¹ Tartīb al-madārik wa taqrīb al-masālik li-ma'rifat a'lām madhhab Mālik, edited by Aḥmad, Bakīr Maḥmūd (Beirut, 1967).Google Scholar

² Abū'l-Faḥl ‘Iyāḍ b. Mūsā b. ‘Iyāḍ al-Yaḥṣubī al-Sabtī, 476/1083–544/1149, cf. Brockelmann, GAL. 1, no. 369 and E. I., II, 602−3. The editor of the Beirut edition makes the significant observation in his biography of the Qāḍī, 1 and 2, p. 23, that ‘the Qāḍī was known for his bias towards the Almoravid state. He it was who led the revolt against the Almohads in Ceuta.’ This fact should be borne in mind when assessing the authenticity of his biography of ‘Abdullāh b. Yāsīn.

³ In this period and context there is, for example, the biography of Abū ‘Imrān alGhafjūmī al-Fāsī, Tartīb al-madārik, ibid. vols. 3 and 4, pp. 702–6.

⁴ cf.Monteil, Vincent, ‘Al-Bakrī’, Bull. de l'IFAN, XXX, série B, I (1968), 59–64.Google ScholarNicholson, R. A., A literary history of the Arabs (London, 1907), 429–30.Google Scholar

⁵ Tartīb al-madārik, 780–2. The edited text is corrupt in places and not always clear. I am grateful to my colleague Jarir Abū Ḥaydar for his opinion and advice on the alternative readings I have proposed.Google Scholar

⁶ Awkād b. Zalwah al-Lamṭī in the Beirut text.

⁷ Gudāla for Jazūla. Rajul is used in the text and not any title such as amīr to suggest a confusion with Yaḥyā b. Ibrāhīm al-Gudālī, who is wholly ignored.

⁸ al-Jawhar b. SKN in the Beirut text.

⁹ The passage is unfortunately obscure at this point, but the main sense is clear.

¹⁰ Gudāla for Jazūla and Lamtūna for Maltūna. Note the straight transfer of abode, no reference to an island, nor a return to Aglū in the Sūs to consult his former master.

¹¹ This title was bestowed on Yusuf b. Tāshfīn by the ‘Abbāsid Caliph. It is Amīr ḥaqq according to the Bayān by Ibn ‘Idhārī.

¹² Yaḥyā b. ‘Umar according to al-Bakrī.

¹³ The Barghuwāṭa, their sect and their Berber Qur' ān are discussed by Henri, Basset, Essai sur la littérature des BerbÈres (Algiers, 1920), 62–3. They were not KHārijites as is sometimes supposed. Their unbelief is confirmed by the Qāḥī 'Iyāḥ.Google Scholar

¹⁴ Neither the editor of the Beirut edition nor the Encyclopaedia of Islam refer to this historical text, the discovery of which could be of major importance to an understanding of the origins of the rise of the Almoravids. The Qāḥī 'Iyāḥ wrote a history of Ceuta. This, it seems, was not his only historical work.

¹⁵ The historical existence of this ‘constructed’ island headquarters has been accepted as gospel by amongst others, Bovill, E. W. in The Golden Trade of the Moors (Oxford, 1962), 72,Google Scholar and Spencer Trimingham, J., A history of Islam in West Africa (Oxford, 1962), 23.Google Scholar

¹⁶ Yūlsuf b. Yaḥyā al-Tādilī, al-Tashawwuf ilā rijāl al-taṣwwuf (Vie des saints du Sud marocain des Ve—VIIe siÈcles de l'hégire). Arabic text published by Faure, A., Institut des Hautes Etudes Marocaines, XII (1958), 424, 425. This personality is numbered 232 among the biographies.Google Scholar

¹⁷ See for example the observations of Jacques Meunié, D., in Cités anciennes de Mauritanie (Paris, 1961), 71, 189. Professor Raymond Mauny shares my scepticism regarding this site.Google Scholar

¹⁸ cf.de Moraes Farias, Paulo Fernando, ‘The Almoravids’, Bull de l'IFAN, XXIX, sér. B., 3–4 (1967), 821–43.Google Scholar

¹⁹ A convenient ‘umbrella term’ to describe the supreme Ṣanhāja chieftainship. Although there is no evidence to show this was the title actually used by the medieval Ṣanhāja, it is to-day general among the Tuareg. Cf. Nicolaisen, Johannes, Ecology and Culture of the Pastoral Tuareg (Copenhagen, 1963), 393–5.Google Scholar

²⁰ Cf. note 4.

²¹ Monteil, , 60.Google Scholar

²² al-Ḥulal al-mawshīya fī dhikr al-akhādr al-Marrākushīya, edited by I. S. Allouche, Rabat, 1936.

