On Pre-Islamic Christian Strophic Poetical Texts in the Koran: A Critical Look at the Work of Gunter Luling

INTRODUCTION

Gunter Luling was born in Varna, Bulgaria, in 1928. The first period of his academic life was spent at the University of Erlangen, Germany, between 1956 and 1961. His original intention was to work as a theologian on the theological themes of the Koran, but owing to the opposition of his professor to this idea, he agreed to work on the theme of the adoption of Greek science by Arab scholarship. Due to continuing bad relations between himself and his supervising professors, he was forced to leave the university in 1961, "without any achievement to show for those years."1

This bad experience, however, was in reality a blessing in disguise, in that it led to him obtaining the post of director of the Goethe Institute at Aleppo, Syria, in the years 1962 to 1965. It was here that he learned what he could not have learned if he had remained in academic life at a German university: a deep acquaintance with the Arabic vernacular. It later turned out that knowing the Arabic vernacular was one of the essential requirements for critically analyzing the text of the Koran. In 1966 Luling seized the opportunity to return to academic life in Germany: first, as assistant professor at the Faculty of Medicine at Erlangen, supervising studies in the history of antique and early medieval Arab medicine, and then transferring at the end of 1967 to the post of assistant professor at the Erlangen Institute of Arabic and Islamic Studies. He was now in a position to freely pursue the dogma-critical approach to the Koran that he had always intended.2

The chief result of Luling's research was that underlying about onethird of the Koran text there are pre-Islamic, antitrinitarian Christian, strophical hymns, dating from the beginning of the sixth century C.E., approximately one hundred years before the time of Muhammad; the postulated background to this discovery being a predominantly Christian central Arabia, including a Christian Mecca, Kaaba, and Quraish. In 1970 Luling submitted a 180-page dissertation on his research as a thesis for a Ph.D. degree. The official reports on the thesis were enthusiastic, stating that: "the proof for strophical ur-texts, which can be reconstructed, can be regarded as established.... The work places the history of the origin of Islam and the history of the dogmatics of early Oriental Christianity into new dimensions," and was consequently awarded the highest possible note of eximum opus. Within a fortnight, however, the tone of the university authorities changed and he was told that he could no longer look forward to the habilitation that an eximum opus entitled him. According to Luling, his ideas had incurred the wrath of the Arabist Anton Spitaler of Munich, the leader of the pure philology school of German Arabic studies. Spitaler refused to give up the conventional method of judging the Koran by the laws of Classical Arabic, and feared that Luling's results would, "certainly turn upside down the hitherto existing notions about the history of the origin, the text and the contents of the Koran." Such was Spitaler's influence that by the end of 1972 Luling was officially dismissed from the University of Erlangen.3

The implementation of this decision, however, could be delayed by legal proceedings against it, and this was the course that Luling took. He was encouraged in this move by the support of the French Koran expert Regis Blachere. After studying Luling's dissertation, Blachere wrote to him: "Highly eminent colleague. In this way (i.e., with Luling's methods) the tradition continues of that German school of scholarship to which we owe so much grateful appreciation. I hope that owing to your generation there will be kept alive a curious speciality of worldwide importance with regard to Islamic studies which alone offers the chance for a global understanding of the problems." Unfortunately, Blachere died a few months after he wrote this letter, but Luling persisted in arguing his case. In 1973 he presented the whole of his collection of restored Christian strophic texts, under the title On the Pre-Islamic Christian Strophe Poetical Texts in the Koran, to the Faculty of Philology at Erlangen as his Habilitationsschrift, essential for the continuation of his academic career. In February 1974 this, too, was voted against by a majority of faculty members, according to Luling, after much politicing on the part of the "pure philology" faction. The decision was disputed in court, but without success, and Luling was finally dismissed from the university in July 1978.4

In 1974, however, he had published his Habilitationsschrift under the title Uber den Ur-Quran, which was well received by many outside the academic establishment in Germany. It is this work, revised and expanded for the English edition under the title On the Pre-Islamic Christian Poetic Texts in the Koran, that forms the basis for the present exposition and critique of Luling's ideas. In addition, Luling has also published a work on the Prophet Muhammad: Die Wiederentdeckung des Propheten Muhammad, The Rediscovery of the Prophet Muhammad, which will be referred to occasionally.

THESES AND METHODOLOGIES

Luling's approach to the text of the Koran is based on four fundamental theses.

Thesis I

About one-third of the present day Koran text, as transmitted by Muslim orthodoxy, contains as a hidden groundlayer an originally pre-Islamic Christian text.

(a) Difficulties in interpreting the old Arabian script.

Every Arabic text today, as for the past fourteen or fifteen centuries, consists of three different layers. First, the basic letter shape, or drawing (rasm). Second, the diacritical points (nuqat), the function of which is to differentiate the otherwise similar letters of the rasm. And third, the strokes for the vowels, (harakat), which are not shown in the basic consonontal text of the rasm. The crucial point here is that in pre-Islamic and early Islamic times the diacritical points were not set at all in everyday writing, because they were not yet known or not commonly accepted. In the Koran text, at least during the first century of its existence, diacritical points were even forbidden, ostensibly out of pious reverence for the holy text. All the older Koran codices that have come down to us display the text without diacritical points and signs, not to speak of the strokes for the vowels, which were introduced even later. This means that finding the originally intended sequence of consonants, and hence the originally intended meaning of the text, becomes something of a lottery, often determined by the preconceptions of the redactor as to what the text ought to mean.5

(b) The difference between strophic poetical structures and prose texts in the Koran as a very important key to the reconstruction of the pre-Islamic Koran hidden in the present-day Koran.

It has always been recognized that in the traditional Koran some suras very clearly contain strophic refrains, although the repetition of these refrains mostly does not follow a perceptible, strict metric measure. This fact has for a long time been taken by some Islamicists as an indication that the Koran originally contained strophic structures, the strict order of which has been destroyed by later redaction of the text. This converted the old strophic or hymnodic speech into prose and changed the wording and meaning of the older material. This apparently occurred when texts were collected and redacted to serve as the holy book of the nascent Islamic community.

According to Luling, uncovering these old strophic texts requires a threefold methodology: (1) philological (grammatical and phraseological), (2) strophe-metrical, and (3) dogma critical; the last two methods having been neglected since the 1930s, this approach represents a fresh start. The special genre of religious hymnody detectable in the groundlayer of the Koran belongs, form historically, to a chain of strophe-poetical traditions reaching from Old Egyptian, Old Testamentarian, preIslamic old Arabian, and Old Jewish models, across the equivalent Byzantine, Coptic, Syriac, Arabic, and Ethiopian hymnody of pre-Islamic and early Islamic times, up to late medieval and modem strophic poetry.

Luling regards it as beyond question that the rich Christian Ethiopian hymnody of the early sixth century goes back in the main to a Christian Coptic original, sometimes word-for-word across hundreds of strophes, the sequence of which is neatly maintained. Frequent misunderstandings of the Coptic original in the Ethiopic translations result from typical misunderstandings of an ambiguous Arabic text minus its diacritical points and vowel strokes. To Luling, this indicates that these voluminous Ethiopic strophic texts of c. 500 c.E., which stem from Coptic sources, must have passed through a central Arabian stage of a likewise voluminous, but now lost, pre-Islamic Christian strophic hymnody. This thesis involves regarding as authentic the strophic poetry of the purportedly preIslamic Imru lul-Qays, and an almost complete recasting of the original Christian hymnody by the earliest Koran text collectors and redactors, who changed poetry into prose and altered its original meaning.

(c) Different languages in the Koran.

The language of the Christian strophic texts that Luling finds underlying the Koran is without diacritical points or vowel strokes, and resembles what he calls "Old Arabian," indicating a tradition of literate education and erudition. This non-Classical, vernacular, though nevertheless literary Christian Arabic had many grammatical equivalents to the language of early Christian literature; a middle-standard, educational, and literate Arabic which served as an intertribal lingua franca, or koine, over the whole Arabian peninsula. Luling regards the grammatical license and rules for medieval strophic poetry, conveyed to us by medieval Arabic philologists, as applying also to the pre-Islamic Christian hymnody in the Koran.

(d) The techniques of the Muslim reinterpretation of the erstwhile Christian ur-Koran hidden in the ground layer of the transmitted orthodox Koran.

(1) The deviating reading of vowels, which would later mean the deviating setting of vowel strokes, but nevertheless maintaining correctly the original groundlayer layer of the Arabic script.

(2) The deviating reading of consonants, which would later mean the deviating setting of diacritical points for ambiguous consonant signs, but nevertheless, out of piety, correctly retaining the rasm groundlayer.

(3) The slight alteration or deformation of the consonontal signs of the groundlayer (rasm), while simultaneously reading alternative consonants and vowels, and eventually changing the diacritical points and vowel strokes.

(4) The omission, addition, or replacement of single characters of the consonontal text of the groundlayer (rasm), of single words, sentences and sections.

(5) Taking advantage of varying or different (vernacular or highstandard) orthography.

(6) Taking advantage of the concurrence of different meanings of one and the same word or word root within Hebrew, Aramaic (Syriac), and vernacular and high-standard Arabic.

(7) Giving a deviating meaning to single words on the basis of vague associations, so that a series of Koranic lexical meanings must be classed as invented.

(8) The disregarding of grammatical and particularly syntactical rules during the initial deviating reading of the erstwhile Christian hymnody.

This last technique, in order to defend the peculiarities of language it produced, led some Arabic grammarians to assert that the content of the Koran is not defined by Arabic grammar, but that Arabic grammar is defined by the Koran. Thus the Koran, in part, not only contradicts the rules of common secular Arabic, but itself, when it says: "We (God) never sent any messenger but with the speech of his people, that he might make (things) clear to them" (XIV.4); and: "This is clear Arabic language" (XVI. 103) In view of these texts, the further explanation of the grammarians that the grammatical peculiarities of the Koran are due to the ecstatic utterances of the Prophet is hardly convincing.

Thesis 2

According to the statements of thesis 1 the transmitted text of the Koran contains four different kinds, or layers, of text.

These four kinds of text are:

First, because historically the oldest, the texts of the pre-Islamic Christian strophic hymnody. These are called "erstwhile Christian texts."

Second, and historically later, the texts of the new Islamic interpretation pressed on the groundlayer of erstwhile Christian texts. These are called "second-sense Koranic texts."

Third, and historically parallel to the second layer, the originally pure Islamic texts; that is to say, texts that are to be attributed to the Prophet Muhammad. These are called "single sense Koranic texts." They make up about two-thirds of the whole Koran, the remaining third being the erstwhile Christian texts transformed into second-sense Koranic texts.

Fourth, and historically post-Muhammad, are the single-sense Koranic texts that have been reinterpreted by Koran editors in such a way that the original ideas of the Prophet have been distorted. These texts are called "texts of the post-Muhammadan Koran editors."

Further analysis of these four different textual layers.

(a) The Islamic scholarly terminology for different layers of Koran text.

(1) al-mutashabihat: "texts that are similar to something else," or "texts that contain different aspects of interpretation." These are the original meanings of the term, but later it was interpreted as referring to "the anthropomorphic expressions of the Koran," probably to divert attention from awkward texts, such as those containing an original Christian hymnody.

(2) al-muhkamat: "texts that are decided, firm, unambiguous." This came to mean the whole of the Koran, but originally referred to the third kind of text listed above, those attributable to the Prophet and without any underlying erstwhile Christian text.

(3) al-mufassal: "texts that are commentary." This term must be a subcategory of al-muhkam since not all muhkam texts are mufassal, but a mufassal text is always muhkam If muhkam denotes the third layer identified above, mufassal must denote the glosses and commentaries inserted into the texts of the first and third layers in order to pin down the eventually preferred meanings of those layers to prevent the original sense from showing through. Such insertions would have been indispensable at the time when diacritical points and vowel strokes were not yet permitted in Koran texts. A clear example of such an inserted commentary is IX.16.

(b) On the characteristically distinct contents of the different kinds of Koran text.

It is in the distinct contents of the different kinds of Koran text that the dogma-critical aspect of Luling's threefold methodology comes into play. He argues that the Islamic intellectual content that was pressed on the erstwhile Christian rasm groundlayer was determined by pagan Arabian conceptions, in fact, by the old Arabian tribal religion of pre-Islamic times, and that it is from this position that these texts are turned against Christian and Jewish dogmatics. Luling regards this classification of these earliest Islamic texts as pagan not as a slight upon them but as a positive evaluation.

Luling sees original Christianity, not that of the later, trinitarian-imperialist Hellenistic church, as the resurfacing of popular pagan/tribal religion from the underground of peasant religion. It was, in fact, the revival of the old Semitic and old Israelite tribal religion at the High Places or Holy Groves, where the cult at the shrines of the primeval Messianic Heroes of the tribes had blossomed. This old religion was persecuted and stamped out by the imperialist Jewish religion of Yahweh, and after a brief revival by Jesus, stamped out again by imperialist Christianity. Luling sees the old religion as having spread very early into tribal society all over Arabia, and even surviving there until the twentieth century.

