It might be useful to distinguish straight away what is new and what is not new in these theses. Muslim scholars of the classical period debated already the question of whether or not there is ‘non-Arabic’ (Aramaic, Persian, etc.) linguistic material in the Qur’an, whereby at least the more broad-minded authorities were content that there was; since God created all languages there is no reason why He should not have used words from different languages in His revelation. Modern linguistic scholarship established, certainly by the middle of the 19th century, that the Arabic language, both in the Qur’an and in other texts, contains a significant number of loan-words from several dialects of Aramaic (Syriac, Babylonian Aramaic, etc). Aramaic was the principal cultural language of the area between the Sinai and the Tigris for more than a millennium and it exercised a considerable influence on all the languages of the region, including the Hebrew of the later portions of the Old Testament. The Arabs participated in the civilisation of the ancient Near East, many of them were Christians or Jews, so there is nothing surprising about the fact that they borrowed heavily from Aramaic. But this does not make Arabic a ‘mixed language’. What is new in Luxenberg’s thesis is the claim that large portions of the Qur’an are not grammatically correct Arabic, but need to be read as Aramaic, inflectional endings and all. The Qur’an is thus not (grammatically) Arabic with Aramaic loan-words, but is composed in a jargon that mixes structural elements of two different languages. We shall examine the plausibility of this thesis in due course.

The second principal component of the author’s argumentation is that, since the later Muslims were unable to understand the Aramaic-Arabic jargon of their sacred book, they were forced arbitrarily to add diacritic signs to the text so as to make it into halfway comprehensible (classical) Arabic, thereby inventing a supposed oral tradition to justify this new reading. To rediscover the ‘original’ meaning we need to disregard the diacritical signs in the traditional text and find some other reading. This line of argument is also not new. It has been pursued in recent years in a series of articles by the North American Arabist J. A. Bellamy as well as in a (particularly bad) book by the German theologian Günter Lüling; strangely, none of these are mentioned in Luxenberg’s bibliography. This too will be discussed in the course of the present review. In any case, a book that announces already in the preface (p. ix) that its author has chosen not to discuss ‘the whole [sic!] of the relevant literature’ because this literature ‘makes hardly any contribution to the new method put forward here’ is one that poses, from the outset, questions about its own scholarly integrity.

But let us look at a few examples of the author’s ‘new method’. Because of the technical linguistic nature of this discussion I will use a consistent Semitist system of transliteration (in bold) and transcription (in italics) for both Syriac and Arabic, a system differing both from the one used by the author of the book under review and from that otherwise followed by this journal.

One of the main planks of Luxenberg’s theory of the ‘Aramaic-Arabic mixed language’ is the contention that in a number of Qur’anic passages the final aleph of an Arabic word stands not for the Arabic accusative ending -an, but for the Aramaic ending of the determinate state (-â in the singular or -ê in the plural). On p. 30 the author discusses Q. 11:24 and Q. 39:29, where the ‘current Qur’an’ (‘der heutige Koran’) has hal yastawiyâni maQalan, ‘are the two similar as an example?’, the last word being an accusative of specification (tamyîz). The author thinks that the meaning is improved if is taken to be a ‘transcription’ of the Syriac plural mtl’ (maQlê) and that the sentence consequently means ‘Are the examples [plural!] similar [dual!]?’. Translated into modern Arabic (‘ins heutige Arabisch übertragen’), the Qur’anic sentence would then (supposedly) be hal yastawiyâni l-maQalâni. Most first-year students of Arabic are sure to know that this is neither classical nor modern Arabic, but simply wrong. But even without this lapsus, it can hardly be claimed that the ‘Syro-Aramaic reading’ offers any improvement in the understanding of the Qur’anic passages.

On p. 37 the author discusses Q. 61:61 innanî hadânî rabbî ’ilâ sirâtin mustaqîmin dînan qiyaman, which, if dînan qiyaman is in fact an accusative of specification, would need to be translated by something like ‘verily, my Lord has directed me to a straight path in accordance with a firm religion’, or, if we assume a mixed construction (hadâ construed first with the preposition ’ilâ and then with double accusative), it could mean ‘..... to a straight path, a firm religion’. Our author’s proposal is that the syntactical difficulty of the latter rendering could be alleviated by taking not as an Arabic accusative but as Syriac dyn’ qym’(dînâ kayyâmâ), which he translates as ‘a firm belief’ (‘feststehender, beständiger Glaube’). But in so doing the author overlooks the fact that, unlike Arabic dînun, Aramaic dînâ does not actually mean ‘belief, religion’, but only ‘judgement, sentence'. Arabic dîn, in the meaning ‘religion’, is not borrowed from Aramaic but from Middle Persian dên (Avestan daênâ-).

