writings on sources in Persian and Arabic, such as the works of Rāzi, Nasafī, and Jāmī, which Liu Zhi named in his text. 18 Both authors wrote at a time when they would have been called Hui or Huihui by non-Muslims, and Wang Daiyu referred to himself as the "old man of the real Hui." When they labeled their thought system, they used the phrases "pure and true," "the Ultimate Way," and "our teaching." Wang Daiyu addressed the problem of terminology explicitly in the introduction to his work. His attitude toward language and translation is clearest in his discussion of Buddhist and Daoist resonances in the text: There is nothing lacking in the classical canon of Islam, but there is no one outside the teaching who knows this. This is because our languages [Chinese vs. Persian/Arabic] are different. I wrote and discussed using their [(Buddhists' and Daoists')] expressions precisely to make our teachings comprehensive. All the borrowed expressions I used were because of my concern to show how the principles work. The expressions do not carry the same meaning, but if I had not borrowed them, how could I make clear that these two doctrines are different from ours? 19 This passage is important for its explicitly functionalist approach to terminological borrowing. It shows that Wang regarded his work as an effort at translation for non-Muslims, rather than the syncretic project that scholars would later see in early Chinese-language Muslim authors.
The linguistic landscape within China's varied Muslim communities in Wang's era is unclear, although we know that religious education was largely conducted in Persian and/or Arabic. The fact that Chinese-language writings about Islam had only recently appeared suggests that Persian remained widely used in the Qing, and with it, presumably a more familiar Islamic terminology, including muslumān and Islām. In Yunnan in the late nineteenth century, textbooks on the Arabic language were written with explanations in Persian, 20 and to this day some Hui use Persian numbers to communicate secretly during business transactions with non-Hui. 21 However, by the twentieth century at the latest, Chinese was the first language for the majority of Muslims in China proper (that is, excluding the Qing colonial conquests of Xinjiang and Tibet). These Muslims began relying on Chinese-language texts of Liu Zhi, Wang Daiyu, and others-texts originally aimed at explaining Islam to non-Muslims using Confucian, Buddhist, and Daoist terminology-for their understanding of Islam. Qingzhen emerged as the common designation for Islam, Qingzhensi (temple of the pure and real) for mosque, and Huihui for Muslim. They were propagated by authors, like Liu Zhi, who were well read in Persian and Arabic Islamic texts and used them as equivalent to the Arabic terms Islām, masjid, and Muslim.
An Arabic revival over the course of the twentieth century eventually brought Sinicized Arabic terms like terms like Yisilan (Islam) and Musilin to prominence, but it is probably safe to say that at times there have been Huihui in China who professed allegiance to Huijiao or Qingzhen, with no awareness of terms etymologically related to Islam and Muslim. By the self-ascription approach, do they count as Muslims? Does their "Hui teaching" count as Islam?
Two possible responses are obvious. One option is to take the words of Hourani and other "whatever-Muslims-say-it-is" supporters in a literal sense, and reject such Huihui claims on the basis that they do not literally call their teachings "Islam." This would confirm Ahmed's argument that the self-ascription approach is merely nominal. It is also a position that no self-ascription supporter has ever described or employed. Certainly Gladney, for example, when faced with the variation among the Hui, did not adopt this kind of extreme phonemic literalism. There is something that feels instinctively absurd about such literal nominalism. One might argue that this is the exposure of a fatal flaw in the self-ascription position. I argue that it exposes something else: self-ascription is something more than simply attaching oneself to a name devoid of meaning. What makes Huihui and Muslims (and Qingzhen and Islam) equivalent is the very history of Islam in China that I outlined above. It is a history of transmission and, in this particular case, efforts at translation that were regarded as crucial to transmission. Selfascription is, among other things, an engagement with the whole of the Islamic past that brought the Huihui to call their "teaching" Qingzhen.
