Modern Textual Criticism of the Koran


Literary Analysis of Koran, Tafsir, and Sira: The Methodologies of John Wansbrough

Andrew Rippin

THAT JUDAISM AND CHRISTIANITY ARE religions “in history” seems to be a commonly accepted notion among many people today. The view is that history is the ‘“proving ground” of these religions, that the intervention of God in the historical sequence of events is the most significant truth attested by these religions. Whether or not this is theologically valid is a question that must he left for those who pursue such questions; what is of interest here are the implications which this view has had for “secular” historical studies and, most importantly here, for the historical study of religion. The idea that these are religions “in history” has led to an emphasis on the desire to discover “what really happened,” ultimately, because of the underlying belief that this discovery would demonstrate the ultimate truth or falsity of the individual religion. Now that may or may not be an appropriate task depending on the particular view of history taken by the historian, but it has led to one important problem in the study of religion—the supposition that the sources available to us to describe the historical foundations of a given religion, most specifically the scriptures, contain within them discernible historical data which can be used to provide positive historical results. In other words, the approach assumes that the motivations of the writers of such sources were the same as the motivations of present-day historians, namely, to record “what really happened.”

Originally published in Approaches to Islam in Religious Studies, edited by Richard C. Martin (Tucson: University of Arizona Press, 1985), pp. 151—63, 227—32. Copyright © 1985 by The Arizona Board of Regents. Reprinted by permission of the University of Arizona Press.

Whether out of theological conviction or merely unconsciously modern scholarship has approached Islam in the same way that it has traditionally treated Judaism and Christianity—as a religion of history, that is as a religion that has a stake in history. Whether this approach is valid or invalid is not the point here. What is relevant is that this view has led to the same sort of attitude toward the sources available in the study of early Islam as that which characterizes the attitude in the study of Judaism and Christianity namely that these sources purport to record (and thus provide us with) an account of “what real happened.” The desire to know what happened in the past is certainly not unreasonable nor is it theoretically an impossible task; Islam most definitely has a history that needs to be recovered. But the desire to achieve positive results must not lead us to ignore the literary qualities of the sources available to us.

Very little material of “neutral” testimonial quality is available for the study of early Islam; sizable quantities of archeological data, numismatic evidence, even datable documents are all very much wanting. Evidence from sources external to the community itself are not plentiful either and the reconstruction of such material into a historical framework is fraught with difficulties. In Hagarism, Patricia Crone and Michael Cook have attempted such a reconstruction and, although they successfully draw attention to the problems involved in the study of Islam, they have not been able to get beyond the limitations inherent in the sources, for they are all of questionable historical authenticity and, more importantly, all are treatises based in polemic. No one has yet expressed the problem better than John Wansbrough. “[C]an a vocabulary of motives be freely extrapolated from a discrete collection of literary stereotypes composed by alien and mostly hostile observers and thereupon employed to describe, even interpret, not merely the overt behavior but also the intellectual and spiritual development of helpless and almost innocent actors?”1 The other sources available to us—the Arabic texts internal to the Muslim community—consist of a limited mass of literature originating at least two centuries after the fact. Such information as this literature contains was written in light of the passage of those two centuries and would indeed seem to have a stake in that very history being recounted. These internal sources intended, after all, to document the basis of faith, the validity of the sacred book, and its evidence of God’s plan for humankind. These sources recorded “Salvation History.”

One brief example may help to clarify the exact dimensions of the problem. Nowhere has the attitude toward the historical character of the sources about the foundation of Islam proved to be more resilient than in the interpretation of the Koran. Muslim exegetes have a category of information available to them called asbab al-nuzul, commonly translated as the “occasions of revelation,” which have been thought by Western students of the Koran to record the historical events concerning the revelation of individual verses of the Koran. Careful analysis of the individual uses of these asba015b in exegesis reveals that their actual significance in individual cases of trying to understand the Koran is limited: the anecdotes are adduced and thus recorded and transmitted in order to provide a narrative situation in which an interpretation of the Koran can be embodied. The material has been recorded within exegesis not for its historical value but for its exegetical value. Yet such basic literary facts about the material are frequently ignored within the study of Islam in the desire to find positive historical results. A good example of what I mean is found in a recent article on Muhammad’s boycott of Mecca; a sabab (occasion) recorded in al-Tabari’s tafsir2 is used to defend and elaborate upon a complicated historical reconstruction about the life of Muhammad.3 The desire for historical results has led to an entire glossing of the problems and limitations of the sources.