Contrary to the portrayal of the central motive of the ur-Islamic movement by later Islamic orthodoxy, depicting it as the defeat of central Arabian paganism, Luling sees the ur-Islamic movement led by Muhammad as a movement away from central Arabian Christian, especially trinitarian conceptions, toward the old Semitic pagan/tribal religion, under the slogan: "Return to the religion of Ismail and the tribes." This shift toward tribal religion is detected in an ur-Islamic reinterpretation of the Hellenistic Christian groundlayer of the Koran text, especially in those sections that fight against the "Garden," the pagan-Arabic Grove of Fertility, condemned by Judaism as the cult at the High Places. These Hellenistic Christian antipagan passages, originally contained in the rasm text of the Koran, have each, with the exception of XVIII.40, been reinterpreted by early orthodox Islamic redactional devices, so that the pagan "Garden" thereafter appears as the revered Islamic "Paradise," with its pagan religious erotic pleasures, differing so markedly from the Christian heaven. Another example of this ur-Islamic shift away from Christian conceptions is that texts of the Christian groundlayer referring to salvation in Christ were reworked to apply to the Koran, or supplied with editorially added negative signs, interpreted as applying to foes of the Prophet.

The Islamic single-sense Koran texts are also considerably influenced by these pagan Arab tribal ideas, and can be identified above all in the urIslamic prose texts pressed on the rasm groundlayer of the erstwhile Christian hymnody. Such single-sense texts were often inserted as glosses and commentaries into the reinterpretation of the Christian texts, and often appear as new, textually neat versions or doublettes of lexically and grammatically clumsy texts, resulting from the original reinterpretation of the strophic texts. Such new versions or doublettes are usually placed at some distance from the clumsy originals, which they repeat in a linguistically neat form. This gives rise to the suspicion that these new neat text versions are not only located far from their clumsy originals in terms of textual space, but also in terms of time of composition. That is to say, they have been composed much later than the original clumsy interference with the rasm, which necessitated their eventual composition for sake of clarification.

(c) On the literary form of the different kinds of Koran text.

The Islamic second- and single-sense Koran texts were originally designed and then handed down as prose texts. In the second-sense texts, as a consequence of the disregard or dropping of the internal rhymes of what was once a Christian strophe, a type of verse or sentence has been developed for which the term "Koranic rhymed prose" came into use in Western scholarship; the later, originally single-sense Koran texts, emulate this type of early Koranic rhymed prose. From the time of the Islamic reinterpretation and elimination of the strophic hymnody onward, each Koranic prose verse, or sentence, now rhymes with its last word, though now far remote from the preceding and next-to-last words of its surrounding long prose verses or sentences; these are now long and monotonous because of the annulment of the internal rhymes. In the latest of the single-sense texts, emulating the oldest so-called Islamic rhymed prose, the sentences became so long that, although each of them finishes with a remote rhyme word, one can hardly speak any longer of a rhyming effect. This underlines the fact that the so-called rhymed prose of the Koran is not a genuine form of poetic art, but a peculiar technique developed from the leveling out of the sophisticated Christian strophic poetry, and the necessity of disguising the strophical structure of the Christian hymnody.

As a result of his research, Luling holds that the equation of the terms "rhymed prose" and saj`, which has been made here and there in Western scholarship, must be abandoned. The saj` is a longer, emphatically pronounced sentence, usually uttered by a pagan soothsayer, consisting of very short syntactic parts, the endings of which are not arranged to rhyme according to a strophe-metrical order, although the saj` shows alliterations and unsystematically scattered rhymes.

In the Koran the saj` comes across only as an Islamic single-sense text, often placed before or after an erstwhile Christian strophic groundlayer, on which a new Islamic interpretation has been pressed. Since Luling attributes the single-sense texts to Muhammad, and they are usually in the form used by soothsayers, he sees this as a confirmation of his thesis that the urIslamic movement led by the Prophet had actually been a movement away from Western, Hellenistic, trinitarian, imperialist Christianity, toward the ur-Semitic tribal religion, symbolized by its progenitors Abraham and Ismail. This use of saj' by the Prophet would explain why his Meccan opponents, who Luling considers to have been trinitarian Christians, considered him to be a Whin or soothsayer. Those sections of the Koran that are hostile to poets, such as XXVI.224-226, are to be attributed to preIslamic Christian invective against the people of the pagan Holy Groves or High Places, which, of course, included the pagan soothsayers.

(d) On the linguistic aspects of the different texts of the Koran.

Lining's linguistic analysis of the different Koran texts yields the result that despite the fact that the canonic Koran is written in what is usually called high-standard or classical Arabic, it in fact contains four different types or standards of Arabic that, insofar as they can be distinguished, follow each other in time, and can be labeled as language strata, or language layers.

(1) The oldest Arabic language layer is that of the erstwhile Christian strophic hymnody. This Arabic, in itself a highly literate vernacular, differs considerably and in several respects from high-stan dard classical Arabic. First of all, the grammatical case endings of nouns and the grammatical modal endings of verbs are either missing, or appear arbitrarily to fill in gaps in the rhythm of the strophe lines, and to supply secondary rhyme endings; all this without regard to the grammatical correctness of such endings. Indeed, it seems very often that the rhyme composer misuses classic grammatical endings intentionally and humorously, as a skillful poetic device. This is a feature of popular and vernacular poetry in many languages, and underlines the literary form of the pre-Islamic strophic poetry. But this vernacular yet literate language not only differs from high-standard classical Arabic in its grammar, but also lexically in its vocabulary. Besides the expected difference in word choice and meaning between vernacular and classical, there are also a lot of Semitic words that have intruded into this, as Luling calls it, pre-Islamic central Arabian Christian Arabic, from neighboring Semitic Christian literatures, especially from contemporary biblical Aramaic, Syriac, and, less frequently, Hebrew. These are words that, although stemming from the same Semitic word root, had in the long-term development of these different Semitic languages assumed slightly different meanings, but sometimes even essentially different ones.

To sum up: the language of the Christian hymnody in the groundlayer of the Koran is grammatically a vernacular Arabic, and moreover, a vernacular language the lexicography of which is thoroughly impregnated by neighboring Semitic Christian literatures. It is therefore absolutely indispensable for research into this ur-Koranic Christian hymnody to use both a grammar of early Christian Arabic,6 and to consult all the available lexica of all the neighboring Semitic Christian literatures.

(2) Second comes the language of the second-sense Koran texts. This language is actually not a language in the true sense of the word, but a selective and unique conception of speech for the occasion, a consequence of the attempt to force a new interpretation on an ambiguous but decisive script. The erratic nature of this speech has been, and is still, being excused unconvincingly, by orthodox Islamic Koran scholarship and by many Western Islamicists, as resulting from the ecstatic state of the Prophet during his experiences of revelation. The explanation, however, of how ecstatic speech was transformed, on the spot, into the text of scripture, is still missing.

(3) The third language is that of the early editorial glosses and commentaries, inserted into or added on to the second-sense texts, in order to pin the new interpretation to the underlying groundlayer of the rasm. This language, with its conspicuous preference for nominal instead of verbal constructions, forms a striking contrast to the literate language of both the underlying Christian hymnody, and old and present-day classical Arabic. This mostly nominal language discloses by its lack of elegance its identity as a nonliterate language; at least as a language without a discernible distinct literary tradition.

(4) The fourth language is that of the late and latest single-sense texts. That is, those originally Islamic texts that are neither immediately nor directly related to the erstwhile Christian rasm groundlayer, but which present larger cohesive complexes of prose texts in classical Arabic. This language, in its phraseology and grammatical sophistication, and in its skilful and therefore elegant appearance, is to be thoroughly distinguished from the language defined in the preceding paragraph.

Based upon information transmitted in Islamic tradition, Luling concludes that this language must be attributed to the Prophet himself or, rather, to the literate secretaries who were charged with recording his pronouncements.

While for languages 1, 2, and 3, it is beyond question that they do not present high-standard classical Arabic and should not originally have been read as such, it still remains an open question whether the fourth and last type of language was originally intended to be so read, including case endings (I'rab). Luling presumes not, but as even this last language has been distorted and reinterpreted by post-Muhammadan Koran editors it is, again, a language especially created for the occasion. This fourth and last language layer is not Luling's major concern, since he is most intent on uncovering the vernacular Christian hymnody of the ground-layer.

Thesis 3

The transmitted Islamic Koran text is the final of several successive editorial revisions.

Three essential motives for the different successive Koran-text editorial processes can be discerned. These processes obviously went on over several decades, and in all probability three successive periods of editorial work on the Koran.

(1) The first and main motive for the editorial reworking of the Koran is dogmatic.

It is the motive for both the Islamic reinterpretation of the erstwhile Christian hymnodies and for the collection and recording of the original Islamic revelation texts. In Luling's view it has its cause in a national-Arabian/pagan-Arabian antagonism toward a central Arabian, and especially Meccan Hellenistic trinitarian Christianity of the Prophet's time. This Christianity had kept good relations with the trinitarian Christianity of all the neighboring nations and, like them, was split into different parties and confessions. Luling considers that this national-Arabian/ tribal-Arabian countermovement against the foreign-dominated Christianity of central Arabia, as not necessarily originating with Muhammad, but that it could have begun even with his grandfather 'Abd al-Muttalib. Likewise, the editorial reworking of the Christian hymnody may have begun before the Prophet's time, being in no way initiated by him, and then added to the text of the Koran after his passing.

This pagan or tribal-Arabian ur-Islamic movement, would have been incited by discord between the different factions and confessions of Christians in Mecca, which had been established there for some two centuries before Muhammad, during which time the Kaaba had been a Christian church. Luling pictures Mecca as containing quarrelsome parties from all the different Christian church denominations of the surrounding countriesMesopotamia, Syria, Palestine, Egypt, and Ethiopia. The Christian groundlayer of the Koran, in the view of Luling, contains an archaic ur-Christian angel-christology, involving the idea that Christ is an angel and therefore a created being. This Judaic or Ebionite Christianity had been condemned as heretical by all the Christian confessions extant at Mecca, who were still controlled by their Hellenistic mother churches. Only the central Arabian heretics (hanif), the few descendants of the earliest ur-Christian and therefore angel-christological communities in centralArabia, continued to stick to their creed. This nontrinitarian, urChristian angel-christology, found by Luling in the strophic hymnody of the ur-Koran, was also eventually abandoned by urIslam in its turn to its own tribal-religious traditions under the slogan: "The religion of Abraham, Ismail and the tribes." This would explain the curiosity that Muhammad was considered both champion and betrayer of hanifdom.

(2) Historically, the second motive for a renewed editorial overworking and enlargement of the Koran comes from the victory of Islam over the Meccan "associators" (mushrikun), whom Luling sees as the Christians of Western trinitarian theology, who deified Jesus as God's son. The effect of this reworking was to tone down the original defection from central Arabian Christianity in order to appease former enemies, and to propagate the idea that there had never been any institutionally established Christian community in central Arabia, least of all in Mecca, the inhabitants of which were thereafter portrayed as pagan idolators.

(3) The third fundamental motive for the editorial reworking of the Koran derives secondarily from the national Arabian/tribal motivation and becomes gradually autonomous. That is, the motive of overworking the Koran text in order to make it read from beginning to end as a high-standard classical Arab prose text. This relatively late linguistic redaction consists, therefore, on the one hand, in the removal of the strophic structure of the original Christian text of the rasm groundlayer, and on the other, in the linguistic reworking of the entire Koran text, including all the second- and single-sense texts, in order to replace the hitherto vernacular reading without case endings, by a high-standard classical Arabic reading with case endings. The time when this redaction took place would be the time when the diacritical consonant points and diacritical vowel strokes had become accepted as permissible in Koran codices.

The motivation for changing the Koran into a classical Arabic prose text, as opposed to a strophic text, was probably to erase the memory of its original Christian contents, in the form of hymnody and responsories; to do this, rhythm and rhyme, the twin pillars of memory, had to be erased. Whether the Prophet himself ever knew and performed the Christian hymnody in its original form is impossible to say for sure, but Luling is inclined to the opinion that he did.

The predilection of rising Islam for the national-Arabian cause brought about a preference for high-standard classical Arabic over the international Arabic vernacular, spoken beyond central Arabia in the surrounding nontribal countries with an urban culture; this "national" high-standard Arabic was pressed on the rasm groundlayer in a gradual process of grammatical and orthographical reworking. While in the earliest phase of developing Islam the lexical and grammatical differences between the Christian Arabic koine and the Old Arabian high-standard language had been used for the reinterpretation of the Koran, it seems that the high-standard Arabic reworking was relatively late, probably long after the death of the Prophet. This exclusively linguistic, as opposed to dogmatic, reworking was not only chronologically the last, but also, in its effect on the meaning of the Koran text, the least significant.

Thesis 4

The presence of the successive layers in the Koran text can be confirmed by material in Muslim tradition that has hitherto gone unnoticed or been misinterpreted.

On the basis of Lining's findings of an ur-Christian and ur-Islamic groundlayer beneath the text of the present-day Koran, the existence of Koranic ideas, thoughts, and expressions in the pre-Islamic Old Arabian poets would cease to be anomalous. It would then appear that the preIslamic poets were indeed influenced by the Koran, but it was the much older Christian ur-Koran contained in the rasm groundlayer of the Islamic Koran. The thesis of Louis Cheikho that the pre-Islamic old Arabian poetry had been largely Christian would then be confirmed, since innu merable pre-Islamic Christian poems would have been reinterpreted using the same editorial techniques as those used on the erstwhile Christian texts of the ur-Koran. Furthermore, on the basis of Lining's uncovering of the history of the formation of the Koran text, Islamic traditions can be read in a new light and exposed to a historicocritical analysis from a new perspective.

EXEMPLIFICATION: UNCOVERING THE HYMNODY

The evidence for the theses outlined above is laid out by Luling in five chapters of over five hundred pages of small print. These contain examples from some seventeen different suras and long footnotes on related topics, together with numerous "excursuses" on related topics. The scholarship is overwhelming, and unless one can retain a sure grasp of the theses being argued for, the average reader is likely to get lost in an endless forest of detail. Because of limitations of space we must confine ourselves to two brief summaries of this material from the first and final chapters.

