Version classiqueVersion mobile
OpenEdition Books

Les non-dits du nom. Onomastique et documents en terres d'Islam

 | 
Christian Müller
, 
Muriel Roiland-Rouabah

Manuscrits et documents

Evidence, methodology and history: on the date and interpretation of an alleged Arabo-Islamic inscription of medieval al-Andalus-Moroco

David J. Wasserstein

Texte intégral

I.

  • 1 This article was written during my tenure of a visiting Fellowship at the Shelby Cullom Davis Cente (...)

1One of the features of a career as varied as it has been fruitful is the interest for and devotion to the character of the sources of our knowledge that Jacqueline Sublet has demonstrated over many years1. This trait appears also in the pleasure that she so visibly took from seeing, on an early visit to London, her name emblazoned everywhere in people’s windows — “Sublet”, announcing flats to let. Both reflect a delight in the details of texts, little words, half-unnoticed expressions which nevertheless mean much for the correct, even the intimate, understanding of the whole. These unnoticed details, each in itself, do not mean much ; it is only in the slow, gradual, erratic accumulation of these details, over many years, across many a study, produced by numberless researchers in the community of scholarship, that we begin, bit by bit, to understand that whole. In the lines that follow, dedicated to a respected colleague and a good friend, I should like to offer an essay, rigorous I hope in her own style, in how to interpret details in our sources.

II.

  • 2 See generally, and excitingly, Reynolds and Wilson 1991. We are probably not yet anywhere near bein (...)

2Our textual knowledge, our written sources from the past reach us not only as products of the workings of chance but also, in the great, the immense majority of cases, in copies. It is worth insisting on this point, for it is a truth whose importance is too rarely recognized. These copies themselves are copies of copies, through a generally unknowable number of “generations” of copies and copyists. The originals — I speak here of the classical and medieval periods of Islamic history — are with very few exceptions lost. This is the case also with virtually all our literary texts from classical Antiquity2. For researchers this is both a problem — how, through the thicket of the errors of many centuries, can we come to know the exact words, the ipsissima verba, of a Plato, a Cicero, a Ṭabarī or a Mutanabbī or an Ibn Ḫaldūn — and at the same time a challenge that makes the careers of many a scholar — in the task of trying to re-constitute the texts as they came forth from the pen of their authors.

  • 3 Including now an administrative document from ancient Egypt containing a note of approval by Cleopa (...)

3However, there is one class of sources that is different : these are the texts which reach us in autographs. Certain texts, tiny in numbers, do reach us in the autographs of their authors3. Similar here is the case of coins, whose inscriptions, perhaps now worn down and scarcely legible, nevertheless without a doubt still say today exactly what they said centuries ago on the day when they were struck. This is true also of pieces of cloth, ṭirāz, ornamented with woven inscriptions including the titles of caliphs and others ; and it is true also of precious works of art, like those wonderful ivory boxes made in Fatimid Sicily or Egypt or in Umayyad and post-Umayyad al-Andalus. And it is true also of inscriptions on stone and wood.

4All of these texts reach us very differently from the rest, the great bulk of manuscripts, just as they emerged from the “pen” of their authors, simply because it is the original documents themselves that survive, not copies (of copies, of copies…). This fact confers on them a special and undeniable value as sources, because the fact that they are the originals means that we are (largely) freed from the need for that special task of textual criticism, attempting to work out the exact wording of the original, and reach at a single bound the real questions, historical rather than philological, represented in the contents of these inscriptions. Except…

5There is yet another class, or sub-class, of these inscriptions. These are inscriptions whose originals actually do not survive or to whose originals we do not have access, temporarily or permanently, but of which we do have transcriptions. One example of this is medieval descriptions, such as one comes across occasionally, of coin issues of which we have no surviving specimens. And there are also stones destroyed (one thinks at once of the Buddhas of Afghanistan, whether they bore inscriptions or not), or lost (one thinks instantly of the treasures of the Archaeological Museum of Baghdad), but which have been described in surviving written texts. For the latter, lost inscriptions, or inscriptions which remain to be found or identified, we have a fine collection in an Arabic manuscript in the Bibliothèque nationale de France in Paris, containing descriptions, which claim to be transcriptions, of a large number of ancient and not so ancient Greek inscriptions from Syria. The manuscript dates from the eighteenth or nineteenth century, the transcriptions of the Greek are visibly the work of someone wholly ignorant of that language and of the script in which it is written, and the contents of the texts thus “transcribed” in consequence almost inaccessible to the reader. We have to hope for the arrival of a researcher gifted in both Greek and Arabic, to say nothing of a visual imagination, who will take an interest in such curiosities.

6Stones are almost indestructible : it takes an effort to render them no longer carriers of their message. As for wood, the situation is rather different. We recall the minbar, commissioned by Nūr al-Dīn and the gift of Saladin, which was the victim of a fire at al-Aqṣā in Jerusalem in 1969. But wood wears away even without much effort, thanks to the passage of time and the touch of many hands. And thus we arrive finally at our cas témoin.

III.

  • 4 Lévi-Provençal 1931.

7In 1931 Évariste Lévi-Provençal published his collection of the Arabic inscriptions of Islamic Spain4. In that work he included one text, numbered 221, which was of Andalusi origin but was once located on the back of the minbar of the Mosque of the Qarawiyyīn in Fes. The original inscription, like the minbar itself, is now long lost, a millennium after its creation, but fortunately the text of the dedicatory inscription itself is preserved, through quotation in a number of literary texts of historical character. And here our problems begin, for the text of the inscription, both in the literary sources and as it is cited by Lévi-Provençal himself, presents certain details which do not agree either with each other or with what we know from other sources on the period, sources which enjoy a good and justifiable reputation for reliability.

