¹ This critical attitude need not deny the historical value of an interpretative approach to these traditions, myths, tales and their variants. Some variants are noted and discussed by Hallam, W. K. R. in J. Afr. Hist., vii (1966), 47–60CrossRefGoogle Scholar. His suggestion that the Bayajida legend may be fitted into a series of events which occurred both north and south of the Sahara in the tenth century A.D. is surely an attempt to be too specific. Standard secondary accounts incorporating these traditions can be found in Hogben, S. J. and Kirk-Greene, A. H. M., The Emirates of Northern Nigeria (London, 1966)Google Scholar; Crowder, M., The Story of Nigeria (London, 1962)Google Scholar; Fage, J. D., A History of West Africa (4th ed., London, 1969)Google Scholar; Trimingham, J. S., A History of Islam in West Africa (London, 1962), esp. 126 fGoogle Scholar. My own acquaintance with the primary written sources is through summaries, paraphrases and translations, which, I am warned, are frequently faulty. There is also – or has been — a considerable body of rural oral tradition in the various regions of Hausaland, much of it apparently contrasting in tone if not in detail too with the more official royal and town histories. Such local traditions are alluded to in some of the secondary literature but rarely presented in detail.

² ‘The beginnings of Hausa society’, in Vansina, J., Mauny, R. and Thomas, L. V. (eds.), The Historian in Tropical Africa (London, 1964), 339–57.Google Scholar

³ Smith, A., ‘Some considerations relating to the formation of states in Hausaland’, J. Hist. Soc. Nigeria, v (1970), 329–46Google Scholar. See also the same author's chapter, ‘The early states of the Central Sudan’, in Ajayi, J. F. A. and Crowder, M. (eds.), History of West Africa, i (2nd ed., London, 1976), esp. 156–8Google Scholar, 177 f. Fisher, H. J. in The Cambridge History of Africa, iii (1977), esp. 292 f.Google Scholar, follows this revisionist line very closely.

⁴ See Last, D. M., The Sokoto Caliphate (London, 1967)Google Scholar; Willis, J. R., ‘Jihad fi sabil Allah – its doctrinal basis in Islam and some aspects of its evolution in nineteenth-century West Africa’, J. Afr. Hist., viii (1967), 395–415CrossRefGoogle Scholar. Scholarship was certainly revered in these jihad societies – higher even than military prowess, so Last opines (in Ajayi, and Crowder, , History of West Africa, ii (1974), 29)Google Scholar. This is interesting as a reflection of an attitude of mind, if hardly an arguable point. And for the pious supposition that literacy and learning helped win battles in the Sudan, see Cambridge History of Africa, v (1976), 140.Google Scholar

⁵ Palmer, H. R. comments in his introduction to the Kano Chronicle (Sudanese Memoirs, iii, Lagos, 1928, 92Google Scholar) that other independent Hausa records and books were doubtless destroyed by the Fulani ‘in their fanatical zeal’. This may be so; but Palmer's assertion rests on negative evidence, and his anti-Fulani bias is notorious. Less informative historically than the Kano Chronicle, but noteworthy nevertheless, is the ‘Song of Bagauda’: see Hiskett, M. in Bull. S.O.A.S., xxvii (1964), 540–67Google Scholar; xxviii (1965), 112–35, 363–85, esp. 365–70. Another important Kano document, ‘The Wangarawa Chroniclè’, is believed to date to the seventeenth century: see Al-Hajj, M., Kano Studies, i, no. 4 (1968), 7 f.Google Scholar

⁶ Since my drafting of this paper there have appeared in this Journal for 1978 two articles reinterpreting various aspects of ‘middle-period’ Hausa history: Lovejoy, P. E., ‘The role of the Wangara in the transformation of the Central Sudan in the fifteenth and sixteenth centuries’, xix, 173–93Google Scholar; and Fuglestad, F., ‘A reconsideration of Hausa history before the jihad’, xix, 319–39.Google Scholar

⁷ Leo Africanus (or his editors) understood this essential fact in the sixteenth century: see the edition of Épaulard, A.et al. (Description de I'Afrique, Paris, 1956) 16Google Scholar. Like other early writers, Leo did not use the term Hausa. He called the language Gobir, probably because of the northerly orientation of his information. He contrasts it with the language of Songhai to the West and that of Bornu (i.e. Kanuri) to the east, without saying anything about local or relict languages. Whether we should conclude from Leo that by 1500 Hausa was the language of the Gobir countryside, or was rather merely a lingua franca extending into those areas with the trade and settlement from eastern Hausaland, is a moot point in view of the observations made below that Gobir should be among the areas of secondary Hausaization.