In the Koran as transmitted by Muslim orthodoxy, the text is divided into numerous sections. These sections are interpreted by Muslim tradition according to an extra-Koranic frame narrative, which assigns them to some purportedly historical incident to give the text a particular meaning. These interpretations are such that they cannot be legitimately derived from the text section alone and in itself on the basis of grammar and lexicography. Indeed, they are often so fantastic that they furnish no parallel, either from non-Koranic or from pre-Koranic Arabic texts.

On the other hand, a word-for-word interpretation of such texts, relying on nothing other than the transmitted rasm text, neglecting even the transmitted pointation, which is in effect the first commentary, yields a cohesive and self-contained text which needs no supplementary elucidation, and certainly no reference to allegedly historical incidents. As a rule, the acribic analysis of the broader rasm text, that is to say, the text sections before and after a dubious section, uncovers one long line of thought, whereas the traditional Islamic interpretation yields several short sections with, as to their content, independent and extremely short lines of thought, usually with totally disconnected ideas.

Luling begins the demonstration of this thesis in chapter 1 with an examination of Suras XCVI and XXX. Sura XCVI is a particularly suitable place to begin, since it is a Sura where the original pre-Islamic Christian sense of the rasm groundlayer can be uncovered without the necessity of changing a single letter of the rasm script. Sura XCVI can also be said to be the pivot upon which the whole edifice of Islam rests, since this text has been taken by Muslim tradition as the first instance of Koranic revelation, the purported historical context of which forms the guarantee of the Koran's transcendental origin. If Sura XCVI turns out to be other than what Muslim tradition has portrayed it as being, that tradition will be shown to be false, indeed, to be a knowing and deliberate falsification, and the Koran to be not a revelation but a man-made text like any other.

According to the traditional Islamic interpretation, Sura XCVI falls into three thematic parts: verses 1-5, 6-8, and 9-19. In the cases of the first and last sections, a frame narrative has been attached to the text that has no basis in the actual words of the texts themselves. In the first and crucial section the frame narrative has the angel Gabriel appearing to Muhammad and appointing him Prophet, first of the Arab nation and then of all mankind, and presenting him with a script and commanding him to read. By this nonKoranic, pseudohistorical narrative, the interpretation of the introductory imperatives: 'igra' "read" (XCVI.1, 3), are pinned down in a definitive fashion. The text immediately following these two imperatives is then taken for the text Gabriel presented to the eyes of the Prophet-to-be to recite.8 Yet the idea that the two commands to read are the only words of Gabriel, and that the rest is the script to be read by Muhammad is not indicated in the text in any way at all; moreover, the actual names "Gabriel" and "Muhammad" are wholly gratuitous and likewise do not appear in the text. The late and spurious nature of the references to Gabriel in Islamic tradition is confirmed by the fact that in the only other places in the Koran where Gabriel is mentioned (VI.97 f.; LVI.4) are so marginal in content as to be classified as glosses added later; also, in Islamic tradition concerning the life of the Prophet, texts that deal with Gabriel display an unproblematic Arabic grammar and lexicography, whereas surrounding passages apparently dealing with facts of history show many such problems.

In a similar fashion, the frame narrative attached to the third and final section of Sura XCVI (verses 9-19), has obviously been concocted to supply an acceptable meaning to an otherwise incomprehensible text. This time it concerns a member of the developing Muslim community, either a slave or even Muhammad himself, who is hindered in the performance of the ritual prayer. Between these two larger sections, 1-5 and 9-19, the short section, 6-8, is of such general significance that it could be attached to either the preceding or following sections, or even to any religious statement whatever. The curious thing is that although this middle section can, because of its general relevance, be understood as a proem to the third section, it is instead joined by Muslim tradition to the preceding section, to which it is much less suited because of its postponed position. It is a strange rule of the Arabic grammarians, in fact derived from this crude Koranic interpretation, that the expression kalld, meaning "not at all," which introduces this middle section, is only ever used as a negation of a preceding sentence, which merely compounds the hardly understandable connection between the middle and final sections of Sura XCVI.

At this point it will help to have before us both a conventional translation of Sura XCVI and Liiling's reconstruction. The following is a simplified English version of Rudi Paret's translation:

(1) Recite in the Name of your Lord who has created

(2) has created man out of an embryo!

(3) Recite! Your Lord is magnanimous as nobody else in the world

(4) [He] who has taught the use of the writing cane

(5) has taught unto man what he didn't know.

(6) Not at all! Man is really rebellious

(7) since he considers himself independent,

(8) however, to your Lord everything returns.

(9) What do you think about him who restrains

(10) a slave when he performs the ritual prayer.

(11) Do you believe that he is following the guidance

(12) or gives the order to be god-fearing.

(13) Do you believe that he pronounces lies and turns away.

(14) Does he not know that God sees.

(15) Not at all! If he doesn't cease we shall definitely grab him by the forelock

(16) a forelock full of lies and sinful.

(17) May he then call for his clique.

(18) We will call up the bailiffs.

(19) Not al all! Do not obey him! Prostrate and approach.

It should be obvious to everyone that this text is, for the most part, incoherent nonsense, and makes a mockery of the Koran's description of itself as "of clear Arabic language." Indeed, what little coherence this text has in translation is only achieved by supplying numerous words that are not in the original Arabic text at all, but drawn from the wholly spurious frame narrative supplied by Islamic tradition. Other illustrious translators, such as Arberry or Bell, only achieve whatever fluency their translations have by the same dubious method. It is impossible to produce a coherent translation because there is no coherence in the original Arabic, and it is precisely this fact that necessitated the invention of the traditional Islamic frame narrative. Muslims themselves had no idea what this text meant until an orthodox interpretation was concocted and spread abroad in Koran commentaries.

According to Luling this incoherence results from the fact that the basic rasm text has been manipulated in order to disguise its original form as a pre-Islamic Christian strophic hymn. Luling's translation and reconstruction follows, arranged in its original strophic form:

The strophic or poetic nature of this text is apparent from the frequent repetition of words, phrases, ideas, and synonyms. Moreover, it now appears as a single text with a theme common to all its parts, rather than three separate texts arbitrarily stitched together. However, there is still much that would strike the ordinary reader as obscure and puzzling. Luling justifies his translation and rearrangement of Sura XCVI in sixty pages of small print in which every word and line is discussed at length. For reasons of space we can only report briefly some of the more important parts of this analysis.

The first thing to be considered is the opening word of Sura XCVI, 'igra', which is usually translated "read," or "recite." It is upon this interpretation that the whole idea of the Koran as a revealed text rests, as well as the idea of Muhammad as its passive recipient. Luling takes his cue for reading 'igra' as "Invoke," rather than "read," or "recite," from the relatively early Arab philologist Abu Ubayda (d. 203/818). Ubayda remarked that the verb qara'a in XCVI.1 has the same meaning as the verb dhakara: "invoke, laud, praise." While in the West, long before Luling, the scholars Gustav Weil (1808-1899) and his contemporary Hartwig Hirschfield had stressed the point that the Hebrew expression gara' be shem Yahwe, "to invoke (with) the name Yahwe," widespread in the Old Testament, must be taken into account in the interpretation of Sura XCVI.

Understanding 'igra' as "invoke," rather than "read" or "recite," becomes plausible when it is realized that in the ancient world reading was invariably reading aloud, so that the distinction between reading and invoking would not have been what it is today. If it was unfamiliar or even totally unknown to an Arab of the sixth and seventh centuries to read silently, the Arabic expression gara'a must have meant "to articulate loudly." One can therefore easily understand the significance of "Articulate loudly the Name of God!" (XCVI.1) as being the equivalent of "Invoke the Name of God!" and, as far as the "name of God" is here a hypostasis, as equivalent to "Invoke God!" At the time of the rise of Islam there must still have existed a self-evident sense of magic in the solemn utterance of words with a religious connotation, which virtually excludes the interpretation of gara'a as a merely aesthetic reciting of something to be stated.

A significant consequence of abandoning the traditional Islamic frame narrative for XCVI.1-5, and taking up Abu Ubayda's interpretation of 'igra' as "invoke (your Lord)," is that there is no longer any need to treat the first five verses as a separate section. One then realizes the striking fact that scattered throughout the whole sura there are a total of seven references that equally, but in different words, express the meaning of "prayer": in XCVI.1 and XCVI.3, gara'a "invoke"; in XCVI.10, sally "perform the ritual prayer," in XCVI.17 and XCVI.18; da`d "call for or intercede in prayer"; in XCVI.19 sajada "prostrate for prayer"; and in the same verse, iqtarib "apprach (God)." These seven verbs with the meaning "to pray," could be the thread that runs through the whole sura as its unifying theme.

A word that cries out for explanation is kalld, "not at all, by no means," which occurs no less than three times in Sura XCVI. Its first appearance at XCVI.6 is senseless, since it cannot be a negation of the preceding section no matter how those verses are interpreted. Against the stipulation of Arabic grammarians and Koran commentators that kalld can only apply to a preceding sentence, Luling, following H. L. Fleischer (1801-1888), treats kalld as a negating particle equally applicable tp preceding as well as following sentences, preferably to the latter. For XCVI.6 this means that a vocal sign from the third layer of the Arabic script must be changed to establish the relation of the negation to the following sentence, so that XCVI.6 will read: "Not at all that man shall be presumptuous," rather than the awkward: "Not at all! Behold: man is presumptuous." This use of kalla is important since it not only makes sense out of the otherwise incoherent Sura XCVI but, as we shall see, performs the same function in Sura LXXIV, the other sura that is sometimes claimed by Islamic tradition as the first to be revealed. The stipulation of the grammarians that kalld can only apply to a preceding sentence thus appears to be deliberately devised to block a return to the erstwhile Christian reading of texts.

The other phrase that draws attention to itself by a triple repetition is a-ra'aita, "have you (ever) seen?" or "do you believe?" which introduces verses 9, 11, and 13. The strophic structure underlying the prose text shines through particularly clearly here, since the three a-ra'aita would make up the anaphoric initial word of three successive strophes. This being the case it would reveal the application of an artistic device that is common in strophic poetry everywhere.

In explanation of the curious reference to a "forelock" in verses 15 and 16, Luling uses the dogma-critical aspect of his threefold methodology. He observes that in the simile "to seize God by the forelock," there appears not only an anthropomorphism offensive to the Islamic understanding of God, but an attitude toward God that is foreign to the whole world of orthodox Islamic religious ideas. It is, however, familiar in Jewish and Christian conceptions of God, and was carried over from the Old Testament to the New and on into evangelical Protestantism. For Luling, pre-Islamic Arabian Christianity, as far as dogma was concerned, had an archaic Jewish-Christian or quasi-Arian character, so that in it we are entitled to take an understanding of prayer as an appropriate pestering of God for granted. The topos of struggling with God in prayer belongs to the oldest Jewish and Christian eschatological attitudes, and thus, on Luling's hypothesis, to the earliest form of the Koran text. The forelock in old Arabian normally had a positive meaning as the seat of honor, so the Koran redactors had to stress that the forelock of the reinterpretation of XCVI.15 had an unfamiliar negative significance instead, just for this special and extraordinary case. The redactors could not make such a curious switch without inserting a gloss or commentary to the word "forelock"; hence, the addition of the spurious "a lying sinful forelock," as verse 16.

The reference to "His high Council" in XCVI. 17, also betrays the Christian origin of the text, in that nddi, "council," has the Koranic synonym mala, which is applied in at least two cases (XXVII.8 and XXVII.69) to the High Council of the archangels. This council is in the Greek Septuagint version of Isa. 9:6, called the megata boula, "the Great Council," where God consults with the highest angels, the highest of which is the Messiah, the angel kataxochen: megalos boulos angelos, "the Angel of the High Council of God." It is in this biblical sense that the word nadiyah, of XCVI.96:17, is to be understood. As a hypostasis, this High Council of Angels appears almost to be a synonym for God, so that to call for the "Council of God," means practically the same as to call for God. In the following line the mysterious word z-zabaniyata, a Koranic hapax legomenon, translated by Paret as "bailiffs of hell," and by Richard Bell as "imps of hell," is reconstructed by Luling to read ar-rabbaniya, "ruling, governing, or powerful angels." This reading is achieved by only the slightest change to the pointation of the rasm, removing one point and setting one additional doubling sign. To reinforce these interpretations, Luling then goes into a lengthy excursus on the significance of the rabbaniyun, the Angels of the High Council of God, in early Near Eastern theology, the Koran, and early Islamic literature.

Lastly, there is a remnant of two words that in the orthodox Koran text are counted as verse 19. Both these words have endings that offer no possibility for a rhyme on long a to fit the previously uncovered strophic order of Sura XCVI. However, these words: wa-sjud wa-qtarib, "and prostrate, and approach," can easily be understood as the title or "headline" of the sura. Historically, the convention was to have a "footline," instead of a "headline," that is, the title was normally given at the end of a poetic or literary piece, and this seems to be the case here. They therefore do not belong to the body of the strophic hymn and should be eliminated along with verse 16. The effort to bring the number of verses up to nineteen in the sura designated as the first to be revealed may be not unconnected to the matter of nineteen in Sura LXXIV. This is not noted by Luling.

This brief but, we hope, representative selection from Luling's case for a Christian strophic hymn underlying Sura XCVI must suffice for the moment; all readers should consult Luling's own text for the argument in full. At this point we must turn from the first to the final chapter of Luling's book for a resume of his analysis of Sura LXXIV.1-30. Suras XCVI and LXXIV.1-30 are related in that they have both been considered by orthodox Muslim tradition as candidates for the first to be revealed; they may also be related for a more controversial reason, as we shall see.