8Here is the text as it appears in Lévi-Provençal :

بسمله تصليه٠ هذا ما امر بعمله الخليفة المنصور سيف الاسلام عبد الله هشام المؤيد بالله اطال الله بقاءه على يد حاجبه عبد الملك المظفر بن محمد المنصور بن ابي عامر وفقهم الله تعالى وذلك في شهرجمادى الاخرة سنة خمس وتسعين وثلاث مائة

9And here is his translation :

…ceci a été fait sur l’ordre du calife victorieux, du glaive de l’Islam, du serviteur d’Allāh, Hišām al-Mu’ayad bi-llāh — qu’Allāh prolonge sa durée ! —, sous la direction de son ḥāǧib ‘Abd al-Malik al-Muẓaffar b. Muḥammad al-Manṣūr Ibn Abī ‘Āmir — qu’Allāh très Haut les assiste! — et cela dans le mois de ǧumādā ii de l’année 395 (15 mars-12 avril 1005).

  • 5 Lévi-Provençal 1931, text n° 221, vol. i, p. 196, n. 3.

10Lévi-Provençal adds5 that he “has restored the date of 395 (instead of 375 : a confusion in the reading of the written forms of sab‘īn and tis‘īn, which are very similar to each other), on the basis of the chronology of the reign of the ‘Āmirid ḥāǧib al-Muẓaffar”. His argument for this is, at least for the moment, simple and irrefutable : the difference between the written forms of the words for “seven” and “nine”, and so for “seventy” and “ninety”, in Arabic is essentially a matter of interpretation of the distribution of scattered dots. And the chronology of the reign of ‘Abd al-Malik b. al-Manṣūr (392/1002-398/1008) indeed demands some such understanding of the date here.

  • 6 Ibid., text n° 221, vol. i, p. 196, n. 2.

11Lévi-Provençal took his text6 (see Lévi-Provençal 1931, ibid., n. 2) from :

(i) The Rawḍ al-Qirṭās of Ibn Abī Zar‘ ; and tells us that from there it is quoted in

(ii) the Zahrat al-Ās of al-Ǧaznā’ī in its entirety ; and in

(iii) the Ǧaḏwat al-Iqtibās of Ibn al-Qāḍī, in part.

We may usefully look at these in turn.

  • 7 See Idris 1971. Idris notes that we still lack a sound critical edition of the Rawḍ.
  • 8 Tornberg, Annales regum mauritaniae a condito idrisidarum imperio ad annum fugae 726 ab Abu-l Hasan (...)

12(i) Abū al-‘Abbās Aḥmad Ibn Abī Zar‘ al-Fāsī, the author of the work entitled al-Anīs al-Muṭrib bi-Rawḍ al-Qirṭās fī aḫbār al-Maġrib wa-ta’rīḫ madīnat Fās, lived in the second half of the seventh/thirteenth century, in the city of Fes itself, where he served as imām, and died between 710/1310 and 720/13207. Lévi-Provençal took the text of the inscription from the edition of the Rawḍ by Tornberg8. Here is the text as it appears in Tornberg :

بسم الله الرحمن الرحيم صلى الله على محمد واله وسلم تسليما هذا ما امر بعمله الخليفة المنصور سيف الاسلام عبد الله هشام المؤيد بالله اطال الله بقاءه على يد حاجبه عبد الملك المظفر بن محمد المنصور بن ابي عامر وفقهم الله تعالى وذلك في شهرجمادى الاخرة سنة خمس وسبعين وثلاث مائة.

13As can be seen, there are a couple of very slight differences between Lévi-Provençal and his source : first, he abbreviates the basmala and the taṣliya, rather than writing them out in full as Tornberg does ; and secondly, as he notes, he has corrected the reading of the decade of the date from “seventy” to “ninety”.

  • 9 In fact, although this edition (which names no editor), published by Dār al-manṣūr li-l-ṭabā‘a wa-l (...)

14A newer edition of the Rawḍ, published at Rabat in 19729, has the same text, but with some other very slight differences :

بسم الله الرحمن الرحيم وصلى الله على سيدنا محمد واله وسلم، هذا ما امر بعمله الخليفة المنصور سيف الاسلام عبد الله هشام المؤيد بالله اطال الله بقاءه، على يد حاجبه عبد الملك المظفر بن محمد المنصور ابن ابي عامر وفقهم الله تعالى، وذلك في شهر جمادى الاخرة سنة خمس وتسعين وثلاثمائة.

15As can be seen, in this edition we do not have quite the same version of the inscription’s wording : the word ṣallā is introduced by the particle wa- ; before the name Muḥammad we find the word sayyidinā ; after the word wa-sallam the newer edition omits the word taslīman ; the word ibn in the expression Ibn Abī ‘Āmir is written here with an initial alif (could this be because the editor took his text from a version in which the word “ibn” came at the start of a line, where convention requires an initial alif in this word ? Or because he saw the “ibn” here as the start of a sort of surname Ibn Abī ‘Āmir , and thus somehow demanding an initial alif?) ; and finally, the date is given in accordance with Lévi-Provençal’s correction, as 395 A.H., not 375.