To understand self-ascription as the mere attachment to a particular combination of phonemes or a particular empty symbol is ahistorical. It fails to consider why people have devoted themselves to a phenomenon and called it Islam. Where did they get the idea to use this "mere name" and why did they find it beneficial to do so? The answer is the entire sweep of the history of Islams. Without the Qur'an and without Muhammad, no one would claim to be Muslim or to devote themselves to Islam (and no one would feel the need to translate these terms into Chinese). Without the Indian Ocean trade, perhaps no one in Indonesia would use these words. Without the beauty of Rumi's poetry, or the miracles of an itinerant ishan, some part of the population of Central Asia would have taken generations longer to embrace something they called "Islam." Everyone had to learn of the existence of the name "Islam" from someone else. The chain of transmission proceeded mostly through Muslims, and in all cases can be traced ultimately to the prophet Muhammad and his companions. The shapes of various Islams were determined by the Muslims from whom new Muslims learned the word "Islam," and from the original insights, transmitted texts, old habits, foreign influences, local wisdoms, and so on of people who convinced (often fellow) Muslims of some part of their understanding of "Islam." The name "Islam" has been driven across the globe by the entire history of Islamic societies (societies in which Muslims were predominant) and brings with it that history, or, if we are interested in some particular slice of the human experience, that body of rituals, that discourse, that accumulation of laws, that literature, and on and on: all of the things we might fear losing with the selfascription argument.
The act of self-ascription also goes beyond reflection of and engagement with the history of transmission, in ways that are particularly visible in religious conversion. Conversion itself is an extreme case, insofar as most Muslims over the last fourteen centuries have inherited rather than adopted their identification as Muslims. In his work on conversion among the Golden Horde of the Russian and Central Asian steppes, Devin DeWeese pointed out unrecognized significance in the adoption of the designation "Muslim." DeWeese was partly arguing against the commonplace that the Islam of Inner Asian Muslims has historically been "'nominal' and superficial," but his insights on the potential meaning of selfascription are valuable in and of themselves: "To call oneself 'Muslim' or by a name whose mention evokes recollection of an islamizer, or of an entire 'sacred history' or genealogy linked to Islamization, is no trivial matter. To adopt a name is to change one's reality, and in this sense there is hardly a deeper 'conversion' than a nominal one." DeWeese goes on to emphasize the difficulty of abandoning old spoken rituals (such as acts of self-ascription), the correspondences between name/form and spiritual power in many Islamic traditions, "Islamic assumptions regarding the sacred power of the external to affect the internal," and the implications of "opening" inherent in communal adoptions of the name "Islam." Mine is a brutal simplification of a sophisticated argument, but the point is that DeWeese argues from within both Islamic and pre-Islamic Inner Asian discourses that selfascription is something much richer than "nominal" would suggest. 22 And what does it take for individuals to present themselves as devoted to a phenomenon called Islam, especially when, in all known cases of such devotion, they believe that phenomenon to include a host of actions beyond a speech-act, for example the assumption of certain loyalties or duties? Beyond a history of transmission or a moment of conversion, there is a present and future of selfascription's continuing effects. Presenting oneself as a Muslim has practical consequences. In a given context, a Muslim is expected to act in certain ways, to consume certain texts, to honor certain loyalties, or to hold certain beliefs. Rather than an empty, symbolic association, self-ascription is an embrace of a world of context-dependent expectations, requirements, responsibilities, privileges, and disadvantages-in short, a certain way of making meaning and relating to others understood as "Islam." We might say that self-ascription is what Ahmed calls, in his discussion of Islam as means and meaning, a "consequential truth." 23 ISLAMIC HISTORY Because self-ascription as a Muslim involves identification with other (selfascribed) Muslims, it implicates a Muslim in shared Islamic representations of the past. This is another of the continuing effects of self-ascription: a Muslim, by virtue of self-ascription, situates her-or himself in some version of Islamic history. Islamic historical traditions have obviously varied greatly across societies in which Muslims are/were predominant. Nonetheless, the great majority of Islamic historical traditions have in common an unusual emphasis on transmission (of the idea of being Muslim and of the willingness to personally embrace that idea), the phenomenon that drives self-ascription and ties all Muslims back to the earliest Islamic community. Thus, the use of transmission to understand the scope of Islam (via self-ascription) is not entirely alien to Islamic historical traditions themselves.
