Full Texts, Split Moons, Eclipsed NarrativesThe Literary History of a Cosmological Miracle
Using techniques of form criticism and discourse analysis, this article aims at developing a better understanding of the famous Quranic verses (Q 54:1–2) on the splitting of the moon, traditionally understood as a miracle of Muhammad. It examines the vast body of biographical traditions and reports and investigates the contributions of Muslim theologians, exegetes, and philosophers. Analysis reveals that this miracle has a rich literary history traceable to an early oral reception of two parallel interpretations of the Quranic text. One was refined through narrative reception and entered the prevalent popular lore and the normative theological depiction of history. A separate section is dedicated to analyzing a single tradition that grew to become an exemplary folktale expressing the communal sectarian sentiments of the transmitters. However, a closer look into pre-Islamic poetry, classical Arabic lexicons, Quranic rhetoric, the Jewish and Christian milieus, and anthropological information provides a deeper insight into the cultural context of the text. It seeks to understand the rather complicated origins of the whole theological-narrative construct. In conclusion, the article proposes a specific reading of the historical origin of these verses, one that predates the hegemony of miraculous interpretation, without committing to a mutually exclusive reading of the possibilities of such origin.
Using the insights of intellectual and social history, this article aims at developing a better understanding of the famous Quranic verses (54:1–2) on the splitting of the moon, which is traditionally understood as a miracle of Muhammad and surrounded with a massive body of commentaries and exegesis. By paying close attention to the links between textual practices and social conflicts, the article examines the vast body of biographical traditions and reports and investigates the contributions of Muslim theologians, exegetes, and philosophers. This miracle, with a rich literary history, occupies a unique place in the prevalent popular lore and the normative theological depiction of history. This reflects its position at the intersection of various social discourses advanced, most notably, by two rival groups: elitist theologians who boasted about their respect for rationalist thinking, and populist traditionalists who took care to consolidate their authority as the guardians of literalism in Islam. However, a closer look into pre-Islamic poetry, classical Arabic lexicons, Quranic rhetoric, Jewish and Christian milieus, and anthropological information provides a deeper insight into the cultural context of [End Page 141] the text. Moreover, the tools of historiographical study reveal the social logic of the texts and their particularities. The body of texts built around the Quranic verses, claiming to provide an interpretation of the their cryptic words, starts to betray the various forces at work in the construction of the miracle narrative and how its many aspects respond to the competing needs of rival groups. Once grasped in light of these considerations, the rather complicated origins of the whole theological narrative construct give way to the ideological mystifications of the relevant lore. The article provides a brief theoretical framework and then a historical background to the relevant debate before proceeding to analyze the material to understand it using the tools provided by the theoretical framework.
The linguistic turn in historiography added unprecedented levels of complexity and anxiety to a discipline that was already showing signs of critical self-awareness; its implications have been “unsettling” (Clark 5). Although not exclusively proposed by scholars who endorse this turn, a widely accepted assumption in the historiography of the time is the insistence that no author can control the meanings of her text; in the words of J. G. A. Pocock (106), “it is a basic rule of historical method that more meanings can be found in a text, or in any document, than the author intended to convey when he wrote it.” This interest in the text reflected the influence of a similar turn in philosophy, particularly represented in the work of Wittgenstein (Clark 29). The question was soon conceived in terms of the tension between text and context. This tension gave rise to two major conceptual errors, as outlined by Quentin Skinner in his seminal article “Meaning and Understanding in the History of Ideas.” First, the excessive interest in understanding the text as containing timeless truths leads to many sorts of myths that reflect a scholar’s penchant to bestow coherence or expect answers when there are none in the corpus being scrutinized. This abstraction away from the text’s specific attributes in time and space ignores the horizon of possibilities and limitations that could not have been overcome by the author in the first place (Skinner 12–24). Second, reducing the text into a mere effect whose sole cause is the context makes one prone to create more confusion than clarity; it so often happens that the context can, with equal validity, provide opposite explanations of the same text (Skinner 46–47). Thus, Skinner argued, it would be much better to take stock of the range of language games available to a given author to attempt a reliable understanding, in a now famous phrase, of her intention “in doing X”—such as producing a particular text (Skinner 45).
This tension between text and context, conceived as such against the background of the linguistic turn, led many proponents of this turn to combine the two [End Page 142] into a grand assemblage of texts. The context, be it social or political, is nothing but a complex text—the whole world being only a linguistic construct (Spiegel 5). The resulting situation is one where the material and textual aspects of the past are flattened into the same phenomenological order. The methodological and conceptual results issuing from such conflation were elegantly summarized and rejected by Gabrielle Spiegel, who astutely observed that “if we want to contextualize texts, we cannot achieve this merely by textualizing the context” (17). The text, she explains, can only be situated; it represents the specific answers of particular individuals to particular situations in time and space. Therefore, what one should do is engage the “social logic” of the text. Only when this social logic is understood does it become possible to grasp the text from the standpoint of literary history. Instead of searching for universal categories, it is better to start from the moment of “inscription,” that is, when the text acquires a fixed meaning in time. From such a starting point, the progress of analysis would allow us to understand the various ideological determinants of the text, for these will become manifest when its varied meanings are fractured in the course of analytical study (Spiegel 27).
The historical tradition, then, is largely shaped by the social configurations in which it appears and which it seeks to describe, for “historiography is a form of thought occasioned by awareness of society’s structures and processes” (Pocock 147). This relation between social configuration and historiography may be studied by analyzing the possible connection between social stratification and the soundness of historical method—concluding that societies with less rigid hierarchies are more likely to produce a sounder historiographical literature (Brown, Hierarchy 324–27). It may also be studied by looking at the relation between a society’s present and its past. In societies that depend heavily on the past, historical awareness is more plural than singular and is socially conditioned in a variety of ways (Pocock 149). The more a society depends on its past, the more complex the problems caused by the challenges posed to traditional expositions, leading to novel explanations of the past. These explanations might be very far from what counts as historical explanations in the normal sense of the term (Pocock 154). In such a context, historization (a series of orderly pictures of events in the past, constructed to fit one author’s interpretation of different elements in current phenomenon) may replace historiography in trying to understand the past of a certain text (Pocock 158).
Based on the foregoing observations, the distinction between documentary and literary texts in the study of history seems highly artificial, imposed by a rather arbitrary standard of classification, because no texts enjoys a transparent [End Page 143] relationship with reality (Clark 125). If texts pertaining to the past are largely shaped by the current social and political forces, the fact that authors were unable to physically and intellectually insulate themselves from broader society makes their texts, regardless of genre, highly conditioned by the said forces. In his perceptive study of Shakespeare, Stephen Greenblatt shows that literary works, no matter how much creativity they reflect, are “the product of collective negotiation and exchange” (vii). Therefore, history cannot be understood as the stable antithesis to the literary text, nor can it be taken to be its stable background (Greenblatt 95). On the contrary, the text appears in history as a result of all these acts of balancing and delicate interactions among various social players. As such, there is no moment of origination of a text; it is fruitless to search for the master hand that “shapes the concentrated social energy into the sublime aesthetic object” (Greenblatt 7). Even the moment of inscription cannot be individual; it is a social moment at which the author commits his ideas to writing (Greenblatt 5). Shakespeare’s plays, seemingly discussing fictional episodes and enacted in the “unreal” setting of theater, were in fact statements about the social, political, and intellectual ferment of England at the time: the nature of political authority, sexual identity, church doctrine, and martial law.
These theoretical considerations are not alien to scholarship on Islamic historiography, and this connection has been made by some of the aforementioned historians. For example, we can turn to Pocock’s remark concerning the similarity of the great Muslim historian Ibn Khaldun (d. 1406) and Vico (d. 1744): the former had access to a great legacy of knowledge about the past whose significance for his society was not obvious, nor were that society’s intellectual techniques adequate to equip this legacy with such significance or function (Pocock 185). As for the relationship between society and historiographical method, Brown clearly states that classical Islam belonged to open societies that he correlates with sounder historiographical practices (Brown, Hierarchy 238–30). Given these and other common considerations, one is tempted to apply the theoretical tools of social and intellectual history to the Islamic tradition, to which we now turn.
The opening two verses of sūrat al-Qamar (the “Chapter on the Moon”; Q 54) read: “The Hour has drawn nigh: the moon is split. Yet if they see a sign they turn away, and they say ‘A continuous sorcery!’”1 The sūrat al-Qamar is generally accepted as part of the revelations belonging to Muhammad’s time in Mecca, that is, after the commencement of his call to faith around 610 CE and before his immigration to Medina in 622 CE. During this period, his followers were relatively few, the [End Page 144] vast majority from underprivileged backgrounds: coming either from lowly tribal origins, poor socioeconomic conditions, or both (Watt 56–83). The whole sūrat al-Qamar betrays typical features of the Quranic text from the Meccan period: apocalyptic, poetic, and allusive (Neuwirth, “Structural” 110–11). A few decades after these verses were first spoken, the generation of the Prophet’s Companions that would have witnessed the purported splitting of the moon died. The inevitable change in social milieu and constitution of the Muslim community, caused by the sweeping conquests of Mesopotamia, Persia, and the Levant, led many of the surviving witnesses away from Medina. In these newly acquired territories, the settlement of Arab tribes, mostly formed of late or nominal converts to Islam constituting the armies of conquest, created a context wherein the vast majority of Muslims had no personal memory of this particular revelation or direct connection with the Prophet. Therefore, one can say that by the end of the seventh century, the text lost its meaning and came to require interpretation.
Henceforth, the term khabar is used to refer to a multitude of texts that narrate a specific event, in particular the collective texts narrating Muhammad’s splitting of the moon. These are found primarily in works of Quranic exegesis, biographies of Muhammad, collections of sayings from him and—less frequently—heresiographies and theological polemics. As noted later, this material must have gone through a period of oral transmission, but it can be assumed to have been written down gradually starting around the end of the seventh century. The generation following that of Muhammad’s Companions, known in Islamic tradition as the Successors, was instrumental in this process, although their written work constituted fragments covering specific questions or answering detailed inquiries more than compiling generic collections or authoring coherent works. This was certainly influenced, among other things, by the scarcity of writing materials, as paper was still unknown to Arabs, who mostly used parchment, bones, and stones to record what they considered worth writing down (Donner 120; see Modaressi 1:xv). The word report is used to refer to any cluster of these texts that share details in terms of content but differ in wording. For each of these texts, the term version is applied.