16These differences are all linguistically meaningless in themselves they make no change to the meaning of the text (except for the last item in the list, concerning the date). However, they do have some meaning if we are concerned, not with the overall meaning of the passage but rather with knowing exactly what appeared in the original inscription. This is a different, a methodologically important question. Regardless of the content of a text, we need, as a prior condition, to be concerned with the correct knowledge and understanding of what it actually says. Who is right about what appeared there ? The copyist of the manuscript(s) upon which Tornberg relied for his edition, or the copyist of the manuscript(s) underlying the newer, 1972, edition ? Idris, as we have seen, pointed to the absence of a critical edition of this text, so, for the moment at least, the correct knowledge of what Ibn Abī Zar‘ wrote in this little passage must remain obscure. And since the minbar itself does not survive, we cannot use that object to confirm the exact wording of what the text of the Rawḍ says here. But it is, at this point, at least striking, and methodologically significant, that there should be disagreement, even if it is only this slight disagreement, about some of the minor details of the contents of such a short, and heavily formulaic, text.

  • 10 Beaumier (trans.), Roudh el-Kartas. Histoire des souverains du Maghreb (Espagne et Maroc) et annale (...)
  • 11 Huici Miranda (trans.), Ibn Abi Zar‘, Rawḍ al-Qirṭās, Valencia, 1964, vol. i, p. 115.

17The passage is translated in Beaumier’s version of the Rawḍ, but with errors, and we can leave this to one side10. It is also translated in the more recent version of the Rawḍ, by Ambrosio Huici Miranda, following the text of Tornberg (including the date of 375)11.

  • 12 Al-Ǧaznā’ī, Zahrat al-Ās, Algiers, 1923, p. 41-42 of Arabic text, with a translation at p. 99 of th (...)

18(ii) Lévi-Provençal tells us that from the Rawḍ this text was copied in its entirety into the Zahrat al-Ās of al-Ǧaznā’ī, a writer of whom little is known, except that he lived in the fourteenth century. Al-Ǧaznā’ī gives the text in the following form12 :

بسم الله الرحمن الرحيم وصلى الله على سيدنا محمد وآله وسلم هذا ما امر به الخليفة المنصور سيف الاسلام عبد الله هشام المويد بالله اطال الله بقاه على يد حاجبه عبد الملك المظفر بن محمد المنصور ابن ابي عامر وفقهم الله تعالى وذلك في ثمان وثمانين وثلاثمائة.

19As can be seen, there are differences between this version and what we have seen so far. The first one worth noting is that instead of saying amara bi-‘amalihi, here the text has simply amara bihi. Otherwise the only difference is in the date : before, the date was uniformly ǧumādā ii 375 (or 395). Now it is 388, and with no month mentioned. This is a radical departure from what we have seen so far, and it makes one wonder, among other things, whether, here at least, al-Ǧaznā’ī is being as slavish a follower of his source, Ibn Abī Zar‘, as used to be thought.

  • 13 Ed. ‘Abd al-Wahhāb ibn Manṣūr, Rabat, 1967, see p. 55.

20There exists a shortened version of this work, entitled Ǧanā Zahrat al-Ās, where the relevant passage occurs too13. There we find the same text as in Ibn Abī Zar‘, but again with some slight differences : ṣallā is introduced without any particle wa- ; sayyidinā does not appear before Muḥammad ; taslīman occurs at the end of the blessing ; instead of sayf al-islām we find sayf al-imām ; and at the end the date is given, as in the Zahrat al-Ās, in the form wa-ḏālika fī sanat ṯamān wa-ṭamānīn wa-ṯalāṯ mi’a.

21Some of these differences are obviously very slight, but the last two are more interesting. Sayf al-imām, in the shorter Ǧanā Zahrat al-Ās, instead of the phrase sayf al-islām of the Rawḍ, and the longer Zahrat al-Ās, seems possibly designed to help compensate for the absence otherwise in the inscription of the word imām ; and the date this time, here as in the Zahrat al-Ās, is completely different, 388. How can we explain these differences ? Can it be that 388 is wrong and 395 (or even 375) is right ? Or vice versa ? Or possibly that none of these dates is right ?

  • 14 Deverdun 1971, vol. iii, p. 814.
  • 15 See also Ben Cheneb and Lévi-Provençal 1922, p. 17, n° 69. Ben Cheneb and Lévi-Provençal list the n (...)
  • 16 Ibn al-Qāḍī, Ǧaḏwat al-Iqtibās, Fes, 1891-1892, p. 32.

22(iii) Our third source is Ibn al-Qāḍī. He was also from Fes, living from 960/1553 to 1025/1616 (al-Maqqarī, the author of the Nafḥ al-Ṭīb, pronounced the funeral oration over him). G. Deverdun, in his entry on him in the Encyclopaedia of Islam, describes him as a polygraph and author of highly regarded biographical collections14. Among these his Ǧaḏwat al-Iqtibās is a collection of Fāsī worthies but also a useful source on the topography of the town. According to Lévi-Provençal, this work was published in Fes in 1309/1891-189215. He offers the following as his text of the inscription16 :

بسم الله الرحمن الرحيم هذا ما امر به الخليفة المنصور سيف الاسلام عبد الله هشام المويد بالله اطال الله بقاءه على يد حاجبه عبد الملك المظفر بن محمد المنصور ابن ابي عامر وفقهم الله تعالى.

23Here we find the basmala, but no taslīm ; and the main text has, instead of bi-‘amalihi, merely bi-hi we have seen this before ; where the Rawḍ and the Zahrat al-Ās have sayf al-islām, and the Ǧanā Zahrat al-Ās has sayf al-imām, we find here sayf al-islām  this might seem to settle where it drew from. However, the text is lithographed in the Maghribī style, in such a way that one can easily see how a casual reader might read it, here or elsewhere, as sayf al-imām ; otherwise the text follows what is in the Rawḍ al-Qirṭās and in the Ǧanā Zahrat al-Ās, but with one big difference : here no date is given at all.