⁸ Infaku'l Maisur as translated/paraphrased by Arnett, E. J., The Rise of the Sokoto Fulani (London, 1922), here p. 11Google Scholar. This observation is echoed a century-and-a-half later by Dr Adamu, Mahdi (Savanna, v, 1976, 5Google Scholar): ‘ It is my working assumption that dialects do not exist in the Hausa language. The language is still, as it has always been, one solid piece.’ The essential point here is doubtless correct: Hausa lacks the degree of regional variation and dialect formation which exists in, say, Yoruba or Akan. Nevertheless, if we employ the term ‘dialect’ as normally understood, Dr Madi's statement needs modification, as Hausa-speakers, students of Hausa and indeed his own discussion agree. In particular the dialects of the Kebbi valley and Sokoto in the west (kabanci/sakkwatanci) are noted for their peculiarities, especially when compared to that of Kano which is commonly regarded nowadays the ‘standard’ form of Hausa.

⁹ Greenberg, J. H., ‘Linguistic evidence for the influence of Kanuri on the Hausa’, J. Afr. Hist., i (1960), 205–12CrossRefGoogle Scholar. As Greenberg hints (205–6), it would have been from this side (Bornu, that is) that the main popular Islamic impact was made upon Hausaland and its towns over time. This is not to dispute necessarily the information in the Kano Chronicle (which continues to impress itself on the general histories) that Islam was ‘introduced’ to Hausaland from Mali (Wangara) in the fourteenth century – and strengthened a century later by the arrival of Fulani mallams and a bigger selection of Muslim literature from the same direction. (For a critical discussion of the chronicle references, see Abdullahi Smith in Ajayi and Crowder, I (2nd ed.), 191; and Fisher, in Cambridge History of Africa, iii, 295–6Google Scholar; as well as Hiskett, M., ‘The historical background to the naturalization of Arabic loan-words in Hausa’, Afr. Lang. Studies, vi (1965), 18–26.Google Scholar) It would indeed be surprising if during and after the heyday of the Mali empire (and again around 1500 when Songhai power pressed strongly against Hausaland) the influence of the schools of the Upper Niger had not been felt. And to suggest that the received view is the result of nineteenth-century Fulani propaganda, emphasizing a western origin for Islam against a Bornuan one, would probably be too facile! Nevertheless, this issue illustrates the fragility of isolated literary references unless the available cultural and linguistic controls are applied – as well as the naivety of the notion that Islam simply ‘arrived’ from a particular direction at a particular time. See also Hiskett, , ‘Song of Bagauda’ (loc. cit., 1965), 369, 382–5Google Scholar, on gradual Islamization; and Lovejoy, ‘Role of the Wangara’, for a recent discussion of this issue.

¹⁰ In particular the Kano Chronicle and Leo Africanus, on whom see note 7 above. (The question of Leo's reliability is not an issue at this level of discussion.) For these and other documentary sources see Hunwick, J., ‘Songhay, Borno and Hausaland in the sixteenth century’, in Ajayi, and Crowder, I (2nd ed.), esp. 274 f.Google Scholar

¹¹ Westermann, D. and Bryan, M. A., The Languages of West Africa (I.A.I., 1952)Google Scholar; Greenberg, J. H., The languages of Africa (Indiana and The Hague, 1963Google Scholar); ‘The Languages of Nigeria by language families’, cyclostyled list compiled by Hoffmann, C. (Ibadan, 1974)Google Scholar; Hansford, K., Bendor-Samuel, J. and Stanford, R., ‘A provisional language map of Nigeria’, Savanna, v (1976), 115–24 with pull-out.Google Scholar

¹² One thinks of Polly Hill's attempt to redress the balance by emphasizing the rural aspect of present Hausaland and the need to understand the rural economy (notably in Hausa, Rural, Cambridge, 1972Google Scholar; and spelt out further in her Kano, Rural, Cambridge, 1977Google Scholar): her observations have considerable implications for the past, and help us to transcend the simplistic ‘city-state’ notion with disproportionate accent on centralized markets and long-distance trade. But this is not to deny the essential cultural and economic role of the towns, both the capital birane (sing, birnt) and the smaller gararuwa (sing, gari) with their markets. Urban–rural complementarity must be seen as a prime factor in the geographical continuity and multi-state system of Hausaland, and dichotomy between town and country would doubtless have been less marked in earlier times when towns were smaller.