John Wansbrough of the School of Oriental and African Studies at the University of London has made a systematic attempt to get beyond the problems involved in trying to understand the beginnings of Islam. In two recent books Wansbrough argues for a critical assessment of the value of the sources from a literary point of view, in order to escape the inherent theological view of the history in the account of Islamic origins. The two works, Quranic Studies: Sources and Methods of Scriptural Interpretation (hereafter QS)4 and The Sectarian Milieu: Content and Composition of Islamic Salvation History (hereafter SM),5 fit together quite logically, although it should be noted that there is some progression of thought between the two works on some specific topics. QS was written between 1968 and 1971 although it was published in 1977; SM was written between 1973 and 1977, but published in 1978. Those following Wansbrough’s numerous reviews will appreciate that his thought has not stopped there either.6Wansbrough emphasizes that the ideas he has put forth in his books are a tentative working out of the problems involved. QS deals primarily with the formation of the Koran along with the witness of exegetical writings (tafsir) to that formation; SM develops the theme of the evolution of Islam further through the traditional biographies of Muhammad (sira and magbazi), and then works through the process of the theological elaboration of Islam as a religious community, examining the questions of authority, identity, and epistemology.

The basic methodological point of Wansbrough’s works is to ask the prime question not usually posed in the study of Islam: What is the evidence? Do we have witnesses to the Muslim accounts of the formation of their own community in any early disinterested sources? The Koran (in the form collected “between two covers” as we know it today) is a good example: What evidence is there for the historical accuracy of the traditional accounts of the compilation of that book shortly after the death of Muhammad? The earliest non-Islamic source testifying to the existence of the Koran appears to stem from the second/eighth century.7 Indeed, early Islamic sources, at least those which do not seem to have as their prime purpose the defense of the integrity of the canon,8 would seem to witness that the text of the Koran may not have been totally fixed until the early part of the third/ninth century.9 Manuscript evidence does not allow for substantially earlier dating either.10

A question for many people still remains however (and the answer to it evidences Wansbrough’s most basic and radical point): Why should we not trust the Muslim sources? Wansbrough’s answer to this is substantially different from other expressions of similar skepticism, for example as argued by John Burton in The Collection of the Koran, where internal contradiction within the Muslim sources is emphasized and then that fact is combined with a postulated explanation of how such contradiction came about.11 No, Wansbrough’s point of departure is more radical: the entire corpus of early Islamic documentation must be viewed as “Salvation History.” What the Koran is trying to evidence, what tafsir sira, and theological writings are trying to explicate, is how the sequence of worldly events centered on the time of Muhammad was directed by God. All the components of Islamic salvation history are meant to witness the same point of faith, namely, an understanding of history that sees God’s role in directing the affairs of humankind. And the difference that makes is substantial. To quote from a work that deals with the same problem but from the biblical perspective, “Salvation history is not an historical account of saving events open to the study of the historian. Salvation history did not happen; it is a literary form which has its own historical context.”12 Salvation history comes down to us in a literary form and must be approached by means appropriate to such: literary analysis.

At the outset it may be appropriate to pay attention to the use of the term “salvation history” in connection with Islam, especially in light of questions raised about its use outside the Christian context in general. Most recently H. W. F. Saggs has pointed to the fact that although the term “salvation” has a clear meaning in Christian thought—“it is the saving of the individual soul from destruction or damnation by sin for eternal life”—its application within Judaism would seem to mean “no more than that God maintained a particular religio-ethnic group in existence when the operation of normal political and social factors might have been expected to result in its extermination.”13 The term “salvation” is ambiguous at best and perhaps only rightfully applied in the Christian case. So that may lead to the question: Are we straitjacketing Islam into a Christian framework by using such a term as “salvation history?” Wansbrough has attempted to make reasonably clear what he means by the term and what it implies to him (e.g., SM ix, 31), making it evident that “salvation ” as such is not the defining characteristic of this history. Indeed Wansbrough suggests (SM 147) that Islamic salvation history may perhaps be more accurately described as “election history” because of the very absence within its early formulation of an eschatological concern. Clearly Wansbrough does not conceive of “salvation” in the term “salvation history” as necessarily laden with its Christian connotations. But further “salvation history” may be taken on a different level simply as a technical term referring to literature involved in documenting what could just as easily be called “sacred history,” that is, the “history” of man’s relationship with God and vice versa. The intellectual baggage of Heilsgeschichte may simply be left behind in favor of reference to the literary genre.