IV.

24Some of the differences between our versions of this text and the difficulties to which they give rise can be dealt with very easily. As we have seen, the difference between sayf al-islām and sayf al-imām is essentially a matter of the reading of a squiggle in the text the choice between the two, from a philological point of view, is a matter of purely personal preference  either is graphically, grammatically, linguistically acceptable. However, from a historical and a logical point of view, things are very different : the correct reading must be sayf al-islām, if only (though not only) because the reading sayf al-imām would make of “the caliph”, who is the subject of the sentence, the “sword of the imām Hišām al-Mu’ayyad”, that is to say, the sword of himself. The correct reading must therefore be sayf al-islām. Similarly, the difference between seventy and ninety in the date is a slight matter, as a reading, and can be dealt with very easily, but it is one of some considerable significance from a historical point of view.

  • 17 For the forms and contents of names and titles in Arabic, see especially Sublet 1991.

25Other problems are not so easy to resolve. The most obvious of these is the presence of the expression “al-Manṣūr” following the word “al-ẖalīfa” at the start of the text in all our versions. What does it mean ? How are we to understand it here ? We could, à la limite, argue that “al-Manṣūr” is not here a formal title but merely a descriptive adjective = “the Victorious”, but the dividing line between the two, adjective and title, in such contexts in Arabic is so slight that this cannot be an objection to what follows17. Written in this way, “al-Manṣūr” can in fact only be interpreted as a title of “the caliph”, who is identified a few words later as Hišām al-Mu’ayyad. However we understand it, though, whether as adjective or as title, the presence of “al-Manṣūr” here constitutes a problem.

  • 18 For an example from fourth/tenth-century Egypt see Bacharach 2006, p. 66 and 149. Here as in the An (...)
  • 19 See Guichard 1995, p. 47-53.

26The problem is two-fold : it lies, first, in the fact that the expression “al-Manṣūr” does not form part of the caliphal protocol of this caliph (that he was never a military leader, far less victorious, is of course of no relevance here) ; and secondly, in the fact that the title “al-Manṣūr” belonged to someone else, Hišām’s ḥāǧib Muḥammad Ibn Abī ‘Āmir, whose name actually occurs a little later in this short text. And thirdly, moreover, as Pierre Guichard has shown, the title “al-Manṣūr”, in this form, is properly a title of people other than caliphs18 ; a caliph who used the title “al-Manṣūr”, in a formal document such as this, would call himself “al-Manṣūr bi-llāh”19. A glance at any of the thousands of surviving coins of the Umayyad caliphs of Cordoba confirms this : a caliphal title includes the element Allāh, with the appropriate preposition or phrase including a preposition. And in formal contexts the element with the word Allāh always appears. Its presence is one aspect of formality, or, in other words, in formal contexts it is a requirement. Here, in all the versions of the text that we have, it is absent. What we have here, in all our versions, is al-Manṣūr, tout court, applied not to Ibn Abī ‘Āmir but to Hišām. However, “Al-Manṣūr”, with or without an element involving the word Allāh, is not known in any case as part of the caliphal protocol of Hišām al-Mu’ayyad. And it is known, in the form “al-Manṣūr” (with no element using the word Allāh), as it occurs here, as the title of Hišām’s ḥāǧib al-Manṣūr, and only, at the time in question, as that. That al-Manṣūr would certainly not have given his own title to the caliph whom he dominated, nor would he have permitted himself to adopt a title that was already in use by that caliph.

27“Al-Manṣūr” here, in this position in this inscription (actually, as we should now begin to recognise, in this alleged inscriptional text) must therefore be wrong. And it is easy to see what is missing. A caliphal protocol includes several elements. They need not all always appear, but there are some that appear very often. Here we find the title ẖalīfa, but we notice the absence of imām, another standard caliphal title. (We note also the absence of amīr al-mu’minīn, but there may be other reasons, such as questions of space, for its absence here). Like amīr al-mu’minīn, imām appears regularly on the coinage. If we replace “al-Manṣūr” here with “al-imām” we get rid of the problem caused by the presence of this cuckoo of a title among the titles of this caliph. But we do more than that : we also restore a missing title to the protocol ; and in addition to that we restore something else : with the word “al-imām” restored, we have a string of expressions  al-ḫalīfa al-imām, sayf al-islām, ‘abd allāh Hišām  which share a common rhyme. Although this is not in itself a conclusive argument in favour of this correction, the popularity of rhymes like this in such contexts certainly supports the overall argument for it.

V.

28This seems to tidy up a hitherto unnoticed problem of detail in our text. We now have a text that appears to make much better sense than our sources, or Lévi-Provençal, permitted. Unfortunately, while this is indeed the case, and helps us to know what the original inscription must have said, it does not resolve all the problems. There remains a major difficulty. As we have seen, Lévi-Provençal realized that there was something wrong with the date as given in our source : 375 could not be right, because ‘Abd al-Malik al-Muẓaffar is named here as ḥāǧib to Hišām. In 375 al-Manṣūr was alive and well and he was the ḥāǧib. ‘Abd al-Malik was very young, and very far from being the ḥāǧib at that point. His father did not die, leaving the succession as ḥāǧib to him, until 392. A date of 375 is therefore impossible. This is why Lévi-Provençal made the correction of the date, a very slight re-reading, rather than a formal correction or emendation, to 395, in the middle of ‘Abd al-Malik’s short reign.

