Missionary, dilettante or visionary?

The source of this surprising statement is the book under review.^[2] If the writer is right, he places a bomb under Islam that is comparable to the effects of Biblical textual criticism to Christianity. Understandably the author's name 'Christoph Luxenberg' is a nom de plume of a professor in Semitic languages at a German university, according to articles in the press.

The statement 'not virgins but grapes' is only a small side step in a book that argues a theory that reaches much further, this theory has hardly enjoyed any attention in the press. According to Luxenberg, the Qur'an was not written in classical Arabic but in a mixed Arabic-Syriac language, the traders' language of Mecca and it was based on Christian liturgical texts. When the final text of the Qur'an was codified, those working on it did not understand the original sense and meaning of this hybrid trading language any more, and they forcefully and randomly turned it into classical Arabic. This gave rise to a lot of misinterpretations. Something like this can only have happened if there was a gap in the oral transmission of the Qur'anic text. That idea is in serious disagreement with the views of both traditional muslims and western scholars of Islam.

The traditional view

Muslims see the Qur'an as insurpassable and inimitable.^[3] This is based on Q 2:23:

If you have doubts about the revelation We have sent down to Our servant, then produce a single sura like it -enlist whatever supporters you have other than God- if you truly (think you can),

Say, 'Even if all mankind and jinn came together to produce something like this Qur'an, they could not produce anything like it, however much they helped each other'.

We know very well that they say, 'It is a man who teaches him,' but the language of the person they allude to is foreign, while this revelation is in clear Arabic (Q 16:103);

So We have revealed an Arabic Qur'an to you, in order that you may warn the capital city and all who live nearby (Q 42:7).

The language of the Qur'an is poetic, terse and sometimes extremely difficult to interpret. During the first centuries of Islam many scholars studied it's text, vocabulary, grammar, style and historical and biographical background in order to estabish how the Qur'an had to be understood. These activities resulted in numerous dictionaries, grammars and extended commentaries, tafsir.

Some critical remarks

In 1923 the al-Azhar university in Egypt issued a standard text that is now used worldwide. This standardisation too had it's reasons because despite Uthman's standardisation, several versions of the text of the Qur'an developed.

Discussions between traditional muslims and western scholars of Islam on this topic can run high. On the side of the faithful it is claimed that these only represent the various Arabic dialects or modes of recitation, the qira'at. All 7 (or 10, or 14) are considered canonical. On the side of scholarship however, differences at the level of meaning are recognised.

A good example are the last three words of Q 2:10. In the Egyptian standard edition these are: بِمَا كَانُو ا يَكْذِبُونَ bima kanu yakdhibuna, 'for their persistent lying'. The standard text is based on the text of imam Asim († 744 AD) as transmitted by imam Hafs († 796 AD). It is used in the whole Islamic world, except in North Africa. Here the text of imam Nafi († 785 AD) as transmitted by imam Warsh († 812 AD) is used. In the latter, the same passage runs like this: بِمَا كَانُو ا يُكَذِّبُونَ bima kanu yukadhdhibuna, 'for what they denied'. 'Lying' or 'denying', there is a subtle difference...^[4]

Not all muslims deny the existence of these differences. A very charming example of the way these are dealt with is Q 5:6: You who believe, when you are about to pray, wash your faces and your hands up to the elbow, wipe your heads, wash your feet up to the ankles. The word أرْجُلَكُمْ arjulakum 'your feet' just like the words for 'face' and 'hands' is written in the accusative case, so it is seen as the object of the verb 'to wash', that only occurs once in this verse. According to the transmissions of the text by Ibn Kathir, Abu Amr, Abu Bakra and Hamza however it is أرْجُلِكُمْ arjulikum. This means the same, but is in the genitive case, just like the word for 'head'. In this version the genitive case is used because of the preposition 'over' (you heads) and because 'feet' has the same case a silent 'over' needs to be understood with '(and wipe over) your feet'. Now the question is: do the feet need to be washed before prayer or is wiping them sufficient? According to some Islamic interpreters both texts are correct, since under normal circumstances people will wash their feet before prayer, but where there is no water, wiping them suffices. The combination of the two different transmissions thus delivers the full revelation as intended by Allah.

Besides these variants early Islamic literature also mentions a lot of alternative readings that do not belong to the canonical texts. According to our sources these are all from Qur'anic texts that were destroyed in the wake of Uthman's standardisation.