This study is an exercise in Quranic exegetical scholarship in the sense of exegetes just noted. Its aim is to analyze various interpretations of the two verses in the Islamic tradition to identify what is, most probably, closest both to the verses’ “original meaning” and their reception by the early Muslim community before the addition of varied layers of later interpretation. The core of my argument consequently follows two complementary tracks. The first track involves comparison of [End Page 145] different reports of the khabar in exegetical sources and ḥadīth compendia to trace its roots. The second track examines the manner in which theology (ʿilm al-kalām) availed itself of the khabar in its attempts to meet the needs of intraconfessional polemics. Both questions prove instrumental in illuminating the context in which the Muslim reception of the khabar developed and the possible motives behind the wide acceptance of a particular interpretation to the detriment of others. This insight allows for an assessment of the trajectory through which the khabar went from being part of the communicative memory of one Muslim generation to occupying a prominent place in the Muslim canon as part of the cultural memory of the whole community.
Scholars have recently worked on the moon-splitting khabar. Most of the scholarship was undertaken by Western academics in the field of Islamic studies, usually employing philological techniques to produce intellectual history narrowly conceived. Some have examined the modern exegetical reception of the miracle in an attempt to highlight and explain the difference between modern and classical exegetes. This approach extends to studying how recent scientific discoveries and technological tools have been exploited to propose novel explanations of the two verses in light of the khabar.2 Other researchers have been more focused on the image of the Prophet as portrayed in the tradition (see the detailed study by Rubin). Both approaches have some commonalities with this article because of the overlap in sources. Even so, the overlap should not mask the divergence between those studies and this one in terms of scope, method, and purpose. This study is an exegetical contribution that relies on diverse sources and approaches, including those concerned with the image of the Prophet and those focused on the formation of various discourses in Islam’s intellectual history.
In its scope, this work does not restrict itself to the significance of the khabar in understanding the development of Muslim perception of the Prophet’s role in the sacred history of previous prophets, nor does it seek to discover the political benefits of promoting the khabar to bestow a religious charisma on the Prophet’s military successes. Ascribing miracles to Muhammad was undoubtedly essential to his growing aura for later Muslims, and splitting the moon proved to be one such miracle—particularly because of its Quranic connection. Muhammad the miracle worker, also a skilled politician and military leader, was necessary to corroborate the image of Muhammad the Prophet. Although it is absolutely necessary to understand how the Prophet’s image influenced—and was influenced by—the reception of the khabar, confining attention to this aspect only may distract scholarship from [End Page 146] rewarding research avenues, such as understanding how the khabar was used by the Muslim community to represent its communal identity through emphasizing a particular image of the past. This study draws on and seeks to synthesize the many efforts that contributed to the development of the khabar, in the process questioning their possible motives, even if they were centered on promoting the Prophet from the status of a mere warner to that of a miracle worker. Therefore, I do not allocate much space to contemporary exegesis, because the study of such works pertains either to the modernist-traditional dichotomy or the study of exegetical lore—both of which fall outside the scope of this study.
As for method, I study the many versions of the different reports of the khabar in detail regarding their body text (matn), with attention to their chains of transmission (isnāds) insofar as these provide information on the origins of various reports. This approach offers valuable insight into the sectarian provenance of certain reports that may refine our assessment of the khabar’s course of development from the perspectives of both narratology and intellectual history; this insight reinforces the exegetical exercise to arrive at the khabar’s earliest “meaning” and reception. By that I do not intend to find the external referent that must have constituted the truth signified by the text, as is the usual practice of earlier exegetes who believed in such a meaning. On the contrary, I try to trace back the possible development of the text as a deed echoing the conflicting social demands. Because language “acquires meaning and authority only within specific social and historical settings” (Spiegel 53), one cannot escape the conclusion that meaning to be inferred will in fact be a function of the earliest such settings and not the true significance of the text. This must all be done without sliding into metaexegesis, as the latter—despite its immense benefits—is more pertinent to studying the formation of Quranic disciplines than to the prevalence of a particular understanding of any given passage.
Theology, one of the primary components of Islamic intellectual history concerned with guarding religious doxa through polemics,3 constitutes a crucial point of divergence between this study and others on the splitting of the moon. I emphasize theology, since it bestows significance on many details that might otherwise seem merely decorative. If the needs of theological disputation are taken into consideration, it becomes apparent that minute details in the reports of the khabar—with their different versions—cannot be ignored, nor can they simply be grounded in the image of the Prophet as independent from polemics. Moreover, concentrating on modern debates about the khabar may distract us from the early classical sources that inspired these seemingly different later debates. Therefore, it is [End Page 147] important to heed the classical theological sources that usually provide readers with clear clues about the reception and analysis of texts. In this regard, these sources offer two advantages. First, they spare us the effort of deciphering what the direct contemporary audience saw in the khabar, and they do so much more effectively than do ḥadīth compendia. Second, they equip readers with a theoretical model that can be applied to other khabars that may be absent from extant theological works. In addition to ḥadīth, exegesis, and theology, relevant to Islamic intellectual history are works of law (fiqh) and belles-lettres (adab), which inform us further about the context of any khabar. These works are usually less grounded in the image of the Prophet in the above-mentioned sense.
As for purpose, this article seeks to investigate the origins of the khabar rather than the course of its development in the sīra and ḥadīth literatures through the classical period. It also seeks to read the khabar’s increasing complexity from theological and narratological perspectives. As such, what might be taken for granted in other studies—even if otherwise convincing—is material for study here; what is omitted elsewhere because of its irrelevance to the purpose is scrutinized here. My overall objective is to move forward in deepening our understanding of the origin and development of the khabar through the concurrence of many methods and approaches.
This section summarizes the relevant findings of previous research and refers readers to their respective venues. Among these findings is the centrality of the Quran for Muslim religious experience since the lifetime of the Prophet and the first generation of Muslims.4 It is accepted that the Quran was compiled at the hands of the committee formed by the caliph ʿUthmān (d. 656), although some scholars propose that the compilation took place long after that time.5 The question of dating non-Quranic material is more problematic, but it seems likely that the earliest written documentation of ḥadīth and sīra, in a limited fashion, dates back to the reign of ʿAbd al-Malik (d. 705).6 It is not a priority here to develop a categorical position on the credibility of traditions or, in particular, on whether they can be trusted to inform us about the times of the Companions or only about the times when the material was written down. It suffices to acknowledge the high likelihood that at least some of the reports ascribed to the Muslims of the first generation can [End Page 148] be trusted to have been uttered by members of this first generation, regardless of whether they are faithfully reflecting what they claim to narrate.
The elevated station of the Quran may be taken to explain why exegesis, in the most basic sense of explaining ambiguous textual occurrences, is probably the oldest form of intellectual activity in traditional Islamic scholarship. Its strong relevance to various aspects of Muslim religious and social life, such as legal issues and stories of past prophets, must have also contributed to its appearance as a more technical discipline. Early sources attest to this connection, as much exegetical material consists of responses to questions of Quranic origin or clarifications of certain verses. This characteristic has caused many scholars—both skeptics and “sanguines”—to argue that exegetical material, in this basic sense, is contemporary or antecedent to other material.7
It follows that the Quran provides historians with content that can assist in understanding the book’s historical context, be it some aspects of Muslim life in Mecca or Medina, the internal and external challenges Muhammad was facing, and the polemics and beliefs of his adversaries in both cities. However, that this Quranic assistance is hampered by the allusive nature of the text and its sensitive position at the heart of Muslim worldview.8 This position makes the Quran extremely unlike other texts in terms of its historiographical utility, because it is subject to heated disputes and polarized biases on sectarian, political, and legal matters.9 Therefore, relying on the Quran for historiographical purposes is a double-edged sword: on one hand, ignoring it may deprive research of a most fruitful source of historical knowledge; on the other hand, extreme caution is warranted in handling its various receptions. This calculated risk is reduced by adopting a criterion of caution, namely, the affinity between the interpretation offered by an exegete and the prima facie reading of the text. Such affinity is an indication that the inevitable influence of the exegete’s biases has been kept to a minimum. A practical manifestation of this affinity is the simplicity of interpretation, since the narratological study of khabars has shown that simplicity is a hallmark of oldness, which reinforces the khabar’s significance in historiography.10 The predictable additions that seek to elaborate on the old text must be carefully handled, avoiding reliance on material that clearly reflects the influence of doctrinal schisms and sectarian divisions. Such biases are best detected in the diverse exegetical methods of traditionalists, theologians, jurists, rhetoricians, mystics, and philosophers: these groups interrogate the Quran to cull an answer that conforms to its approach—a method antithetical to the purposes of this study. [End Page 149]
Miracles as a Polemical Requirement
The religious history of the ancient Near East had long settled on miracles as an essential aspect of a charismatic leader’s call to faith, especially when associated with messianic undertones (Novakovic 108). It comes as no surprise that the Prophet’s initial audience and the non-Muslims encountered in the course of the conquests after his death asked for proof to accept Muhammad’s message. In light of the Quranic text and the surrounding context, the expected proof must have been a miracle similar to those ascribed to Jesus and Moses, whose followers formed the bulk of the disputants with the Muslims before the issue grew into an intra-Muslim discussion as well. This expectation is attested by the fact that supernatural feats were seen as necessary pieces of evidence for any claimant to prophethood or supernatural qualities.11 The Quran, despite its clear admission that the Meccans did ask Muhammad for miracles, does not unequivocally claim that he delivered any; indeed, in some places the text hints that Muhammad contented himself with the Quran and took his adversaries’ failure to imitate it as a compelling proof of his prophethood.12 It follows that the need to ascribe a miracle to Muhammad was felt from early times, maybe even at the inception of his call to Islam, although the means of meeting this need developed in a complicated manner that reflects the course of religious polemics and communal evolution.
The quest by early Muslim theologians for a prophetic miracle must have been given added urgency by the ambiguity of the Quran concerning such miracles, especially in view of its clarity in reporting the miracles of other prophets and the need for Muslims to defend their faith. The quest seems not to have yielded fully satisfactory results, and thus the Quran remained the foremost miracle in the theological works on prophethood by theologians of various affiliations (ʿAbd al-Jabbār 16:151; Abū Yaʿlā al-Farrāʾ, Muʿtamad 157;Baqillānī, Tamhīd 132, 141; Juwaynī 345; Murtaḍā, Dhakhīra 364), followed by other miracles, such as the yearning of the trunk for Muhammad, his conversation with beasts, and trees greeting him. Reports of these miracles are usually found in works of sīra, maghāzī, and ḥadīth, as well as the signs of prophethood genre (dalāʾil al-nubuwwa), all of which—as has been noted—are later and less significant than the Quran.13 Many traditional scholars, particularly ḥadīth critics, deemed most of these reports weak and tried to locate more reliable ones to document the Prophet’s miracles (see, e.g., Ibn Ḥajar 6:433).