Literary analysis of salvation history has been fully developed within biblical and Mishnaic studies; the works of Bultmann and Neusner are obvious prime examples.14 All such works start from the proposition that the literary records of salvation history, although presenting themselves as being contemporary with the events they describe, actually belong to a period well after such events, which suggests that they have been written according to later points of view in order to fit the purposes of that later time. The actual “history” in the sense of “what really happened” has become totally subsumed within later interpretation and is virtually, if not totally, inextricable from it. The question of whether or not there is an underlying “grain of historical truth” may be thought to be of some concern here, namely, whether or not there must have been some sort of historical event or impetus out of which traditions grew and which therefore forms the kernel of the narrative. But the real problem here is that even if one admits the existence of such a “kernel” of history, is it ever possible to identify and extract that information? Wansbrough implies in his work that he feels that it is not, at least for the most part.15 The records we have are the existential records of the thought and faith of later generations.

This basic insight into the nature of the sources is not totally new to the study of Islam. Goldziher, and Schacht even more so, understood that traditional sayings attributed to Muhammad and used to support a given legal or doctrinal position within Islam actually derived from a much later period, from times when these legal or doctrinal positions were searching for support with the body of material called the sunna. It has become characteristic of Islamic studies after Joseph Schacht, however, either to water down or to ignore totally the implications of such insights. This was clear to Schacht himself toward the end of his life.

One thing disturbs me, however. That is the danger that the results achieved by the Islamic scholars, at a great effort, in the present generation, instead of being developed and being made the starting point for new scholarly progress might, by a kind of intellectual laziness, be gradually whittled down and deprived of their real significance, or even be turned inside out by those who themselves had taken no part in achieving them. This has happened in the past to the work of Goldziher... and it has happened again, recently, with regard to the conclusions of the history of Islamic law achieved by critical scholarship.16

The “intellectual laziness” is, it seems to me, a counterpart to the desire to produce positive historical results—to satisfy that internal yearning to assert “what really happened.”17 The works of three people can be cited as the most obvious examples of the latter trend: in “The materials used by Ibn Ishāq,” W. M Watt distorts the work of Schacht and attempts to use Schacht’s results against the latter’s own position,18 and in the works of Sezgin19 and Abbott20 elaborate schemes are set forth to contradict the insights of Schacht, but are based on no tangible evidence. Sezgin especially displays an overt tendency to date works to the earliest possible historical period, with no apparent justification thereof, for example, in the case of the works of Ibn ‘Abbas.21

Wansbrough’s argument, however, is that we do not know and probably never can know what really happened; all we can know is what later people believed happened, as has been recorded in salvation history. Literary analysis of such sources will reveal to us the components with which those people worked in order to produce their accounts and define exactly what it was that they were arguing, but literary analysis will not tell us what happened (although the possibility of historical implications of such studies cannot be, and certainly are not within Wansbrough’s work, totally ignored).

The point of Islamic salvation history as it has come down to us today, Wansbrough argues, is more specific than merely to evidence the belief in the reality of the theophany; it is to formulate, by adopting and adapting from a well-established pool of Judeo-Christian religious themes, a specifically Arabian religious identity, the inception of which could be placed in seventh-century Arabia. At the beginning of QS Wansbrough brings forth a multitude of evidences from the Koran which point to the idea that the very notions in that book demand that they be put within the total Judeo-Christian context, for example, the prophetic line ending in the Seal of the Prophets, the sequence of scriptures, the notion of the destroyed communities, and the common narrative motifs. This notion of extrapolation is, in a sense, the methodological presupposition that Wansbrough sets out to prove within his books by posing the question: If we assume this, does the data fit? At the same time he asks: What additional evidence appears in the process of the analysis to corroborate the presupposition and to define it more clearly?22 This kind of approach to the material is similar to that of Harry A. Wolfson in his use of the scientific method of “conjecture and verification.” 23 So the question raised by some critics concerning whether it is accurate to view Islam as an extension of the Judeo-Christian tradition cannot be considered valid until the evidence and the conclusions put forth in Wansbrough’s works have been weighed. The point must always be: Is the presupposition supported by the analysis of the data? To attack the presupposition as invalid is to miss the entire point. To evaluate the work one must participate within its methodological presuppositions and evaluate the final results.


