Published online by Cambridge University Press: 22 January 2009
It is generally accepted that Ibn-Ḥawqal crossed the Sahara and visited Awdaghost in 951/2 to be the first Arab geographer who reached the gates of bilād al-Sūdān. This is inferred from what he says about a cheque he saw in Awdaghost. A critical analysis of the references to this cheque suggests that it was a bill of debt owed by a trader from Sijilmāsa resident at Awdaghost to another trader from Sijilmāsa. The transactions between the two were part of the trans-Saharan trade, with one partner resident at Awdaghost and the other at Sijilmāsa. The bill must have been held by the creditor, whom Ibn-Ḥawqal could have met at Sijilmāsa.
It was from this man that Ibn-Ḥawqal recorded the information about the ruler of the Ṣanhāja in the southern Sahara, and probably also about his relations with Ghana and other Sudanese polities. Another valuable piece of information, about the abandoned route from Ghana to Egypt, was collected in the oases of Egypt. Also, the detailed list of Berber tribes inhabiting the area between Tadmekka and Air could have been obtained in North Africa, rather than at Awdaghost, because communications in the Sahara were more frequent between north and south than between west and east.
Ibn-Ḥawqal's description of the Sahara is stereotypic and adds nothing to what earlier geographers had to say, and the same applies to his words concerning Awdaghost itself. Towns and kingdoms in the Sudan mentioned by Ibn-Ḥawqal had already been recorded earlier by al-Yaʻqūbī, Ibn al-Faqīh and al-Masʻūdī. For a traveller who crossed the Sahara, Ibn-Ḥawqal's information is rather poor, but if he collected his information north of the Sahara, as this paper suggests, his labours as an inquisitive geographer are praiseworthy.
1 On the development of Muslim geography, see Kramers, J. H., ‘Geography and commerce’, in The Legacy of Islam, ed. by Arnold, T. and Guillaume, A. (London, 1931);Google ScholarAhmad, Nafis, Muslim Contribution to Geography (Lahore, 1947);Google ScholarAhmad, S. Maqbul, ‘Djughrāfiyā’, in The Encyclopaedia of Islam, new edition, III, 575–87.Google Scholar
2 Quoted by al-Mas'ūdi, , Murūj al-dhahab wa-ma'ādin al-jawhar (Les prairies d'or), text and trans. by de Meynard, C. Barbier and de Courteille, Pavet, IV (Paris, 1861–1977), 37–9.Google Scholar
3 al-Khwāizmī, , Kitāb ṣūrat al-arḍ, ed. von Mžik, Hans (Leipzig, 1926), 6.Google Scholar
4 Al-Ya'qūbī, , Kitāb al-buldān, ed. de Goeje, M. J. (Lugduni Batavorum, 1892; BGA VII), 345;Google Scholaridem, Les pays, trans. Wiet, G. (Cairo, 1937), 205.Google Scholar
5 References to the Sudan are available in two works by al-Ya'qūbī. He first wrote his historical work (Kitāb al-Ta'rīkh, ed. Houtsma, M. Th., Lugduni Batavorum, 1883), which he completed probably in 872. Later, when he was in Egypt, he wrote the geographical work (K. al-buldān, quoted above), which he completed in 891.Google Scholar (See Brockelmann's, C. article ‘al-Ya'qūbī’, in The Encyclopaedia of Islam, 1st ed. vol. IV;Google ScholarWiet's, G. Introduction in Les pays, op. cit. viii–ix.)Google Scholar It is significant that the long list of kingdoms and peoples in the Sudan appears in al-Ya'qūbī's earlier work (K. al-Ta'rīkh), which he probably wrote when he was still in the east.