However, even if one avoids traditional approaches, one should not dismiss in historiographical work all reports that contain fictional (in this case supernatural) [End Page 150] elements. That the foundations of historiography will be destroyed by literary analysis is, in the words of Julie Scott Mesiami (149), a “pseudo-problem.” Instead, a safer approach is to include these reports in the material to be studied and separate their supernatural characteristics from their other aspects to retain the kernel of truth that might exist in the overall material (see Günther 174; Leder 312).
The Reception of Q 54:1–2: General Appraisal
The moon is much celebrated in the Quran; it is frequently singled out as a sign of God. By it, people can learn about their sacred feasts and perform religious rituals, and its eclipse at the end of time signals the advent of the Hour (Varisco). A whole sūra is named after the moon, and scholars studied the rulings and beliefs concerning it, particularly in legal matters.
This volubility relates to the normal moon, be it a crescent or a full moon, eclipsed or not. However, when the moon is split, jurists are reduced to silence, leaving the floor to exegetes, traditionists, and theologians. The reason for this is that in the opening verses of sūrat al-Qamar, the Quran comes closest to affirming that Muhammad performed a miracle, although it still stops short of a categorical declaration (Ālūsī 27:77). If the Quran is approached from the perspective of literal understanding, the text can be taken to signify a miracle by the Prophet, especially since this reading is supported by many reports to the same effect narrated on the authority of Companions. Moreover, such a reading of the verses meets the theological need to prove that the Prophet indeed performed a miracle.
Even an incomplete survey reveals the general trajectory of scholarly views on the splitting of the moon. On the basis of an overview of opinions that does not get into the details of reports, a recent researcher sees a movement from the eschatological to the miraculous: the splitting of the moon, originally pereceived as a sign of the impending end of times, came to be understood and remembered as a miracle that proved Muhammad’s prophetic status (Rubin 57). The existence of widespread agreement on this view can be gleaned from the fact that some exegetes and traditionists claim it is a matter of consensus that the moon was split in Mecca before the migration to Medina (Abū Ḥayyān al-Andalusī 8:171; Ṭūsī, Tibyān 9:442), and others even make it the first sign of Muhammad’s message (Al-Ḥākim al-Naysābūrī 2:472). Early classical theologians (before the eleventh century CE) [End Page 151] also almost unfailingly list the splitting of the moon among the Prophet’s miracles, sometimes placing it at the top of the list—except for the Quran.14 It is also noteworthy that the story of Baqillānī’s (d. 1013) trip to meet the Byzantine emperor features a religious debate between the two that starts with the emperor inquiring about the Muslim belief in the splitting of the moon.15 Regardless of its authenticity, this story attests to the spread of this belief among early classical theologians.
Turning to the more skeptical position taken by recent scholarship does not reveal dramatically different outcomes: a meticulous study of the versions of some reports of the khabar, particularly those whose isnāds meet at the Baṣran Shuʿba b. al-Ḥajjāj (d. 776), led one author to conclude that Shuʿba is probably responsible for the report’s wording, but the general meaning is attributable to the work of an earlier storyteller (qāṣṣ) from the Umayyad period who had attempted to explain the two Quranic verses in question (Juynboll 221). Another researcher ascribed the khabar to an incorrect explanation of these verses by the Companion Ibn Masʿūd (d. 653) (Nöldeke 1:122). The conclusion that the khabar is old is supported further by the fact that even extreme skeptics accept that the exegetical material antedates ḥadīth, which applies to our case: the khabar appears first in early works of exegesis such as those by Mujāhid (d. 722), Muqātil (d. 767), and ʿAbd al-Razzāq (d. 827).16
The Interpretation of Q 54:1–2: Three Views
Despite the spread of the khabar and many claims concerning its attestation, there is no consensus that the moon was indeed split. Three views on the question can be identified, though they command unequal numbers of followers and have various provenances.
the metaphorical reading
According to later sources, a few (unfortunately unnamed) scholars are said to have argued that the phrase “the moon is split” is a metaphor for the clarity of the matter (wuḍūḥ al-amr). The sources provide scant information concerning this opinion, mostly recounting it on the authority of al-Rāghib al-Iṣfahānī (d. 1011). [End Page 152] Later scholars, who nonetheless rejected the validity of this interpretation, located occurrences of such metaphorical usage of the image of the moon in classical Arabic poetry (Abū Ḥayyān al-Andalusī 8:171; Qurṭubī 17:126; Al-Rāghib al-Iṣfahānī 1:459; Shawkānī 5:120; Zabīdī 13:252).
Among the akhbār (Arabic plural; sing. khabar) used by traditionists and exegetes to support the claim that the moon was split in Mecca (the third view, discussed later) is another khabar that is actually better understood as a metaphor. This is the khabar (henceforth Ḥudhayfa’s khabar)17 of a Friday sermon delivered by the Companion Ḥudhayfa b. al-Yammān (d. 656) during his tenure as governor of Madāʾin. The sole narrator, and also an eyewitness, of the sermon is Abū ʿAbd al-Raḥmān al-Sulamī (d. 693), who attended the prayer with his father. Quoting “the moon is split,” Ḥudhayfa proceeded to assert that “today is the arena and tomorrow is the race” (al-yawm al-miḍmār wa-ghadan al-sibāq). This statement confused the young Sulamī, who asked his father whether a race was to be held the day after, upon which he was made to understand that the real race was the race in deeds. The following Friday, Ḥudhayfa delivered a similar sermon, adding that Paradise is the prize of those who beat others in the race (ʿAbd al-Razzāq, Muṣannaf 3:193 [only Ḥudhayfa’s sermon]; al-Ḥākim al-Naysābūrī 4:609; Ibn ʿAbd al-Barr 1:323; Ibn Abī Shayba 8:200 [with slight differences]; Ibn Kathīr, Tafsīr 4:280; Ṭabarī 27:114; Thaʿlabī 9:161).
By showing the narrator’s inability to grasp the metaphorical meaning of Ḥudhayfa’s reference to a race, the text of Ḥudhayfa’s khabar opens the possibility that the same problem may have affected the narrator’s understanding of the significance of the Quranic phrase. The addition of “indeed” (qad) to the wording of the verse could be seen as an affirmation of the situation intended by the metaphor, not an attempt to stress the factuality of the moon having been split. In any case, comparing different versions of Ḥudhayfa’s khabar shows that some include an addition that seeks to date the splitting to “the time of your Prophet” (ʿahd nabiyyikum) (Abū Ḥayyān al-Andalusī 8:171; Qurṭubī 17:126; Zamakhsharī 4:36). This addition, once accepted, undermines the metaphorical interpretation of Ḥudhayfa’s khabar and supports a literal one.
the eschatological reading
The second view does not support a metaphorical reading of the verses but proposes that they refer to a real splitting of the moon that will take place at the end of time, [End Page 153] heralding the advent of the Hour. This view is supported by more scholars than the previous view; its advocates include Successors and early authorities, such as the archetypical pietist of early Islam al-Ḥasan al-Baṣrī (d. 728) (Abū Ḥayyān al-Andalusī 8:171; Qurṭubī 17:126; Shawkānī 5:120; Ṭabrisī 9:310; Ṭūsī, Tibyān, 9:443) and ʿAṭāʾ b. Abī Muslim (d. 752) (Ibn al-Jawzī 7:242; Shawkānī 5:120; Ṭabrisī 9:310; Thaʿlabī 9:160). It is also ascribed to the later Muʿtazilī theologians al-Naẓẓām (d. between 825 and 835) ([Pseudo-]ʿAbd al-Jabbār 1:55; Ibn Qutayba, 70) and Abū al-Qāsim al-Kaʿbī al-Balkhī (d. 931),18 but its ascription to some other theologians seems problematic.19
the realist reading
The third view, shared by most traditionists, exegetes, and theologians, is that the moon was split during the Prophet’s time in Mecca. Beyond this commonality, there are disagreements on the details. Although the event is not mentioned by Ibn Isḥāq (d. 767) or Ibn Saʿd (d. 845), the common ground is preserved in early Quran commentaries, muṣannafs, musnads, and ṣaḥīḥs. Later authorities disagree as to the extent of acceptance of this khabar, claiming variously that it is a matter of unanimity among the Companions (Samʿānī 5:307; Shawkānī 5:120) and exegetes (Rāzī 29:28), that it is prevalent (mutawātir) (Ibn Kathīr, Tafsīr 4:280; Shawkānī 5:120), or that it is confirmed by the testimony of reliable witnesses (Qurṭubī 17:126). The disagreement is one of degree, not of principle.
The old provenance of the khabar makes it merit closer investigation of its different reports, with the goal of unveiling the various contexts that might have concurred to produce reports about the splitting of the moon that are much richer and more significant than the sparse common ground. Although it is possible to ascribe the origins of the khabar to the work of storytellers, a closer look at the question opens up new possibilities, especially once the scope of attention is widened to encompass other historical data and different matns and isnāds.
The Various Reports: Typology and Scrutiny
It seems most likely that the khabar originated as an explanation for the Quran’s reference to what must have seemed like a very curious event. In addition, a simple wording of the common-ground view amounts to an almost literal rephrasing of [End Page 154]
Click for larger view
View full resolution
[End Page 155]
the verses; this extreme similitude indicates strong affinity between the text of the Quran and that of the khabar. However, the absence of the khabar from both Ibn Isḥāq’s and Ibn Saʿd’s works suggests that they did not know of a miracle involving splitting the moon, especially given the perfect convenience of the khabar for the purposes of their books and Ibn Isḥāq’s reputation for laxity in verifying the reliability of his sources.20
If we accept the likely conclusion that the Quran is the origin of the khabar, the different reports—regardless of their isnāds and versions—may be classified according to their similarity to the Quranic text. This produces a division of reports into three types distinguished by increasing complexity: simple, composite, and complex. Each type contains further subdivisions, with reports of lesser and greater complexity. Note that this division does not claim that simple reports are less biased than composite—let alone complex—ones. Rather, it assumes that each interpretation reflects the particular concerns of the exegete who opted for it based on a specific connection between this interpretation and the Quranic text, whether the connection is grounded in the prima facie reading of the text or in an imagined narrative context.21 Table 1 lists the various types of additions found in the reports.