Early Islamic linguists, and since the 19^th century also western scholars of Islam, have discovered loanwords in the Qur'an derived from various languages, mainly from Syriac. In the 7^th century this was the lingua franca of the Middle East, besides Greek, that was mainly spoken in the Byzantine empire. Mecca, Muhammad's home city was a trade settlement and Muhammad himself worked in the caravan trade for years. It is unthinkable that he had no knowledge of Syriac. So it is not surprising that Syriac loanwords are present in the Qur'an.

Die Syro-Aramäische Lesart

As 'difficult' he defines those passages that have been recognised as such by western translators or that have been called so by Tabari (839-923 AD) in his extensive tafsir.

A number of possibilities that could lead to a solution are then checked:

1. A plausible explanation indicated by Tabari himself, but overlooked by western translators;
2. A plausible explanation unknown to Tabari in the Lisan, the most extensive Arabic dictionary (there were no dictionaries yet in Tabari's time);
3. An unchanged reading of the Arabic, looking into the possibilities for it actually being a Syriac word;
4. A different placement of the diacritical dots (meaning the use of different consonants) that might result in another Arabic word;
5. A different placement of the diacritical dots that might lead to another Syriac word;
6. A literal translation of the Arabic into Syriac in order to see whether a Syriac expression or phrase has been literaly translated into Arabic. This is called a morphological calque (the German Fernseher is a morfological calque of television);
7. A correct Arabic expression, the meaning of which has now been lost, which may still have been preserved in old Syriac literature and lexica;
8. A correct Arabic expression written in Arabic script, but in Syriac orthography, which might thus be easily misunderstood.

That thesis can hardly be summarised in short, but the type of reasoning that Luxenberg uses, can be illustrated with a few examples that the interested layman can follow.

Lion or lame donkey?

Hook or alif?

In Q 68:13: coarse, and on top of all that, ill-bred, the word عُتُلًّ 'utull features. This is translated in various ways: 'ignoble', 'violent', 'greedy', in Shakir's, Yusufali's and Pickthal's translations respectively, are just some examples. Instead of عتل 'utull Luxenberg proposes عال 'al. The same Arabic word, or derivations of it, occur elsewhere in the Qur'an (10:83, 23:46, 38:75 and 44:31). They are translated as 'domineering', 'arrogance', 'high and mighty' and 'tyrant'.^[8] It fits perfectly in this passage and it is correct Arabic, which 'utull isn't, according to Luxenberg. Here the ﺘ t (without the dots, just like the nun in the previous example) was allegedly used to indicate an ا a. This was not understood in later times and the dots were added, so it became a 't'.^[9]

Seven Sleepers

Luxenberg suggests two mistakes: the م m at the end of the word is a misreading of د d. In the oldest script that was used for the Qur'an, hidjazi, this is a very likely mistake. Secondly, instead of the ﻴ i an ا a needs to be read. This latter exchange we have seen twice before. Thus the word becomes: الرقاد ar-ruqad 'sleep': do you find the Companions in the cave and the sleep so wondrous, among all Our other signs?^[11]

Mecca or Bakka?

Luxenberg proposes a Syriac reading: ܬܝܟܗ taykeh. This means 'which he has demarcated'. When this is written in Arabic letters in rasm, so without the diacritical dots: تيكه, taykeh, it is identical with ببكه bibakkah. This would indeed give a translation that seems more logical: The first House (of worship) to be established for mankind was the one which He has demarcated. It is a blessed place; a source of guidance for all people.^[12]

The changes that Luxenberg suggests, aren't limited to single words. Sometimes he rereads entire phrases and comes up with a reading that is more closely related to what non-muslims often consider 'the sources of the Qur'an'.

Abraham's sacrifice

فَلَمَّا أسْلَمَا و تَلَّهُ لِلْجَبِينِ fa-lamma aslama wa tallahu li-l-jabini, 'When they had both submitted to God, and he had laid his son down on his face,' (Q 37:103)

The word aslama is explained by Tabari in three different ways: 'to agree' (Abraham and Ishmael), 'submit' (to God's will) and 'give' (Abraham his son, the son himself to God). This already indicates a problem. A wide spectrum of possibilities can be found in various translations: 'to surrender', 'to submit', 'prononcer le Salam', 'resign'. Now, the previous verse already states that Ishmael submits himself and it sounds a bit odd to repeat that in this one. To complete the problem other, non-canonical, variants have been recorded in early Islamic literature: fa-lamma sallama.^[13] This reading is used by Luxenberg, because it agrees with Syriac ܫܠܡܘ shlemu '(when they) were ready'.