Sura LXXIV.1-30 is of primary importance in Luling's analysis, since it not only exemplifies all the techniques of his methodology but brings to the fore his contentions about the nature of the Christianity he finds underlying the Koran. According to Luling, the pre-Islamic Christian hymnody contained in the Koran is theologically and dogmatically highly heretical, and unacceptable to the official trinitarian Christianity contemporary with the rise of Islam. With regard to the central issue of understanding the person of Christ in his relation to God, the pre-Islamic Christian hymnody confesses the blunt heresy that Christ is created. For Luling, this so-called heresy is in fact the original truth of Christianity, and the victorious trinitarian creed, its falsification. Luling maintains that Sura LXXIV. 1-30 contains an extraordinary pre-Islamic hymn in which, in the form of a formulaic confession of faith, there occurs the explicit statement that God created Jesus Christ. This pre-Islamic hymnody is therefore of inestimable value as a residue of ur-Christianity, and the archaic biblical dogmatics of central Arabia in pre-Islamic times, which, again according to Luling, formed the starting point for the new world religion of Islam.

As a whole, Sura LXXIV contains fifty-six verses in its orthodox recension. The restriction to verses 1-30 is justified by the fact that verse 31 is acknowledged, even by Islamic scholarship, to be a late insertion. Even the layman can recognize this from the length of verse 31 in comparison with the preceding and following verses. While verses 1-30 have an average of about four to six words each, verse 31 has no less than sixty-six words, and the following verses again small quantities. So verse 31 is quite obviously a commentary on the enigmatic number nineteen, mentioned in the previous verse: "Over it are Nineteen." Nevertheless, this over-length commentary in verse 31 is not an insertion into a large pre-Islamic text unit, because the second half of Sura LXXIV, although of a seemingly similar structure to 1-30, is in fact different in kind. Sura LXXIV.32-56 is what Luling calls a single-sense Islamic text, made up of phrases that occur elsewhere in the Koran in slightly different form, sometimes as reinterpreted sentences of pre-Islamic Christian hymn texts. The possibility cannot be ruled out that LXXIV.32-56 contains some scattered fragments of the hymn in LXXIV.1-30, since that is only the introductory part of a larger text, but the last part of Sura LXXIV is too damaged for a fruitful reconstruction.

At this point it will help to have the text of Sura LXXIV.1-31 before us in its canonical form. The following translation is Luling's depending on Bell and Arberry:

(31) We have appointed only angels to be masters of the fire, and their number we have appointed only as a trial for the unbelievers, that those who were given the Book may have certainty, and that those who believe may increase in belief, and that those who were given the Book and those who believe may not be in doubt, and that those in whose hearts there is sickness, and the unbelievers, may say: "What did God intend by this as a similitude?" So God leads astray whomsoever He will, and he guides whomsoever He will; and none knows the hosts of thy Lord but He. And it is naught but a Reminder to mortals.

As with Sura XCVI, it is the introductory words of Sura LXXIV that led it to be considered a candidate for a first revelation; in both cases this interpretation rests on a reinterpretation, or even a misrepresentation, of the underlying rasm text. A further parallel is that the traditional interpretation of these two suras derives from extra-Koranic frame narratives that have only a minimal basis in the text itself. Sura LXXIV is divided into different sections, seemingly without any coherence as to their content, each of them receiving its general idea from its respective frame narrative adduced in the commentaries of later orthodox Islamic Koran scholarship. The two basic sections of Sura LXXIV.1-30 are 1-10 and 11-30, each with its own frame narrative; the most curious of these is that attached to the crucial second section. This frame narrative has it that God himself utters threats to a Meccan foe of the Prophet named al Walid ibn al Mughirah, a distinguished, extraordinarily rich Qurayshite, friendly to Christians. Not much is known of this person other than that he is very rich and, in addition to some daughters, has no less than seven sons; all this conjured from the contents of verses 12 and 13, in the orthodox form of the text.

Luling's point of departure for a christological reading of Sura LXXIV. 1-30 lies in verses 11-17. As in Sura XCVI.6, the key to uncovering the original meaning of the rasm text is to ignore the arbitrary rule of classical Arabic grammar that kalla, "not at all," can only relate to a preceding sentence, even if there is no preceding sentence, and never to a following sentence. As was seen in the analysis of Sura XCVI, this rule seems to have been deliberately devised to prevent a return to the original reading of the text. The annulment of this theoretical rule, which has practically no support in vernacular Arabic, results in the change of kalld inahu, "Not at all behold him ... ," to kalld annahu, "Not at all that he . . . ," which requires no change in the rasm and results in a reverse of meaning; this provides the clue for a reconstruction of the whole context about Jesus Christ, closely corresponding to the usual Christian phraseology. This means that Sura LXXIV.16, instead of reading: "Nay! He has rebelled against Our signs" (Paret), will read, "Not at all that He was rebellious against Our signs." Paret's translation, "he has rebelled," is simply wrong, and the correct translation provides a christological topos: the suffering servant of Isa. 52:13, 53:7, and 53:12; Phil. 2:6-11; and Heb. 5:1-10.

This reading is confirmed by the following verse, LXXIV.17, which in its orthodox form is an exemplary case of a fantastic interpretation with no lead in the text. This verse consists of only two words: sa-'urhiquhu sa'udan, and the key word is sa'udan, "a height, hill, ascending path." According to orthodox Koran scholarship this refers to the central mountain of hell, up which sinners will be rushed until, after seventy years, they fall off in order to begin the process again, but none of this can be derived from the accompanying verbal stem r-h-q.

The solution to the problem lies again in interpreting both these words as literally as possible, but with the freedom of changing the punctuation, which is a late phenomenon and as such highly dubious. The rasm of sa`udan [slwd] is comparatively unproblematic, because there are no reasonable variants on the basis of another punctuation, so its general significance of "height, hill, ascending path," remains valid; there only needs to be found a plausible christological topos involving the key word "height." This emerges when the first word is read as sa-'azhagahu, rather than sa-'urhiquhu, which can be done with no change of the rasm. This means, quoting Lane's lexicon word for word: "He [God] caused his soul to go forth, pass forth, or depart." If this meaning is added to the second word, we get the sentence for LXXIV.17: "finally He [God] made him pass through death to height," which is the christological topos of Christ ascending, or being elevated after his death. Moreover, the word salad in LXXIV. 17, is a technical term in the Arabic-speaking church: the Ascension of Christ in the Christian Arabic Church was from of old called salad ar-rabb, "the ascending of the Lord," and the feast of the Ascension was known as ad as sa`ud

The traditional interpretation of LXXIV.11-15 has a series of statements by God expressed in the first person singular, supposedly referring to some anonymous evildoer, the details of which are supplied by the highly imaginative extra-Koranic frame narrative; this in itself is highly suspicious. It seems likely that this person was only invented to give the transmitted rasm text, within the very limited range of deviating punctuation possible, a fundamentally different meaning because the original christological content had to be eliminated.

The second half of the transmitted text of LXXIV.I I reads: wa man khalaqtu wahidan, and is interpreted as: "(the evil doer) whom I (God) created as a single being." This is nonsense, and does not even correspond grammatically to the text as it stands. The first person singular here is almost certainly due to the Koran editors, and to get back to the original christological hymnody needs to be changed to the third person: wa man khalaqahu wahidan, "and whom He (God) created (him) as a unique being." This makes it easier to understand the first half of LXXIV.11, which traditionally reads: dharni, "let me go, dismiss me," but which Luling reads as: dhara'ani, "He has created Me," making dhara'a a synonym for khalaqa, "to create"; this has precedence elsewhere in the Koran at Suras VI.136; VII.179; XVI.13; XXIII.79; XLIL11; and LXVII.24. Sura LXXIV.11 now reads: "He has created me and whom he has created as a unique being," a clear christological statement.

This grammatical reconstruction of LXXIV.11 is justified on dogmacritical grounds by the fact that in most of the Christian creeds there is the custom of first mentioning God's creation of the cosmos in general and man in particular, and only afterward to mention Jesus Christ. In the case of Sura 74:11 if. we have, according to Luling, an ur-Christian, angelchristological creed. Whereas in the trinitarian creeds, the statement about Christ's creation cannot occur at all, because Christ is identified with the creator; the ur-Christian angelogical creed, preserved uniquely in LXXIV.11 ff., has the two statements following each other. God created Jesus Christ as the lastly added remedy for the entire creation, so that Christ might become, as God's suffering servant, the salvation of the world. This explains why, even in the old creeds of trinitarian Christianity, Jesus Christ is only mentioned after the creation of the cosmos and mankind. The ur-Christian creed of Sura LXXIV.11 if. would be the original form of all Christian creeds before the triumph of trinitarianism. Luling regards the designation of the Quraish of Mecca as mushrikun, "associators," as evidence that they had converted to a Hellenistic-Christian trinitarian creed, and had rejected the pre-Islamic angel-christological hymnody. The early Islamic movement abandoned faith in Jesus Christ as a central issue and reinterpreted the hymnody to fit a wholly Islamic Koran. The only people who remained faithful to the pre-Islamic hymnody were the hanifs, or heretics, who were eliminated by victorious Islam.

In a similar manner to those analysed above, verses 12-15 of Sura LXXIV can be reconstructed to read as parts of a pre-Islamic, angelchristological hymn, and the reader is referred to Luling's text for the lengthy and detailed arguments. After verses 11-17, Luling turns his attention to section LXXIV.18-30, and in particular to the crucial section LXXIV.26-30, which concerns Christ's descent into hell.

Attention is first drawn to a conspicuous inconsistency in the text of LXXIV.18-30, in that in the closely following verses 25 and 29, the word bashar, "flesh," is in the first instance given its normal significance of transitory human being, while in the second it is given the context: "and is scorching the flesh (in hell)." The latter is especially suspicious in that the rasm, lawwahatun, interpreted as yielding the verbal predicate "is scorching," cannot mean this since the verb l-w-h always means elsewhere in Arabic, apart from this single Koranic instance, "to shine, to beam," and has nothing to do with fire and burning heat. The only plausible alternative reading of the rasm is as la-wahatun, "verily, an oasis," so that the whole of LXXIV.29 would read: la-wahatan li-l-bashar, "verily, an oasis for the transitory flesh." This is a very suitable expression for the underworld, the place of the dead until the day of Last Judgment. It also unites the two instances of the word bashar, in verses 25 and 29, with a single meaning.

If there is no reference to hellfire in LXXIV.29, we are led to suspect that the word saqar in LXXIV.26 does not refer to hellfire either, since the spurious nature of LXXIV.29 is only constructed to provide an answer to the question: "What has let thee know what saqar is?" in LXXIV.27.; if saqar does not mean: "the heat of hellfire," as the traditional interpretation of the rasm would have it, then the accompanying verb s-l-w will not mean: "to roast, to stew." The word saqar holds a key position in the interpretation of the section LXXIV.26-30, and thus of the whole section LXXIV.1-30; Luling devotes some fifteen pages of fine print to its analysis, of which we can only give an indication here.

Saqar, with the meaning: "hellfire, heat of hell," as assigned to it by Islamic tradition, has no certain origin in Arabic or any other Semitic lan guage. If Luling is right in his theory that the original rasm of Sura LXXIV.1-30 contained a pre-Islamic christological hymn, and that that hymn concerned Christ's descent into hell, then it is reasonable to assume that the word saqar had some connection with the idea of an underworld or realm of the dead, rather than simply of hell or hellfire. This underworld would be hell in the older sense of a receptacle for the dead, good as well as bad; a realm ruled by Death, into which Christ would bring salvation and the sacraments of the New Covenant for the benefit of the deceased of earlier, pre-Christian times. If this line is taken one stumbles upon remarkable connections and coincidences in the Arabic language, which are at the same time probably etymological relations.

Luling suggests that the Koranic word saqar, which he interprets as meaning "underworld" or "place of the dead," is an irregular derivation from the tenth verbal stem of the verb q-r-r, "to be stationary," "to be cool/cold," an appropriate description of both the dead and the place where they reside. The direct phonetic parallel to saqar in Hebrew is shaqar, which probably originally meant the womb as the underworld, the receptacle of the dead as well as the place of rebirth and resurrection. As such this is an archaic messianic concept, originally centred in the cult of the heroic ancestors at the High Places, which as burial sites represented a womb. This was a wholly positive idea of the underworld, the notion of eternal punishment in hellfire being an innovation of the newly invented monotheisms, which sought to destroy the old religion of the High Places and turn the saqar idea into its opposite: deceit, delusion, illusion, hallucination, lie. The positive idea of the underworld as womb and place of resurrection can also be traced in Akkadian, Assyrian, and Babylonian, where it has no connotation of heat or hellfire, but, on the contrary, that of coolness.

So the reconstructed meaning LXXIV.27 now reads: "And what has taught you what the underworld is?" which makes more sense of the following enigmatic verse LXXIV.28 when translated: "it (the underworld) doesn't let remain and doesn't let go"; it doesn't leave anyone remaining on earth and doesn't let anyone go until the final judgment. In conformity with these reconstructions, LXXIV.26: sa-aslahu saqara, can now be translated: "So He (God) has finally exposed him (Christ) to the underworld," the verbal root s-l-y being given its core meaning of: "to turn oneself or something toward something," as in the verb salla, to pray.