¹³ See Ballard, J. A., ‘Historical inferences from the linguistic geography of the Nigerian middle belt’, Africa, xli (1971), 294–305.CrossRefGoogle Scholar

¹⁴ Sutton, J. E. G., ‘Kufena and its archaeology’Google Scholar, and Bala, C. D., ‘The archaeology of the Zaria region inselbergs’, Zaria Archaeology Papers, 15 and 16 (cyclostyled, 1977)Google Scholar. Greenberg, J. H., The Influence of Islam on a Sudanese Religion (New York, 1946)Google Scholar, remains an important contribution to the cultural and religious background of rural Hausaland.

¹⁵ Assuming that the kingdom of Zamfara mentioned by Leo about 1500 was in roughly its present position and was also Hausa or Hausaizing, this fourteenth/fifteenth-century suggested dating must be the later limit. Trimingham, (History, 129, 240)Google Scholar seems to regard Zamfara as an established powerful state no later than A.D. 1300, but he does not specify his evidence for this view. Presumably he had access to the king-lists and other information in the cyclostyled Sokoto Provincial Gazetteer (compiled by Harris, P. G., c. 1938Google Scholar) and to some of Krieger's, K. work. In Geschichte von Zamfara (Baessler-Archiv, N. F., Beiheft I, Berlin, 1958Google Scholar), Krieger notes (p. 9) that the Bodleian MS of the twelfth-century Kitab Rujar of Idrisi substitutes the name Zamfara for Semegonda. Was this done by a late and over-zealous scribe (one who had read Leo)?

¹⁶ If Lovejoy is correct in identifying Leo's ‘Guangara’ with Kebbi, (‘Role of the Wangara’, 182–3)Google Scholar – an interesting suggestion if rather cavalierly proposed – then we have further reason to admit a fair degree of Hausaization in the west by 1500, especially since Leo (p. 16) places Guangara in the Hausa-speaking, not Songhai-speaking, zone. But if we continue to equate Guangara with Katsina Laka or some other part of Hausaland – see note 75 – the issue does not arise. Lovejoy's hypothesis might be considered consistent with Fisher's arguments (see note 20, below) on the unreliability of Leo Africanus and doubts concerning the supposed conquest of Hausaland by the Songhai empire in Askia Muhammad's time, in that much of Leo's muddled information could perhaps relate to the feats of Kebbi in its assertive period in the very early sixteenth century.

¹⁷ Notably by Fisher, in Cambridge History of Africa, iii, 294 fGoogle Scholar. Compare Smith in Ajayi and Crowder, i (2nd ed.), 190 f. Lovejoy and Fuglested (see note 6) carry the discussion further.

¹⁸ Hiskett, , ‘The historical background’, esp. 20.Google Scholar

¹⁹ Travels in Asia and Africa, 1325–1354 (ed. Gibb, H. A. R., London, 1929), esp. 336Google Scholar. For recent archaeological corroboration of the Tagadda copper-workings, see Bernus, S. and Gouletquer, P., ‘Du cuivre au sel’, J. Soc. Africanistes, xlvi (1976), 7–68.CrossRefGoogle Scholar

²⁰ See Fisher, H. J., ‘Leo Africanus and the Songhay conquest of Hausaland’, Int. J. Afr. Hist. Stud., xi (1978), 86–112.CrossRefGoogle Scholar

²¹ Sudanese Memoirs, iii, iii; part of the passage is reprinted and annotated in Hodgkin, T., Nigerian Perspectives (2nd ed., London, 1975), 113–15.Google Scholar

²² Jaggar, P. H. in Savanna, ii (1973), 24–5Google Scholar; Krieger, K., ‘Notizen zur Eisengewinnung der Hausa’, Zeitschrift für Ethnologie, lxxxviii, 1963, 18–31Google Scholar. Also in Adar in the Sahel: Echard, N. in J. Soc. Africanistes, xxxv (1965), 353 f.CrossRefGoogle Scholar

²³ Sutton, J. E. G., ‘Iron-working around Zaria’, Zaria Archaeology Paper 8 (1976).Google Scholar

²⁴ Sudanese Memoirs, iii, 97–9.

²⁵ I am indebted to Alhaji Garba Na-Dama of Gusau and Ahmadu Bello University for discussing with me his research findings on Zamfara.