But the reinterpretation isn't done yet, because li-l-jabini 'on the forehead' also has it's difficulties. Tabari reads jabinan 'two temples' and concludes the forehead must be meant, since that is in between the two. It is easier to refer to jabin, a word that in Hebrew (גבינא), Syriac (ܓܒܝܢܐ) and Arabic (جبين) means'eyebrow'. This however seems a funny way to indicate a forehead. Luxenberg suggests to read the word without the diacritical dot under the first letter حبين habbin. This agrees with Syriac ܚܒܢ habbin 'the firewood' (literally: 'the burning ones').

Then he interprets tallahu as Syriac ܬܠܐ tla, which means 'to tie (down)', and has nothing to do with صرعه sara'ahu 'throw down', the word that Arabic commentators use to explain the passage.

Luxenberg's rereading is completed when he translates the preposition li- as 'upon', 'on top of': When they were finished, and he had tied him down on the firewood.^[14]

This reading doesn't differ vastly from the classical one, but it does have one important advantage: It fits the biblical story, as it was known in the 7^th century Syriac version, much better.

A lot of Luxenberg's arguments are built up like dominoes. If the first one falls, the rest has to come down with it. That is a weakness on the one hand, but on the other hand this type of domino-reasoning consistently delivers a reading that agrees much better with what are sometimes regarded as the Christian sources of the Qur'an.

Virgins or grapes?

If this reading is accepted, hur 'in cannot refer to virgins any more. Furthermore, the way in which hur 'in is traditionally translated requires some idiomatic acrobatics. Literally hur is a plural of the female adjective 'white' (حوراء hawra'). So it could refer to white women, but the object might just as well be a female word in the grammatical sense only. The word 'in is traditionally seen as the plural of the word for 'eye' and is translated by 'wide-eyed'. It is however not a usual plural and it only occurs in the phrase hur 'in. So in a sense it is a hapax. The literal translation 'wide-eyed whites' would then be a description of the virgins in paradise. In English translations this rather too literal choice of words is rendered as 'fair ones with wide, lovely eyes' (Pickthal), 'fair women with beautiful, big and lustrous eyes' (Yusufali) and even just 'Houris pure, beautiful ones' (Shakir).

Luxenberg doesn't deny that hur can mean'white' and 'in 'eye', but he proposes, through Syriac, a different reading of bi hur 'in as a consequence of the changed context: 'among/under fine/crystal clear whites'. That cryptic phrase works more or less in the same way that 'big cheese' can describe an important person in English. Luxenberg finds parallels for the metonymic use of the word 'white' in the sense of 'grape', both in Arabic and in Syriac. The 'eye' in his view is a metaphor to describe 'the appearance' of something. For this too he manages to find expressions in both languages, like 'the "eye" of a man' meaning 'his appearance' and 'the "eye" of something' in the sense of 'it's preciousness'.^[16]

This reading too requires some acrobatics in vocabulary, but Luxenberg succeeds in reinterpreting all 8 other passages in which the virgins feature, as well as the 3 passages that deal with the (male) youths in paradise (Q 52:24, 56:17-19 and 76:19). All these other 11 reinterpretations are consistent with his first rereading of Q 44:54.^[17] The added advantage, thinks Luxenberg, is that the texts about the 'grapes of paradise' not just fit the descriptions of paradise in 7^th century Syriac Christian texts much better; they also liberate the Qur'an from, in his eyes, shamefully erotic imagery of the hereafter.

Luxenberg reinterprets about 57 passages in his book. The conclusion he draws from this has several 'layers':

1. The application of his method delivers a more fitting interpretation for difficult passages than the traditional way of reading the Qur'an;
2. This proves that the Qur'an was written in a mixed Arabic-Syriac language, probably the traders' language of Mecca;
3. In view of the type of mistakes that were made, there must have been a hiatus in the oral transmission of the Qur'an;
4. Given the content of the 'improved readings' the conclusion that the Qur'an was based on Christian liturgical texts is justified, if it wasn't a Christian text to begin with.

Reactions

The first reactions on Islamic sites on the internet were exclusively based on the article in Newsweek. So only the virgins or grapes figured in them. Mostly these reactions were rather polemic and generally lacked any solid reasoning. The widespread notion among muslims that 'the west' or 'the orientalists' were just out to slander Islam and ideas about 'anti-islamic propaganda' played a more important role than the simple facts. Only Shibli Zaman, a writer on an Islamic apologetic website, has some knowledge of Arabic, Hebrew and Syriac. He emphatically proves that the literal meaning of the word hur is 'white' and 'in means 'eye' and that they don't mean 'grapes'. He doesn't realise however that Luxenberg doesn't deny these literal meanings at all. Zaman clearly didn't read the book.