On pp. 39ff. the author connects the problematic Qur’anic term hanîfun with Aramaic hanpâ, ‘pagan’, and specifically with the Pauline doctrine of Abraham as the paradigm of salvation for the gentiles. I have recently argued along similar lines in a lecture delivered in the summer of the year 2000 and eventually published in Bulletin of the School of Oriental and African Studies 65 (2002), pp. 16-25, but differently from ‘Luxenberg’ I did not fail to mention that the same suggestion had been made long ago both by Margoliouth and by Ahrens, nor did I commit the absurdity of claiming (as our author does on p. 39) that Arabic is a ‘Wiedergabe’ of Syriac hnp’, despite the fact that the Arabic form has an -i-, of which there is no trace in Syriac.

But in the eyes of our author, the Aramaic suffixes -â and -ê are ‘represented’ in the Qur’an not only by alif, but also by ha’. Thus [p. 34] Arabic (xalîfatun) is ‘the phonetic transcription’ of Syriac hlyp’ (hlîfâ). Unfortunately, no reason is given for why, in this ‘phonetic transcription’, the Aramaic laryngeal h is not ‘transcribed’ by the phonetically identical Arabic laryngeal h, but by x.

On p. 35 the author discusses the Qur’anic word for ‘angels’ (plural), for which the traditional reading is malâ’ikatun. The author thinks that this is really the Syriac word for ‘angels’, which he spells, in Syriac script, (correctly) as ml’k’, and which he transcribes (wrongly) as malâkê; in fact, the correct Syriac vocalisation is malaxê (the first aleph being left over from the older form *mal’ax- ) In any case, neither the Syriac spelling, nor the correct vocalisation, nor even the author’s erroneous vocalisation explains the -y- of the Arabic plural. The author then goes on to claim that the postulated ‘Syro-Aramaic pronunciation’ of the Qur’anic plural is made certain (‘gesichert’) by the ‘modern Arabic’ of the Near East malâykê. This is a big jumble. In fact, the Arabic singular mal’akun or malakun is in all likelihood borrowed from Aramaic mal’ax- or malax-, but the plural malâ’ikatun is a perfectly regular Arabic formation, and is represented graphically by , with the usual Qur’anic defective spelling of internal -a-. The cited ‘modern Arabic’ (more correctly Levantine) form is the expected dialectal reflex of the classical pausal form malâ’ika(h), with palatalisation (’imalah) of the final -a to -e (I see little justification for the transcription with long -e), and has nothing to do with the Syriac plural malaxê.

But once the ‘mixed-language’ status of the Qur’an has been postulated, the author evidently thinks it possible to take any Arabic word that vaguely resembles something in Syriac and to determine its meaning not from the Arabic but from the Syriac lexicon. Thus on pp. 196ff. the very ordinary Arabic verb daraba, ‘to beat’, is quite arbitrarily said to derive from the Syriac verb traf, which, among other things, means ‘to beat, to move, to shake (wings), etc.’ Brockelmann, Lexicon Syriacum, p. 290, compares the Arabic verb tarafa, ‘to repel’. It seems unlikely that the Aramaic root should also have anything to do with Arabic daraba; the correspondences d/t and b/p(f) are certainly not the norm in Semitic cognates and would be perhaps even more surprising in the case of a loan-word. But this difficulty does not stop the author from assigning the meanings of the Syriac word to the various occurrence of daraba in the Qur’an.

Then, on p. 283 the author claims that the Arabic verb taga, ‘to rebel, tyrannise, etc.’ has, apart from the secondary , nothing Arabic about it, but is a ‘borrowing’ from Syriac tʿa. He then picks out of a Syriac dictionary the meaning ‘to forget’ and assigns this to the Qur’anic instances of taga. But the fact that the Arabic root has gayn where the Aramaic has ʿayin shows very clearly that the Arabic word is not borrowed from Aramaic, but that they are good Semitic cognates. Anyway, the usual meaning of Syriac tʿa is ‘to err, to be led into error, etc.’, although it can also mean ‘to forget’. So even if the Arabic verb were a borrowing from Syriac there would still be nothing compelling about the new meaning assigned to it by our author.