Charles Adams has summed up the common feeling of many students of the Koran in the following words: “Such matters as the formation of the Koran text, the chronology of the materials assembled in the text, the history of the text, variant readings, the relation of the Koran to prior literature, and a host of other issues of this kind have been investigated thoroughly.” 24 Wansbrough, however, has made it clear that we have really only scratched the surface of these studies. All previous studies, he states, have involved an acquiescence to the normative data of the tradition and are characterized by “a distinctly positivist method: serious concern to discover and to describe the state of affairs after the appearance of Islam among the Arabs...” (SM 2). What Wansbrough has done has been to bring to the study of Islam and the Koran the same healthy skepticism developed within modern biblical studies (and modern historical studies in general) in order to supplant such positivism. At this point it is worth noting that the highly praised work of Richard Bell,25 although supposedly using the biblical methodology consequent on the Documentary Hypothesis, has, in fact, progressed not one iota beyond implicit notions in the traditional accounts of the revelation and the collection of the Koran; he took the ideas of serial revelation and the collection after the death of Muhammad (the common notions accepted by most Western students of the Koran) and applied them literally to the text of the Koran. However, the primary purpose of employing modern biblical methodologies must be to free oneself from age-old presuppositions and to apply new ones. This Bell did not do; in fact, he worked wholly within the presuppositions of the Islamic tradition. Wansbrough’s claim that “as a document susceptible of analysis by the instruments and techniques of Biblical criticism [the Koran] is virtually unknown” (QS ix) can certainly not be questioned least of all by adducing the work of Bell.26

Wansbrough has also tried to show a way to free the study of the Koran from the uniquely fundamentalist27 trend of the vast majority of modern treatments of the book in which the idea of an “original” meaning or intention is pursued relentlessly but ultimately meaninglessly. Such a position in scholarship has been reached especially because of two factors inherent in previous methodologies in the study of Islam. One, the basic historico-philological approach to Islam has become trapped by the consequence of narrow specialization on the part of its proponents. For the most part, there are few scholars active today who can move with equal agility throughout the entire Western religious framework and its necessary languages.28 Scholars have come to feel that competent knowledge of Arabic and of seventh-century Arabia are sufficient in and of themselves to understand the rise of Islam.29 No different are the views of such people as Serjeant, who attempt to champion the notion of the influence of pre-Islamic southern Arabia on Islam, but do so to the virtual exclusion of the Jewish element in the population there.30

The philological method has been affected also by a second method which within itself has produced the stagnation of Islamic studies within its fundamentalist framework. The irenic approach, which according to Charles Adams aims toward “the greater appreciation of Islamic religiousness and the fostering of a new attitude toward it,”31 has led to the unfortunate result of a reluctance on the part of many scholars to follow all the way through with their insights and results. The basic problem of an approach to Islam that is concerned “to understand the faith of other men”32 is confronted when that approach tries to come to grips with the historical dimensions of a faith that conceives itself as having a stake in that very history.33 The irenic approach to Islam, it would seem, in order to remain true to the “faith of other men,” is doomed most of all to avoid asking the basic question: How do we know?