  • 20 It is preserved in Ibn al-Ḫaṭīb and Ibn ‘Iḏārī; for translation and comments see Wasserstein 1993, (...)
  • 21 See Ibn ‘Iḏārī, al-Bayān al-Muġrib, vol. iii, p. 15, 16. It would be good to be able to turn to the (...)

29This correction should have solved the problem, and, in combination with the correction suggested above, restored the record of our inscription to its original state. However, things are rarely so simple. Al-Manṣūr indeed died in 392, leaving his son ‘Abd al-Malik to inherit the post and title of ḥāǧib, dominating the caliph Hišām al-Mu’ayyad as his father had done before him. But although he acquired the post and title of ḥāǧib in 392, he did not gain the title of “al-Muẓaffar” at that time. That title appears in this record of our inscription too. ‘Abd al-Malik waited for several years, until he had won a great victory, to adopt the title of “victorious” (Muẓaffar) as his laqab. He did not do this, in fact, until muḥarram 398/October 1007, barely a year before his death. We happen to possess the text of the decree by which the caliph granted him this title20. We know the date of it not because it is in the text of the document itself, but because we are told that it was produced “five ḥaǧǧ’s (= years) and three months” after ‘Abd al-Malik began to reign, and “at the start of the year 39821”. The name and title of the caliph Hišām al-Mu’ayyad of course appear uniformly whatever the date of the inscription as he was caliph from 366/976 right through until after the death of ‘Abd al-Malik. If we accept Lévi-Provençal’s reading of 395, therefore, in this inscription, we are left to explain how ‘Abd al-Malik could have erected an inscription in 395 which included a title for himself that he acquired (or, more precisely, caused the caliph to confer on him) only some three years later. This difficulty is not easy to resolve. The only possible explanation must locate the inscription in the year 398.

  • 22 For the rather involved history of the minbar of the mosque of the Andalusiyyīn see Terrasse 1957, (...)

30This minbar is not the only ‘Āmirid gift to Fes. If we move from the place where this inscription was originally to be found, the Mosque of the Qarawiyyīn in Fes, some half a kilometer away to the Mosque of the Andalusiyyīn, in the same city, we find, still standing (but covered now in green paint), another minbar, a gift of al-Manṣūr. As the minbar in this case is still available, and published, we can compare what its inscription says with what we have seen from the Mosque of the Qarawiyyīn22.

31The wording is rather similar :

بسم الله الرحمن الرحيم هاذا ما امر بعمله الحاجب المنصور سيف دولة الامام عبد الله هشام المؤيد بالله اطال الله بقاه ابو عامر محمد ابن ابي عامر وفقه الله في شهر جماد الاخر سنة خمس و٠وثلاث. ٠٠٠

32The translation of G.S. Colin, in Terrasse, is as follows :

  • 23 See G.S. Colin, in Terrasse, n.d., p. 5-6, planches liv-lvi.

Au nom de Dieu, le Clément, le Miséricordieux ! Ceci est ce qu’a ordonné de faire le maire du palais aidé de Dieu (Al-Manṣūr), glaive de l’empire de l’imām et esclave de Dieu Hišām, l’Assisté de Dieu (puisse Dieu prolonger ses jours !), Abū ‘Āmir Muḥammad fils d’Abū-‘Āmir (puisse Dieu le bien inspirer !), dans le mois de ǧumād ii de l’année 3…523.

33The similarities, like the differences, are apparent. Could some of the problems in these versions of the text that we have looked at so far derive from one or more of our authors’ reading of this other inscription, and his being influenced by its formulaic character and language as he tried to decipher what must undoubtedly (as we know from surviving examples) have been a complicated script on the other inscription ? By his time, too, our inscription was probably rather worn. Our earliest source for the text of the inscription on the minbar of the mosque of the Qarawiyyīn, Ibn Abī Zar‘, lived some three hundred years after the creation of the minbar and its inscription. This is clearly a possibility, if one impossible of demonstration. But, possible or not, it offers little or no help with the problem presented by the dates that we have.

34The dates are in fact the key to the solution of this little problem. What we have in our transmission is not an inscription, but a series of versions of a literary text providing a record of an inscription that is lost. In that sense, this text is no different from any other literary text. It is not a piece of wood with an original inscription on it, but a version in fact, several versions of what (is alleged) was inscribed on a piece of wood which happens not to survive. It may or may not give us the text, or versions of the text, of a genuine original inscription. If, however, there was such an inscription, and if it was a genuine product of the man named in it, ‘Abd al-Malik al-Muẓaffar, then it can only be a product of the time when he bore the title “al-Muẓaffar”. Since he acquired that title only in 398 (and died at the very beginning of the following year), any inscription bearing his name and this title cannot antedate 398. Both simple logic and the norms of bureaucratic and honorific formality impose this conclusion. This means that, if there is a genuine inscription behind the records that we have, its date can only have been 398 (as al-Muẓaffar died on 16 ṣafar, or in the middle of the second month, of 399, that year can probably be left out of consideration here quite safely). This reminds us of the date 388 in the versions of al-Ǧaznā’ī but it should do no more than remind us of it, for that writer did not give us otherwise any better a version of the inscription than the others, and indeed he also omits the month of the year in question.