We are now in a position to consider some of the ramifications of the controversial verse LXXIV.30: `alayha tis`ata `ashara: "Over it are Nineteen." As noted above, Sura LXXIV.1-30 is composed in a very homogeneous form, in that every verse has the same rhythmic style and approximately the same length of, on average, three to four words, which in itself is a good indication that it was originally a strophic text. This makes the contrast between LXXIV.1-30 and the following verse LXXIV.31, which is a prose text taking up half a page, so conspicuous that it could not fail to draw the attention of commentators. Even traditional Islamic Koran scholarship has classified LXXIV.31 as a late insertion into an earlier text; late and early in that context referring of course to Medinan and Meccan. Luling proposes that this traditional classification should be abandoned in favor of the contrast: "pre-Islamic Christian strophic texts," and "Islamic texts." However, the essential point to grasp about the juxtaposition of verses LXXIV.30 and 31 is that, although they are quite clearly different types of text, it is also quite clear that the latter is a commentary on the former. This raises all sorts of questions about how this could have come about.

The traditional interpretation of Sura LXXIV.31 is that it is a commentary on the immediately preceding words of LXXIV.30, yet in order to make sense of that text it utilizes the frame narrative of verses 11-30, which, if Luling is correct, is a late interpretational device of those wishing to disguise the meaning of the original rasm. Since verse 31 twice mentions "those who were given the Book," presumably meaning the Koran, when, according to the traditional picture of events, the Koran was not assembled into a book until long after the Prophet's death, we have in this verse a major anomaly. Why would the Prophet, at Medina or anywhere else, receive a revelation about a book that did not yet exist, even a particular verse of that book?

The whole idea of a book, and that book being the Koran, arises out of the text of verses 25 and 26: "This is nothing but magic made impressive. This is nothing but human speech," which is taken by the commentary of verse 31 as an insult by the protagonist of the frame narrative to the revelatory utterances of the Prophet, for which he is condemned to roast in hellfire (saqar). But if Wing's interpretation of saqar is correct there is no hellfire and no roasting, only an underworld. Moreover, there is still no explanation for the appearance of the number nineteen. Even if nineteen refers to angels who watch over the underworld or over hellfire, why would "their number" be a "trial for unbelievers," and why would it give certainty to those that have the book, and cause the unbelievers to say: "What did God intend by this similitude?"

According to Luling the number nineteen arises from a misreading of the rasm of LXXIV.30, the true meaning of which must be sought in conformity with the whole preceding section, 1-29, and in particular to the section 26-29, which concerns Christ's descent into the underworld. Most probably LXXIV.30 should be seen as still part of the answer to the question set in LXXIV.27: "and what has let thee know what saqar is?" Verses 28 and 29 are the first two answers to that question, and verse 30 would be a third.

Everyone whose job it is to read old Arabic manuscripts, which are very poorly supplied with diacritical dots, will have been in the situation of being unable to decide whether to read tis`a, "nine," or sab`a, "seven," because these two numbers, when written in words, are indistinguishable in a rasm without dots. This means that the original rasm of LXXIV.30 could have read "On it are seventeen," rather than the traditional "On it are Nineteen," and apparently there is at least one record of a variant reading of this kind.9 It is therefore quite possible that there could have existed in verse LXXIV.30 solely the number seven, so that the following word, which is written separately, need not have been ten, extending the preceding seven to become seventeen, but a noun designating something that is counted by the preceding number seven. So the text of LXXIV.30 may have originally read: "on it there are seven x," where x stands for something expressed by a noun hidden behind the rasm of `ashara. It so happens that there is a very plausible candidate for this hidden noun, bearing in mind Luling's translation of saqar as "underworld."

It was a common idea in the old Orient as well as in Islam that the underworld had seven sections or levels, corresponding to the seven heavens above, and that these seven levels had seven gates. Referring to "Gehenna," which originally meant the underworld, Sura XV.44 reads: "it has seven gates to every one of which belongs a section partitioned off." This brings to mind the fact that the Semitic and Arabic word `ashar, "ten," very much resembles in both its script and its sound the Hebrew word sha'ar, "gate." It would not be exceptional if the Hebrew word sha'ar had an offshoot in pre-Islamic old Arabic, where the Arabic word would have had the form saar, or sa'r. The difference between the alternative words "ten" and "gate," consists in Arabic script of the metathesis of only one consonant: s-r I-s-r. It therefore appears quite possible that the early Koran editors, in the course of their reinterpretation of the whole section LXXIV.1-30, "corrected" the original rasm of the Hebrew loanword "gate" to get "ten." So the original text of LXXIV.30 may well have read "On it are seven gates," rather than "On it are Nineteen."

On this reading of LXXIV.30 the number nineteen arises solely from a deliberate misreading of the original rasm, which could have been read as either seventeen or nineteen. Apparently it was sometimes read as seventeen, but this was dropped in favor of nineteen, presumably, if Luling's diagnosis is correct, to further remove it from the original seven of the rasm. The commentary in the following verse 31 is read by Luling as a threat against those who still remembered the Christian hymn and the number seven hidden in the original rasm, but this hardly seems an adequate explanation for all the nuances of the text of verse 31.

Even if Luling is correct about the substitution of nineteen for seven or seventeen, why would the redactors deliberately draw attention to it by incorporating such an obvious and anomalous commentary into the text? Presumably the "correction" of the rasm of verse 30 was one of the first editorial acts perpetrated on the material that eventually formed the Koran, perhaps even before the life of Muhammad, whatever dates are give to that life. The commentary, however, is quite obviously late, and must have been concocted by those engaged in one of the final stages in the production of the canonical Koran, a time when the Koran was already some form of book. Verse 31 appears to be there to draw attention to some feature of that book that would cause problems for unbelievers and convince believers of its supernatural nature. This feature is of such a kind that it can be called a "similitude." Of what could the number nineteen possibly be a similitude? There appears to be only one plausible answer: nineteen is a similitude of the book's supposed author, the One God, Allah.

By the abjad system of substituting numbers for letters, nineteen is the number of the Arabic word wahid, meaning one. If the Koran was formed in a sectarian milieu of competing forms of Christianity in which the trinity was a bone of contention, the oneness of the One God of emer gent Islam would have been its distinguishing feature. What better similitude of this One God could there be than that his revealed book be permeated throughout by recurring patterns of the number nineteen. This would be especially telling in a milieu where it was well-known that Jewish and Christian books were based on numbers revelatory of their nature; for instance, Ecclesiastes was based on the number fifty-six, the number of the word hebel, "vanity." This feature of the holy book of the Muslims may have been most prominent and pervasive in its early recensions, only the ruins of such a scheme being apparent in the Koran as we have it today. The idea of nineteen in the Koran has lingered in the corners of the Muslim mind for centuries, and has become a matter for dispute in recent times, but is too detailed to be pursued here.'0

Having shown in some detail Luling's method of reconstruction of LXXIV.11-17 and 26-30 in order to show the underlying pre-Islamic Christian creed in the form of a strophic hymn, the remaining text LXXIV.1-10 and 18-25 needs less attention. This is because the main points that identify this underlying text as a Christian angelogical creed have already been made, and the remaining sections, although not without puzzling elements in need of explanation, simply confirm the preceding analysis; the reader is referred to Luling's own text for his punctilious line-by-line and word-by-word argumentation.

We are now in a position to present Lining's reconstruction of the whole pre-Islamic hymn present in LXXIV.1-30:

The reader should be aware that there is a caesura before the last line of each strophe, since this was sung by the lay audience as a response to the preceding lines which were usually sung by a soloist priest or deacon. It should also be noted how the content of the hymn is arranged by strophe. Strophe 1 being a prelude admonishing the congregation to be vigilant; strophe 2, a reminder of doomsday; strophe 3, a creedlike assessment of God's preparing Christ for the salvation of mankind; strophe 4, a description of Christ's Passion and death as decided by God; strophes 5 and 6, a description of Christ in Gethsemane giving up his will to the will of God; and strophe 7, a description of the nature of the underworld..

THE HISTORICAL CONTEXT

The question of whether Luling's uncovering of a pre-Islamic Christian hymnody underlying about one-third of the Koran is in the end convincing, must be left to specialists in the field of Arabic language and philology. However, these linguistic and textual problems are only half of Luling's thesis about the origin of Islam. Just as important is the historical and geographical context in which he envisages this adaptation of Christian texts to have taken place.

His position might be characterized as at once traditional and revisionist. He is traditional in that he accepts without question the Muslim account of the origin of Islam in central Arabia and the Hijaz towns of Mecca and Medina. He also accepts that a major role in the birth of Islam and the Koran was played by a prophet called Muhammad, who had roughly the dates and a good part of the biography assigned to him by Muslim tradition. The most radical difference Luling has with this traditional picture is that he sees practically the whole of the Arabian peninsula as thoroughly Christianized, even to the extent of Muhammad's own tribe of the Quraysh being Christian. This view of the religious allegiance of the people of central Arabia is, in a sense, forced upon him by trying to reconcile the fact of the Koran having Christian hymns hidden in its rasm, with the fact of its having originated in central Arabia or the Hijaz. Where could these Christian hymns have come from if there was not a widespread Christian population at hand already making use of them?

To bolster this assumption of a thoroughly Christianized central Arabia, even to the extent Mecca being a center of Christianity for some two centuries before the birth of Muhammad, Luling has to make the most of every piece of Muslim tradition that can be interpreted in a way that would support his thesis. Chief among these are the stories that the Kaaba at Mecca contained biblical pictures, including Jesus and Mary, as well as architectural features that Luling interprets as indicating a Christian church oriented toward Jerusalem. He also reads the Koranic term mushrikun, meaning "those who adjoin a companion to God," as referring not to the idol-worshiping polytheists of the jahiliyya, as Muslim tradition would have it, but to trinitarian Christians. Now, all these items of information may be significant for assessing the origin of Islam, but none of them necessitate the interpretation given them by Luling as backing his thesis of a Christianized cental Arabia; indeed, the very opposite may well be the case.

The inherent implausibility of Luling's assumption of a widespread Christianity in central Arabia in pre-Islamic times is shown by the paucity of contemporary scholarship that he can adduce in support of the idea. Apart from some German Protestant theologians, he only mentions J. S. Trimingham as insisting on the existence of organized Christian communities in that area in pre-Islamic times, but if Trimingham's book on the subject is actually consulted nothing substantial is found that supports Luling's thesis.

In discussing the presence of monotheism in West Arabia, Trimingham remarks that: "Christianity was non-existent among the Arabs of western Arabia south of the Judham tribes." In a chapter headed "Christians in the Hijaz," after describing the history of Mecca according to the Muslim sources, plus its geographical location, he concludes that "these factors are sufficient to explain why Christianity in any of its available forms could have no influence upon its inhabitants." I 1 The following section is promisingly entitled "Christians in Central Arabia," but is concerned mostly with the nomad alliance state of Kinda. The main body of Kinda lived in the Hadramawt, but early in the fourth century some sections separated and moved into the Najd area of central Arabia among clans of Ma'add, with whom they formed alliances; because of a subsequent alliance with the Byzantines it is assumed that the Kinda confederation must have adopted some form of Christianity. This would at least appear to be true of the ruling clan, judging by an inscription found in a convent church at Hira on the Euphrates.12 Hira is over six hundred miles from Mecca and can hardly be called central Arabia.

The other scholar to whom Luling refers is Irfan Shahid. In his Byzantium and the Arabs in the Fifth Century, Shahid makes the remark, "Places with distinctly Christian association, such as Maqbarat al Nasara, the cemetery of the Christians, are attested in Mecca in later Islamic sources and these could not possibly have been fabricated." 13 To which one is inclined to reply: Why not? When so much other material in Muslim tradition has been either invented or transposed, why not a Christian cemetery at Mecca? The source of the information is Azragi's "Meccan Reports," Akhbar Makka, assembled some two hundred years after the supposed lifetime of the Prophet and published a hundred years after that, so there is no reason to think it any more trustworthy than any other literature produced at the same time. Moreover, if it is true, as many scholars believe, that much of the material referring to Mecca in the southern Hijaz has been transferred there from somewhere much farther north, then the whole picture changes. The elements of Muslim tradition that Luling attributes to a Christian milieu in the midst of which Islam emerged, need not indicate the Hijaz and central Arabia at all, but a place where we know for a fact, rather than unsubstantiated hypothesis, that such a society existed.

These considerations bring us to the crucial point that is likely to make Luling's thesis unacceptable to the majority of scholars. It is not that the posibility of a Christian hymnody hidden in the rasm of parts of the Koran is inherently unlikely or unacceptable, but that the historical context of this possibility, as postulated by Luling, is simply too farfetched and involves too many unverifiable presuppositions. There is also the problem that all those who adopt a skeptical attitude to a body of tradition cannot at the same time use that tradition to reconstruct "what really happened," since the criteria for selection from that tradition will then be entirely arbitrary; whatever supports the favored hypothesis is treated as true and genuine tradition, while the rest is ignored or dismissed as irrelevant fabrication. Also, a conscious burying or falsification of history that is generally known to be true, as Luling's hypothesis requires, seems inherently less likely than a gradual, unconscious, construction of a tradition that no one knows to be false.

It is now well-known that the consensus of scholarly opinion on the origins of Islam underwent something of a seismic shift in the last quarter of the twentieth century.14 This change of view was initiated by John Wansbrough in his Quranic Studies, published in 1977; behind the scenes Wansbrough had also been a crucial influence on Patricia Crone and Michael Cook, who published their Hagarism: The Making of the Islamic World in the same year. These two books made a radical break with the traditional picture of the origins of Islam. The shift is one of both place and time, in that the location for that origin is no longer the Hijaz in the lifetime traditionally attributed to Muhammad, but the Fertile Crescent some time after his death and subsequent to the Arab conquests.