Wansbrough’s analysis of the basic character of the Koran reveals his assessment of the extent of the problem involved in the use of these two approaches to the Koran. Wansbrough isolates four major motifs of the Koranic message all from the “traditional stock of monotheistic imagery” (QS 1): divine retribution, sign, exile, and covenant. These motifs, Wansbrough notes, are “repeatedly signalled but seldom developed” (QS 1), a fact which leads him to emphasize throughout his works one of his major insights concerning the Koran: its “referential” style.34 The audience of the Koran is presumed able to fill in the missing details of the narrative, much as is true of a work such as the Talmud, where knowledge of the appropriate biblical citation is assumed or supplied by only a few words. Only later when “Islam” as an entity with a fixed and stable identity (based on a political structure) comes into being after the Arabs’ expansion out of their original home, does the Koranic material become detached from its original intellectual environment and need written explication—explication that is provided in tafsir and sira.

Two of the examples discussed by Wansbrough will clarify his notion of the referential character of the Koran. Most evident is the example of Joseph and the mention of the “other brother” in S. 12:59 (see QS 134, SM 24-25), parallel to the biblical account in Genesis 42:3-13; knowledge of this latter story is assumed on the part of the Koranic audience, for within the Koran no previous mention has been made of Benjamin and his being left at home due to Jacob’s fears for his safety. Joseph’s statement in the Koran, “Bring me a brother of yours from your father,” comes out of nowhere within the context of the Koran, although not if one comes first with a knowledge of the biblical story. The second example is one which deals with Abraham’s willingness to sacrifice his son and the removal within the Koran of the dramatic impact contained in the biblical story, where the son does not know that he is the one to be offered (SM 24). The question is far more complex because the Jewish exegetical tradition may play a role here; the study of Geza Vermes35 makes it clear that many Jewish (and Christian) traditions adjust the story to let Isaac know he is to be sacrificed well before the actual event in order to emphasize the willingness of Isaac to offer himself. The Jewish exegetical tradition is referential as well; it already assumes that the basic story of the sacrifice is clear to its audience and that the significance of Abraham in the story will be evident to all who read the Bible; thus the emphasis is on Isaac but certainly not to the exclusion of the role of Abraham. The position of the Koran is similar. The knowledge of the biblical story is assumed; reference is made to developed traditions concerning the sacrifice. The referential character of the Koran should make clear the insufficiency of an approach to the Koran which looks at the so-called “exclusively Arabian” (whatever that may be!) character of the book and tries to ignore the total Judeo-Christian background.36

The notion of the referential style of the Koran also leads Wansbrough to the supposition that what we are dealing with in Islam is a sectarian movement fully within the Judeo-Christian “Sectarian Milieu” (QS 20, also SM 45). The parallels between Koranic and Qumranic literature, while not necessarily displaying an interdependency, do demonstrate a similar process of biblical-textual elaboration and adaptation to sectarian purposes.37 The inner workings of the sectarian milieu are to be seen in both literary traditions.

The Koran as a document, according to Wansbrough, then, is composed of such referential passages developed within the framework of Judeo-Christian sectarian polemics, put together by means of literary convention (for example the use of qul [“say”] (QS 12 ff., also 47—48), narrative conventions (QS 18 ff.), and the conjunction of parallel versions of stories called by Wansbrough “variant traditions” (QS 20 ff.), which were perhaps produced from a single original tradition by means of variation through oral transmission within the context of liturgical usage (QS 17). Here, clearly enough, a variety of individual methods (e.g., form analysis, oral formulaic analysis) which have been worked out in fields outside Islam, primarily the Bible, are used by Wansbrough in his analysis of the nature of the Muslim scripture.

Significantly, Wansbrough’s analysis reveals that the Koran is not merely “a calque of earlier fixed forms” (QS 33), that is, it does not merely seek to reproduce the Bible in Arabic, adapted to Arabia. For one thing, the Koran does not follow the fulfillment motif set as a precedent by the New Testament and its use of the Hebrew Bible. Rather, and, indeed, because of the situation of polemic from which the Koran derives, there is a clear attempt made to separate the Koran from the Mosaic revelation through such means as the mode of the revelation and the emphasis on the Arabian language of the Koran.