Thus far I have used "history" in the lay sense of everything that happened in the past, but for the remainder of this essay I will reserve that word for a narrower phenomenon. For this purpose, I borrow Greg Dening's definition of history as the past "transformed into texts-texts written down, texts spoken, texts caught in the forms of material things" and "the texted past for which we have a cultural poetic." 24 Even accepting this definition of history alongside the self-ascription definition of Islam, there are still numerous possible meanings of "Islamic history." The phrase could indicate the history of Muslims' pasts, the history of Islams past, histories that Muslims deem Islamic, histories created by Muslims, histories of Muslims' pasts created by Muslims, histories of societies in which Muslims have been predominant, and so on. The openness of the term is reflected in the multiplicity of ways it is used. And this is to ignore formulations based on other understandings of "history": Islamic history often denotes the professional academic study of the pasts of societies in which Muslims have been predominant. Without advocating for the superiority of any one of these framings, here I will use "Islamic history" to denote the vast collectivity of histories (texted pasts) created by Muslims, and "Islamic histories" to denote narrower bodies of texts or individual texts created by Muslims. This is roughly what has sometimes been covered under the study of "Islamic historiography" or "Islamicate historiography." 25 If Muslims are people who identify as Muslims, and Islam is whatever they say it is, then Islamic history as I have framed it is an extraordinarily large, diverse, and geographically widespread phenomenon, including oral performances of the West African epic of Sundiata, the Huihui Yuanlai (Origins of the Chinese Muslims) of China, historical essays in the Nation of Islam's official US newspaper, oral accounts of Sunan Kalijaga from Indonesia, 26 the Alexander Romance, and perhaps Shahab Ahmad's What is Islam?, not to mention more widely recognized Islamic histories such as Ṭabari's chronicle, dynastic histories such as the Saljuqnama, Persian local histories, and the Qur'ān. This is a more expansive conception of Islamic history than what is found in most Englishlanguage scholarship. The current academic historiography of Muslim societies is overwhelmingly dominated by cultural products of the Middle East, as is the modern academic field of the history of Muslim societies. Marshall Hodgson's influential Venture of Islam presented Muslim societies beyond the Middle East as "Islamicate" rather than Islamic, a term that has gained wide currency despite its essentializing assumptions. Much of Shahab Ahmed's energy was aimed at breaking the Middle Eastern monopoly on Islam, but his own conception is rooted in a "Balkans-to-Bengal complex" that excludes enormous populations of Muslims with rich historical traditions. Outside of the realm of historiography, expansive understandings of Muslim and Islam are not rare. 27 However, to date there have been few efforts to survey the whole historiography of self-ascribed Muslims. The raw population figures for Muslims who are thereby excluded from most images of Islamic history are staggering: approximately half a billion Muslims-one third of the global Muslim population-live in Southeast Asia or sub-Saharan Africa.
It is clear that there is no essential character to Islamic history in the broad formulation I apply here (nor, I would argue, in any other), but that does not mean it is impossible to generalize. We may not be able to identify traits shared universally among such histories, but we can find traits that appear with remarkable frequency in that vast corpus of Islamic histories. One of these is a common guiding question: how did Islam arrive here (wherever that may be)? The answers are different in every place, but they are all histories of transmission, of the movement (and contestation) of self-ascription across space and time, with all of its attendant texts, beliefs, loyalties, practices, and so on, Islamic history is, most commonly, a history of arrival.