35Two questions remain. First, was there really such an inscription as these texts allege ? And secondly, do our sources transmit or enable us to re-construct a correct version of the text of such an inscription ? As to the first, one feature of the history behind this text and of the text itself encourages the belief that there was such an inscription. This is the reference to ‘Abd al-Malik as “al-Muẓaffar”. He acquired this title so late in his career and his life, and held it for so short a time before his death, that anyone forging or inventing, or anyone mis-reading, such a text involving this title for him would have needed to possess remarkable knowledge of the details of his reign in order to incorporate the title into what he wrote. An opposing argument would hold that a clumsy forger or a poor reader would put a random date and any set of titles in his reading, but as against that it can be urged that a clumsy forger or a poor reader would be more likely to ascribe his text to someone more famous than a dictator who ruled for only six short years, someone like, say, al-Manṣūr. The likelihood is, therefore, that we have here to do with a (poor) set of records of an authentic original inscription.

36What of the second question, the original wording of the inscription, the ipsissima verba of ‘Abd al-Malik ? Here too, despite the problems imported by our sources, we can be fairly sure that we know at least the central contents of the inscription’s text. We cannot know the precise written form of the introductory formulas  the basmala etc.  in the original inscription in the wood, but this is of minor importance. Much more importantly, we can know, with some high degree of certainty, the text of the central message of the inscription. It must have told us something very close to the following :

... هذا ما امر بعمله الخليفة الامام سيف الاسلام عبد الله هشام المؤيد بالله اطال الله بقاءه على يد حاجبه عبد الملك المظفر بن محمد المنصور بن ابي عامر وفقهم الله تعالى وذلك في [شهرجمادى الاخرة] سنة ثمان وتسعين وثلاث مائة.

  • 24 It is striking that the month here (though not the year), taken from the Rawḍ, should be the same a (...)

37This is what the caliph, the imām, the Sword of Islam, the servant of God, Hišām al-Mu’ayyad bi-llāh, may God lengthen his life, ordered to be made by his ḥāǧib ‘Abd al-Malik al-Muẓaffar b. Muḥammad al-Manṣūr Ibn Abī ‘Āmir, may God give them success, and this was in [the month of ǧumādā ii,] in the year 398/[February-March] 100824.

38One last question calls for consideration : we have three sources for this text, and three different versions of when it is to be dated : ǧumādā ii 375 (in the Rawḍ, “corrected” to 395), 388 (in al-Ǧaznā’ī), and no date (in Ibn al-Qāḍī). As we have seen, none of these is correct. Where might 375 and 388 have come from ? It is tempting to see the first date, ǧumādā ii 375, as influenced by the date in the inscription that we have looked at from the mosque of the Andalusiyyīn, and with it some of the textual formulae as well, in particular the occurrence of the expression “al-Manṣūr” near the start. If this is indeed the case, and the inscription thus described is in fact genuine, then the explanation for this set of errors is presumably to be seen as an attempt by whoever transmitted the text to our source to make sense of a text that he could not read successfully by using the text of another inscription that he could read nearby. 388, more simply, looks like a (failed) correction of that date, unless it is a mis-reading of the decade in the inscription itself, in which case possibly the month is to be rejected as belonging to the dating 375, itself an importation from the inscription on the minbar of the mosque of the Andalusiyyīn. The absence of a date in Ibn al-Qāḍī may in its turn reflect a recognition of the historical difficulty presented by the two dates recorded in his sources.

VI.

  • 25 See for these events, Lévi-Provençal 1950, p. 268-272. Lévi-Provençal places the minbar and its ins (...)
  • 26 As is done still by Cambazard-Amahan 1989, p. 27-28.

39Such a re-dating of an obscure inscription on a long-lost minbar of which we know only through literary quotations may appear of little significance. The reality is otherwise. Why was this minbar given to the mosque of the Qarawiyyīn ? As we have seen, al-Ǧaznā’ī tells us that the date in the inscription on this minbar was 388 (without a month). This year appears to offer a highly convenient dating for the minbar. Lévi-Provençal describes in some detail the activities of ‘Abd al-Malik in Fes and in Morocco in general at this time. The dictator al-Manṣūr had found it necessary to deal with Zīrī b. ‘Aṭīya and restore Umayyad-‘Āmirid rule in Morocco. A campaign under the leadership of al-Manṣūr himself, his son ‘Abd al-Malik, and the Slav general Wāḍiḥ succeeded in driving Zīrī out of Fes and re-establishing Umayyad rule. ‘Abd al-Malik entered Fes in triumph at the end of šawwāl 388/October 998. A decree naming him as viceroy was read out publicly on the last Friday in ḏū l-qa‘da 388/18 November 998, from the minbar of the mosque of the Qarawiyyīn in Fes25. What more natural than that we should associate the inscription on this minbar with the reading out of this decree26. The fact that the victory, and even more so the investiture of ‘Abd al-Malik as viceroy, occurred so late in the year need not of itself invalidate a dating of such an inscription to the year 388 but it would make one worry, to put it in understated terms, about a dating to ǧumādā ii of that year, the month that we find in the version of the inscription’s text transmitted by Ibn Abī Zar‘, as that falls long before ḏū l-qa‘da. In any case, as we have seen, as the inscription names ‘Abd al-Malik as al-Muẓaffar, it cannot be from this time at all, and must be from 398.

  • 27 Lévi-Provençal 1950, p. 287-289.

40In 398, we have an even better occasion for the donation of such a minbar to the mosque of the Qarawiyyīn in Fes. By now, al-Manṣūr was dead ; ‘Abd al-Malik was ruling in his own right as ḥāǧib to Hišām al-Mu’ayyad ; he had just won a great victory over the Christians in the Iberian “Peninsula, at Clunia ; and as a result he had been granted, or had taken, the title “al-Muẓaffar”27. What more natural in this context than that he should use the moment to mark not only the victory but also his own new titulature ? One way in which this could best and, in terms of the creation of ‘Āmirid self-image, most appropriately be done was by giving a major mosque a new minbar, with a suitable dedicatory inscription recording his name and titles as the donor.