²⁶ There is no ‘tsetse line’, but rather a number of tsetse-infested areas both within Hausaland and further south. The factors — climatic, vegetational, human, etc – determining this are complex. The boundaries moreover are unstable. See Ford, J., The Role of the Trypanosomiases in African Ecology (Oxford, 1971), ch. 21 f.Google Scholar

²⁷ On Fulani economy and movement, see Stenning, D. J., J.R.A.I., lxxxvii (1957), 57–73.Google Scholar

²⁸ The passage in the Kano Chronicle referred to above (n. 21) includes among the new developments occurring in the mid-fifteenth century the arrival of Fulani scholars in Hausaland and Bornu. Even if this be roughly correct chronologically, it does not necessarily imply the migration of cattle Fulani to northern Nigeria by that time. Nor does it rule it out.

²⁹ John Lavers of Kano informs me that comparative study of the Chadic languages as a whole suggests that the early speakers of this language family were very dependent on fish in the old Chad basin. Though that would have been several millennia before the emergence of Hausa specifically, this observation may have a slight bearing on the culture and economy of the early Hausa in the river valleys east of Kano. See also note 33 below.

³⁰ See Law, R., ‘Horses, firearms and political power in pre-colonial West Africa’, Past and Present, lxxii (1976), 112–32CrossRefGoogle Scholar. This article argues that heavy cavalry techniques were introduced to the Sudan belt about the fourteenth century. Interestingly the next two centuries are the period which witnessed the consolidation of larger and more formal states in Hausaland and Bornu and the erection of big city walls enclosing large areas. More modest town walls may have been built earlier. See Sutton, J. E. G., ‘The walls of Zaria and Kugena’, Zaria Archaeology Paper iiGoogle Scholar; and Moody, H. L. B., The Walls and Gates of Kano City (Federal Department of Antiquities, Lagos, c. 1970)Google Scholar. For Bornu, Bivar, A. D. H. and Shinnie, P. L., ‘Old Kanuri capitals’, J. Afr. Hist., iii (1962), 1–10CrossRefGoogle Scholar; G. Connah, ibid, xvii (1976), 349–50. While the rise of Bornu in the fifteenth century doubtless affected the maintenance and development of town defences in Hausaland – and Bornu was renowned for its cavalry – this is unlikely to explain the origin of Hausa town walls.

³¹ Notes 9 and 37 are relevant here.

³² See above, and the works cited in note 3.

³³ See Sutton, J. E. G., ‘The aquatic civilization of middle Africa’, J. Afr. Hist, xv (1974), 527–46.CrossRefGoogle Scholar

³⁴ It is true that Nicolaisen, J.(Structures politiques et sociales des Touareg de l'Aïr et de l'Ahaggar, IFAN-Niamey, 1962, trans, from English, 19Google Scholar) refers to an old agricultural black and Hausa-speaking population in Aïr, which has now adopted Tuareg but retains part of its original vocabulary. Unfortunately the evidence for its having been Hausa is not elaborated. An alternative, Nilo-Saharan, speculation is offered below.

³⁵ This process would have begun, I suggest, some five-thousand years ago, Chadic emerging (from the Afroasiatic super-family and also through interaction with Nilo-Saharan fishing populations) as an adaptation to the waters and swamps of Lake Mega-Chad as the harshness of the Sahara became progressively more acute. (See Sutton, ‘The aquatic civilization’.) From about 1000 B.C. some of these Chadic communities, finding the lake declining further, would have turned towards farming the swamps and valleys to the west and south-west. (The Daima sequence is interesting here: see Connah, G. in J. Afr. Hist., xvii, 1976, 321–52.CrossRefGoogle Scholar) This would have been the western division of Chadic from which Hausa later emerged. So, desiccation is in a way relevant, but not for Hausa specifically, nor as Smith imagines it.

³⁶ Smith, A., ‘Some considerations…’ (1970), 331.Google Scholar

³⁷ The example of Kanuri, developing as the dominant language of Bornu in the savanna to the east of Hausaland, may be instructive. The divergence of Kanuri from Kanembu is connected historically with the movements from Kanem, lying east of Lake Chad, to Bornu on the western side around the fourteenth century. While this process may in fact have begun a little earlier, it seems that Kanuri has been developing and spreading as an independent tongue for a rather shorter period than has Hausa. Nevertheless, the divergence of Kanuri from Kanembu is (so I am informed) probably no less than that between eastern and western Hausa dialects.

³⁸ This is the ‘ I.A.S. I’ of Hoffman (see note II above), building on Greenberg's scheme. The speakers of these languages seem to equate roughly if not exclusively with the ‘Gwari’ of Hausa tradition. However, the specific Gwari (or better ‘Gbari’) language and people (see next note) do not belong to this ‘Plateau’ group. There has been some terminological confusion over time.