At the birth of Islam, the guiding question of arrival was answered by the Qur'an, the first Islamic history, but within a few generations Muslims found themselves chronologically and often spatially distant enough from the events of the Qur'an that they needed new answers. The hadith literature, accounts of the deeds and words of the prophet Muhammad and his companions, answered this obliquely through the isnads (chains of transmission) that accompanied the hadith. Biographical compendia that helped establish the authority of hadith transmitters provided richer descriptions of the community's links back to the Prophet. As time passed and Islam spread, genealogies often did the heavy lifting, especially where Muslims have been eager to claim descent from the prophet Muhammad or one of his prominent family members. 28 Where Islamic genealogies were rare or deemphasized, other historical literatures filled the gap. In Persia and urban Central Asia, local histories often claimed authority and legitimacy for cities or regions by connecting them to the birthplaces of Islam. 29 In places such as the Inner Asian steppe and Java, conversion narratives gained remarkable popularity, distilling the arrival of Islam into the story of one great man's act of self-ascription. 30 In others, such as nineteenth-century Hausaland or the Tarim Basin of the eighteenth century, tales of holy war proliferated, in which the reader's ancestors are figured as unbelievers, Islam's reluctant hosts rather than Muslim guests. 31 In all of these cases, Islamic histories trace the very paths expressed in self-ascriptive definitions of Islam, the multifarious, intertwined branches of transmission of the idea that something called Islam exists, and that people can align themselves with it. CONCLUSION At the most practical level, the self-ascriptive approach to delimiting Islamic history has already been implicitly adopted by the community of historians who see their business as "Islamic history," at least as far as one can judge from the aggregate of these historians' more narrowly focused, regionally specific works. The mass of books and articles that frame their subject as part of "Islamic history" represents a cultural geography that maps neatly onto the range of communities that consider themselves Muslim. Even if some historians may, for example, divide the world into a truly Islamic Middle East and a secondary "Islamicate" sphere, when we look at who among professional historians present their work as Islamic history, we find representation for all self-ascribed Muslims, from Siberia to the Philippines to Detroit.
Is this category of "Islamic history" worth talking about, not just for the value of its contents (the creative output of Muslims over the centuries), but in terms of this category? What would be lost if we used other categories for Muslims' "texted pasts"? To begin with, there are good practical reasons to continue debating the nature of Islamic history, to map the contours of Islamic history, and to offer Islamic approaches to history as comparative examples. The salience of the notion of "Islam" in the world today makes an exploration of a category called "Islamic history" of inherent value. There are simply too many world-shaping decisions being made based on (mis)understandings of Islam and its histories to ignore framings of the past in terms of "Islamic history." Moreover, if we adopt the self-ascriptive understanding of Islam, then "Islamic history," when used as it is here, can be seen as an emic reflection of the process and importance of selfascription. It is valuable as an emic explanation for the contradiction apparent in etic definitions, that is, the mutual exclusivity of so many Muslims' understandings of Islam. It is a script for the lives of communities, future-oriented as much as past-oriented. And if "Islamic history" is used in the more common sense of "all the things that happened in Muslim societies," it encompasses the very process of the spread and reproduction of self-ascription.
Islamic history is at once delimited by the diachronic unfolding of transmission and the account of that unfolding. Moreover, Muslims' attention to transmission emerges in any view of the aggregate of Islamic histories as a predominant trait. Those transmissions of Muslim identity occur most often within Islamic societies, across generations and among members of the same communities. They are facilitated and reshaped by mechanisms that include translation, contestation, and historical production by Muslims for Muslims. More rarely, those transmissions take place across great social distances, sometimes taking the form of conversions, and even involving arrivals of Muslims in non-Muslim contexts. In this sense Islamic history is a history characterized by an extraordinary emphasis on arrivals and transmissions, it is created by arrival and transmission, and its defining, self-ascribing feature, at least as proposed here, is generated by arrivals and transmissions. In Islamic contexts this is particularly important because the notion of transmission is itself a prominent part of that which is transmitted, that is to say an emphasis on traceability back to some vision of original Muslims is so often a part of Islamic histories. It is only by setting aside this "contingent working-out" of Muslim identities across time 32 that essentialisms creep back into the study of Islamic history, and it is by highlighting them that we might begin to answer the question "what is Islamic history?" in ways that are more productive than reductive.