² Michael Gomez, who preceded us in demonstrating the opportunities for major reconsiderations in Songhay history, has argued in his stimulating article, ‘Timbuktu under imperial Songhay: a reconsideration of autonomy’ (J. Afr. Hist., xxxi [1990], 5–24)Google Scholar, for a substantial degree of control over Timbuktu by the askiya dynasty, and for relatively little Timbuktu influence over Gao, the hub of Songhay power. By and large, his argument and ours do not overlap directly: his canvas embraces the entire sixteenth century, ours barely a decade, and his interests are more with such matters as taxation, while we concentrate on specifically religious practice and belief. There are, however, a few points at which his interpretation and ours do clearly diverge, and these will be noted at the appropriate stages in this paper.

³ See below, pp. 84–7.

⁴ T/S, 119/72, says the askiya ‘in the second year of the tenth century went on pilgrimage in the month of Ṣafar’; the author adds, ‘and God knows’, as if he were not quite certain. Ṣafar is the second month of the year. The editor and translator, O. Houdas, takes this to mean 901 a.h., giving a departure date in October/November 1495, just above thirty months after the coup. This seems too early. The phrase ‘in the second year’ very probably means 902, a departure in October/November 1496. One version of the T/F, 124 n. 3, not in the printed Arabic, gives ‘the third year’, a departure (if this means 903) in September/October 1497. See also Saad, Elias N., Social History of Timbuktu: The Role of Muslim Scholars and Notables, 1400–1900 (New York, 1983), 46–7.Google Scholar For T/S and T/F, which recur throughout the article, see the Note on Sources at the end.

⁵ For the return, the T/S, 121/73, gives the last month of ‘the third year’, which Houdas identifies as 902, a return in July/August 1497; as with the departure, the intention of the Arabic is more likely to be a year later. The T/F, 131/68, likewise gives a return at the end of the third year, which Houdas and Delafosse here identify (correctly in our view) as 1498. John O. Hunwick accepts July/August 1498 for the return (Sharī'a in Songhay: The Replies of al-Maghīlī to the Questions of Askia al-Ḥājj Muḥammad [Oxford University Press for the British Academy, 1985], 41).Google Scholar It seems likely that the T/F departure year, ‘the third year’, appearing in one Ms only, is wrong, for it would require a round trip within a single year, and would miss the pilgrimage proper, in the twelfth month, entirely.

⁶ Hunwick, , Sharī'a, 41Google Scholar, suggests this; there is no firm evidence.

⁷ Ibid. 41; he adds, ‘there is no internal indication of when or where’ the Replies were written.

⁸ See below, p. 80 and n. 85.

⁹ See Hunwick, , Sharī'a, 42Google Scholar, for dates and discussion.

¹⁰ T/S, 132–4/80–1, and T/F, 148–9/78 and 155/82, both give the Festival of Sacrifices in the thirty-fifth year. If this is 935, T/F correctly locates the deposition in 1529; the editors of T/S opt for 1528.

¹¹ Cissoko, S.-M., Tombouctou et l'empire Songhay (Dakar/Abidjan, 1975), 75.Google Scholar

¹² Kaba, Lansiné, ‘The pen, the sword, and the crown: Islam and revolution in Songhay reconsidered, 1464–1493’, J. Afr. Hist., XXV (1984), 256.Google Scholar See also Cissoko, , Tombouctou, 75, 76Google Scholar; Saad, , Social History, 45, 46Google Scholar; Mauny, , ‘Soudan occidental’, 198Google Scholar; Triaud, J.-L., Islam et sociétés soudanaises au moyen-âge (Paris, 1973), 155;Google ScholarZouber, M. A., Ahmad Baba de Tombouctou (Paris, 1977), the Introduction.Google Scholar