³⁹ The recent northward push of Gbari to meet with Hausa has resulted in the separation of the westerly surviving ‘Plateau’ languages (Kamuku, Kambari, Dakarkari, Reshe, etc.) from the rest. Ballard, ‘Historical inferences’, supplies useful observations and comments on this linguistic situation and its historical implications.

⁴⁰ The opening section of Alkali, M. Bello, A Hausa Community in Crisis: Kebbi in the Nineteenth Century (M.A. thesis, Ahmadu Bello University, 1969Google Scholar) is useful here: see esp. pp. 26–9, 44–7, on the riverain Sorko, who were originally Songhaic-speaking but in time assimilated into western Hausa society. (See also my paper ‘The aquatic factor in northern Nigerian history’ to the seminar, Kano on ‘The Economic History of the Savanna’ in 1976Google Scholar: publication expected.) Songhai speech may have been given a boost in this region during the ascendancy of the Songhai empire and the status of Kebbi as its south-eastern province at the turn of the sixteenth century. But Bello Alkali's assumption that the rural population of Kebbi was basically Hausa before this period seems to me unsupported. It may be inspired by the Kebbi claim to be the ‘Original Hausa-speaking people’: O. and Temple, C. L., Notes on the Tribes … of Northern Nigeria (2nd ed., 1922), 404 nGoogle Scholar. Apparently certain traditions of the old Kebbi kingdom – including this claim of Hausa originality – have been maintained by the small new Kebbi kingdom (the Argungu emirate) which managed to preserve an independence of sorts from Sokoto and Gwandu in the nineteenth century.

⁴¹ Muhammad Bello records the tradition that Kebbi was born of a Katsina mother and a Songhai father (Arnett, 13). More specifically Kanta, the warrior-king to whom Kebbi's independence of Songhai is usually attributed, is thought by some to have hailed from central Hausaland, perhaps Katsina-Laka (see Sudanese Memoirs, iii, 76). The independence which Kebbi won in the early sixteenth century and its position as a powerful kingdom for the next hundred years or more between the Songhai and Bornu spheres is summarized by Hunwick in Ajayi and Crowder, i (2nd ed.), 283–5. For preliminary archaeological investigations of Kebbi forts and capitals (lying close to the Kebbi river slightly west of Sokoto town), see Obayemi, Ade and Sutton, J. E. G., Zaria Archaeology Papers, 2 and 5 (1976).Google Scholar

⁴² Urvoy, Y., Histoire des populations du Soudan central (Paris, 1936)Google Scholar; Last, D. M., The Sokoto Caliphate, ix.Google Scholar

⁴³ Smith, A., ‘Hausaland’, 332–3Google Scholar; Urvoy, op.cit.

⁴⁴ See Sutton, , ‘The aquatic civilization.’ Interestingly, Lovejoy (‘Role of the Wangara’, 182, citing works not available to me)Google Scholar mentions Songhai dialects persisting in Aïr until the present. Whether this is ancient Songhaic, or simply the effect of Songhai influences in the middle of the present millennium, remains unclear to me.

⁴⁵ In Infaku'l Maisur: Arnett, 9–10. For here and below one can compare another English version of this work in Denham, D. and Clapperton, H., Narrative of Travels and Discoveries in Northern and Central Africa, 1822–24, ii (London, 1826), esp. 161–4.Google Scholar

⁴⁶ Arnett, 11–12.

⁴⁷ In Raudhât'ul Afkâri: Palmer, H. R., ‘Western Sudan history’ J. Afr. Soc., xv (1916), 264–6Google Scholar. (Palmer attributes the work to Bello: this is corrected by Last, 249, and Smith, loc. cit. 336).

⁴⁸ Azben – or more correctly Abzin – in Hausa folklore. It may be significant that while both Bello and Abd al-Qadir state that this region was once inhabited by black Gobirawa, neither insists that they spoke Hausa then or there. Nor does the Kano Chronicle, which simply records the tradition that ‘the Azbenawa came to Gobir’ about the fifteenth century (Sudanese Memoirs, iii, iii). Palmer, (‘Western Sudan history’, 53Google Scholar) reads more into this when he claims ‘that Gobirawa or Gobir Hausas controlled the Asben Oasis’ previously. This reading has misled later scholars. Lovejoy in his recent article (‘Role of the Wangara’, 185) takes these Gobirawa to have been Tuareg. This interpretation is possible but unlikely, I think, in view of the traditions later recorded at Sokoto. Surprisingly, Lovejoy seems unaware of his unorthodoxy here – and two pages later, contradictorily, he repeats the standard line of Tuareg nomads enforcing Gobir's evacuation of Aïr for the south (and, by misplacing the footnote indicator, appears to attribute this assumption, erroneously, to Leo).