The first type of reports add almost nothing to the literal text of the two verses; in the cases in which such additions are encountered, they merely specify the time (during the Prophet’s life) or place (Mecca) in which the event occurred. Sometimes the reaction of Muhammad’s adversaries, who accused him of sorcery, is also reported. These reports are simple because they could be considered primitive receptions of the text, although this does not necessarily strip them of polemic tenor. A polemic attitude is evident in the fact that the only benefit of such reports—superfluous to the Quranic verses—is to deny the metaphorical or eschatological interpretation, which suggests that the audience was aware of such interpretations and the purpose of these reports was to distance the audience from them. Simple reports survive in early sources; this documentation and their strong affinity with the Quranic text thus concur to support the opinion that these reports are extremely old, but this does not mean that they are the only old reports, as seen in the discussion of composite reports. A typical simple report is “The moon was split during the time of the Messenger of God” (inshaqqa al-qamar ʿalā ʿahd rasūl Allāh) (Abū Yaʿlā al-Mawṣilī 5:306; Ibn Ḥanbal 3:275; Muslim 8:133; Ṭayālisī 265; Tirmidhī 5:71). [End Page 156]
In the composite reports, the Quranic text is interpreted beyond the prima facie reading. These reports are characterized by many significant details that accrue from the concurrence of the literal reading and the extra-Quranic material the exegete incorporates. The additions vary depending on their position in the report and their connection with the purpose of the exegete or transmitter.
The difference in complexity between composite and simple reports does not necessarily mean that the latter predate the former. Instead, it can be argued that the khabar was, from the earliest phases, received in two distinct ways, developed along separate courses. If one continues to assume that nonsimple readings of the text must have grown out of simple ones, composite reports must be taken as acts of reinterpretation. But if it is assumed, as I do here, that two distinct readings existed, the diversity of reports can be seen as evidence of parallel interpretations. Parallel interpretation is preferable to reinterpretation as the explanatory hypothesis because the latter ignores the fact that in religious traditions the transmission dynamics are characterized by ruptures and rediscoveries in a way that resists being reduced to the simple handing down of conscious memories (Assmann, Religion 26).
Of the additions that emphasize the splitting of the moon as a miracle the Prophet produced, two kinds can be noted: contextual additions and graphic additions. Although these correspond roughly to the distinction between scenic and narrative memories (see Assmann, Religion 2–3), they are different in that they are not assumed to belong to different registers of memory or consciousnness.
Contextual additions aim to create a narrative course; they transform the splitting of the moon from an extraordinary incident to one whose plot lends itself much more efficiently to polemics. The first such addition is that the moon was split after Muhammad’s adversaries had asked him for a sign. At this relatively early point of his call, his adversaries were almost solely his Meccan tribesmen, particularly tribal leaders and the wealthy commercial elite. Among other things, they accused him of being a sorcerer (Kermani and Graham 125). Thus the first addition serves a theological purpose inasmuch as it transforms a mere account into a polemical device that can be used to prove Muhammad’s prophethood. It also provides a context for the Quranic description of the reaction of Muhammad’s adversaries: “Yet if they see a sign they turn away, and they say ‘A continuous sorcery!’” The word sign (āya) used in this verse has many meanings in Quranic usage, so the addition highlights a particular meaning, namely, a miracle. This addition is very old and [End Page 157] has been part of the Muslim reception of the khabar since the earliest times, as attested by early sources.22
The second contextual addition is that the Prophet solicited the testimony of the audience, supposedly a mixed group of believers and unbelievers, that the moon had indeed been split. It shows that the exegete or transmitter is aware of the existence of a contrary reception of the khabar. In fact, asking those present to bear witness to an event that would have been overwhelmingly clear is best seen as a message addressed to those who deny the event because of their absence from the scene. While using the language of narrating a past interaction, this line of argumentation seems more concerned with a similar reaction among its present audience. The Companion who reports the request for testimony—mostly Ibn Masʿūd—is then in a position of authority, testifying to support the prophetic miracle in the face of Muslims who chose not to accept it. By invoking the authority of the esteemed Companion Ibn Masʿūd, this addition shows that its audience is a crowd that held this Companion in high regard and still harbored doubts about the occurrence of a miraculous splitting of the moon. This suggests it is an Iraqi crowd in the city of Kūfa where Ibn Masʿūd settled following the conquests (Gilliot, “Creation” 45). This addition is early, too.23 It also appears in a more specific version in which the Prophet asks another Companion, Abū Bakr, to be his witness (Ṭabarī 27:115). This further reinforces the polemic edge and highlights the excellence of Abū Bakr. The message of this version is similar to the message in a version of the story of Muhammad’s night journey (isrāʾ). In that version, Abū Bakr is presented as the only person accepting Muhammad’s claims of having traveled to Jerusalem and back to Mecca overnight. Abū Bakr relied only on his great faith in the Prophet’s veracity to fully assent to the latter’s claim, which earns him the epithet “the Truthful” or “the Trusting” (al-ṣiddīq) (Athamina). The strong polemic edge of this version is best understood as part of the strong effort to defend and vindicate Abū Bakr: after all, he was Islam’s first caliph, and his disputed credentials in the eyes of his detractors gave rise to bitter political and sectarian debates in later Islamic history. Some scholars even argue that the dispute over Abū Bakr’s qualifications formed the earliest discourse of Sunni–Shiʿi polemics in Islam (Afsaruddin 272).
The third contextual addition is located between the first two. It shows the Prophet stipulating that his adversaries believe in his message before he proceeds to perform the miracle. The benefit of this addition is to emphasize their stubbornness, which explains their rejection of Islam in the eyes of later generations. Another [End Page 158] function of the addition is to show the similarity between Muhammad’s plight and the suffering of previous prophets in calling their recalcitrant peoples to the path of God. This addition might be later than the first two, as it appears in later sources.24 In addition, it presents Muhammad as someone who will deliver a miracle only after making sure he will not do so in vain; this characterization reflects a theologically more nuanced awareness than that found in the first two additions, as it echoes the discussion about the significance and legitimacy of soliciting miracles without purpose (iqtirāḥ al-muʿjizāt) (see, e.g., Murtaḍā, Dhakhīra 384; Rāzī 21:59).
The fourth contextual addition describes in detail the reaction of Muhammad’s adversaries. In addition to judging the miracle to be an act of sorcery, they escalate their response by suggesting that they wait to ask people who were traveling outside Mecca at the time about what they saw in their respective destinations to make sure Muhammad’s sorcery did not reach far beyond Mecca. This addition appears in an early source (Ṭayālisī 38), which makes it unlikely that it postdates the simple reports. In some versions of this addition, the scope of the proposed investigation is widened from travelers to “travelers to all destinations” (Iṣfahānī 1:281; similarly Bayhaqī 2:266). The thrust of the argument in these later versions is clearly directed against Muslim theologians who denied the reality of the splitting of the moon by pointing to the absence of astronomical observations attesting to it in different countries.
Graphic additions revolve around describing how the moon was split; as such, their purpose is to dismiss any doubt concerning the miraculous nature of the event. The first graphic addition provides a description of the split moon; it mostly consists of phrases that qualify the verb “was split” (inshaqqa), thus reporting that the moon was rent into two “halves” (niṣfān), “parts” (firqatān), or “pieces” (shiqqatān) (ʿAbd al-Razzāq, Tafsīr 3:257; ʿAbd b. Ḥumayd 356; Abū Yaʿlā al-Mawṣilī 5:463; Bayhaqī 2:265; Bukhārī, Ṣaḥīḥ 6:53; Ibn Ḥanbal 4:81; al-Ḥākim al-Naysābūrī 2:471–72; Nasāʾī 6:476; Nuʿaym b. Ḥammād 367; Muqātil 3:96, 296; Muslim 8:133; Ṭabarī 27:112; Tirmidhī 5:72), in addition to other, less common, terms that were gradually incorporated into the khabar (al-Ḥākim al-Naysābūrī 2:472 [falqatān]; Iṣfahānī 1:280 [qamarān]; Ibn Kathīr, Sīra 2:117 [juzʾān]; Muslim 8:132). The preponderance of this addition in early sources, together with its strong affinity with the remaining text of the khabar, suggest its old provenance. Nevertheless, it is unclear whether it was a later qualification of the simple report or a parallel development along a separate course that gave rise to the composite interpretation.
In the second graphic addition the location of one of the two parts is curiously [End Page 159] described: it was hidden by the mountain (satarahā al-jabal) or—in an alternate variation—went behind the mountain (dhahabat khalf al-jabal). This description leaves open the possibility that the event was a natural phenomenon, with the moon blocked by the mountaintop, which casts doubt on its miraculous nature. This addition may also be as old as the khabar itself, which places it in the context of a particular reception that did not emphasize the event’s miraculous nature. It is worth noting that most of the isnāds of this addition converge at Ibn Masʿūd, who, at the time the verses were revealed, was old enough to witness the event, unlike other narrators (Ibn Ḥajar 7:139).
The third graphic addition is ascribed almost exclusively to Ibn Masʿūd, who is said to have seen the mountain between the parts of the split moon.25 This addition is likely to be an affirmation of the first of the graphic additions, aiming to dispel any doubt that the moon was divided into separate pieces. It may be seen as a response to the second addition, if one takes the reception of the khabar to have developed along one course; or it can be a product of a reception that took it to signify a miracle from the outset.
The fourth graphic addition betrays the concern of the exegetes or transmitters for showing the scale of the event and documenting its details by recording the names of relevant places. Therefore, this addition lists the names of the two mountains on top of which the separate pieces of the moon stood.26 The names given are variously al-Suwaydāʾ and al-Khandama,27 or Abū Qubays and Quʿayqiʿān (Ibn al-Jawzī 7:242; Iṣfahānī 1:279; Samʿānī 5:307; Qurṭubī 17:127), or Abū Qubays and al-Suwaydāʾ,28 or al-Ṣafā and al-Marwa (Iṣfahānī 1:280). Some later traditional scholars, aware of the difficulty caused by the discrepancy in the names, tried to reconcile these versions of the same report (Ṣāliḥī 9:432). However, it remains the case that this addition is likely to be later than the third graphic addition in the development of the khabar, whether the reception of the khabar developed along one course or two.