²³ Cf. al-Bayyina, an Arabic cultural journal, published in Rabat, , 3 (1962), 65–72. Extracts from the manuscript text of this Moroccan work are cited in an article devoted to Abü ‘Imrān al-Ghafjūmī al-Fāsī by ‘Abd al-Qādir Zammāma.Google Scholar

²⁴ The following diagram illustrates the relationship of the Ṣanhāja aristocracy as far as it can be gleaned from the Arabic sources: In proposing a solution to the varied and complex alternative possibilities with regard to the personalities of Yaḥyā b. Ibrāhīm, al-Jawhar b. Sakkum, Yaḥā b. ‘Umar, and others, I am particularly grateful to my colleague Dr Humphrey Fisher for his suggestions.

²⁵ Some historians believed that his death took place before the ‘island retreat’ episode, others that it occurred after that, about the year 440/1048. If the latter date should prove correct it would confirm a lengthy stay of ‘Abdullāh b. Yāsīn during the lifetime of this chief, possibly some ten years among the Gudāla.

²⁶ This would have given Yaḥyā b. ‘Umar a strong electoral position. For a description of the system, cf. Johannes Nicolaisen, 396, 478, 486, and his article on the political system of pastoral Tuareg in Aīr, and Ahaggar, in Folk, 1 (1959), 90–1 and 115–27. Professor Nicolaisen is of the opinion that the northern Tuareg originally preferred sister's Sons to succeed as chiefs. This was the custom of the Massūfa, who were one of the Almoravid ‘drum-groups’, when lbn Baṭṭūṭa visited Waläta in 1352.Google Scholar

²⁷ Dirāsa ḥawla kitāb al-ḥulal al-mawshiya fī dhikr al-akhbār al-Marrākushīya wa ahammīyatihi fī tārīkh al-Murābiṭīn wal-Muwabḥḥīn, in the journal Taṭwān, 5 (1960), (11), 149. This article has many shrewd observations on Almoravid history, and it foreshadows recent ideas of Western scholars. Of special interest is this author's clearly argued hypothesis of an evolution, in the context of tafsīr, of the title of al-Murābiṭūn. This followed definable stages determined by historical events. The last stage he suggests (pp. 13/151) is that of the rule of Yūsuf b. Tāshfin in Spain, and he quotes ‘Abd alWāḥid b. ‘Alī al-Tamīmī al-Marrākushī, al-Mu'jib fi talkhīṣ akhār al-Maghrib, 1939 edition, 163, concerning Yūsuf, likewise this same author on ‘Alī (p. 171). “All b. Yūsuf took command after his death, and he adopted the title of his father Amīr al-Muslimin. He named his companions al-Muābiṭūn. He followed the customary practice of his father in the furtherance of the jihād, casting fear in the foe, and the protection of the country.’ Aḥmad Mukhtār al-‘Ibādī then concludes: ‘From all that has gone before it is our view–though one with an element of caution and reservation–that the real elucidation [tafsīr] of the word al-Murābiṭūn began after the Lamtūna had assumed the leadership of the Ṣanhāja—that is to say since this term had become a title bestowed upon his followers by their leader in its metaphorical sense [mujāhidūn, those steadfast in the path of God]. As for the specific explanation—by the ribāṭ built by ‘Abduliāh b. Yāsīn in Gudāla territory—it appears that it had lost its meaning through the withdrawal of the Gudāla from obedience to Ibn Yāsin and to the Jam¯’a of the Murābiṭūn.’Google Scholar

²⁸ Cf. note. 4.

²⁹ Abū Muḥammad Muṣṭafā b. Ḥasan b. Sinān b. Aḥmad al-Ḥusaynī al-Hāshimī, known as al-Jannābī, the author of al-Baḥr al-zakhkhār, cf. Fagnan, E., Extraits inédits relatifs au Maghreb (Algiers, 1924), 354–6, and Oxford MS Pococke, 176, folios 396–7.Google Scholar

³⁰ Cf. Paulo, Farias, 801–8, 813–17.Google Scholar

³¹ Cf. Paulo, Farias, 861, no. 3 of his conclusions.Google Scholar

³² Kitāb al-‘ibar; (Beirut, 1956–61), 583,Google Scholar and tr. of the Muqaddima by Rosenthal, F. (London, 1958), chapter iii, section 51, Ṣūfī opinions about the Mahdī, 196–7. Ibn khaldūn is regrettably inconsistent. While admitting that most of the western Sahara was virtually terra incognita, he proceeds elsewhere to interpret the ‘island ribāṫ’ as though he was familiar with the course of the Sūdānic Nile. He only succeeds thereby in displaying his ignorance of the seasonal rainfall and Saharan topography.Google Scholar

³³ Monteil, , 59.Google Scholar

³⁴ Monteil, , 64.Google Scholar

³⁵ Robertson Smith, W., The Religion of the Semites (London, 1894), 429 and 439.Google Scholar

³⁶ Monteil, , 64.Google Scholar

³⁷ Henri, Basset, Essai sur la littérature des BerbÈres, 61–100. This splendid survey is, in the opinion of my colleague Dr J. Bynon, the only reference work on this subject.Google Scholar

³⁸ See my article on shhaykh Mā al-' Aynayn in Bull. SOAS, XXXI, 2 (1968), 357−9 and 362, poem 45, verses 9–11.