¹³ Gomez, ‘Timbuktu’, 16, approves, remarking that ‘Kaba has effectively demonstrated that the combined interests of the merchant and scholarly communities led to the ouster of the Sunnis and to the ascendancy of Muḥammad Ture’. Gomez goes on to argue that the return to Timbuktu in 1493, the year in which Muḥammad seized power, of those who had fled the city in 1488 suggests that Timbuktu support for Muḥammad antedated the actual coup. The evidence seems to us not to bear quite such a clear construction. According to the T/F (94–6/48–9), Sunni Ali cleared the city utterly in 1488, save for those, seriously ill or blind, abandoned by the rest of the population, ordering all the rest to Hawkiyi, perhaps on the south bank of the river (T/F 95 n. i). The T/S (116/70), in a much leaner account, says that the people of Timbuktu entered Hawkiyi in 1488 and stayed there for five years, a very unobtrusive way of indicating their return. The T/F, it seems, does not even mention the return. The return of those expelled by Sunni Ali's threats, after his death, seems natural enough: if it were really a significant marker of Timbuktu's involvement in revolutionary activity at this very early stage, we might expect the chroniclers, very much in favour of both Timbuktu and the first askiya, to bring out these implications of the homecoming a bit more.

¹⁴ Hiskett, M., The Development of Islam in West Africa (London, 1981), 38;Google Scholar see also Hunwick, , Sharī'a, 28.Google Scholar

¹⁵ Batran, A. A., ‘A contribution to the biography of Shaykh Muḥammad ibn ‘Abd al-Karīm al-Maghīlī al-Tilimsānī’, J. Afr. Hist., XIV (1973), 392.Google Scholar See also Hodgkin, T., Nigerian Perspectives (London, 1960), 90 n. 5;Google ScholarNorris, H. T., The Arab Conquest of the Western Sahara (Harlow, England and Beirut, 1986), 122Google Scholar; Clarke, P. B., West Africa and Islam (London, 1982), 50;Google Scholar and Gwarzo, H. I., ‘The life and teachings of al-Maghīlī with particular reference to the Saharan Jewish community’ (Ph.D. thesis, S.O.A.S., University of London, 1972), 266–7.Google Scholar

¹⁶ Quotations from Norris, , Arab Conquest, 122Google Scholar, and his The Tuaregs: Their Islamic Legacy and its Diffusion in the Sahel (Warminster, 1975), 199Google Scholar; see also Barth, Heinrich, Travels and Discoveries in North and Central Africa…in the years 1849–1855 (3 vols.) (London, 1965), I, 310Google Scholar, and Gwarzo, , ‘Life and teachings’, 266–7.Google Scholar

¹⁷ See T/F, 14–15/11–12, for decline under the first askiya, and for further decline later, see for example T/F, 230/126.

¹⁸ Hiskett, , Development, 38.Google Scholar And Levtzion, N., ‘The western Maghrib and Sudan’, in Oliver, Roland, (ed.), The Cambridge History of Africa, Vol. 3. From c. 1050 to c. 1600 (Cambridge 1977), 429;Google ScholarPardo, A., ‘The Songhay empire under Sonni Ali and Askia Muḥammad: a study in comparisons and contrasts’, in McCall, D. F. and Bennett, N. R. (eds.), Aspects of West African Islam (Boston, 1971), 53.Google Scholar

¹⁹ Levtzion, , ‘Western Maghrib’, 418.Google Scholar

²⁰ Norris, , The Tuaregs, 42Google Scholar; see also Pardo, , ‘The Songhay empire’, 51.Google Scholar

²¹ 116–17/71–2.

²² See, for an example of general complaint, 103/64, and, more interestingly for our purposes, 115/70, a passage discussed in more detail below, n. 45. This second passage describes prayer against Sunni Ali, performed in Arabia, and adds that from this moment the king's power declined and soon came to an end. Since, however, this was in 1487, Sonni Ali in fact enjoyed a further five warlike and triumphant years.

²³ Hunwick, , Sharī'a, 24, 25, 26.Google Scholar The English language italics are our own.

²⁴ See the discussion and references in Ibid. 24. In a personal communication of 12 March 1991, Hunwick [commented]: ‘… I doubt that the Askiya was seen as an Islamic “saviour” in the first instance…’, thus concurring with our hypothesis that, at the very beginning, MuḤammad was not regarded as an Islamic hero (though we believe he may perhaps have had some rather utilitarian ritual support from the Gao Islamic establishment).