The key to the thinking behind this new view of things is to be found in the following passage from Quranic Studies:

The frequently adduced view that the text of revelation was easily understood by those who had witnessed its first utterance, as well as by their immediate successors, but by later generations would not be, is in my opinion not merely ingenuous, but belied by the many stories of early efforts towards the interpretation of scripture associated with the figures of `Umar b. Khattab and `Abdallah b. Abbas. Whatever the reasons for production of those stories, it seems hardly possible that at the beginning of the third/ninth century the Muslim community had to be reminded of what it had once known. Tafsir traditions, like traditions in every other field, reflect a single impulse: to demonstrate the Hijazi origins of Islam [our italics].15

In other words, the whole Muslim tradition of the origin of the Koran, including both the "occasions of revelation," (asbab an nuzul), and their meaning, is a fabrication of the ninth century. The purpose behind this literary effort was to convince both Muslims and non-Muslims that Islam had its origin not in the civilized region of the Fertile Crescent conquered by the Arab tribes, as contemporaries might presumably have supposed, but in the Arab homeland desert region of the Hijaz. This set Islam apart, both geographically and theologically, from the Judaism and Christianity with which it obviously had so much in common.

The final divine revelation had descended not upon the corrupted Hellenes of the north, but upon the unspoiled Arabs of the pristine desert of the south, the final chosen people whose divinely guided destiny it was to rule the civilized world in the name of the one true God, Allah.

Because so much of the Koran and Hadith is obviously dependent upon a historical context in which the presence of Jews and Christians is taken for granted, it was necessary for Islamic apologists to populate the Hijaz with Jewish and Christian communities for which there is little or no evidence outside Muslim literature; Luling's attempt to fill central Arabia and the Hijaz with Christians is just a continuation of this project into modern times. Remarking on this phenomenon with regard to the necessity for postulating Jewish influence on the collection of and commentary upon the Koran, Wansbrough says:

Some scholars ... have been excessively generous in their assessment of the documentary value of Islamic source materials for the existence and cultural significance(!) of Jewish communities in the Hijaz about which Jewish sources are themselves silent. References in Rabbinic literature to Arabia are of remarkably little worth for purposes of historical reconstruction, and especially for the Hijaz in the sixth and seventh centuries. The incompatibility of Islamic and Jewish sources was only partially neutralised, but the tyranny of the "Hijazi origins of Islam" fully demonstrated, by insistence upon a major Jewish immigration into Central Arabia.16

Luling's insistence on a major Christian immigration into central Arabia is just as undocumented and unlikely, and just as much a product of a dogmatic insistence upon the "Hijazi origins of Islam." In this matter Luling takes Muslim tradition wholly at face value, regardless of how much he wishes to undermine and call it into question in other respects.

If a plausible Mecca is necessary to account for all the material in the Muslim sources, a suitable candidate would be Moka, a town in Arabia Petraea mentioned by Ptolemy; not far distant was a "Kaaba" toward which Muslims prayed in early times, as witnessed by Jacob of Edessa.'7 Such locations in northwest Arabia, far north of Mecca in the Hijaz, provide a suitable environment for the origin of Islam, which obviates the necessity to postulate large-scale migrations of Jews and Christians into Arabia. Indeed, anywhere in the arc of the Fertile Crescent would provide a suitable location for the notorious foreign vocabulary of the Koran, which indicates familiarity with a cosmopolitan environment far removed from the southern Hijaz.18

However this may be, there is a sense in which the geographical location of the original Mecca or Medina is irrelevant, in that the only fact about which there can be no doubt is that in the second quarter of the seventh century there was a large-scale occupation of the Hellenic Middle East by Arabs. The crucial point at issue is the exact nature of this event: was it simply an act of political opportunism, taking advantage of an exhausted opponent, or was it driven by some kind of religious zeal? The first alternative is favored by Wansbrough, who sees the development of Islam as wholly subsequent to the establishment of a religiously unspecific polity: "Both the quantity and quality of source materials would seem to support the proposition that the elaboration of Islam was not contemporary with but posterior to the Arab occupation of the Fertile Crescent and beyond."19 Both the Muslim and Christian accounts of "what happened" are read by Wansbrough as indicating "the persistence of Judaeo-Christian sectarianism in the Fertile Crescent under Arab political hegemony, the establishment of a modus vivendi between the new authority and the indigenous communities, and the distillation of a doctrinal precipitate (a common denominator) acceptable initially to an academic elite, eventually an emblem of submission (islam) to political authority."20 Thus, one of the few "facts" that we have about early Islamic history is that "the religious movement later identified with the state begins as the sectarian expression of a scholarly elite."21

If the "doctrinal precipitate" that we know as Islam first appears to us in documents produced by a literary elite c. 800 c.E., it seems reasonable to suppose that there were earlier documents that have not survived. "What really happened" may not be what the literary elite would have liked us to think happened, but whatever the exact process, the emergence of Islam involved a gestation of at least 170 years, which left traces other than documents. It is perhaps significant that such documentary evidence as survives from the Sufyanid period (661-684) makes no mention of Muhammad as the messenger of God, and the coinage invokes Allah without mentioning his prophet, but by the Marwanid period (684-744) coinage was being struck which identified Muhammad as rasal Allah, and thereafter reference to Muhammad as messenger of God becomes standard on Arab coins. It was the caliph cAbd al Malik (685-705) who devised the classical solution for Arabic/Islamic coinage, when he produced a coin with no images, only Arabic script containing a distinctly Islamic message: "There is no God but God and Muhammad is his messenger who He sent with guidance and the religion of Truth to make it supreme over all others whether the polytheists like it or not."23 Since the caliph was in effect the state, the crystallization of the idea of Muhammad as messenger of God and Islam as a distinct religion had been adopted by that institution within seventy years of its foundation. At the same time, there was at least an incipient version of a collection of texts that was eventually canonized as the Koran, as witnessed by the inscriptions on the Dome of the Rock, also completed in the reign of 'Abd al Malik.

None of this is, of course, sufficient to validate the picture of the origin of Islam that was propagated by the literary elite of the ninth century, but does indicate that more history is available to us than can be derived from documents. This issue constitutes the main difference between the approaches of Wansbrough, on the one hand, and Crone and Cook, on the other, to the origins of Islam. While Wansbrough eschews any attempt at a reconstruction of "what really happened," beyond such gnomic utterances as those quoted above, Crone and Cook put forward a positive thesis that contradicts the traditional account and lays itself open to refutation and critical attack; but even Wansbrough's sparse assertions are vulnerable to the hard facts of architecture and numismatics.

Once the traditional picture of the origin of Islam is laid aside the problem resolves itself into two questions: (1) Did Islam as a religious phenomenon begin before or after the Arab conquests? (2) Was it a development from or against Judaism or Christianity or both? Crone and Cook would answer the first question "before," and the second "from Judaism"; Wansbrough would answer the first question "after," and the second question, probably, "from and against both"; Luling would answer the first question "before," and the second question "from and against Christianity." When scholars who have spent their lifetimes examining the same source material in the original languages can come up with such contradictory conclusions, the problem is obviously not easily solved, and is perhaps susceptible to whatever bias is brought to the subject. The old Pascalian adage may apply here as elsewhere: There is always enough evidence for those who want to believe (whatever), and never enough for those who do not.

If the Koranic texts on the Dome of the Rock and the inscriptions on the coinage are taken as our hardest evidence for what counted as "Islam" at the end of the seventh century, there is a case for taking the Koran and the Prophet as fundamental elements of the faith from the beginning, remembering that "Koran" simply means "reading" or "recitation," and that rasul Muhammad could be an office that never had an incumbent. The danger is that given the existence of these ideological shibboleths we read into them, at the seventh-century stage, all that the scholarly elite of the ninth century would like us to believe about them. It has been argued, for instance, that because the inscriptions on the Dome of the Rock resemble phrases and short passages from the canonical Koran, that text must have existed in the reign of cAbd al Malik as a perfect exemplar from which it was permitted to deviate in order to form an argument against Christians.25 The logic of this assertion is hard to follow. Why could not a fixed text have been assembled later from already existing fluid and fragmentary texts? All the inscriptions on the Dome of the Rock and the coinage demonstrate is that Koranic sentiments and phraseology were in circulation at the end of the seventh century. This is hardly surprising and does not contradict Wansbrough, since he never denied the existence of Koranic logia from an early period. Apart from these considerations it is obviously impermissible to argue from the existence of fragments of a text to the existence of the whole, especially when those fragments are either polemical or of a type used in prayer and thus most likely of all to have existed from the beginning.

If the sentiments of the Dome of the Rock and coinage inscriptions are anti-Christian, in their assertion of one God as a refutation of "polytheists" in the sense of trinitarians, this could be in favor of monotheist Judaism, as Crone and Cook might argue, or of a nontrinitarian aboriginal Christianity, as argued by Luling. Whichever the case, it could have begun either before or after the conquests, either in northern Arabia or the Fertile Crescent. What the role of Muhammad might have been in such a development is, again, a matter of choice. In Wansbrough's account of postconquest developments, the Prophet appears as little more than a cipher, a more or less inevitable feature of any religious evolution from Middle Eastern monotheism, necessary as a channel for revelation and as a moral example for the nascent community. Basing themselves on Syriac and Greek sources contemporary with the conquests, Crone and Cook have no doubts about Muhammad's historicity.26 Indeed, Crone seems most enamored of the traditional, not to say Romantic, image of the Prophet: "There is no doubt that Muhammad lived in the 620s and 630s A.D., that he fought in wars, and that he had followers some of whose names are likely to have been preserved,"27 and even goes as far as to declare that "it is a fact that whichever way the origins of Islam are explained, Islamic civilization is the only one to begin in the mind of a single man."28 In this respect, if in no other, Crone is in agreement with Luling, the difference resides in their wholly opposed conceptions of the nature of ur-Islam.

As already noted, the companion volume to Luling's On the UrKoran is his Rediscovery of the Prophet Muhammad: A Criticism of the "Christian " Orient. We have already had occasion to mention most of the features of Luling's "rediscovered" Muhammad when illustrating his reconstruction of the Christian hymnody in the text of the Koran. The chief feature of Luling's newly discovered Prophet is that he is the messenger for an Ebionite Judaeo-Christianity, which is taken as the ur-Christianity of Christ himself. This ur-Christianity is also an ur-Judaism in the form of a cult of the Hero Messiah at the groves on the High Places of the northern kingdom of Israel, a cult known to Lining's Muhammad as the religion of Abraham, Ishmael, and the tribes; this ur-Christianity is opposed to the Hellenistic trinitarian Christianity of the imperial Roman Church. Both these Christianities are imagined by Luling as existing in central Arabia and the Hijaz during Muhammad's lifetime.

Now, in comparison with the views of Wansbrough, Crone, and Cook, this is obviously a very different kind of thesis; one not so much drawn from the sources usually utilised for establishing the origins of Islam, as one born elsewhere and brought to those sources seeking confirmation wherever it can be found. It is in fact a thesis that brings with it a lot of baggage carried over from a German Protestant, liberal theological study of the Old and New Testaments, and a psychologically driven reading of anthropology, linguistics, and world religions; a thesis born of an obssessional program extraneous to Islam as such. Luling's Muhammad is, in effect, the embodiment and protagonist of his own preoccupations; not so much a "rediscovery" as a reinvention in his own image.

Extraordinary as it might seem, the mission for world peace and understanding of Muhammad in seventh-century Arabia turns out to be the same as that of Gunter Luling at the beginning of the third millenium: the abandonment of the imperialistic orthodoxies of Judaism, Christianity, and Islam, and a return to a prehistoric henotheism. Even more extraordinary, Luling appears to expect modern Muslims to welcome his rediscovery of this previously unknown Muhammad, along with his uncovering of Christian hymnody at the base of their holy book, and embrace his vision of a new/old religion as spiritual inspiration in the new millenium, even though it is not a new millenium for them. For a man who has spent a large part of his life in the Middle East, this shows a surprising naivete about the mentality of Muslims.

Luling's view of Muhammad and the origin of Islam is not only opposed to the traditional account, but also to the "revisionist" hypotheses of Wansbrough, Crone, Cook, and Hawting. In a lengthy preface written for the English edition of his book on the ur-Koran, Luling makes an attempt to refute the revisionist approach to the origins of Islam. In general terms, the revisionists are castigated for dismissing the huge body of Islamic literature, dating from 800 C.E. and later, as historically worthless, rather than painstakingly sifting through it, using the source-separating and text-separating methods of linguistic, literary, historical, and theologico-dogmatic text criticism. These are the methods, presumably, that enabled Luling to come to his equally revisionist conclusions about Muhammad and the origins of Islam; the difference being that whereas Luling questions everything in the Muslim sources apart from their overall geographical and chronological framework, the anglophone revisionists question this, too. The problem with the methods advocated by Luling is that the researcher tends to end up with evidence for whatever prejudice he starts out with. If he takes it for granted that the traditional Islamic account of events must be basically true, he will find evidence for that; if, like Luling, he assumes there has been a massive coverup of what really happened, he finds evidence for that, and so on. Luling manages to find evidence for both the traditional account with regard to the framework, and his own idiosyncratic version of revisionism, as it suits him.

A contemporary scholar whose methods Luling would probably approve is Michael Lecker,29 who has sifted throgh mountains of detailed source material about early Medina in search of "hidden pearls" of historical truth, the assumption being that it all has to be either true or fabricated, since "it would be absurd to argue that the fine, detailed information we have on certain aspects of early Islamic Medina is unusable or entirely the outcome of later inventions."30 Unfortunately, as Patricia Crone has pointed out,31 Lecker's "hidden pearls" of historical truth turn out to be nothing of the kind, but simply endless contradictory anecdotes that make no difference to the overall picture of events. Anecdote piled upon anecdote is in fact what makes up the bulk of historical material on early Islam, all carried by the basic framework of a Hijazi origin at Mecca and Medina, if that goes the rest goes with it, at least in that context.