Bibliographie

Arabic Sources

Ibn Abī Zar‘ al-Fāsī, Abū l-‘Abbās Aḥmad, al-Anīs al-Muṭrib bi-Rawḍ al-Qirṭās fī aḫbār al-Maġrib wa-ta’rīḫ madīnat Fās, Dār al-manṣūr lil-ṭabā‘a wal-warāqa, Rabat, 1972.

Roudh el-Kartas. Histoire des souverains du Maghreb (Espagne et Maroc) et annales de la ville de Fès, Beaumier A. (trans.), Paris, imprimerie impériale, 1860.

Ibn Abī Zar‘, Rawḍ al-Qirṭās, Valencia (“Textos medievales”, 12), Huici Miranda, A. (trans.), 1964.

Ǧaznā’ī (Al-), Zahrat al-Ās, ed. A. Bel, Algiers, J. Carbonel, 1923.

Ǧanā Zahrat al-Ās, ed. ‘Abd al-Wahhāb ibn Manṣūr, Rabat, 1967.

Ibn al-Qāḍī, Ǧaḏwat al-Iqtibās, Fes, 1309/1891-92.

Ibn ‘Iḏārī, al-Bayān al-Muġrib, ed. E. Lévi-Provençal, Paris, Librairie orientaliste Paul Geuthner, 1930.

Tornberg, C.J. (ed.), Annales regum mauritaniae a condito idrisidarum imperio ad annum fugae 726 ab Abu-l Hasan Ali ben Abd Allah Ibn Abi Zer’ Fesano… conscriptos, 2 vol., Uppsala, 1843-1846.

Secondary Works

Bacharach, Jere L., 2006 : Islamic history through coins : an analysis and catalogue of tenth-century Ikhshidid coinage, Cairo and New York, The American University in Cairo Press.

Ben Cheneb, M. and Lévi-Provençal, E., 1922 : Essai de répertoire chronologique des éditions de Fès, Alger, Ancienne Maison Bastide-Jourdan Jules Carbonel.

Blair, Sheila S., 1998 : Islamic Inscriptions, Edinburgh, Edinburgh University Press.

Bloom, Jonathan, 1989 : Minaret, Symbol of Islam, Oxford, Oxford University Press (“Oxford Studies in Islamic Art”, vii).

Cambazard-Amahan, Catherine, 1989 : Le décor sur bois dans l’architecture de Fès, Époques almoravide, almohade et début mérinide, Paris, Éditions du C.N.R.S.

Colin, G.S., n.d. : in Terrasse Henri, La Mosquée des Andalous à Fès, Paris, Les Éditions d’art et d’histoire, Publications de l’Institut des Hautes-Études Marocaines, t. xxxviii.

De l’empire romain aux villes impériales, 6000 ans d’art au Maroc, Paris, Musée du Petit Palais.

Deverdun, G., 1971 : “Ibn al-Qāḍī, Šihāb al-Dīn Abū l-‘Abbās al-Miknāsī”, Encyclopedia of Islam, 2nd ed., vol. iii, p. 814.

Dodds, Jerrilynnm D. (ed.), 1992 : Al-Andalus, The Art of Islamic Spain, New York, Metropolitan Museum of Art.

Guichard, Pierre, 1995 : “Al-Manṣūr ou al-Manṣūr bi-Llāh ? Les laqab/s des Amirides d’après la numismatique et les documents officiels”, Archéologie islamique 5, p. 47-53.

Hamidullah, Muhammad, 1985 : Six originaux des lettres diplomatiques du Prophète de l’Islam, Paris, Éditions Tougui.

Hoenerbach, W., 1970 : Islamische Geschichte Spaniens, Übersetzung der A‘māl al-A‘lām und ergänzender Texte, Zurich/Stuttgart, p. 195-197, 555-556.

Idris, H.R., 1971 : “Ibn Abī Zar‘” in Encyclopaedia of Islam, 2nd ed., vol. iii, p. 694.

Lévi-Provençal, E., 1931 : Inscriptions arabes d’Espagne, 2 vol. (vol. i : texte, vol. ii : planches), Leiden, Brill, and Paris, E. Larose.

1950 : Histoire de l’Espagne musulmane, 2, Le Califat umaiyade de Cordoue (912-1031), Paris, Maisonneuve et Larose.

Reynolds, L.D. and Wilson, N.G., 1991 : Scribes and Scholars : a guide to the transmission of Greek and Latin literature, Oxford, Clarendon Press, and New York, Oxford University Press, 3rd ed. ; first published 1968.

Sublet, Jacqueline, 1991 : Le Voile du Nom, essai sur le nom propre arabe, Paris, Presses universitaires de France.

Terrasse, Henri, 1957 : Minbars anciens du Maroc, in Mélanges d’histoire et d’archéologie de l’occident musulman, t. ii, Hommage à Georges Marçais, Algier, Imprimerie officielle.

Van Minnen, P., 2000 : “An official act of Cleopatra (with a subscription in her own hand)”, Ancient Society 30, p. 29-34.

Walker, S., and Higgs, P. (ed.), 2001 : Cleopatra of Egypt, From History to Myth, London, British Museum Publications.

Wasserstein, D., 1993 : The Caliphate in the West. An Islamic Political Institution in the Iberian Peninsula, Oxford, Clarendon Press.