The sources relate that the moon was split twice, which would be an addition worth studying (ʿAbd al-Razzāq, Tafsīr 3:257; ʿAbd b. Ḥumayd 356; Abū Yaʿlā al-Mawṣilī 5:463; al-Ḥākim al-Naysābūrī 2:472; Ibn Ḥanbal 3:165; Muslim 8:133; Nasāʾī 6:476; Nuʿaym b. Ḥammād 367; Ṭabarī 27:116; Tirmidhī 5:72). However, it is very likely that this addition is due to misreading a word such as firqatān, taking it to be marratān. This view is further supported by the fact that later traditionists tried to interpret the word marratān to mean two parts and not two times, affirming that the moon was split only once.29 [End Page 160]
Any report of the khabar containing additions, contextual or graphic, must be counted among the composite type. The reason for this judgment is that all these additions aim to perfect the khabar and fortify it against doubt concerning its relevance as proof of Muhammad’s prophethood. The additions do not seem to go beyond this purpose, even when they are all found in a single report: the core remains the actual splitting of the moon, and the rest is supporting details.
Early sources of various genres—Quranic commentaries, musnads, muṣannafs, ṣaḥīḥs, and muʿjams—are replete with composite reports. This profusion shows that this kind of report did not come into existence much later than the simple type; it is even possible that some of its less composite forms—such as those containing the first contextual addition or the first graphic addition—were born with the khabar itself. In any case, composite reports seem to have gained widespread acceptance by the time of the inception of historical writing in Islam. This is why it cannot be assumed that composite reports are cases of reinterpretation; the evidence is plentiful enough to suggest a case of parallel interpretation. Muslim reception of the khabar provides further support for the thesis of reinterpretation: if interpretation depends on the audience, then it may have been divorced from the text since the initial reception, which makes for a good case in favor of interpretation, not reinterpretation.30
Reports of the complex type are best treated separately, because they do not aim to enrich the Quranic text with additions that elaborate on the details and significance of the event as a miracle. Instead, these reports inject into the khabar additions that are independent from—or only weakly connected to—the function of the described miracle. Examples of this type include additions in which the sectarian biases of the exegete or transmitter are so overwhelming as to make the Prophet’s miracle just one aspect of the khabar among many others. Some complex reports contain additions that are contextual or graphic in nature but so entangled in the detailed framework of the complex report that classifying them as composite would entail an act of decontextualization. One such case is an addition relating that the moon was split when Muhammad raised his head; this is a graphic addition that aims to show reverence to the Prophet. Its source, however, is a report full of details aimed at demonstrating the superiority of the Shīʿa, thus overlaying a common composite report with a particular sectarian attitude.31 Despite the importance of [End Page 161] complex reports, their sources in the first four Islamic centuries are almost exclusively works that pertain to sectarian polemics, which undermines their relevance for the purposes of this study.
A prominent example of a complex report is one ascribed by the early Nuṣayrī scholar al-Khuṣaybī (d. 958 or 969) to Muḥammad b. ʿAlī al-Bāqir (d. 733) on the authority of Jābir al-Juʿfī (d. 745), the famous gnostic figure in Kūfan Shīʿism.32 The report specifies that the Prophet’s adversaries were Qurayshīs of the clans of Umayya, Taym, and ʿAdiyy—an allusion to people usually presented in a negative light in Shīʿī contexts. It also claims that the Prophet gave ʿAlī an influential role in performing the miracle, putting the two in the context of previous prophets and their legatees (waṣīs), such as Moses and Yūshaʿ b. Nūn.33 The report also describes, in a very dramatic tone, how the two parts of the split moon fell on the mountains of Mecca. The miracle itself is clearly taken for granted, and the more critical interest of the author is to praise certain figures and vilify others.
Theological Challenges and Responses
I noted earlier that theologians were divided into two camps: the first affirmed the past reality of the splitting of the moon, whereas the second took the verses to refer to apocalyptic events. Although al-Ḥasan al-Baṣrī and ʿAṭāʾ—neither of whom was a theologian in the later technical sense of the term—are said to have supported the latter view, the sources are silent on their reasons. However, it is possible to glean the justifications provided by theologians for this position.
Two main theses provide the deep theological foundation for denying the actual reality of the moon’s splitting. The first is an epistemic consideration about the impossibility of suppressing reports of significant events in the absence of sufficient motivation to do so.34 The splitting of the moon would be an extraordinary occurrence; thus, there would necessarily be a plethora of reports attesting to its reality. Since this is not the case, the event cannot have taken place. The second is a moral consideration about the need for a binding proof that could lead all those who are subject to the law (mukallafūn) to believe in God. Although theologians disagree on how to argue for the necessity of such a proof, whether by way of reason (ʿaql) or through revelation (samʿ),35 neither applies to the splitting of the moon, since it does not present itself to all subjects of the law with an equal degree of clarity. [End Page 162]
The sources concur almost unanimously that the most vehement denier of the moon-splitting miracle was Ibrāhīm al-Naẓẓām, who went as far as to blatantly call Ibn Masʿūd a liar for claiming to have seen it. In reference to the moral justification, Naẓẓām argued that miracles, inasmuch as they are proofs of prophethood, cannot be shown to a group of people and concealed from the rest.36 With his Muʿtazilī view of divine justice, Naẓẓām must have found it easier to discredit a Companion than to doubt God’s justice. This position aligns well with his insistence on divine justice to the point of claiming that God is unable to commit acts of injustice;37 it is also compatible with his bitter criticism of Companions on a number of issues (Van Ess, “Al-Naẓẓām”).
Nevertheless, Naẓẓām’s position is not confined to the moral justification. He also summons the epistemic justification by noting that the splitting of the moon was not mentioned in poetry, nor was it used in fixing calendars, nor did it make any non-Muslim convert to Islam; he even hints that it was not an argument that Muslims used to defend Muhammad’s prophethood (Ibn Qutayba 71). Just like Naẓẓām’s moral justification is in accordance with his views on divine justice, his epistemic justification dovetails with his views on knowledge and the transmission of reports; the latter views were so extreme as to have been dubbed “arrogant rationalism” in recent scholarship, and they left him with very few followers (Van Ess, Flowering 172). It seems that his disciple al-Jāḥiẓ (d. 868) followed him on this issue, adding that if the moon were split then calendars would have become irregular due to confusion regarding the stations of the moon. For al-Jāḥiẓ, the Quranic verses refer to the splitting of the moon during the ultimate apocalypse (Marzūqī 86).
For these theologians, the relative paucity of reports about the splitting of the moon and the small number of initial transmitters meant that the event had not taken place. Although this position appears to have attracted a group of Muʿtazilīs, it did not gain popularity outside their circles; it is even likely that the various adversaries of the Muʿtazilīs took this position to be a weak point on which to criticize them.38 Although Ibn Qutayba’s attacks on Naẓẓām suggest that believing in the splitting of the moon was not common in his time, two generations after Naẓẓām,39 it may nonetheless be inferred that affirming the occurrence of this miracle had become too famous a position to be ignored by the beginning of the tenth century. The social dynamics of religious learning can be seen in this discussion. The position of some Muʿtazilī masters reveals a haughty attitude toward other religious scholars, since they base their dismissal of the khabar on the claim that it was forged by people they describe dismissively as purveyors of biographies, [End Page 163] stories, and accounts of military expeditions (min tawlīd aṣḥāb al-siyar wa-muʾallifī al-maghāzī wa-nāqilī al-āthār) (Mufīd, Fuṣūl 20–21).
In addition, the philosophers (falāsifa) are also reported to have denied the splitting of the moon, though little explanation is provided of their reasons for doing so. The scanty available material, contained in the works of their adversaries, suggests that the position of the philosophers was based on the impossibility of the destruction and reconstruction of the heavens; thus unlike that of the theologians, the justification of the philosophers is neither moral nor epistemic but ontological (Ibn Ḥajar 7:141; Rāzī 29:28). Ibn Sīnā is reported to have written a commentary, now lost, on the verses concerning the splitting of the moon, but the authenticity of this ascription cannot be ascertained. What can be inferred from such reports is that the issue was being discussed in high intellectual circles in the sectarian milieus in which Ibn Sīnā flourished.40
The preceding exposition shows that the metaphorical and eschatological views of the moon-splitting khabar, both of which deny the reality of the event, soon came to be limited to a minority, whereas the third view, that the Prophet did indeed split the moon, prevailed. Proponents of the third view were aware that the objections leveled against their position—particularly those based on theological justifications—could not be ignored. Therefore, many theologians and exegetes attempted to answer these objections, and some continue to do so.41
Within the ranks of the Muʿtazilīs, a few theologians took it on themselves to renounce the view of Naẓẓām and his supporters. However, it seems that his view was treated differently because of its bluntness; his supporters’ positions were usually more open to interpretation, bringing them closer to the third, “mainstream” view. The strategy that was employed to counter Naẓẓām’s view consists of demonstrating its moral invalidity. Thus if the splitting of the moon is to be read as an apocalyptic sign, it becomes futile to ascribe any moral value to it, since the end of time brings with it the end of all moral obligation for humanity. Not content with the alleged moral futility of Naẓẓām’s alternative, his critics proceeded to turn this moral threat into an opportunity: the fact that the event was seen only by a few people in a restricted locality can be seen a proper moral measure intended to preempt the possibility of charlatans in faraway places claiming it as a miracle for themselves. As for Naẓẓām’s epistemic justification, it was argued that people sometimes fail to observe even a total lunar eclipse, so it is more likely that they overlook a swift event such as the splitting of the moon; alternatively, some witnesses may have neglected to report the event, contenting themselves with the testimony of Ibn [End Page 164] Masʿūd and the Quranic mention of it.42 Other Muʿtazilīs took their apologetics even further, claiming that what their early masters had denied was not the reality of the event but its validity as a proof of Muhammad’s prophethood for those who were not physically present; but if one establishes belief in Islam through other ways, such as through the Quran, it becomes obligatory to subscribe to the belief in the reality of this miracle.43
Muʿtazilī theologians thus share the views of other theologians in defending the occurrence of this miracle, the only difference being the former’s need to justify the unpalatable positions of their early masters. Because the remembrance of an event must be enriched with the meaningfulness of a significant truth (Assmann, Cultural Memory 24), it is paradoxical that the apologists, regardless of their doctrinal affiliation, base themselves more often than not on slighting the significance of the event due to the lack of proper attestation. Qualifiers proliferate, leaving the impression that these putative defenders of the prophetic miracle are essentially issuing disclaimers. Once more, the lines of sectarian division are blurred, with Ashʿarī, Muʿtazilī, Imāmī, and Ḥanbalī theologians—among others—joining forces to suggest as many justifications as possible, hoping to explain the relative lack of documentation of a supposedly huge celestial phenomenon. They suggest, inter alia, that the reality of the moon’s splitting can be inferred only through investigation, so belief in it is inaccessible to anyone who is unable to carry out this investigation; that the Companions’ failure to discredit the khabar’s narrators is evidence of its reliability; that few people ponder the heavens; that the moon may appear only to some people; that the split lasted a mere moment; that only those arguing with the Prophet were alerted to the miracle, while the rest were asleep; that people cover themselves and go indoors during the night and thus did not see the event; that clouds could have concealed the split moon; and even that some observers might have thought it was simply an eclipse ([Pseudo-]ʿAbd al-Jabbār 1:56; Jaṣṣāṣ 3:552; Ghazālī, Mustaṣfā 114; Ibn Ḥajar 7:141; ʿIyāḍ, Shifā 1:283; Rāzī 29:28; Ṣāliḥī 9:43; Ṭabrisī 9:310; Ṭūsī, Tibyān 9:443). In case these arguments fail to accomplish the desired end, the theologians’ discussion reverts to raw polemics: they remind their opponents (who are supposedly non-Muslims) that this miracle is as well attested as are those of Moses and Jesus (Bāqillānī, Tamhīd 139). The last remark is very indicative, as it reveals the ultimate motive behind the theological ascription of miracles to the Prophet—namely, meeting the challenge of other religions. This bears an interesting resemblance to the case in Shakespearean negotiations when it comes to handling the question of exorcism: the official position, that is, the [End Page 165] affirmation of the reality of exorcism, is emptied out by the semblance of loyally confirming it, thus invoking Macheray’s “inernal distantiation” (Greenblatt 102).