Luling also objects to the revisionist thesis on the grounds that it shows a superficial understanding of such concepts as monotheism, polytheism, Judaism, and Christianity, as well as plain ignorance of the religio-histor- ical situation in central Arabia, so that it has to be conjectured that a religion such as Islam could not possibly have originated there. A further complaint is that the "mainstream of oriental monotheism" (Wansbrough), out of which Islam is supposed to have emerged, is chiefly identified by the revisionists with Judaism rather than Christianity; the classic texts of C. C. Torrey and A. Katsch on Judaism on Islam,32 are said to have been long since disproved, but we are not told how, where, when, or by whom. Gerald Hawting is especially singled out for attention, seemingly because in his book on idolatry and early Islam,33 he is seen as having trespassed on Luling's patch without giving him due recognition, especially with regard to his insistence upon a Christian presence at Mecca and elswhere in central Arabia. Hawting lists Luling's book on Muhammad in his bibliography, but does not refer to him in the text, probably because he had already published his views on Luling's theories almost twenty years previously, when they first appeared in their German editions.

In his review of The Rediscovery of the Prophet Muhammad, Hawting acknowledges that Luling has made several important points with which he can agree:

for example, the drawing of attention (pp. 153 ff.) to the tension which is observable in Muslim sanctuary traditions and practises between, on the one hand, the concept of the sanctuary as something which is entered and within which rituals are performed and, on the other, the idea that the sanctuary is closed and venerated externally, seems perceptive and important. I also share Luling's views on the importance of the traditions about the area of the sanctuary called al hijr (pp. 133 ff.) and the curiously unconvincing nature of the reports about the idols said to have been destroyed by the Prophet at the time of the fath (pp. 168 ff.). Again, I find attractive the argument (pp. 183 ff.) that words like shirk and mushrikun may, in the Qur'an and elsewhere, be understood as polemic and not as references to polytheists or idolators in the literal sense. More generally, his view that the Qur'an can be seen as containing material which cannot all be attributed to one individual is, when expressed in this way, in keeping with recent Qur'anic studies, and the similarity between Muslim prophetology and what we know of the prophetology of JudaeoChristianity, which Luling adopts from H. J. Schoeps, is striking and, as Schoeps presents it, persuasive.34

However, the way in which these points are elaborated and developed into a wider thesis show "significant weaknesses in argumentation and in method."35

Now, Luling's thesis about the origin of Islam, no less than that of the revisionists, involves the rewriting history; what, up to now, is generally thought to have happened, is not what either party thinks really happened. The main difference between them being that Luling is convinced he has found the truth; there is only one right way of reading the evidence, his way. In Wing's writing arguments are always "proven," or "clearly demonstrated," and counterarguments are long since "disproved," and unworthy of attention; nothing is put forward as a hypothesis for consideration, but as established fact resulting from an infallible method, while those who disagree are either ignorant or ill-motivated. Luling accuses the revisionists, Hawting in particular, of dismissing Muslim tradition as unreliable while at the same time relying on it when it suits them. But this is precisely what he does himself. According to Luling there has been a massive rewriting of history by the Muslim traditionists themselves, but he still relies on that tradition to confirm his favored theses about Christians in Mecca, the appearance and use of the Kaaba, Muhammad's biography, and so on. Moreover, if both parties are guilty of at once dismissing and using the tradition as it suits them, Luling is the more guilty in that he assumes that the truth of "what really happened" was generally known, but that a massive amnesia was somehow imposed by "orthodox" Islam, whereas the thesis of the revisionists is that "what really happened" was not known, which is precisely the circumstance that favored the invention of so many conflicting stories. The traditional Islamic picture of the origin of Islam was not a massive rewriting of events that were common knowledge, but the pious filling of an embarrassing void.

The problem with Luling's general method of arguing was well diagnosed by Hawting as a failure to distinguiush between evidence which unambiguously says what he wants it to say, of which there is hardly any, and evidence that can somehow be made to support his hypotheses:

Having argued, for instance, that the mushrikun were not really polytheists (and there is no way of proving this), Luling then goes on to show that they were Hellenistic, trinitarian and iconodule Christians. His proofs for this, though, seem to be entirely external, in the sense that they depend on other arguments, which themselves start from questionable presuppositions; the identity of the mushrikun is not clearly or persuasively shown from the information pertaining to them in the Muslim traditions.... Again, the traditions about al-Hijr do not themselves lead to Luling's conclusion that it was once the apse of a church.

This conclusion ... is rather imposed on the material and rests on ideas which depend on other evidence and the inferences drawn from it. In short, it seems to me, that there is a failure to maintain a proper relationship between the author's ideas and the evidence, and one is involved in a sort of circular argument where conclusions about one body of evidence constantly depend upon conclusions about another.36

If these strictures apply to Luling's arguments for the historical background of early Islam, they must apply also to his reasoning on the nature of the material he thinks he finds in the rasm of the ur-Koran. As Hawting goes on to say, describing Luling's methods of restoring this supposed urKoran:

In general the method of restoration consists of the reinterpretation of certain key words, slight changes in the vocalisation or consonant structure of the text which has come down to us, and sometimes in the relationship which has traditionally been accepted between certain sentences or clauses. Again, it seems to me that the argument is essentially circular and that since there is no way of controlling or checking the recomposed Ur-Qur'an, there is a danger that it will be recomposed to suit one's own preconceptions about what one will find in it.37

In short, Luling's case for finding Christian hymnody in the ur-Koran is unverified and unverifiable. Perhaps this is overstating the case against. The material in the Koran has to have come from somewhere and, as we have seen, Hawting himself admits that it cannot all be attributed to one individual. If Luling is right about central Arabia being full of Christians, or, more likely, if the revisionists are right about the milieu of early Islam being the predominantly Christian Hellenistic north, the likelihood surely is that at least some of the underlying material in the text of the Koran would be of Christian origin. There is evidence that this could have been the case that does not involve any of Lining's dubious speculations and circular arguments, as we hope to indicate in the conclusion.

Without his grand plans for the spiritual future of mankind, and without his assumption of a large Christian presence in central Arabia, Luling's main thesis on the ur-Koran is not implausible. The main problem is his insistence upon Mecca and the Hijaz as the location where it all happened. Once again the crucial objections to his postion have been well stated by Hawting when he points out that Luling

accepts that all of the material which Muslim tradition preserves and applies to the Ka'ba at Mecca did in fact originate with regard to the Meccan Ka'ba.

In other words, he accepts the view that in the time of the Prophet Islam took over the Meccan sanctuary, and that all the many reports, often difficult to understand and sometimes contradictory, about the develoment of the sanctuary in the early Muslim period, in fact refer to the sanctuary at Mecca. If one is starting from a sceptical position, then there is no reason why one should accept this, and it seems to me to make more sense to envisage that some of the sanctuary material did not originally relate to the Meccan sanctuary but that it has been adapted by Muslim tradition and made to apply to Mecca at a secondary stage.

Again, the general point is that if one believes, as Luling does, that Muslim tradition has deliberately falsified the historical record, why should we accept the main framework which that tradition provides?38

This is not only the main argument against Luling, it is the main revisionist argument against the traditional Muslim picture of the origin of Islam, and all the Western scholarship that has taken it at face value.

A final point worth noting is one upon which Luling is more radical than the revisionists. In his newly composed preface to the English edition of his book on the ur-Koran, he criticizes Wansbrough for taking the Islamic tradition seriously with regard to there having been an oral tradition in the transmission of Koran texts. Basing himself on the work of Fritz Kernow and others, Luling declares that there was in principle no oral tradition at all, neither for old Arabic poetry nor for the Koran, and that Wansbrough, in using this traditional idea in his revisionist theory, goes against his own principles. If Luling is right about this it would explain the role of the qussas, the storytellers of early Islam, who appear to have had a free hand in inventing the provenance and meaning of Koran texts; they could say whatever they liked because there was no one who knew any better in order to contradict them. Once a text was known to be "Koran" it would be remembered for use in prayers, sermons, and polemic, but it was not remembered as issuing from the mouth of "the Prophet."

CONCLUSION

That there was a pre-Islamic Koran is something about which Luling and the revisionists can agree. Speculating on the reasons for there being so much that is obscure in the text of the Koran, not only to us but also to the earliest Muslim commentators, Michael Cook proposes two possible explanations:

One is to suppose that the materials which make up the Koran did not become generally available as a scripture until several decades after the Prophet's death, with the result that by the time this happened, memory of the original meaning of the material had been lost. The other is to speculate that much of what found its way into the Koran was already old by the time of Muhammad [our italics]. The two approaches do not exclude each other. Each has its attractions, and each has its problemsnotably the need to reject much of what our narrative sources tell us.39

As far as the revisionists are concerned, what most needs to be rejected from the narrative sources is the origin of Islam at Mecca and Medina in the Hijaz.

Remaining true to the fundamental thesis of Hagarism, Cook draws attention to several factors that indicate a non-Hijazi region as the back ground for Islam. The Koran often speaks of seafaring, when there is no record in the sira of Muhammad ever having been to sea. In this connection it is curious to note that the most plausible etymology of "Quraysh," the Prophet's tribe, is that it means "shark," or at least some kind of fish.40 Why would a central Arabian tribe be named after a fish? There is also the curious fact that XXXVII.133-38 refers to the site of the destruction of Sodom and Gomorrah (Genesis 19), which is traditionally located on the borders of Palestine, as a place that "you," presumably meaning Muhammad, pass by "in the morning and the night." Why would he do this if he was centered in the Hijaz? It is from this same region that we have the important testimony of the fifth-century Christian historian Sozomenus, who tells of a group of Arabs who had come into contact with Jews and learned from them their biblical descent from Abraham's son Ishmael, and who thereafter adopted Jewish laws and customs.41 As Cook acknowledges, this would not amount to Islam as we know it, but it provides a plausible location for the acquisition of a fusion of monotheism and Arab identity, long before the Koran and the Prophet.

In conformity with the drift of these facts is the archaeological evidence from the Negev desert, marshaled by Y. D. Nevo and J. Koren. They note that, despite extensive excavations in the Syro-Jordanian desert, the Arabian peninsula, and especially the Hijaz, no remains of pagan sanctuaries from the sixth and seventh centuries have been found, and that in fact there are very few indications that the Hijaz was much inhabited in the fifth to seventh centuries C.E., and certainly not by any people using any form of classical Arabic. On the other hand, there is evidence for an active pagan cult in the central Negev that existed from Nabatean times down to the start of the `Abbasid period in the second half of the second/eighth century. Moreover, many features of the material remains of this cult resemble the Muslim descriptions of the shrine at Mecca, while there is also literary and epigraphical evidence of a monotheistic cult of Abraham in the Negev. The largest pagan sanctuary at Sde Boqer shared its site with a sect of monotheistic Arabs whose many rock inscriptions in classical Arabic in the Kufic script reveal a Judaeo-Christian flavor. These features conform to elements in the traditional account of early Islam, for which they could well have served as models, and then transferred to Mecca and the Hijaz in the literary sources, as suggested by Hawting and others.42

We have already seen how Luling cites J. S. Trimingham's Christianity Among the Arabs in Pre-Islamic Times as confirming his thesis of widespread Christianity in the Hijaz and central Arabia, and how, in fact, it does nothing of the kind; not only that, but this work in reality provides much evidence for the revisionist thesis of a northern origin for Islam.

To take a few points more or less at random, we may note the following: "The particular form Allah need not be thought of as an Arab contraction of al-`ilah, but as a proper name for God, coming through Aramaic, as the absolute state of Alaha, the most used form in Syriac (see Payne Smith, Thesaurus Syriacus, I. col. 195-96). This was the name of God among Nabateans and other north-Arabian inscription writers. Its usage in compound names like whb'lh necessitates vocalising as Wahb Allah, "Gift of God," not "gift of the god," showing the divine name to be Allah and not a contraction of al-'ilah."43

While discussing the spread of Christianity in the Syro-Arab region, Trimingham refers to two tribes that are mentioned in a number of inscriptions: the Khaseteroi and Audenoi of the villages of Merdoch and Rimea in the Hawran east of Palestine. Of the Khaseteroi he says: "It has been suggested that they were a clan of the Azd who migrated to southern Syria around A.D. 200 under chiefs Jafna, Mujalid, `Aws and Khazraj (clan names). The last two groups of `Aws and Khazraj returned later to settle in the Hijaz, but other clans stayed in Syria to make their reappearance as the Banu Ghassan."44 How do we know any of this? If it is only from Muslim sources, it must be regarded as suspect. The 'Aws and Khazraj were parties to the notorious "al Aqaba" and "Constitution of Medina" agreements. Why could they simply have never moved from the Hawran region on the borders of Palestine, and this be the real location where the "Constitution of Medina" was contracted, "Medina" being any town in the region?