³⁹ In my opinion the explanation for the title given in the Bayān and the Ḥulal almawshīya, cf. p. II of the latter, where 'Abdullāh b. Yāsīn named the Lamtūna thus on the basis of a tough battle fought with great fortitude against non-Muslim Barbers near, or in the Adrār, is not particularly convincing and has a bias suggesting its ‘discovery’ after the great conquests. He may have used it to praise the Lamtūna on such an occasion, but clearly the name and title were known among the Ṣanhāja long before this episode, if it ever took place. Nevertheless, it is not irrelevant to draw attention to a remark made by Henri Lhote in his Les Touaregs du Hoggar, Payot (Paris, 1944), 321—'les hommes se réunissaient, discutaient de l'expédition projetée et prononçaient une formule sacramentelle… qui les liait les uns aux autres ‘;the italics are mine] pendant toute la durée de l'expédition et pendant laquelle ils devaient se porter une aide mutuelle.’

⁴⁰ This is particularly true of Rawḍ al-Qirtāṣ, for example, where Yaḥyā b. Ibrāhīm al-Gudālī is told by ‘Abdullāh b. Yasin that he should only have four wives and that he should divorce the other five. Ṣanhāja society, whether Tuareg or Moor, is fundamentally monogamous, however lax its Islam may have been. ‘Companions’ were tolerated, divorce was easy, but polygamy of this kind is quite out of keeping with what is known of the social life and matrilineal character of the Berber Saharan nomads.

⁴¹ It may be presumed that these fatāwā included his most personal rulings, such as his habit of taking one third of property of suspect provenance, asserting that this ‘purified’ the rest and made ownership of it lawful. What is the explanation for this unusual rule? Is it ‘islamization’ of a practice such as the Tuareg tribute named ennehet? If foreign Tuareg, paying an annual tribute called tiouse to the Ahaggar Amenukal, were raided by people of the Ahaggar, the Amenukal had a right to this ennehet tribute, about one- third of the booty taken. It was paid to him by all raiders whether nobles or vassals. Might it be viewed as a tendency to regard all property as inherently ‘impure’? See note 42.

⁴² The later tenets of the followers and successors of the Córdoban ascetic Ibn Masarra, A.D. 883–931, cf. Miguel Asín Palacios, Abenmasarra y su escuela, Madrid, 1914, 98–105. The doctrines are summarized on page 106. The influence of Massarid ideas are discussed on pages 107−29.Google Scholar

⁴³ Monteil, , 63 (para 25).Google Scholar

⁴⁴ Cf. the article on Ṣalāt in the Shorter Encyclopaedia of Islam, Brill, 1953, paragraph IV, 495−6.Google Scholar

⁴⁵ Paulo Farias, 811−20. My own view is that this author's arguments for a revolution of Ṣanhāja battle technique are overstated in view of known traditional Tuareg tactics. Nevertheless the religious inspiration and the ideology of Abdullāh b. Yāsīn in their fighting seems beyond any dispute.

⁴⁶ Monteil, , 62 and 63 (para. 23).Google Scholar

⁴⁷ Nicolaisen, Johannes, Ecology and Culture of the Pastoral Tuareg, 114–15.Google Scholar

⁴⁸ Cf.Farias, Paulo, 809, 810, and the Arabic text, of the Muwaṭṭa', printed by the Maydān al-Azhar, 1343 A.H., the third section of the commentary by al-Ṣuyūti, entitled Tanwīr al-lḥzawālik, 138.Google Scholar

⁴⁹ Cf. Farias, Paulo, 851–62.Google Scholar

⁵⁰ This is from approximately 465/1072 to 489/1096, the date of his death.

⁵¹ Kitāb al-'ibar, Beirut, , 1956–61, VI (1959), 376.Google Scholar

⁵² In 451/1059, according to Ibn ℑaldūn.

⁵³ I suggest that Ubān (?) should read Wajjāj, cf. Tartib al-madārik, 780.

⁵⁴ Ibid. 780.

⁵⁵ These reforms included the abolition of illegal taxes (maghārim and mukūs) and the levying of the ṣadaqa. According to the Rawḍ al-Qirtāṣ, musical instruments were destroyed and wineshops burned.