So successful were the proponents of the realist reading that scholars found it easier to defend it with this long list of disclaimers than to reject it based on rational arguments or for reasons of weak transmission. At the same time, these arguments reveal the strength of the objections posed to the realist view on moral and epistemic grounds that endured as pillars of theological thinking. Naẓẓām was indeed an iconoclast in this regard, going against the tide of popular piety. As for other, later theologians, the need to balance theological speculation and scriptural-literal dictates became a major concern, though strongly inclined in favor of the latter in case of conflict. These theologians faced a dilemma that required either of two solutions or both. The first option was to sacrifice the clarity of the miracle, since it was hard to deny it took place. But it was even harder to claim that it did take place in a clear manner but was not copiously reported, for this claim might eventually jeopardize the probity of prevalence as an epistemic tool. The second possibility was to sacrifice the universality of the miracle, since it was witnessed by so few. This sacrifice protects the moral foundation that all subjects of the law must base their belief on proof, in this case the miracle, that is accessible to them.
An Alternative Reading
The foregoing analysis leads to the conclusion that the Quranic text gave rise to the discussion concerning the splitting of the moon and this discussion is old, given its strong affinity with the text and its mention in early sources. Moreover, simple and composite reports of the event probably existed prior to the early sources in which they survive, in contrast to complex reports, which were compiled at a later stage.
The abundance of reports about a particular event and the prevalence of one view over others merit special investigation. Although the need of the community to ascribe miracles to its Prophet and the fact that the event is mentioned in the Quran are important reasons behind such attention, they are not peculiar to this khabar. It is possible to approach the question by comparing these two verses with similar Quranic occurrences and placing them in the context of seventh-century Arabia, in addition to interrogating the various interpretations of different exegetes. This might shed light on a more convincing reading of the text and allow for a better understanding of the motives behind the distinct attention paid to the khabar by [End Page 166] early Muslims. Attempting a more convincing reading would also help in assessing the respective likelihoods of the two distinct courses of reception: reinterpretation and parallel interpretation.
The first verse is different from most apocalyptic verses in that it does not contain a conditional clause (unlike, e.g., Q 82:1, Q 99:1, and Q 74:8), nor does it contain a time clause (in contrast to Q 89:23, Q 82:19, Q 86:9, and others). Even though this difference does not alone necessitate a different interpretation of this verse, it is very likely that its peculiar wording would draw the attention of an audience for whom the Quran was the heart of religious experience. The syntax of the verse suggests that the described event actually took place, rather than a future occurrence like those alluded to in the other apocalyptic verses.
This conclusion is further supported by the context of the three opening verses: the movement from the past tense (verse 1) to the conditional (verse 2) and then back to the past tense (verse 3) suggests that the condition mentioned has been met, an implication frequently noted in the works of exegetes concerned with the Quran’s rhetorical aspects (Ibn Ḥajar 7:141; Jaṣṣāṣ 3:552; Nawawī 17:143; Qurṭubī 17:126; Ṣāliḥī 9:432; Samʿānī 5:307; Ṭabāṭabāʾī 19:55; Ṭūsī, Tibyān, 9:443; Zamakhsharī 4:36). This syntactical peculiarity supports the interpretation of these verses to mean that Muhammad’s adversaries did see a sign, then considered it an act of sorcery.
The Quran usually mentions the moon in association with the sun in verses that extoll the greatness of God’s creation and in those that pertain to the end times (Varisco). The sun and the moon are occasionally mentioned separately, as in twhere the Quran describes the state of the moon when the Hour draws nigh. There, the advent of the Day of Judgment is related to a lunar eclipse. Whereas Q 54:1 speaks of the moon splitting and the Hour approaching, Q 75:7–10 binds together the advent of the Hour and the eclipse of the moon: “But when the sight is dazed, and the moon is eclipsed, and the sun and moon are brought together, upon that day man shall say, ‘Whither to flee?’”
The sources provide ample material attesting to the spread of apocalyptic awareness among the first generations of Muslims, many of whom anticipated the end of the world. This awareness is mostly indicated by the currency of books on the apocalypse (malāḥim, fitan, ashrāṭ al-sāʿa) and relevant parts in the corpus of traditions.44 It comes as no surprise, then, that in such an atmosphere people would hasten to embrace an apocalyptic interpretation of phenomena. A lunar eclipse is no exception, since the splitting of the moon can be taken to refer to an eclipse that will take place when the Hour draws nigh, especially given that [End Page 167] “whatever will come to be is well-nigh” (kull āt qarīb).45 Indeed, sources preserve the view, though as an anonymous one, that the splitting of the moon does refer to an eclipse (Samʿānī 5:307).
Referring to the splitting of the moon as a sign (āya) does not contradict reading the splitting as a metaphor for eclipse. Early works of ḥadīth and fiqh attest to such reading, as in the prayer of signs (ṣalāt al-āyāt) ordained to be performed on solar and lunar eclipses (ʿAbd al-Razzāq, Muṣannaf 3:103; Abū Yaʿlā 9:271; Bukhārī, Ṣaḥīḥ 2:23; Ibn Abī Shayba 2:355; Ibn Ḥanbal 6:76; Muslim 3:27; Nasāʾī 3:130; Tirmidhī 2:36), since both heavenly bodies are signs of God by which His servants should be edified (Ibn Ḥanbal 5:16; Ibn Ḥibbān 7:102). The Quran contains many references to signs (āyāt) that are not technically miracles,46 which provides an additional clue that the first recipients of the Quranic text could easily have taken the “sign” of the splitting of the moon to refer to something other than a miracle. If one consults early lexical sources, one can even argue that the most common understandings of the term have nothing to do with miracles.47
Some pre-Islamic Arabs believed in the influence of eclipses on worldly events and human life. Oracles (kuhhān) used to claim superior power based on their alleged ability to influence heavenly bodies. They also based their predictions on solar and lunar eclipses. It appears that panic prevailed whenever these bodies changed appearance, for the oracles believed that such changes indicated the death of dwellers in heaven (halāk man fī al-samāʾ) (ʿAlī 6:804); this reaction to eclipses is by no means peculiar to pre-Islamic oracles, as it has been observed as far as China, where eclipses invoked the performance of memory rituals (Assmann, Religion 13–14). These Arabian beliefs are evidently akin to sensing the end of the world upon observing grand celestial phenomena. Also indicative of the moon’s great significance in sorcery is that women, in making spells to repel men, are said to have taken oaths on the vanishing (ufūl) of the moon (ʿAlī 6:749). It is thus no wonder that the Prophet’s adversaries considered the lunar eclipse a work of sorcery and accused him of being an oracle, as mentioned in these and other verses of the Quran. This analysis also finds support in the observation, couched in the language of an anthropological remark, that the infidels used to react to the lunar eclipse by saying that it was the work of sorcerers (ahl al-shirk idhā kusifa al-qamar yaqūlūna hādhā ʿamal al-saḥara) (Ṭabarī 27:114).
This analysis shows that paying close attention to the prima facie reading of the text, taking into account certain beliefs held by the pre-Islamic Arabs and early Muslims, and consulting various works of Quranic exegesis and lexicons provide [End Page 168] insights that necessitate neither accepting the reports of a miracle nor dismissing all source material as purely imaginative.
Traditional scholars—that is, Muslim scholars applying the premodern approaches to the Quran—agree that sūrat al-Qamar (Q 54) is a Meccan sūra,48 with some specifying its date of revelation as around five years before the hijra (Ibn Ḥajar 6:464, without specifying his source). In modern scholarship, it has been judged to belong to the second Meccan period, that is, the fifth and sixth years after Muhammad’s initial reception of revelation (Nöldeke 1:69–70). These two estimates are close, though neither claims decisive accuracy. It thus seems plausible that a lunar eclipse took place during Muhammad’s Meccan period, probably around six years before his migration to Medina. His adversaries reacted by calling it an act of sorcery; Muhammad, however, asserted that the phenomenon was a sign from God that called for reflection, as attested by the two verses.