Only much later western scholars of Islam entrusted their ideas on the subject to paper and these are not all positive. François de Blois points to a couple of grammatical mistakes in Luxenberg's book and calls him a 'dilettante'.^[18] His grasp of Syriac is limited to knowlegde of dictionaries and in his Arabic he makes mistakes that are typical for the Arabs of the Middle East. The latter is corroborated by the professor of Islam at Leyden University Hans Jansen: Luxenberg is not a professor at a German university, he is a Lebanese Christian. This would explain Luxenberg's 'Christian agenda': Christians from the Middle East have been involved in harsh religious debates with muslims for centuries.

Angelika Neuwirth, a scholar of Islam from Berlin, dedicates a few words to Luxenberg in an article and mainly emphasises the lack of interdisciplinary research. He does indeed raise many questions that could be answered by other disciplines, but it's not entirely fair to blame this on Luxenberg. Her remark that his conclusions are not fully justified by the results of his research is much more important. A word in Arabic looking similar to a Syriac word is to be expected, since the two are closely related languages. The use of loanwords also doesn't automatically mean that the unchanged and full meaning is borrowed as well. If qur'an was indeed derived from Syriac Qeryana 'lectionary', that still does not mean that the Qur'an actually was a Christian lectionary, at most it had a comparable function: text to be recited.^[19]

Assessment

It looks like a solid scholarly volume: often footnotes take a larger part of the page than the text and some pages are even completely filled with notes. No one can put it to Luxenberg that he doesn't document his claims. A nice advantage is his custom to provide almost every Arabic, Hebrew and Syriac word with a transcription. This way the layman can still get an idea of his line of thought.

Whether Luxenberg's readings are better than the traditional ones can to some extent even be decided by someone who does not speak Arabic by looking at the new translations he offers. At times these interpretations seem more logical, but some others seem to make very little difference. Rereading the otherwise unknown place-name Bakka to: 'that which he has demarcated' seems a good candidate for a succesful reinterpretation. But whether unbelievers shy away from the Qur'an 'like asses from a lion' or 'from a lame donkey' seems to make very little difference. On the one hand the text becomes more ironic, on the other hand: how disrespectful is it to compare the Qur'an to a crippled ass? It's only advantage is that you don't need to assume an Ethiopian loanword that cannot be found. Furthermore the word for 'lion' قسورة qaswara has a س sin, while the Syriac word that it should be based on has a ܨ tsade in it's root, which normally is a ص sad in Arabic. So this is a leap of faith, and Luxenberg needs a dialectal form to make it.

Even a succesful rereading like تيكه taykeh 'that he has demarcated' instead of ببكه bibakkah 'in Bakka' falls victim to the same method. Compare ببكه bibakkah with بمكه bimakkah 'in Mecca'. It is quite conceivable that a carelessly executed ﻤ is confused with a ﺒ, especially if the latter already existed in the shape with a diacritical dot. The whole problem with 'Bakka' has evaporated, and the translation of the passage remains unchanged.

Where Luxenberg tries to conform the passage about Abraham's sacrifice more correctly to biblical sources, he needs three pages to argue that the preposition li can be translated in the way he wishes: When they were finished, and he had tied him down on (li-) the firewood. This preposition generally does not mean 'on top of', 'upon'.

He brings on a plethora of arguments: first he produces a Hebraism which results in a second alternative translation: When they were finished and he had tied him down as (li-) a burnt offering; a Syriac passage where li is used in the sense of 'on' and finally an Arabic quote from the Qur'an: Q 7:143 When his Lord revealed himself to the mountain.^[20] For those who know the biblical story of God's revelation to Moses on mount Sinai, this text seems to need another preposition: When his Lord revealed himself on the mountain. But this explanation presupposes exactly what it tries to prove: namely that the Qur'an had Christian precursors.

One would expect that the correct reconstruction of a mistaken text has a certain self-evidence to it, that it speaks for itself. The philological tugging and squeezing that Luxenberg applies, doesn't plead in favour of that, on the contrary, the three explanations are partky mutually exclusive. It's still unclear whether li- is now Hebrew, Syriac or Arabic.

Luxenberg's main problem however is that his line of reasoning doesn't follow the simple and strict method that he set out at the beginning of his book.