Notes

1 This article was written during my tenure of a visiting Fellowship at the Shelby Cullom Davis Center for Historical Studies in the Department of History at Princeton University, in 2008-2009. I am grateful to the Center for inviting me and to Vanderbilt University whose granting of a sabbatical year made my visit to Princeton possible, as well as to the readers for this volume.

2 See generally, and excitingly, Reynolds and Wilson 1991. We are probably not yet anywhere near being able to have such a work for Arabic culture.

3 Including now an administrative document from ancient Egypt containing a note of approval by Cleopatra, in her own hand. See van Minnen 2000, p. 29-34; the papyrus document in question (from the collection of the Ägyptisches Museum und Papyrussammlung in Berlin) was displayed at the exhibition “Cleopatra of Egypt: from History to Myth”, at the British Museum in London in 2001, and is illustrated, with Cleopatra’s endorsement enlarged, in Walker and Higgs 2001, p. 180; it is item 188 (of the year 33 B.C.E.) in the catalogue. The document provides presumptive, if not conclusive, proof of the ability of at least one member of the Ptolemaic dynasty to write. Hamidullah 1985, « avec une introduction à l’origine de l’écriture arabe », is less plausible.

4 Lévi-Provençal 1931.

5 Lévi-Provençal 1931, text n° 221, vol. i, p. 196, n. 3.

6 Ibid., text n° 221, vol. i, p. 196, n. 2.

7 See Idris 1971. Idris notes that we still lack a sound critical edition of the Rawḍ.

8 Tornberg, Annales regum mauritaniae a condito idrisidarum imperio ad annum fugae 726 ab Abu-l Hasan Ali ben Abd Allah Ibn Abi Zer’ Fesano …conscriptos, Uppsala 1843-1846, vol. i, p. 32-33.

9 In fact, although this edition (which names no editor), published by Dār al-manṣūr li-l-ṭabā‘a wa-l-warāqa, in Rabat, says 1972 on its title-page, on the cover it has 1973.

10 Beaumier (trans.), Roudh el-Kartas. Histoire des souverains du Maghreb (Espagne et Maroc) et annales de la ville de Fès, Paris, 1860, p. 73.

11 Huici Miranda (trans.), Ibn Abi Zar‘, Rawḍ al-Qirṭās, Valencia, 1964, vol. i, p. 115.

12 Al-Ǧaznā’ī, Zahrat al-Ās, Algiers, 1923, p. 41-42 of Arabic text, with a translation at p. 99 of the French section.

13 Ed. ‘Abd al-Wahhāb ibn Manṣūr, Rabat, 1967, see p. 55.

14 Deverdun 1971, vol. iii, p. 814.

15 See also Ben Cheneb and Lévi-Provençal 1922, p. 17, n° 69. Ben Cheneb and Lévi-Provençal list the number of pages as 358, whereas the copy that I used in the Firestone Library at Princeton is described in the catalogue (its call number there is 2271.409387.349) as having only 355; the difference seems to lie in some extra unnumbered pages at the end of the Princeton copy. I am indebted to Ms Joyce E. Bell of the Firestone Library for help with the location and dating of Princeton’s copy of this work.

16 Ibn al-Qāḍī, Ǧaḏwat al-Iqtibās, Fes, 1891-1892, p. 32.

17 For the forms and contents of names and titles in Arabic, see especially Sublet 1991.

18 For an example from fourth/tenth-century Egypt see Bacharach 2006, p. 66 and 149. Here as in the Andalusi case the title is borne by a non-caliphal ruler.

19 See Guichard 1995, p. 47-53.

20 It is preserved in Ibn al-Ḫaṭīb and Ibn ‘Iḏārī; for translation and comments see Wasserstein 1993, p. 20-21; see also Hoenerbach 1970, p. 195-197, 555-556; Lévi-Provençal 1950, vol. ii, p. 281.

21 See Ibn ‘Iḏārī, al-Bayān al-Muġrib, vol. iii, p. 15, 16. It would be good to be able to turn to the coins and confirm this chronology by showing that the new title appears for the first time only in the year 398, but al-Muẓaffar, like his father before him, saw the wisdom of mentioning only his given name on the coins, along with the title of ḥāǧib.

22 For the rather involved history of the minbar of the mosque of the Andalusiyyīn see Terrasse 1957, p. 159167; his argument is resumed in Bloom 1989, p. 109-112 (where, in n. 42, the year 3[7]5 is mis-printed as 3[6]5); De l’empire romain aux villes impériales, 6000 ans d’art au Maroc, Paris, Musée du Petit Palais, p. 188-191; Dodds 1992, p. 249-251, n° 41; Blair 1998, p. 130-139.

23 See G.S. Colin, in Terrasse, n.d., p. 5-6, planches liv-lvi.

24 It is striking that the month here (though not the year), taken from the Rawḍ, should be the same as that in the inscription just quoted, from the mosque of the Andalusiyyīn. If our inscription is not in fact authentic, then the mention of this same month in this work could be an indication of the source on which the person who in that case invented the text in our literary sources drew. I put the month in square brackets, as it is in the Rawḍ, but in fact, if the inscription text is genuine, but the month an invention, it could belong to any month in the year 398 following ‘Abd al-Malik’s assumption of the title “al-Muẓaffar” in muḥarram of that year.

25 See for these events, Lévi-Provençal 1950, p. 268-272. Lévi-Provençal places the minbar and its inscription, with al-Ǧaznā’ī, in this year.

26 As is done still by Cambazard-Amahan 1989, p. 27-28.

27 Lévi-Provençal 1950, p. 287-289.

Auteur

Vanderbilt University

© Presses de l’Ifpo, 2013

Conditions d’utilisation : http://www.openedition.org/6540