Canonization and stabilization of the text of the Koran go hand in hand with the formation of the community, according to Wansbrough (QS 51). A final fixed text of the scripture was not required, nor was it totally feasible, before political power was firmly controlled; thus the end of the second/eighth century becomes a likely historical moment for the gathering together of oral tradition and liturgical elements leading to the emergence of the fixed canon of scripture and the emergence of the actual concept “Islam.”38 This time period, Wansbrough several times points out, coincides with the recorded rise of literary Arabic.39 Further evidence for this position is derived in QS from a “typological analysis” of tafsir (see QS 44 and ch. 4). The basic inspiration and thrust of Wansbrough’s approach may once again perhaps be traced to modern biblical studies; such people as Geza Vermes and Raphael Loewe have been drawing attention to the need to stop plundering exegetical works, Greek and Aramaic translations of the Bible, and so forth, in order to find support, somewhere, for one’s own argument. Rather, they suggest, such works must be studied as a whole with attention given to the historical context of their writing and to their literary context.40 The analysis of Koranic tafsir literature into five genres—haggadic, halakhic, masoretic, rhetorical, and allegorical—once again sets the basic insight. The genres display an approximate chronological development in the above sequence and display a historically growing concern with the textual integrity of scripture and then with the community function of scripture.41

The sira, while partly exegetical as Wansbrough explains in QS, has a much greater role in Islam: it is the narrative witness to the Islamic version of salvation history. Most significant here is Wansbrough’s analysis (SM ch. 1) of much of the contents of the sirainto elaborations of twenty-three polemical motifs traditional to the Near Eastern sectarian milieu—items such as the prognosis of Muhammad in Jewish scripture, the Jewish rejection of that prognosis, the role of Abraham and Jesus in sectarian soteriology, and resurrection. All these themes are elaborated within a narrative framework set in seventh-century Arabia but yet all are themes that had been argued so many times before among sectarian groups in the Near East. The analysis of the sira underlines Wansbrough’s main contention in both of his books that “by its own express testimony, the Islamic kerygma [is] an articulation... of the biblical dispensation, and can only thus be assessed” (SM 45).

It will be clear to any attentive reader of QS and SM that Wansbrough’s work still leaves much to be done with the basic data in order to work out fully the implications of his kind of study. Close and detailed analyses of the many texts involved are still needed in order to demonstrate and, indeed, to assess the validity of his approach. What Wansbrough has accomplished, it seems to me, is to point to a new direction that Islamic studies could take in order to revitalize itself; Wansbrough has marked a path in broad outlines, but the road must still be cleared.42

Several reviewers have seized upon (and, indeed, Wansbrough himself has emphasized the point)43 various statements in QS with regard to methods determining one’s results. I.J. Boullata, in his review of QS, put the matter this way: “To quote him from page 91 ‘Results are, after all, as much conditioned by method as by material.’ If this is true and his material is given credence in spite of its selectivity, there remains a big question about his method and the extent to which it conditioned his results.”44 Fazlur Rahman in his book Major Themes of the Koran makes a similar point about method.

My disagreements with Wansbrough are so numerous that they are probably best understood only by reading both this book and his. (I do, however, concur with at least one of his points: “The kind of analysis undertaken will in no small measure determine the results!” [p. 21]) I do believe that this kind of study [i.e., the comparison with Judaism and/or Christianity] can be enormously useful, though we have to return to Geiger and Hirschfeld [! not Speyer?!] to see just how useful it can be when done properly.45

What does Rahman mean here concerning method? Does he mean to imply: Well, Wansbrough has a method and that has been his downfall; I have no method so I have imposed nothing upon the material? I doubt that Rahman wants to urge methodological naivete. More likely, perhaps, Rahman means: Wansbrough has his method and I have mine, but mine is right. That the methods which Rahman (and virtually every other student in the field) imposes upon his study happen to be, for the most part, the traditional theologico-historical methods is a fact that needs to be recognized, just as does the fact that Wansbrough imposes literary methods. If the study of Islam is to remain a scholarly endeavor and retain some sense of intellectual integrity, then it must, first, become methodologically aware and, second, be prepared to consider the validity of other methods of approach to the subject. This means that Islamic studies must differentiate between the truth claims of the religion itself and the intellectual claims of various methods, for ultimate “truth” is not susceptible to methodological procedures. To remain within the search for the “true” meaning of Islam and not to be prepared to free oneself from, for example, the priority of history46 within the study of Islam, will surely sound the death-knell for a potentially vital and vibrant endeavor of human intellectual activity.

If you find an error please notify us in the comments. Thank you!