⁴⁹ See, for instance, Adamu's, Mahdi map of the ‘historical boundaries of Hausaland’ in Savanna, v (1976), 4Google Scholar, in which the surface area of Hausaland is doubled in order to accommodate Aïr and Asuda, the ‘pre-15th century capital of Gobir’. It is noteworthy that Heinrich Barth, who visited the much reduced town of ‘Asodi’ in the mid-nineteenth century (Travels, i, 375–6), said nothing about its ruins being associated with Gobir.

⁵⁰ See Hunwick's account in Ajayi and Crowder, 2nd ed., 281. Palmer, (Sudanese Memoirs, iii, 84–5Google Scholar) seems to prefer an early, twelfth-century, date for the expulsion from Aïr and apparently sees the move as rather sudden. But in his commentary on the ‘Asben Record’ (The Bornu Sahara and Sudan, London, 1936, 73) a later date is impliedGoogle Scholar. That Gobir may not be entirely Hausa in ethnic origin, its rulers being perhaps of noble Tuareg or Zaghawa extraction, is hinted by Hogben and Kirk-Greene (188). Smith, Abdullahi (in Ajayi and Crowder, 2nd ed., 186Google Scholar) notes this in passing.

⁵¹ This was written before seeing Fuglestad's article ‘A reconsideration of Hausa history’. Following a remark by Barth, (Travels, 1, 336)Google Scholar and also information collected by Landeroin (of the ‘Mission Tilho’ of 1906–9), he suggests that it was not the Gobirawa as a whole but simply their ruling family which came south from Aïr. This is going one step further than the hints just noted. It remains consistent with the Kano Chronicle's statement which (pace Palmer) need not imply that any Hausa as such moved from Aïr (Azben) to Gobir. But if we are right in assuming the existence of some Nilo-Saharan-speaking populations in Aïr and Tudu into the present millennium, then the whole issue is rather more subtle.

⁵² Bello (in Arnett, 12) puts the point succinctly.

⁵³ For the Sokoto angle, see Last, , The Sokoto Caliphate; and Johnston, H. A. S., The Fulani Empire of Sokoto (London, 1967).Google Scholar

⁵⁴ For some initial probings see Zaria Archaeology Papers, 1, 2 (Obayemi) and 5, 11, 16 (Sutton). ZAP, 5, on Kebbi includes descriptions and discussion of the fortifications of Surame and Gungu. Bello had seen these ruins and was impressed (Arnett, 15). Maybe they influenced the building of Sokoto's walls which, if somewhat less substantial, were constructed in rather similar style.

⁵⁵ In Ajayi and Crowder, 2nd ed., 185. See also Urvoy, Y., ‘Chroniques d'Agades’, J. Soc. Africanistes, iv (1934), 145–77CrossRefGoogle Scholar, notably Ms. D and Urvoy's annotation. (This paper was available to me in Miss A. E. Owoyele's English translation in Zaria.)

⁵⁶ Bello in Arnett, 12; Abd al-Qadir in Palmer (1916), 265. The implication is clear that the Gobirawa's pedigree is not black, but that they came to the Sudan, via Aïr, from Egypt. It is interesting that the jihadists should have preserved – and enhanced with an Islamic (cum-Coptic) twist – the propaganda of the state which they accused of infidelity or apostasy and after a long struggle overthrew. Contradictorily, in discussing Aïr, Bello (Arnett, 9–10) emphasizes the distinction between the original black ‘Gobirawa’ inhabitants and the Tuareg who overcame and expelled them. Last (lxvi) discusses the sources which Bello and Abd al-Qadir may have used.

⁵⁷ Y. B. Usman usefully reviews this subject in an unpublished paper given at the ‘Borno Semina’ of the Departments of History (Zaria and Kano) of Ahmadu Bello University in 1973. The caliphal role which the mais of Bornu seem to have enjoyed in the Central Sudan during the two centuries or more preceding the jihad is an obviously related aspect.

⁵⁸ See note 9 above, and Greenberg's article on Kanuri influence; also Smith, Abdullahi, ‘Some notes on the history of Zazzau under the Hausa kings’, in Mortimore, M. J., Zaria and Its Region (Ahmadu Bello University, 1970), 82–101.Google Scholar

⁵⁹ Abd al-Qadir, loc. cit.; Smith (1970), 336.

⁶⁰ Épaulard (ed.), 9, 16, 448, 472–3.

⁶¹ Gibb (ed.), 336.