I shall quote one last example of the author’s ‘Syro-Aramaic reading’ of the Qur’anic text. In Q. 96:19 the last word of the sura is (i)qtarib, which has until now always been understand to mean ‘draw near’ (imperative). But our author [p. 296] thinks it means ‘take part in the eucharist’ (‘nimm an der Abendmahlliturgie teil’), since iqtaraba is ‘without doubt borrowed’ (‘ohne Zweifel .... entlehnt’) from the Syriac verb eQkarrab, which besides meaning ‘to draw near’, also means more specifically ‘to (draw near to the altar to) receive the eucharist’. In support of this he quotes (on p. 298, in the wake of some editorial mishap twice) a passage from the Kitabu 1-'agani in which the Arabic verb taqarraba is used unambiguously to mean ‘receive the (Christian) eucharist’. But this alleged confirmation scuppers the author’s argument. The (actually well-known) Christian Arabic technical term taqarraba is indeed a calque on Syriac eQkarrab, with the same stem formation, i.e., D-stem with prefix t(a)-. There is no good reason to assume that the same Syriac verb was ‘borrowed’ a second time as the (differently formed) stem iqtaraba.

The examples that I have quoted could be expanded manyfold, but they are perhaps enough. They illustrate what is actually the less controversial, or in any case less fantastical part of the author’s line of argument, the part, namely, in which he applies his ‘Syro-Aramaic reading’ to the actual traditional text of the Qur'an. But this book goes a lot further. Having established (as he thinks) that the Qur’an is composed in an Aramaic-Arabic ‘mixed language’ the author proceeds to juggle the diacritic points of the traditional text to create an entirely new Qur’an which he then attempts to decipher with the help of his (as we have observed, often very shaky) knowledge of Syriac. I do not really think that there is very much point in discussing this aspect of the book. There is no doubt that, without the diacritical points, the Qur’an is indeed an extremely obscure work and that the possibility of repointing affords virtually limitless opportunities to reinterpret the scripture, in Arabic or in any other language that one chooses. I think, however, that any reader who wants to take the trouble to plough through Luxenberg’s ‘new reading’ of any of the numerous passages discussed in this book will concede that the ‘new reading’ does not actually make better sense than a straight classical Arabic reading of the traditional text. It is a reading that is potentially attractive only in its novelty, or shall I say its perversity, not in that it sheds any light on the meaning of the book or on the history of Islam.

It is necessary, in conclusion to say a little about the authorship, or rather the non-authorship, the pseudonymity of this book. An article published in the New York Times on 2nd March 2002 (and subsequently broadly disseminated in the internet) referred to this book as the work of ‘Christoph Luxenberg, a scholar of ancient Semitic languages in Germany’. It is, I think, sufficiently clear from this review that the person in question is not ‘a scholar of ancient Semitic languages’. He is someone who evidently speaks some Arabic dialect, has a passable, but not flawless command of classical Arabic, knows enough Syriac so as to be able to consult a dictionary, but is innocent of any real understanding of the methodology of comparative Semitic linguistics. His book is not a work of scholarship but of dilettantism.

The NYT article goes on to state that ‘Christoph Luxenberg is a pseudonym’, to compare him with Salman Rushdie, Naguib Mahfouz and Suliman Bashear and to talk about ‘threatened violence as well as the widespread reluctance on United States college campuses to criticize other cultures’. I am not sure what precisely the author means with ‘in Germany’. According to my information, ‘Christoph Luxenberg’ is not a German but a Lebanese Christian. It is thus not a question of some intrepid philologist, pouring over dusty books in obscure languages somewhere in the provinces of Germany and then having to publish his results under a pseudonym so as to avoid the death threats of rabid Muslim extremists, in short an ivory-tower Rushdie. Let us not exaggerate the state of academic freedom in what we still like to call our Western democracies. No European or North American scholar of linguistics, even of Arabic linguistics, needs to conceal his (or her) identity, nor does he (or she) really have any right to do so. These matters must be discussed in public. In the Near East things are, of course, very different.

Back To The Text Of The Qur'an