6 There are two editions of Ibn-Ḥawqal's Kitāb ṣūrat al-arḍ: the first edition was prepared by M. J. de Goeje (Lugduni Batavorum, 1873), based on three MSS from Leiden, Paris, and Oxford; a second edition, which replaced the first edition as vol. 11 of the BGA, was prepared by J. H. Kramers (Lugduni Batavorum, 1938–1939) following the discovery of a better and more complete MS at the Old Seray Library in Istanbul. Kramers incorporated in his edition passages from de Goeje's edition which are not to be found in the Istanbul MS. Such is the case of the paragraph quoted here, which appears in de Goeje's edition on p. 42 and is reproduced in Kramers's edition on p. 61. (Throughout this paper the references are to Kramers's edition, unless otherwise indicated.) Following his edition of the Arabic text, Kramers prepared a French translation, which was completed and published after his death by Wiet, G. (Configuration de laterre (Beirut and Paris, 1964), 58). The French translation of this passage reads: ‘J’ai vu une reconnaissance de dette de Muhammad ibn Abi Sa'dun, d'Audaghust, contresignée etc.’ (my italics).Google Scholar
7 Text, 99; translation, 97−8. The translation here is somewhat misleading as it goes: ‘J’ai vu à Audaghust une reconnaissance de dette par laquelle un négociant d'Audaghust [the phrase ‘who was himself of the people of Sijilmāsa’ is omitted here] se reconnaissait débiteur envers un habitant de Sidjilmasa d'une somme de 42,000 dinars, etc.’ Because of the importance of this text it may be appropriate to quote the version of de Goeje's edition (p. 70):
‘I saw a bill in Awdaghost certifying a debt owed by a man from the people of Sijilmāsa to another man of the same town on the sum of 40,000 dinars, etc.’
8 Text, 100; compare the French translation (98): ‘J'ai entendu raconter par Abu Ishaq …, qui était à Audaghust, le bénéficiaire du chèque dont j'ai parlé.’ In de Goeje's edition this informant is referred to as Abū-Isḥāq Ibrāhīm ibn Faragha Shaghlahu, without associating him with the famous cheque (71).
9 Pérès, H., ‘Relations entre le Tafilelt et le Soudan à travers le Sahara du XIIIe au XIVe siècle’, Mélanges géographiques et orientalistes offerts à E.-F. Gautier (Tours, 1937), 409–54.Google Scholar
10 Text, 83, 99; translation, 79, 97.
11 Compare my translation of Ibn-Ḥlawqal's text (pp. 61 and 100) above with the French translation as quoted in notes 6 and 8 above. The disagreement in the translation shows the different meanings which may be attributed to the phrase bi-Awdaghost, according to the context and its position in the sentence. The disagreement of the MSS makes quite plausible the suggestion that the phrase bi-Awdaghost was inserted by mistake or misplaced.
12 Ibn-Baṭṭūṭa, , Tuḥfat al-nuẓār fī gharā'ib al-amṣār wa-'ajā'ib al-asfār (Voyages), Arabic text and Fr. trans. by Défrémery, C. and Sanguinetti, B. R., IV (Paris, 1922), 377–83.Google Scholar
13 Text, 103; translation, 101.
14 al-Ya'qūbī, K. al-buldān, 360 (Les pays, 226); al-Hamadhānī, Ibn al-Faqīh, Mukhtaṣar kitāb al-buldān, ed. de Goeje, M. J. (Lugduni Batavorum, 1885; BGA v), 81, 87;Google Scholaral-Iṣṭkhrī, , Kitāb al-masālik wa'l-mamālik, ed. de Goeje, M. J. (Lugduni Batavorum, 1870; BGA 1), 45.Google Scholar
15 Text, 101−7; translation, 99–105.
16 Lewicki, T., ‘A propos d'une liste de tribus berbères d'Ibn Hawkqal’, Folia Orientalia, 1, 128–35;Google ScholarLhote, H., Les Touaregs du Hoggar (Paris, 1955), 126–7.Google Scholar
17 G. Wiet's introduction to the French translation of Ibn-Ḥawqal, xii.
18 Text, 101–2; translation, 99–100.
19 Ibn–Baṭṭūṭa, , op. cit. 377.Google Scholar
20 Text, 92; translation, 90. His description of towns in the Maghrib which he visited is always short, but the few notes which characterize every town clearly suggest direct observation (e.g. text, 66–83; translation, 62–80).