²⁵ T/F, 98–9/51. The invocation, by another but unspecified aggrieved holy man, of divine wrath against Sunni Ali, is also mentioned here.

²⁶ See above, pp. 65–6, and pp. 84–7 below.

²⁷ T/F, 102–6/53–5.

²⁸ Levtzion, N., ‘A seventeenth-century chronicle by Ibn al-Mukhtār: a critical study of the Ta'rīkh al-fattāsh’, Bulletin of the School of Oriental and African Studies, XXXIV (1971), 571–93.CrossRefGoogle Scholar

²⁹ Ibid. 576, 584.

³⁰ See below, p. 86.

³¹ Hunwick, , Sharī'a, 13Google Scholar; T/S, 119/73; T/F, 124/65.

³² Hunwick, , Sharī'a, 24.Google Scholar

³³ Hunwick, (Sharī'a, 13)Google Scholar points out that the only religious office in Gao mentioned for the sunni period is that of the khaṭīb, or preacher, and that under the askiya's the khaṭīb continued prominent. Two khaṭīb's are mentioned (both, significantly it seems to us, in the T/F, 197/106–7, 268/149–50) as having been also the qāḍī, and, as there is nowhere any separate mention of a qāḍī of Gao (but cf. n. 35 below), it seems likely that there the two offices were held by the same man. If, as appears to have been the case, the rank of khaṭīb carried more weight in Gao than that of qāḍī, this may have had a symbolic significance: the qāḍī may have represented a degree of independence vis-à-vis the ruler which, though acceptable in Timbuktu, was not tolerable at the royal capital, Gao, while contrariwise the khaṭīb, giving his weekly sermon in the name of the ruler, provided exactly the right element of subordinate support to the state. Hunwick remarks that several khaṭīb's under the askiya's may, by their names, be identified as of western, Malian origin.

³⁴ 25–6/16 and 125–6/65.

³⁵ Levtzion, , ‘Chronicle’, 592.Google Scholar

³⁶ T/F, 25 and nn. 7–8 (French)/16 line 15 (Arabic), and 125–6 (French)/65 line 3 from the bottom (Arabic). One was apparently a Fulani from Masina, the other a Mandingo, both thus of western origin. The second man is given the title qāḍī. Was he qāḍī of Gao? If so, is he the exception to Hunwick's rule (see n. 33 above) that no qāḍī of Gao is mentioned as an official distinct from the khaṭīb? or perhaps he was himself also the khaṭīb?

³⁷ T/F, 212–13/116–17.

³⁸ T/F, 217/119.

³⁹ This seems more likely than that he was correctly a member of that entourage and incorrectly added to Askiya Dāwūd's suite, since Dāwūd's advisers included, alongside Gao Zakariyā, the sons of two other members of the pilgrimage party. These apparently intimate links between Askiya Muḥammad's clerical companions on ḥājj in the 1490s, and the advisers of Askiya Dāwūd between 50 and 80 or more years later, suggest that the central government was continuing to work closely with the Gao religious establishment, even though Timbuktu had regained centre stage religiously.

⁴⁰ See also pp. 69–76 above, and n. 44 below.

⁴¹ Levtzion, , ‘Chronicle’, 576, 584.Google Scholar

⁴² T/S, 119–20,/72–3.

⁴³ T/F, 125/65.

⁴⁴ Unless it be Hawgāru, whose name may derive from Hoggar, indicating a Saharan origin. The T/F mentions also, among the askiya's companions, four aqyāl kuray, which Hunwick translates (in the letter of 12 March 1991) as ‘petty kings of the whites– Tuareg/Ṣanhāja chiefs’. Such a Saharan representation, if it is this, is less likely to indicate an askiyan link with Timbuktu at this stage than to reflect the rivalry amongst diverse Saharan tribes and clients, some grouped around Timbuktu, others around Gao. It is perhaps significant, too, that the petty kings appear here without religious specification. We are indebted to Harry Norris for this line of argument.