When discussing the heads of the Tanukh tribe, Trimingham mentions Jadhima (or Judhaima) al Waddah al Abrash, who, Tabarl (1.750) reports Ibn Kalb! as saying, ruled the territory that "lay between Hira, Anbar, Baqqa, Hit, and district."45 Hira, Anbar, and Hit all lay along the Euphrates, but the location of Baqqa appears to be unknown. Could this be the location of the Baqqa mentioned in the Koran (111.90), which most exegetes are eager to identify with Mecca in the Hijaz, but which really lay somewhere between Anbar and Hit, on the Euphrates? It may also be significant that the Arabic script known as Kufic was in common usage among the Ibad, the Arab Christians of al Iraq al Arabi. Tabari (Ta'rakh, I. 206) reports that when the Arabs of that region were asked, "From whom did you learn letters?" they replied, "We learned the script from the lyad." Trimingham goes on to say: "It is claimed that the script through which the Qur'an was preserved was taken from the Christian Arabs of Hira and Anbar, carried to Duma and from there to Mecca."46 Perhaps, in reality, the last part of that journey was unnecessary.

The Syriac influence on the Koran has long been noted. A crude word count of the foreign vocabulary shows a predominance of Syriac words, closely followed by Aramaic;47 the latter was the Middle Eastern lingua franca, and Syriac the language of the semi-independent Syriac church, split into the factions of Monophysites and Nestorians. Of particular significance is the influence of Syriac on the Koran's religious terminology, almost all of which is derived from that language, including the words for God, soul, last judgment, salvation, sacrifice, resurrection, heaven, angel, priest, Christ, and prayer; this influence also appears in such theological expressions as "light upon light," and upon semibiblical quotations and biblical events and facts. In comparison, the Jewish influence upon the religious vocabulary of the Koran is negligible. Alphonse Mingana, who is the source of these observations, also remarks on the fact that, given the proximity and alleged intimacy between the Hijaz and Abyssinia, "the only Ethiopic religious influence on the style of the Kur'an is in the word hawariyun, `Apostles.' "48

The Syriac church regarded itself as descended from Saint Thomas rather than Saint Peter, and had many distinctive features that may account for the odd character attributed to Christianity in the Koran. From the beginning the church emphasized that there is one God, the God of the Old Testament, who sent the Messiah. Doctrinally it was divided between the factions of the Nestorians and Monophysites, regarded as heresies by the parent Greek church. Monophysitism, the idea that Christ had one nature, not two as defined by the Council of Chalcedon (451), produced, in the territory of the Arab tribe of the Banu Ghassan, a reaction in the form of a farther heresy, tritheism, the idea that since the divine nature belongs to each of the hypostases (Father, Son, and Holy Ghost) there are in reality three gods. This could have contributed to the Koranic notion that Christians worship three gods (e.g., V.116).49 This point gains credi bility when the widespread presence of the Diatessaron in the Syriac church is taken into account.

The Diatessaron (Gr. "through [the] four [gospels]") is a gospel harmony created about the year 172 by the Mesopotamian scholar Tatian.50 Its importance lies in the fact that it is the earliest extant collection of second-century gospel texts, and is probably the form in which the gospels first appeared in Syriac, Latin, Armenian, and Georgian. The importance of this work for the study of the Koran rests on the fact that, in addition to the four canonical gospels, it incorporated Judeo-Christian elements, most probably from the Gospel of the Hebrews.51 It is from this source incorporated in the Diatessaron that the Koranic notion of tritheism in the form of God, Jesus, and Mary most probably derives, since it is in this gospel that an assimilation of Mary to the Holy Spirit took place. Origen, in a commentary on the Gospel of the Hebrews, tells us that it depicts Jesus as saying: "Even now did my mother the Holy Spirit take me by one of mine hairs, and carried me away unto the great mountain Thabar." Such an expression probably derived from the fact that in Hebrew the word for "spirit" is feminine.52

The Diatessaron sets sections of Matthew, Mark, and Luke into John's framework. It begins with the Johannine prologue about the Word, continues with Luke's account of the angelic annunciations to Zechariah and Mary, and follows with Matthew's report of how Mary was found to be pregnant by the Holy Spirit. This is similar to the Koran's view of Jesus as God's Word, which he casts into Mary, and with the apparent identification of the angel Gabriel with the Holy Spirit. It is also worthy of note that the Diatessaron says of Mary that no man had known her, rather than that she did not know a man (Luke 1:34), making the male the active partner, reflected in 111.47 and XIX.20.53

Familiarity with the Diatessaron in the hands of Nestorian and Monophysite Christians may well have given the impression to nonChristian Arabs that the gospel (injil) was one book. In addition, it would explain several features of the Koranic view of the Old and New Testaments, particularly the fact that the Old Testament personages are, with few exceptions, of the patriarchal period. Adam, Abel, Noah, Abraham, Lot, Isaac, Jacob, Moses, David, Elijah, Elisha, and Jonah, are all mentioned in the Koran and the Diatessaron with either complete identity of spelling or a close relationship.54

Knowledge of the Diatessaron, or even the canonical gospels in their Syriac interpretation, is not of course sufficient to account for all the features of Jesus and Christianity to be found in the Koran; much is obviously drawn from the apocryphal infancy gospels such as the Protevangelium of James, the Infancy Story of Thomas, and the Gospel of PseudoMatthew. Nor is it possible to say that the influence is more Monophysite than Nestorian. As Neal Robinson has said of the Koran: "In agreement with the Nestorians it stresses the full humanity of Jesus and Mary. In agreement with the Monophysites it emphasises that God is One. In opposition to both it rejects all Trinitarian language and all talk of divine Sonship no matter how it is understood."55 What is clear, however, is that the milieu in which much of the Koran took shape was strongly Christian, albeit of a somewhat less than orthodox nature.

Such a conclusion is hardly new or radical.56 In the days of the old scholarship, before the appearance of the revisionists, when the Hijazi origin of Islam and the traditional biography of the Prophet was more or less taken for granted, all such conjecture as those above would have been accompanied by speculations as to where, when, and how Muhammad could have acquired all the diverse and obscure information in the Koran, usually on his trading trips to the north. Gunter Luling, in a move that is at once revisionist and traditional, solves the problem by shifting all those influences into the Hijaz, making Mecca a cosmopolitan center for practically every religious faction existing in the seventh-century Middle East. Since the evidence for this is either nonexistent or susceptible to alternative explanations, the solution of the revisionists appears to hold the field.

Without the Hijaz, and without the Prophet as sole source for the Koran text, many new hypotheses become possible, including that of preIslamic hymnody in the rasm of the ur-Koran. If the milieu of Koran formation was antiorthodox Christian, as we have seen that it probably was, Luling could even be right that the Christian element underlying the Koran was not straight forward Hellenic orthodoxy, but was antitrinitarian and considered Jesus as both created and angelic. The provenance of such views, however, is already known in the arc of the Fertile Crescent. To postulate an unverified and unverifiable location in the Hijaz appears an unnecessary extravagance that jeopardises the acceptance of valuable insights. Luling needs the revisionists, and the revisionist should regard Luling as one of their own.

NOTES

1. See G. Luling, "Preconditions for the Scholarly Criticism of the Koran and Islam, with some Autobiographical Remarks," Journal of Higher Criticism 3, no.I (spring 1996): 81.

2. Ibid., pp. 89-90.

3. Ibid., p. 95.

4. Ibid., pp. 96-99.

5. See also M. Cook, The Koran: A Very Short Introduction (Oxford, Oxford University Press, 2000), chap. 7, and Y. H. Safadi, Islamic Calligraphy (London: Thames & Hudson, 1978).

6. First and foremost that of Joshua Blau, A Grammar of Christian Arabic (Louvain, 1966), based mainly on South Palestinian texts of the first millenium.

7. Also Noldeke, Geschichte des Qorans, I, pp. 78 ff.

8. See Guillaume, trans. The Life of Muhammad (Oxford: Oxford University Press, 1978), pp. 104-107.

9. See Ibn Hisham, Sira, ed. Wustenfeld, vol. 2, p. 67, lines 4-16. Ibn Hisham quotes this "seventeen" as a tradition of ahl at tafsir, "people of Koran exegesis." The late Koran commentary Tafsir al Jaldlain also refers to this variant reading "seventeen" in Sura LXXIV.30.

10. For a preliminary examination of some of the problems involved in the idea of nineteen in the Koran see Ibn al Rawandi, Islamic Mysticism: A Secular Perspective (Amherst, N.Y.: Prometheus Books, 2000), pp. 201-24.

11. J. S. Trimingham, Christianity Among the Arabs in Pre-Islamic Times (London: Longman, 1979), pp. 249-58.

12. Ibid., pp. 276-77.

13. Irfan Shahid, Byzantium and the Arabs in the Fifth Century (Washington, D.C.: Dumbarton Oaks, 1989), p. 387.

14. For a brief overview of these developments see Ibn al Rawandi, "Origins of Islam: A Critical look at the Sources," in The Quest for the Historical Muhammad, ed. Ibn Warraq (Amherst, N.Y.: Prometheus Books, 2000, pp. 89-124.

15. J. Wansbrough, Quranic Studies: Sources and Methods of Scriptural Interpretation (Oxford: Oxford University Press, 1977), p. 179.

16. Ibid., p. 51.

17. See P. Crone, Meccan Trade and the Rise of Islam (Oxford, Blackwells, 1987), pp. 134-37, and P. Crone and M. Cook, Hagarism: The Making of the Islamic World (Cambridge: Cambridge University Press, 1977), pp. 24, 173 n. 30.

18. See A. Jeffrey, The Foreign Vocabulary of the Qur'an (Baroda: Oriental Institute, 1938).

19. J. Wansbrough, The Sectarian Milieu: Context and Composition of Islamic Salvation History (Oxford Oxford University Press, 1978), p. 99.

20. Ibid., p. 127.

21. Ibid., p. 124.

22. See P. Crone and M. Hinds, God's Caliph: Religious Authority in the First Century of Islam (Cambridge: Cambridge University Press, 1978), p. 99.

23. See F. Robinson, ed., The Cambridge Illustrated History of the Islamic World (Cambridge: Cambridge University Press), 1996, p. 13.

24. For Wansbrough's view of Crone and Cook see the remarks on Hagarism in The Sectarian Milieu, pp. 116-19, and the review of same in The Bulletin of the School of Oriental and African Studies 41 (1978): 155-56; also the review of Meccan Trade, in BSOAS 52 (1989): 339-40. For a reply, see Michael Cook's review of The Sectarian Milieu, in Journal of the Royal Asiatic Society (1980): 180-82.

25. See Whelan, E., "Forgotten Witness: evidence for the early codification of the Qur'an," Journal of the American Oriental Society (January-March 1998): 5.

26. See M. Cook, Muhammad (Oxford: Oxford University Press, 1983), p. 74, and Crone and Cook, Hagarism, chap. 1.

27. P. Crone, Slaves on Horses: The Evolution of the Islamic Polity (Cambridge: Cambridge University Press, 1980), p. 15.

28. Ibid., p. 23.

29. M. Lecker, Muslims, Jews and Pagans: Studies on Early Islamic Medina (Leiden: E. J. Brill, 1995).

30. Ibid., p. 49.

31. See P. Crone, review of Muslims, Jews and Pagans, JSS 42 (1997): 182-85

32. See C. C. Torrey, Jewish Foundations of Islam (New York: Ktav Publishing House, 1967), and A. Katsch, Judaism in Islam (New York: New York University Press, 1954).

33. G. R. Hawting, The Idea of Idolatry and the Emergence of Islam: From Polemic to History (Cambridge, Cambridge University Press, 1999).

34. G. R. Hawting, review of Die Widerentdeckung des Propheten Muhammad, JSS 27 (1982): 109.

35. Ibid., p. 110.

36. Ibid., pp. 110-11.

37. Ibid., p. 111.

38. Ibid., p. 112.

39. Cook, The Koran: A Very Short Introduction, pp. 137-38.

40. See Jeffrey, The Foreign Vocabulary of the Qur'an, pp. 236-37.

41. See Shahid, Byzantium and the Arabs in the Fifth Century, pp. 167-80, 332-49.

42. See Y. D. Nevo and J. Koren, "The Origins of the Muslim Descriptions of the Jahili Meccan Sanctuary," Journal of Near Eastern Studies 49 (1990): 23-44.

43. Trimingham, Christianity Among the Arabs, p. 251 n. 14.

44. Ibid., p. 78.

45. Ibid., p. 154.

46. Ibid., p. 227.

47. See Jeffrey, The Foreign Vocabulary of the Qur'an, pp. 298-311. Alphonse Mingana estimated that "taking the number 100 as a unit of the foreign influences on the style and terminology of the Kur'an Ethiopic would represent about 5 per cent. of the total, Hebrew about 10 per cent., the Graeco-Roman languages about 10 per cent., Persian about 5 per cent., and Syriac (including Aramaic and Palestinian Syriac) about 70 per cent." "Syriac Influences on the Style of the Kur'an," Bulletin of the John Rylands Library 11 (1927): 80 (chapter 3.1 of this volume.)

48. Ibid., p. 87.

49. Trimingham, Christianity Among the Arabs, pp. 183-84.

50. See the essay "Tatian's Diatessaron," by W. L. Pearson, in Ancient Christian Gospels: Their History and Development, ed. H. Koester (London: SCM Press, 1990), pp. 403-30.

51. Ibid., p. 412 n. 2.

52. M. R. James, The Apocryphal New Testament (Oxford: Clarendon Press, 1969), p. 2. See also G. Parrinder, Jesus in the Qur'an (London: Sheldon Press, 1965), pp. 134-37.

53. See N. Robinson, Christ in Islam and Christianity (New York: SUNY, 1991), pp. 18 ff.

54. See J. Bowman, "The Debt of Islam to Monophysite Christianity," Nederlands Theologisch Tijdschrift 19 (1964-65): 177-201.

55. N. Robinson, Christ in Islam and Christianity, p.20.

56. See R. Bell, The Origins of Islam in Its Christian Environment (New York: Macmillan & Co., 1926).

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