The book would be much easier to follow if some examples had been clarified along the lines of this method. Which alternatives did Tabari give? What possibilities are there in the Lisan? Which Syriac words could offer an alternative? What other readings were possible with a different vowelisation? Which solutions were offered by rearranging the diacritical dots? This way the reader is also acquainted with alternatives that didn't make it and the reasons why.

Now we are confronted with only positive results. And those positive results are ordered in a discourse that has it's own structure and goal. Already at the beginning of his book Luxenberg creates the impression that the Qur'an was actually a reworked Christian text by putting forward that Qur'an actually means 'Lectionary', a collection of texts to be read in the Christian liturgy. Then follow examples of Syriac loanwords in the Qur'an. The last of which clearly indicate the influence of Christian Syriac texts. The end of the book features the reinterpretation of two entire suras, one of which (sura 108) is traced back to 1 Peter 5:8-9. This consistently Christian rereading of Qur'anic passages and the method of reasoning feeds the suspicion that Luxenberg is arguing towards a preset conclusion: the Qur'an is (actually) a Christian text. Given his arguments that seems a far reaching conclusion, too far even.

Tracing the Qur'an to a Christian source raises other objections. The number of parallels between the Qur'an and Jewish sources like the Targum for example are quite large. The Targum are Aramaic translations of the Hebrew Bible. For purposes of clarification and teaching, these were often expanded with non-biblical material.

The story of king Solomon and the queen of Sheba in the Qur'an for example (sura 27), agrees very well with the account that the Targum Sheni gives of the book of Esther (I.13), much better than with the Biblical version.^[21]

There are also philological indications for 'Jewish sources'. The Qur'anic word for 'hell' is jahannam. This is a Hebrew loanword (gehinnom). Had it been borrowed from Syriac the last 'm' would be missing. The occurrence of these kind of derivations indicates a much more complicated development of the Qur'an than Luxenberg supposes.

The Bahira-legend

Muslims know the story of the Christian monk Bahira, who recognised Muhammad in his early youth as 'the seal of the prophets', the last prophet. Christians have reworked that story and the (very old) wonder at the Christian character of the Qur'an to a Christian counterpart of the Bahira-story, in which Bahira wrote the Qur'an as a Christian text, with a few adaptations especially intended for Arabs. He then sent the book to Muhammad on the horns of a cow. The Qur'an is therefore also called 'the book of the cow'. That seems a bizarre detail. The second sura of the Qur'an is called al Baqara 'the Cow' and it is known that in some early Qur'anic collections, that of Ibn Masud for example, the first sura was missing. These Qur'ans indeed opened with 'the Cow'.

With his conclusions Luxenberg, without mentioning it himself, comes very close to a modern variant of this legend, which probably developed in the 8^th century. So the suspicion that his approach causes is not entirely unfounded. It looks like a new step in an age old apologetic tradition, that is felt by muslims to be 'anti-islamic'.

It is striking that a comparable legend developed among Jewish circles. 'Ten Jewish Sages' would have written the Qur'an according to this story, and this one too originated from the need to explain the great similarities between the Qur'an and Jewish sources.

Sanaa

One of the peculiarities discovered so far, is the use of the 'hook' of the letters ﺒ b, ﺘ t, ﺜ th, ﻨ n en ﻴ y to indicate the letter for which later the ا alif was used. For example: in some early Sanaa codices in Q 40:3: لاَ إلَهَ إلاَّ la 'ilaha 'illa 'There is no god but Him' the Arabic word اله 'ilaha is written as اليه that's with an 'i', if you read it in the classical way. This peculiarity could explain why Christian and Jewish words and names like 'Abraham', 'Satan' and 'Torah' in Arabic suddenly get an extra 'i': 'Ibrahim', 'Shaitaan' and 'Tawrija'.^[22] This discovery was published in 1999 but is unknown to Luxenberg, just as he seems to be unaware of much of the other literature on the subject. But even he, as we saw above, regularly argues that the 'hook' in early Qur'anic manuscripts must have been used for a long 'a'.

Certainly not everything Luxenberg writes is nonsense or too far-fetched, but quite a few of his theories are doubtful and motivated too much by a Christian apologetic agenda. Even his greatest critics admit he touches on a field of research that was touched on by others before and that deserves more attention. However, this needs to be done with a strictly scientific approach. In fact, his investigations should be done again, taking into account all the scholarly work that Luxenberg doesn't seem to know.

It is to be hoped that such research will be done without any apologetic agenda or anti-islamic sentiments in the background; and wouldn't it be nice if the results would keep people from hijacking a plane and in good spirits throw themselves into an inferno.

Notes