This proposed reading of the event explains the proliferation of reports concerning the splitting of the moon without necessarily contradicting the conclusion that the composite reports—which specify that the moon was split into two parts— are in fact a creation of storytellers. It also offers a reasonable explanation for the complete absence of the khabar from the works of Ibn Isḥāq and Ibn Saʿd: if no miracle took place, it is not surprising that the event was not reported. This reading is further supported by the fact that Ibn Isḥāq dedicates a lengthy section of his work to listing the miracles the Meccans suggested Muhammad perform; the text enumerates many things the Quran reports (the gushing forth of a wellspring, the falling apart of the heavens, the summoning of angels), but it makes no mention of splitting the moon. Ibn Isḥāq notes that the Prophet did not deliver any of the requested miracles, and his wording suggests a rejection of the notion that Muhammad would perform miracles in response to a challenge from his opponents (as assumed by the first contextual addition discussed above). The same applies to the commentary on Ibn Isḥāq’s text by Suhaylī (d. 1185) (Suhaylī 2:47).
It remains to find support for the proposed reading from external evidence, for its compatibility with the Quranic text and its cultural context do not suffice to justify its prioritization over the reports found in traditional sources or the interpretations of recent scholarship. Two such external clues are available.
First, there are two versions of one report of the khabar, both on the authority of Ibn Jurayj (d. 767), in al-Muṣannaf of ʿAbd al-Razzāq (d. 826) and al-Muʿjam al-kabīr of al-Ṭabarānī (d. 971). The first version is ascribed to ʿIkrima (d. 723), a client of Ibn ʿAbbās (d. 687) and the second to Ibn ʿAbbās himself. Either way, it [End Page 169] must be an early report, since the historical authenticity of the Muṣannaf material, particularly that narrated on the authority of Ibn Jurayj, has been convincingly shown (Motzki, “Muṣannaf”). Thus, it is safe to assume that the content of these extremely similar versions was established by the seventh century. The text, as it appears in the Muṣannaf, reads as follows:
ʿAbd al-Razzāq, on the authority of Ibn Jurayj, who said: ʿAmr b. Dīnar informed me, from ʿIkrima, the client of Ibn ʿAbbās, that the moon was eclipsed during the lifetime of the Messenger of God. They said: the moon has been subjected to an act of sorcery. The Prophet said: “The Hour has drawn nigh: the moon is split.”
The report of the khabar in this version is self-evident. It is almost a simple report, without any additions to the words of the Quran. Its only divergence lies in its use of the word eclipsed (kusifa) instead of split (inshaqqa), which is still valid since the visual aspect of the eclipse can be described in terms of splitting, especially when one recalls that some of the graphic descriptions of the event portrayed the moon as if it were eclipsed. This is probably what made Ibn Kathīr (d. 1373) propose—while still expressing his discomfort—that the moon was both eclipsed and split (Ibn Kathīr, Tarīkh 6:8).
Second, astronomical tables show that the moon underwent several partial eclipses that were visible in Mecca between 614 and 621.50
Some scholars have noted the possibility of an eclipse, even endorsing it and citing ʿAbd al-Razzāq’s version of the khabar (see Rubin 47; see also Görke 101). But they have not closely investigated the relationship between the wording of the Quranic text and this version of the khabar, nor have they highlighted the theological significance of various reports. The core of the present study is investigating the relationship between the Quran and early akhbār, concentrating on the polemical needs that gave rise to narrative construction.
Conclusion: Narrative and Historical Truth(s)
Islamic culture developed a sophisticated historiographical literature, producing subgenres almost unknown to other historiographical traditions (Khalidi 56–57). This reflects the extent of its dependence on the past, best captured in the observation [End Page 170] that sunna (Arabic for “custom, tradition”), which describes the almost unanimous Muslim view of most material about Muhammad, is in fact an expression of normativity grounded in past recollection. It is also the formal self-proclaimed epithet of the vast majority of Muslims, known as Sunnis in common discourse; the long version is ahl al-sunna wa-l-jamāʿa (Those Preserving the Custom and Adhering to the Community). Following Pocock’s formula concerning the direct proportionality of past dependency and nonnormal historical explanation, one expects to find a higher frequency of such explanations when certain traditional expositions of the past are met with challenges in the Muslim community (Pocock 145).
The specific discussion of the two verses on the splitting of the moon is a case in point. It is clear that a discussion of the Quranic text ensued after the Prophet’s death, and that its relation to the Prophet, and the relation of those discussing it, was a highly charged question in a community of faith. The most probable scenario is that the khabar was initially received in ways that engendered two different primary understandings. In turn, these two understandings were reinterpreted along separate courses of development, giving rise to parallel interpretations. The first of these understandings is based on a view of the event as nonmiraculous, whereas the second sees it as miraculous. The first understanding survives in the version of the khabar related on the authority of ʿIkrima, operating most likely within the Arabian Peninsula; the second is rooted in Ibn Masʿūd’s Iraqi reception. Regardless of which version is the prototype of the latter understanding, it must be noted that the khabar as related by Ibn Masʿūd is the richest in narrative features, for the narrator is not detached from it but is at its heart. Perhaps he or one of his close surrogates, speaking on his behalf, gave it its original form, which then developed further in later composite reports.51
Later, the dominant explanation of the two verses came to be the miraculous explanation. This explanation is far from historical in the normal sense of the term, in accordance with Pocock’s formula. It is not difficult to grasp why it ended up this way: the dependence of the community on its past is best exemplified in a moment where the sacred divine word meets with a miracle believed to be the proof of Muhammad’s veracity. This moment is an anchor of Muslim identity, which heightens to a maximum the need to address any challenge to its authenticity as a traditional exposition of Islamic past. The social negotiations and the competing needs of various discourses therefore understandably favored such a miraculous explanation, which becomes more dominant with time. Even long after the defeat of elitist theologians, who from the beginning expressed skepticism about the [End Page 171] miraculous explanation, their challenges were still heeded by many later, less skeptical theologians. That is, the force of these early skeptical positions was still felt and had to be reckoned with in acts of calibration and balancing, generating a curious situation in which these later theologians were in fact emptying out the miraculous scenario by loyally confirming it, echoing Greenblatt’s (126) observation regarding the position of Shakespearean plays on exorcism. This long history of social conflicts is brought to light when the meanings of the text are fractured, giving away its many ideological mystifications formed due to the contributions of various discourses: skeptical theologians posing clever challenges, populist traditionalists guarding the literality of the miracle reports, and sectarian polemicists promoting their favorite personalities. That the dominant miraculous explanation is more Iraqi centered than the other is also a telling indicator, for it was in Iraq that most of extra-Quranic material was circulated and codified (Donner 55–60); it was in Iraq that the struggle between these social forces, intellectual traditions and political-sectarian parties crystallized (Hodgson 1:241–80); and finally, it was in Iraq that the rich mythological lore of the ancient Near East was first made available to Muslims (Cobb 252).
To gain more insight into the perspective provided by Ibn Masʿūd’s reports, it might be useful to draw on recent psychological and neurological research on the working of memory. This research stresses the malleability of human memory, and the prime example of this malleability is the great difficulty we experience in distinguishing between two kinds of “memories.” On one hand, we have genuine memories of external events, but on the other, we also have “memories” that we may have inadvertently borrowed from the accounts of other people or that have been simply suggested to us by an external agent. These two types of memories have been called “historical truth” and “narrative truth,” respectively—apt terms from the perspective of the present study. Experiments have shown success in implanting false memories by suggesting to a subject that he or she has experienced a fictitious event, thereby creating a narrative truth for that subject. When the suggestion is compounded by the presence of an authority figure and accompanied by vivid sensory imagery and strong emotion, there may be no way of distinguishing narrative truth from historical truth. It has even been proposed that our truths are frequently purely narrative, leading to the conclusion that human memory is dialogic: it arises from both individual direct experience and the interaction of many minds that shape memories by borrowing and suggesting them to each other (see Sacks). These findings are extremely helpful for understanding Ibn Masʿūd’s reception of the khabar. It is easy to detect the factors that contribute to creating [End Page 172] narrative truth and its replacement of any possible historical truth. The authority figure is no less than the Prophet himself; his involvement need not have taken the form of suggesting the event but could simply have consisted of uttering the verses that set Ibn Masʿūd’s mind on the track to constructing a memory of the miracle. The emotional and sensory factors are also evident in Ibn Masʿūd’s deep commitment to his faith, the panic of the Meccan crowd, and the splendor of an eclipsed moon. Although the dialogic aspect of memory might have also played a role in developing the story further for Ibn Masʿūd, its function was probably more influential when the narrative began to be formulated in its more elaborate versions, including the ones provided by storytellers.
This is not to determine with certainty the identity of the originator of the khabar. For this, as argued by Greenblatt, is a fruitless quest. The social energy, surely pregnant with the charisma of a great religious leader and grand historic conquest, was too much to be concentrated by one person. But it can be inferred that the moment of inscription, inevitably a social moment (Greenblatt 5), was the product of negotiating the new alignments of power and identity within the emerging Muslim community and empire: who knew the scared text best, who was closer to the Prophet, who witnessed the miraculous firsthand, and in whose hands was the body of religious teaching most reliably centered? Thus despite the possibility of Ibn Masʿūd being the first narrator to provide this explanation of the two verses, he serves less as the historical person who did so and more as the figure credited, by later generations, with the “moment of inscription.” As such, his moment is that of “choice, decision and action that creates the social reality of the text” (Spiegel 25–26). [End Page 173]
Hussein Abdulsater (Ph.D. religious studies, Yale 2013) works on Muslim ethical and theological discourses and on the formation of group identities and self-understanding in classical Islam. In particular, he studies how such themes were expressed in humanistic disciplines such as historiography and literature, in addition to theology and Quranic exegesis. He has published articles on the development of Shiʿism, sixteenth-century Quranic exegesis, freedom of conscience in classical Islamic theology, human nature in Islamic theology, and primordial human existence in the Qurʾān. His book, Shiʿi Doctrine, Muʿtazili Theology, was published by Edinburgh University Press in 2017.
1. Unless otherwise noted, Arberry’s translation of the Quran is used throughout the article.
5. See Gilliot, “Creation” 45, and Donner 60. Donner argues that the Quranic text crystallized before the First Civil War, without specifying an exact date. See also Burton 186, who proposes that the Quran was collected during the Prophet’s lifetime, while still arguing that other material was created later based on the Quranic text. For the skeptical position see Rippin, “Literary Analysis” 161, and Wansbrough 45, who put the date at the end of the eighth century.
6. See the summary based on extant documents, the rise of isnāds, and the appearance of a distinct religious identity given in Donner 120. Harald Motzki analyzed the traditions and reports related to early Meccan fiqh and concluded that the isnāds can be trusted to inform us about the first Islamic century; Motzki, Origins 287. That the isnāds appeared during the time of the Companions is the view of Aʿzami 246, and Brown, Hadith 18, among others. Juynboll 19 argues that the systematization of isnāds was done at the hands of Zuhrī (d. 742), whereas Schacht 37 puts this process in the first half of the second century of Islam.