⁴⁵ For example, Ahmad bin ‘Umar bin Muḥammad Aqīt, a very senior man indeed, set out in 890/1485, and returned during the fitnah of Sunni Ali (T/S, 115/70). The use of the term fitnah, unrest or disturbance, here is interesting, since fitnah is usually reserved for inter-Muslim disputes: this T/S usage may be another instance, perhaps, of the unwillingness of even Sunni Ali's most severe critics amongst the Timbuktu establishment to go the whole way and call him an outright unbeliever. The T/S here adds that, in 892/1487, Sunni Ali's name was mentioned at Arafah in the presence of the jurisconsult ‘Abd al-Jabbār-Koko (Arabic K.ku), who then cursed him. Someone was there – Arafah is a crucial station in the pilgrimage rites – who knew Sunni Ali, but probably not a Timbuktu scholar. The failure to name him suggests ignorance on the part of the author of the T/S, or – rather farfetchedly – deliberate separation from non-Timbuktu (presumably Gao) clerics. ‘Abd al-Jabbār-Koko is unmentioned elsewhere in either chronicle; we do not know who he was. ‘Abd al-Jabbār was the name of al-Maghīlīʾs son, but such a link-up is highly implausible, and the name not unusual. We do not know the resonance, if any, of Koko; possibly it sounds more Sudanese than Saharan or Arab.

⁴⁶ Saad, , Social History, 46.Google Scholar

⁴⁷ Another hint of Islamic distance between Gao and Timbuktu, much later than our immediate concern in this section, occurs in the elaborate and sometimes fantastic story of the sharīf, Ahmad as-Saqlī, allegedly sent to Songhay in 925/1519 by the sharīf of Mecca, in response to the askiya's request on pilgrimage more than twenty years earlier. Ahmad as-Saqlī went first to Timbuktu (T/F, 27/17 ff.), but was later taken by the askiya, by force, to Gao (38/23). Ahmad, his tale adorned with dreams, jinn, minor miracles and the like, evidently belonged, or was remembered or imagined to have belonged, rather more to the Gao than to the Timbuktu mode. The T/F material on Aḥmad is all nineteenth-century addition, but much of it probably comes from earlier records, oral and/or written. Aḥmad as-Saqlī is historical, named several times in the T/S, but only with reference to two daughters of his and to subsequent descendants. The contrast between this relative neglect, and his high-profile role in the T/F, is striking.

⁴⁸ The imām's work is entitled Naṣīḥat ahl al-sūdān, ‘Advice to the people of the Sudan’. Both this work, and its author, are known only from eighteenth- and nineteenth-century quotations. For full details, see Hunwick, John O., ‘Askia al-Ḥājj Muḥammad and his successors: the account of al-Imām al-Takrūrī’, Sudanic Africa (Bergen), 1 (1990), 85–9.Google Scholar For the name Takrūr, see ‘Umar al-Naqar, ‘Takrūr; the history of a name’, J. Afr. Hist., x (1969), 365–74.Google Scholar

⁴⁹ See above, p. 68.

⁵⁰ Hunwick, , ‘Al-Imām al-Takrūrī’, 89.Google Scholar

⁶¹ It is also possible that the anonymous imām had in mind particularly the askiya's generosity towards the clerics of Gao; the adjective al-takrūrī, though the name Takrūr, as a geographical indicator, is often loosely used, might well indicate that the author was part of a Sudanese Muslim network rather than a Saharan. The title of the work, advice to the people of the Sūdān, points in the same direction. It is also just possible that the askiya's openhandedness towards divines and academics, described here, relates specifically to his behaviour whilst on pilgrimage.

⁵² For a detailed study of al-Maghīlī, with many references, see Hunwick, Sharī'a, ch. 2 and passim.

⁵³ See above, n. 5.

⁵⁴ But see Gwarzo, ‘Life and times’. The late Dr Gwarzo was Grand Kadi of Kano State, in Nigeria.

⁵⁵ Hunwick, , Sharī'a, 28.Google Scholar

⁵⁶ T/S, 65/39, 66/40, 67/41, 69–70/43, 75/46.

⁵⁷ T/F, 132/69.