⁶² The standard list retold nowadays is: Daura, Kano, Zazzau, Gobir, Katsina, Rano and Garun Gabas (Biram). This is based on the writings of Sokoto (Arnett, 12; Palmer, ‘Western Sudan history’, 264–5) and the Daura, girgam (Sudanese Memoirs, iii, 134)Google Scholar. Hodgkin, , Nigerian Perspectives, 74–6Google Scholar, provides some useful annotation.

⁶³ Temple (406) includes Zamfara, as well as Gobir, in the Hausa bakwai. This is clearly a western list, if not specifically a pastoral Fulani classification of the sedentary indigenes. It is balanced by excluding Garun Gabas. But Temple‘s further comment – ‘some say that Auyo should be substituted for Gobir’ – is an acknowledgement that the Bornu marches, but not necessarily Garun Gabas itself, deserve to be represented among the ‘original’ components of Hausaland.

⁶⁴ This is an imprecise gloss on Bello (Arnett, 13). Zamfara came under Sokoto not Gwandu, but Bello included it in the banza bakwai (pace Temple, previous note). Moreover, to the east of old Hausaland several emirates created by the jihadists also came under Sokoto. These included the patently non-Hausa region of Adamawa, as well as the Hausaizing or semi-Hausa ones of Bauchi and the Bornu marches. Moreover, as Dr Robin Law has pointed out to me, if we accept that Bello's Infaku'l Maisur was written as early as 1812 and has come down to us unamended (see Last, xxx f), it predates the final settlement of the division of the Fulani Empire between Sokoto and Gwandu, and in particular the creation of the Ilorin emirate in northern Yorubaland where it eclipsed Oyo.

⁶⁵ Arnett, 12; Denham and Clapperton, iii, 162.

⁶⁶ Palmer, , Sudanese Memoirs, iii, 87Google Scholar. This is a fantastic collation of African and Near Eastern folklore, in which six of the seven are irrelevant.

⁶⁷ Cambridge History of Africa, iii, 151, 142.

⁶⁸ The Kilwa Chronicle as translated by Freeman-Grenville, G. S. P. in The East African Coast: select documents (Oxford, 1962), 35Google Scholar. For commentary see Chittick, N., J. Afr. Hist, vi (1965), 275 fCrossRefGoogle Scholar; and Pouwels, R. L., Azania, ix (1974), 65 fCrossRefGoogle Scholar. The latter argues a Shi'i, or more specifically Ismaili, connotation in the emphasis on seven here.

⁶⁹ Note also closer at hand the seven tribes of the (greater) Yoruba nation established by the grandsons of Oduduwa: Rev. Johnson, Samuel, History of the Yorubas (Lagos, 1921), 7–8Google Scholar. This seems to be the anti-Muslim — possibly anti-Fulani-Hausa — response.

⁷⁰ ‘Some considerations’, 336. Independent confirmation that the idea of Hausa being a conglomeration of seven was well established by the seventeenth century at the latest is provided in the Turkish writing of Çelebi (see the translated extracts in Hodgkin, , Nigerian Perspectives, 185Google Scholar). Çelebi's information had clearly derived from Bornu, and may thus help strengthen the suggestion that this image of Hausaland is a Bornuan one.

⁷¹ Hallam, (note I above); Hogben and Kirk-Greene, 149Google Scholar; Sudanese Memoirs, III, 135.

⁷² The Biram story and the inclusion of Garun Gabas in the Hausa bakwai is doubtless symbolic for the Bornu–Hausa borderland as a whole. This is a region of old small states with modest towns, around the headwaters of rivers or close to granite hills in some instances, typical probably of the early centuries of Hausa society. Trimingham, (History of Islam, 129–30) makes a perceptive observation here.Google Scholar

⁷³ Temple (406) did include it: see note 63.

⁷⁴ For the ‘originality’ of Kebbi, see notes 40 and 41 above. Abd al-Qadir (Palmer, 263) and Bello (Arnett, 13) go further to disprove the Kebbi claim, stating that Kanta was a slave of the Fulani. Whatever the sources for assertions of this sort, they were doubtless useful to Sokoto in its propaganda war with the new Kebbi state of Argungu.