7. See Crone 203, who considers much of the information about early Islamic times to be of exegetical origin, which makes it unreliable in her judgment. Burton, Introduction 181, takes ḥadīth to have sprung from ancient exegesis. Much of the material ascribed to the first generation of Muslims is also exegetical; see, for example, Aʿzami 28.
10. See Günther 175, who argues that the longer reports preserved in canonical ḥadīth collections are in fact the result of modifications to shorter, older ones; thus the possibility of a “basic meaning” or “authentic kernel” in these reports cannot be ruled out. Leder 304 considers it obvious that the gradual elaboration of a khabar is a product of the transmission of narrative material. See also Speight 270, who studies certain traditions based on form criticism, concluding that the simpler traditions gave rise to the more complex ones.
12. The Quran’s lack of clarity on the question gave rise to disagreement among scholars; see, for example, Gril, who argues that the Quran does not admit of any miracle by Muhammad, whereas Thomas 203 states that the Quranic position is unclear; see also Khūʾī 103, who argues elaborately in support of the traditional Muslim view that the Prophet did perform miracles documented by the Quran and traditions.
13. See, for example, Iṣfahānī, Bayhaqī,, and ʿIyāḍ (d. 1149); see also the sections named after the content of such reports in the chapter on prophethood in ʿAbd al-Razzāq and Dārimī, the chapter on the Prophet’s miracles in the Ṣaḥīḥs of Muslim and Ibn Ḥibbān, and Ibn Isḥāq’s texts in Suhaylī 2:160 and 3:77.
14. ʿAbd al-Jabbār, Mughnī 15:215; Abū Yaʿlā al-Farrāʾ, Muʿtamad 156; Baqillānī, Tamhīd 132; Juwaynī 353; Murtaḍā, Dhakhīra 404. See how this miracle is used by Ibn Qiba al-Rāzī (d. before 931) in his refutation of Zaydism in Ṣadūq, Kamāl 78. The miracle is also discussed at length in an early work whose ascription to ʿAbd al-Jabbār is nevertheless dubious; [Pseudo-]ʿAbd al-Jabbār 1:55. On the problems of ascription and related issues, see Ansari.
16. Mujāhid, Tafsīr 633; Muqātil 3:96. The tafsīrs ascribed to Mujāhid and Muqātil were compiled after their authors’ deaths, but still as early as the second century; see Gilliot, “Exegesis,” and, more specifically, Leemhuis 19–27 (for Mujāhid) and Sinai 117 (for Muqātil).
17. Ḥudhayfa’s khabar is not part of the corpus of reports narrating the splitting of the moon as an event; rather, it is used to support a certain interpretation of the khabar on Muhammad’s splitting of the moon. Therefore, it cannot be counted among the many reports that describe the event directly and thus constitutes not a report but a separate khabar.
18. Ṭabrisī 9:310; Ṭūsī, Tibyān 9:443. Rubin 58 takes him to be Abū Zayd al-Balkhī (d. 934). Nevertheless, Ṭūsī (d. 1068) usually refers to Abū al-Qāsim al-Balkhī whenever he uses Balkhī unqualified, because (1) the latter also authored a work of Quranic exegesis (el Omari 104); (2) Ṭūsī frequently compares Balkhī’s views to those of the Baṣran Muʿtazilīs; and (3) al-Sharīf al-Raḍī (d. 1015) and Ṭabrisī (d. 1153) both ascribe to Abū al-Qāsim al-Balkhī what Ṭūsī ascribes to Balkhī (see Raḍī 202, 227, and Ṭūsī, Tibyān 2:543, 559; also see Ṭūsī, Tibyān 4:182, and Ṭabrisī 4:93).
19. It seems that Qurṭubī 17:126 is the only source to ascribe this view to al-Māwardī (d. 1058) and al-Qushayrī (d. 1074). But both authors support the third view; Māwardī, Nukat 5:409; Qushayrī 3:493.
21. This position agrees with Rippin, “Construction” 186, who argues that distinguishing between the categories of the literal and the imaginative or “the organic and the fictive” ignores the dexterity of exegetes in using interpretative tools to relate their exegesis to the Quranic text.
22. ʿAbd b. Ḥumayd 356; ʿAbd al-Razzāq, Tafsīr 3:257; with almost the same isnād in Nuʿaym b. Ḥammād 367; Abū Yaʿlā al-Mawṣilī 5:463; Bukhārī, Ṣaḥīḥ 4:186; al-Ḥākim al-Naysābūrī 2:472; Ibn Ḥanbal 3:165, 220; Muqātil 3:296; Muslim 8:133; Nasāʾī 6:476; Ṭabarī 27:111, 116; Tirmidhī 5:72.
23. Bukhārī, Ṣaḥīḥ 4:186, 243, 6:52; Ibn Ḥanbal 1:377 (with an addition, ḥattā naẓarū ilayh), 456; Ibn Ḥibbān 14:421; Muslim 8:132; Nasāʾī 6:476 (without Ibn Masʿūd reporting having been there nor that it was in Minā); Nuʿaym b. Ḥammād, 367; Ṭabarī 27:112; Ṭayālisī 257; Tirmidhī 5:71.
24. The first occurrence of this addition seems to be in Iṣfahānī 1:281, on the authority of Ibn ʿAbbās, with an isnād that is considered weak by Ibn Ḥajar (d. 1449) 7:139. The same isnād appears in Ibn Kathīr, Sīra 2:116. The addition appears, also on the authority of Ibn ʿAbbās, in Ḥalabī 1:491; Ibn al-Jawzī 7:242; Qurṭubī 17:127; and Ṭabrisī 9:310.
29. Compare with Ibn Ḥajar’s remarks 7:139. Some versions mention that the moon was split “into two pieces twice” (shiqqatān marratān); ʿAbd al-Razzāq, Tafsīr 3:257; Bayhaqī, Dalāʾil; al-Ḥākim al-Naysābūrī, 2:471. This, however, can be taken as an attempt to combine different versions that each contained only one of the two words shiqqatān and marratān.
31. Qummī 2:341. It must be noted here that Ibn Kathīr (d. 1373) mentions that the moon was split at the Prophet’s gesture; Ibn Kathīr, Tārīkh 3:159; Ibn Kathīr, Sīra 2:121. But he specifies neither his isnād nor the source of this addition. I could not find it in early ḥadīth works or in the Dalāʾil, either Iṣfahānī’s or Bayhaqī’s. If it were to be found with an isnād or in early sources, the addition could be considered a graphic addition in composite narratives.
34. On this principle in classical theology, common to both Ashʿarīs and Muʿtazilīs, see Abd al-Jabbār 15:405; Abū Yaʿlā al-Farrāʾ, ʿUdda 3:852; Bāqillānī, Tamhīd 382; ʿJuwaynī 420; Murtaḍā, Dhakhīra 472.
36. [Pseudo-]ʿAbd al-Jabbār 1:55; Ibn Qutayba 70, also concerning Ibn Masʿūd. Caution is needed regarding the phrasing of Naẓẓām’s view as presented by Ibn Qutayba (d. 889), because it is intended to defame Naẓẓām. Ṭūsī, Rasāʾil 122, relates that Naẓẓām considered the splitting of the moon impossible, without specifying the reasons for the latter’s judgment.
41. See recently Kharsān passim. for elaborate traditional responses to objections.
42. [Pseudo-]ʿAbd al-Jabbār 1:56; Marzūqī, 85. The latter’s Muʿtazilī leanings can be inferred from his discussion of the knowledge of God, specifically his assertion that it is based on the obligation to express gratitude; see Marzūqī 86.
43. ʿAbd al-Jabbār 16:152; see ʿAbd al-Qāhir al-Baghdādī 162, since ʿAbd al-Jabbār’s arguments apply to ʿAbd al-Qāhir’s text. If ʿAbd al-Qāhir’s work was written after al-Mughnī, this indicates that Muʿtazilīs had already become infamous for denying this miracle.
44. See, for example, Nuʿaym b. Ḥammād (d. 843) and the chapters on malāḥim, fitan, and ashrāṭ al-sāʿa in ʿAbd al-Razzāq’s and Ibn Abī Shayba’s Muṣannafs; Bukhārī’s and Muslim’s Ṣaḥīḥs; and Ibn Māja’s, Abū Dāwūd’s, and Tirmidhī’s Sunans. See also Cook 312, who argues that Islam is in general an apocalyptic religion and proposes that the apocalyptic material motivated early Muslims to conquests.
46. Compare the following entries, all related to āya, in the Encyclopaedia of the Qurʾān: Abrahamov; Neuwirth, “Verses” and “Form and Structure of the Quran.”
47. See Farāhīdī 8:441, where “miracle” is not mentioned among the meanings of āya and the entry is confined to the meanings of a “Quranic verse” and a “sign.” ʿAskarī 368, mentions only that an āya is a “fixed sign.” Neither Ibn Fāris 1:167 nor Jawharī 6:2275 mentions “miracle” among the meanings. Ibn Manẓūr 14:62, and Zabīdī 19:181, note that the āyāt of God are His wonders (ʿajāʾib), without specifying their authority for this meaning.
49. An almost verbatim version is found in Ṭabarānī, al-Muʿjam al-kabīr 11:200. The isnād, however, is different: [Aḥmad b. ʿAmr al-Bazzār ← Muḥammad b. Yaḥyā al-Qaṭīʿī (maybe al-Qaṭaʿī; see Mizzī, Tahdhīb 26:608) ← Muḥammad b. Bakr ← Ibn Jurayj ← ʿAmr b. Dīnār ← ʿIkrima ← Ibn ʿAbbās]. See also Ṭabarānī, al-Muʿjam al-awsaṭ 8:175, where the isnād is [Mūsā b. Zakariyyā ← Muḥammad b. Yaḥyā al-Qaṭīʿī ← Muḥammad b. Bakr al-Barsānī ← Ibn Jurayj ← ʿAmr b. Dīnār ← ʿIkrima ← Ibn ʿAbbās], but the body text has “the sun was eclipsed” (kusifat al-shams).
50. See the eclipse website of the National Aeronautics and Space Administration (NASA) at http://eclipse.gsfc.nasa.gov/LEcat5/LE0601-0700.html.