⁵⁸ For example, Hunwick, , Sharī'a, 41 n. 4.Google Scholar

⁵⁹ T/F, 15/12, 22–3/15.

⁶⁰ Ibn ‘Askar, Dawḥat al-nāshir, cited in Hunwick, , Sharī'a, 41.Google Scholar Hunwick, in his letter of 12 March 1991, is rather dismissive of Ibn ‘Askar: ‘He seems to have completely lost track of al-Maghīlī once the latter had returned from Morocco to Tuwāt and makes no mention of his trans-Saharan peregrinations… Ibn ‘Askar was a northern Moroccan faqīh of rather limited horizons and defective literary attainments. Tuwāt was no doubt the limit of his horizons and while he may have heard of Timbuktu he had probably never heard of Gao. He… gets his geography (to say nothing of his knowledge of the state of Islamization of the area) confused.’ This cautionary summary must be kept in mind, when we come later to Ibn ‘Askar on the Jewish question (see below, p. 80).

⁶¹ Aḥmad Bābā was writing in 1596. His composition, Nayl al-ibtihāj bi-taṭrīz al-dībāj, published in the margins of Ibn Farhūn al-Ya'murī, Al-dībāj al-mudhahhab fī ma'rifat a'yān ‘ulamāʾ al-madhhab (Cairo, 1351/1932–1933, new edition c. 1975).Google Scholar We have not yet seen the Arabic text, and take the quoted English (p. 331 in the Arabic) from Hunwick's letter of 12 March 1991 (all brackets are his). See also his Sharī'a, 38, 41, 42. Aḥmad Bābā, who spent many years in exile in Morocco, might have learnt there of the visit; if so, this would reinforce still further the idea of censorship, or a news black-out, at the Timbuktu end.

⁶² Gomez, (‘Timbuktu’, 22)Google Scholar, pursuing his argument that the influence of Timbuktu over the askiya dynasty and over Gao has hitherto been exaggerated, both in the Timbuktu chronicles, and in secondary literature drawing heavily upon those chronicles, suggests that the clerical advisers whom the askiya sought out were not those from Timbuktu. ‘Rather, there was al-Maghīlī himself, while the Askia had consulted with al-Suyūtī in Cairo during his pilgrimage. But his principal advisers were in Gao…’ Our argument concurs with Gomez, insofar as the initial predominance of Gao clerics is concerned, certainly at the time of the askiya's departure on pilgrimage, and presumably in the years between the coup and the pilgrimage. But we doubt whether the askiya could have continued in close harmony with the Gao clerics during the ascendancy of al-Maghīlī in the askiya's favour. (Gomez opening words in this portion of his article are these: ‘ Al-Maghīlī did stress to the Askia the need to be well advised. Ibn al-Mukhtar [the T/F– the French page reference should be 114–15, not 118] records that the Askia was obedient to this admonition.’ The unwary reader might infer from this that the T/F here mentions al-Maghīlī, which it does not; and indeed, despite the rococo language concerning the first askiya's piety, al-Maghīlī might well have blanched at the extravagant largesse heaped by the ruler upon his clerics.) And we doubt, too, whether the religious pre-eminence of Gao was fully restored after al-Maghīlī's departure: all the names of Gao notables mentioned by Gomez here are rather early (but see n. 39 above).

⁶³ T/S, 35–6/20–1, 42/25; Hunwick, , Sharī'a, 14–15.Google Scholar

⁶⁴ T/S, 36/21.

⁶⁵ Hunwick, , Sharī'a, 16Google Scholar & n., citing Hopkins, J. F. P. and Levtzion, N. (eds.), Corpus of Early Arabic Sources/or West African History (Cambridge, 1981), 178.Google Scholar It is greatly to be regretted that this quintessentially important book has been allowed to go out of print.

⁶⁶ In Hopkins, and Levtzion, (eds.), Corpus, 282.Google Scholar

⁶⁷ Ibid. 290.

⁶⁸ ‘Ṣanhājah scholars of Timbuctoo’, Bulletin of the School of Oriental and African Studies, xxx (1967), 634–40.Google Scholar

⁶⁹ Ibid. 636.