21 al-Ya'qūbī, K. al-buldān, 360 (Les pays, 226).
22 Text, 100; translation, 98.
23 Text, 101; translation, 99.
24 Text, 9–10; translation, 9–10.
25 al-Iṣṭakhrī, op. cit. 4–5.Google Scholar
26 Kramers, J. H., ‘La question Balkhi–Istakhri–Ibn Hauqal’, Acta Orientalia, x, 9–30;Google Scholar Wiet's introduction to Ibn-Ḥawqal, x–xi, xiv–xv; Ahmad, S. Maqbul, op. cit. 582.Google Scholar
27 Text, 92; translation, 90.
28 al-Ya'qūbī, , K. al-Ta'rikh, 219–20;Google Scholaridem, K. al-buldän, 345 (Les pays, 205); al-Faqīh, Ibn, op. cit. 68;Google Scholaral-Mas'ūdī, , op. cit. III, 2;Google Scholaridem, Akhbār al-zamān, quoted from MSS in Monumenta Cartographica Africae et Aegypti by Kamal, Youssouf (Leiden, 1932), III, 629.Google Scholar
29 The name Kanem, which was known to earlier geographers, does not appear as such in Ibn-Ḥawqal's text. Instead there is a name (K.Z.M.), which is unidentified and does not occur again in the geographical literature. I would therefore incline to suggest that should be read, i.e. Kanem.
30 Text, 61, 64; translation, 58, 61.
31 al-Bakrī, , Al-masālik wa'l-mamālik; kitāb al-mughrib fī dhikr bilād Ifrīqīya wa'lMaghrib, ed. de Slane, M. G. (Algiers, 1911), 177;Google Scholaridem, Description de l'Afrique septentrionale, trans. de Slane, M. G. (Algiers, 1913), 331–2.Google Scholar
32 Text, 92; translation, 91.
33 al-Ya'qūbī, , K. al-Ta'rīkh, 219–20;Google Scholaral-Mas'ūdi, , Murūj al-dhahab, III, 2.Google Scholar
34 al-Mas'ūdī, , Akhbār al-zamān, 629.Google Scholar
35 al-Ya'qūbī, , K. al-buldān, 345 (Les pays, 205).Google Scholar
36 Text, 61; translation, 58. In the 8708 al-Ya'qūbī (op. cit. 359; Les pays, 225) did not record a flourishing trade in Sijilmāsa. The people of Sijilmāsa, he says, are mixed, most of them are Berbers, and among the latter the Ṣanhāja are the majority.
37 al-Faqīh, Ibn, op. cit. 68.Google ScholarMauny, R. (Tableau géographique de l'ouest africain au moyen òge (Dakar, 1961), 139) identifies Maranda with the water-point Marandet, south of Agades.Google ScholarLewicki, T. (‘A propos du nom de l'oasis de Koufra chez lea géographes arabes du XIe et du XIIe siècle’, J. Afr. Hist. VI (1965), 296) reads the name as Marinda, and seeks to identify it with one of the Toubou tribes, which lived in the region of Air before the fifteenth–sixteenth century. Both scholars, however, suggest that the route from Gao passed over Air, and then to Tibesti, where T. Lewicki (loc. cit.) identifies the location of the Marāwa.Google Scholar
38 al-Iṣṭakhrī, , op. cit. 52.Google Scholar
39 Text, 153; translation, 151. On these oases, see also Leo, Africanus, Description de l'Afrique (Paris, 1956), II, 457–8.Google Scholar
40 Trimingham, J. S., A History of Islam in West Africa (London, 1962), 22 n.Google Scholar
41 An implicit disappointment with Ibn-Ḥawqal's information may be traced in Trimingham (op. cit. I) and Mauny, (op. cit. 71–2, 215).Google Scholar