⁷⁰ See above, pp. 74–5.

⁷¹ Hunwick, John O., ‘A new source for the biography of Ahmad Bābā al-Tinbuktī (1556–1627)’, Bulletin of the School of Oriental and African Studies, XXVII (1964), 581 n. 126.Google Scholar Norris also mentions the occasional use of al-Masūfī as nisbah among the Aqīts (‘Scholars’, 636).

⁷² T/S, 67/41. Here the French misprints al-ʿĀqib as al-'Aqīt. He is the only Messoufite in the index, though there is at least one more, 62/38.

⁷³ T/S, 58/35–6; Norris, ‘Scholars’, 636; Hunwick, , Sharī'a, 108.Google Scholar

⁷⁴ Hunwick, , Shari'a, 86–7Google Scholar; the first brackets are Hunwick's, the second ours. We return to the Masūfah portion of the sixth question later, in our conclusion; for references to its later citation and interpretation, see below, p. 90.

⁷⁵ Ibid. 106.

⁷⁶ It is just conceivable, though highly unlikely, that the askiya is here looking back to military action in which he was involved as a loyal officer still of Sunni Ali. But the words are those of a supreme ruler: ‘I stayed my hand from them.’

⁷⁷ Hunwick, , Sharī'a, 87 n. 3.Google Scholar Notice the clear implication of the Replies text that the askiya had been in consultation with the clerics of Timbuktu (for it is improbable that the clerics of Gao, predominantly western Sudanese, would have evinced such concern about Masufah nomads). We do not wish to suggest that the askiya, in seeking first the support of Gao clerics, and later of al-Maghīlī, had ever entirely cut himself off from what was for him, in the long run, the only viable religious option – scholarly, but unreformed, Timbuktu. This tiny detail, that the askiya was confused by clerical advice concerning the Masūfah, also weighs somewhat against Gomez’ thesis, that Timbuktu exercised little influence over even the first askiya.

⁷⁸ Ibid. 88 (Hunwick's brackets).

⁷⁹ A critical edition and translation of al-Maghīlīʿs major work in this connection is included in Gwarzo's 1972 thesis, ‘Life and teachings’; this is a crucial source (see n. 15 above). See also Hunwick, John O., ‘Al-Maghīlī and the Jews of Tuwāt: the demise of a community’, Studia Islamica, LXI (1985), 155–83CrossRefGoogle Scholar; and his ‘The rights of dhimmīs to maintain a place of worship: a 15th century fatwā from Tlemcen’, Al-Qanṭara (Madrid), xii (1991), 133–55Google Scholar, an article dedicated to the memory of Dr Gwarzo. Also Hunwick's, Sharī'a, for example 32, 35.Google Scholar

⁸⁰ T/F, 119–21/62–3; Levtzion, , ‘Chronicle’, 592.Google Scholar

⁸¹ l'Africain, Jean-Léon, Description de l'Afrique, ed. and trans. Épaulard, A. et al. (3 vols.) (Paris, 1956), II, 468;Google Scholar our translation from the French.

⁸² This quotation is taken from Hunwick, , ‘Rights’, 150–1 and n. 37.Google Scholar The Arabic is in Muḥammad bin ‘Askar al-Shafshāwanī, Daw ḥ at al-nāshir li-ma ḥāsin man kāna bi'l-maghrib mirt mashāyikh al-qarn al-'āshir, ed. Ḥijjī, Muḥammad (Rabat, 1397/1977), 96.Google Scholar See also Hunwick, , Sharī'a, 42 and 141.Google Scholar

⁸³ Hunwick, , Sharī'a, 41Google Scholar; see also above, p. 74 and n. 60.

⁸⁴ See, for example, Fisher, H. J., ‘Leo Africanus and the Songhay conquest of Hausaland’, Int. J. of Afr. Hist. Sṭudies, XI (1978), 86–112.CrossRefGoogle Scholar

⁸⁵ Hunwick, , Sharī'a, 42Google Scholar, thinks a local feud, with Muslims and Jews on both sides, a more likely explanation for the death.

⁸⁶ Hunwick, , Sharīa, 42.Google Scholar

⁸⁷ Ahmad Bābā, Nayl (see above, n. 61), cited in Ibid. 42. Another version of the same incident is recounted by oral traditions of the Kunta Arabs: al-Maghīlī demanded of the askiya that he send a force to Tuat to kill all the inhabitants; the king suggested that only the guilty ones be punished, and on al-Maghīlī's insistence upon the massacre of the whole town he refused, whereupon al-Maghīlī left for Katsina in indignation (Batran, ‘Contribution’, 393; see above, n. 15). Although we might not accept these facts as reliable in detail, this version also expresses the difference in political interest between the askiya and al-Maghīlī. The Kunta traditions, as quoted by Batran, do not indicate to what extent the ruler might have been under the influence of Timbuktu in this instance.

⁸⁸ Saying this, we assume that other Songhay towns had a Tuati and/or Jewish population, for which however we have no confirmation.

⁸⁹ For full details and references, see Hunwick, John O., ‘Al-ʿAqib al-Anuṣammanīʾs replies to the questions of Askiya al-Ḥājj Muḥammad: the surviving fragment’, Sudanic Africa, II (1991), 139–63.Google Scholar

⁹⁰ Nayl, 217–18; quoted in Hunwick, , ‘A1-'Anuṣammanīʾ, 139–40, n. i.Google Scholar

⁹¹ 67/41; see also n. 72 above.

⁹² For a sample of al-Suyūṭīʾs opinions, directly relevant for our general area, see Hunwick, John O., ‘Notes on a late fifteenth-century document concerning “al-Takrūr”’, in Allen, Christopher and Johnson, R.W. (eds.), African Perspectives (Cambridge, 1970), 7–33.Google Scholar

⁹³ Saad, , Social History, 66.Google Scholar We are not sure how far Saad here refers specifically to al-Anuṣammanī, who was a Saharan, not a trans-Saharan, master; and is the ‘particular generation’ that of al-Anuṣammanī himself, or of the chroniclers busy in the seventeenth century?

⁹⁴ Hunwick, , ‘Al-Anuṣammanī’, 159.Google Scholar

⁹⁵ Ibid. 161.

⁹⁶ Ibid. 160.

⁹⁷ Ibid. 161.

⁹⁸ Ibid. 160–1.

⁹⁹ Readers in Britain will notice the parallel with the present Prime Minister, John Major, now his own man, after having been a Thatcher acolyte.

¹⁰⁰ Gomez, ‘Timbuktu’, does explore this further field, though without our central religious focus. Gomez summarizes the various early options open to the first askiya, accurately enough in our view: ‘in the first seven years of his reign, Askia Muḥammad solidified his position by transferring the spiritual basis of authority from Songhay magic to the baraka of the ḥājj, won polemical support for his coup from the illustrious al-Maghīlī, and considerably fortified his alliance with the Timbuktu scholarly community via the elevation of the Sankore faction (16)’. Where we differ with his interpretation is that he seems to present all this as a kind of unified package deal, where we see hotly contested points of difference, boycotts and black-outs, and mutually exclusive choices, with some of the askiya's manoeuvres weakening rather than solidifying his position.

¹⁰¹ ‘The role of the Wangara’, J. Afr. Hist., xix (1978), 176–7.Google Scholar

¹⁰² An interesting illustration of Timbuktu/Tuat shared interests in trade survives from the time of Aḥmad Bābā, when merchants in Tuat sought his advice about the morality of the slave trade (at least insofar as professedly Muslim slaves were involved); see Barbour, Bernard and Jacobs, Michelle, ‘The Mi'raj: a legal treatise on slavery by Ahmad Baba’, in Willis, J. R. (ed.), Slaves and Slavery in Muslim Africa, (London, 1985, 2 vols.): vol. I, Islam and the Ideology of Enslavement, 125–59.Google Scholar

¹⁰³ Hunwick, , Sharī'a, 71, p. 16Google Scholar, line 5 of the Arabic. Gomez, ‘Timbuktu’, citing this passage, suggests that the askiya claimed ‘to have dethroned’ Sunni Ali, but the text does not seem to bear this gloss.

¹⁰⁴ 235/152 line 10.

¹⁰⁵ T/S, 116/71. There is a problem here, inasmuch as in November all small rivers in the area are dry; Rouch, J.-P., Contribution à l'histoire des Songhay (Mémoires de l' I.F.A.N., No. 29) (Dakar, 1953), 185–6.Google Scholar

¹⁰⁶ 98–9/51. A similar story is recounted by an oral tradition of the Zarma-Songhay, but without the causal sequence of the prayers of the persecuted leading to the death of the persecutor; Laya, Dioulde, Traditions historiques Zarma-Songhais (Colloque Fondation SCOA) (Niamey, 1977), 39–41.Google Scholar

¹⁰⁷ Sonni Ali Ber (Études nigériennes no. 40) (Niamey, 1977), 133.Google Scholar

¹⁰⁸ Contribution, 187–8. Here the assassination takes place on the festival prayer ground, in the midst of the festival prayers, as if to emphasize how heinous the event was by Muslim standards – the askiya is clearly the killer, but whether Sunni Ali himself, or his son and very temporary successor, was the victim is not so explicit.

¹⁰⁹ T/F, 56/33, 212–13/116–17; in Dāwūd's case, a penitential visit to the Prophet's tomb is specified, and formal pilgrimage is not mentioned.

¹¹⁰ Gamal Abdul Nasser, President of Egypt, urged a change in our understanding of the pilgrimage, which should no longer ‘be regarded as only a ticket of admission into Paradise after a long life, or as a means of buying forgiveness after a merry one’ (Egypt's Liberation: the Philosophy of the Revolution [Washington, D.C., 1955], 112).Google Scholar

¹¹¹ Rouch, (Contribution, 194)Google Scholar cites an oral tradition which says that the askiya set out for Mecca pour y obtenir le pardon d'Allah, ‘in order to obtain there the forgiveness of God’. But it would be rash to fit this into the theme of penitential pilgrimage, since we do not know the resonance of the words and concepts lying behind the French, and this particular tradition does take outlandish liberties with other aspects of the askiya's ḥajj.

¹¹² T/F, 126–8/66–7.

¹¹³ Rouch, , Contribution, 183–4.Google Scholar

¹¹⁴ See above, n. 45.

¹¹⁵ Hunwick, , Sharī'a, 78Google Scholar: jihād against those who ‘have confounded the truth with falsehood in such a way as to mislead many of the ignorant Muslims so that they become unbelievers without realizing it’ is more worthy than that against ‘unbelievers whom no Muslim would imitate’.

¹¹⁶ T/F, 117/61. Gomez takes a tentative step in our direction here, suggesting that the circumstances of this unusual request may perhaps have concerned ‘the way in which Muḥammad Toure had seized the throne’ (‘Timbuktu’, 18).

¹¹⁷ See above, p. 73.

¹¹⁸ See above, pp. 67–8.

¹¹⁹ For further indications of the revolution's western links, see note 101 above.

¹²⁰ T/S, 121–2/74.

¹²¹ For a brief discussion of the terms ‘mixed’ and ‘reformed’, see Fisher, Humphrey J., ‘The juggernaut's apologia: conversion to Islam in black Africa’, Africa, LV (1985).Google Scholar

¹²² See above, pp. 77–9.

¹²³ F. H. Masri (ed.), (Khartoum University Press/Oxford University Press, 1978), 134; this is the longest of three references to al-Maghīlī in the Bayān.

¹²⁴ For a full translation of al-Gunahānīʿs apologia, see Appendix 2 in Norris’ Conquest (n. 15 above). See also Fisher, H. J., ‘The vanquished voice: Shaykh Muḥammad's defence: a case study from the western Sahara’, in Henige, David and McCaskie, T. C., (eds.), West African Economic and Social History: Studies in Memory of Marion Johnson (Madison, 1990), 47–61.Google Scholar

¹²⁵ See above, n. 5.