⁷⁵ Mentioned by Leo Africanus, 478–9. The name may well be a form of ‘Wangara’ implying Mande (or at least western) influence. Though Leo's account is not entirely clear on the position of this kingdom, it seems to place it roughly in what is now west-central Hausaland. Trimingham, (History, 128)Google Scholar and Urvoy (op. cit., 230–31 and maps), following Palmer, (S.M., III, 76Google Scholar), identify it with Katsina Laka. But Lhote (in Epaulard, 478 n) suggests a possible northerly situation (near modern Maradi in Niger). For discussion see Hunwick in Ajayi and Crowder, I (2nd ed.), 276. More recently Fuglestad (‘Reconsideration’) has tried to argue the problem away by regarding Leo's ‘Guangara’ as simply another mention of Katsina, whereas Lovejoy sees it as actually applying to Kebbi (see note 16). Guangara – as well as Kebbi – is again noted in the sixteenth-century work of d'Anania, which is based on Leo and other sources: see Lange, D. and Berthoud, S., in Cahiers d'histoire mondiale, xiv (1972), 289–351, esp. 332.Google Scholar

⁷⁶ Rano has, it is true, a fairly complex wall system. But much of this may date only to the nineteenth century. It has moreover been noted for its industries, especially cloth manufacture. Yet several other small towns of this south Kano region were also important for cotton-production, weaving and indigo-dyeing. The fame of Rano itself for this may have been exaggerated academically for the simple reason that every one of the bakwai needed to be noted for something – Zazzau for extent of territory, Kano for wealth, Daura for seniority, Gobir for war, etc. Abd al-Qadir in the early nineteenth century had nothing to tell his readers about Rano (or ‘ Naru’) save to explain that it was not far from Zaria and close to Kano, as though they were unlikely to know where it was (Palmer, ‘Western Sudan history’, 266). On Bello's map (Denham and Clapperton, ii, facing p. 109) there is some confusion: ‘Na-roo’ is placed as in Abd al-Qadir, whereas ‘Ranoo’ is marked between Agades and Gobir! In 1850 Barth observed that Rano, was ‘not mentioned hitherto by any traveller’ (Travels, ii, 73)Google Scholar. The distinction between the powerful Hausa states and the obscure traditional sites had occurred much earlier, apparently by mid-millennium. This is the implication of Leo's account of the region between Songhai and Bornu where he includes seven kingdoms – Gobir, Agades, Kano, Katsina, Zazzau, Zamfara and Guangara (Épaulard, , Description, 10, 16, 472 f.Google Scholar), noting how one language was known in five of them (that is, all but Agades, clearly not Hausa, and Zamfara, perhaps excluded through oversight or ignorance). Leo's silence on Daura, Rano and Garun Gabas is doubtless significant. The same inclusions and omissions are found in d'Anania (see previous note). But he also notes Kebbi (p. 342), though somewhat out of place, whose independent rise occurred just too late for Leo's notice (pace Lovejoy's speculation: note 16, above). Interestingly d'Anania mentions also two places in the Hausa—Bornu borderland: Shira and, further north, Tsotsebaki (p. 334).

⁷⁷ Smith, M. G. (‘The beginnings of Hausa society’, 342–3Google Scholar) lists a number of such old centres in ‘core Hausaland’ which were later overshadowed as Kano, Zazzau and Katsina consolidated themselves as territorial states.

⁷⁸ Fuglestad (‘Reconsideration’) argues that it would have been rather different in form as well as scale.

⁷⁹ See above, especially note 3. Dr Fisher, Humphrey puts Abdullahi Smith's argument succinctly: ‘The theme here is thus not the origin of a people, but the way in which a society already in place developed new forms and institutions’ (Cambridge History of Africa, iii, 292Google Scholar). Compare Smith (1970), 330–31.

⁸⁰ Dr Adamu, Mahdi goes some way towards raising the issue, though dealing with a fairly late period and the external aspect, in ‘The spread of Hausa culture in West Africa’, Savanna, v (1976), 3–13Google Scholar. So does I. A. Tahir, ‘The dynamics of Hausanization [sic]: an overview of an assimilative process’, paper presented at the ‘Culture Seminar’ in Zaria, 1977. (Both these papers came to my notice after drafting the present article.)

⁸¹ The sources for this exploratory discussion of the Middle Belt are mostly secondary and cited above: see especially notes 1 and 3; and Abdullahi Smith in Mortimore, Zaria and its Region; and for ‘South Zaria’ and Smith, Abuja M. G., Government in Zazzau (London, 1960Google Scholar). On Kwararafa and the question of its relationships, Meek, C. K., A Sudanese Kingdom (London, 1931), 25 f.Google Scholar, and Low, V. N., Three Nigerian Emirates (Northwestern U.P., 1972), 81–6Google Scholar, contain helpful suggestions. I have also benefited from informal discussions and unpublished seminar papers at Ahmadu Bello University. Some broader cultural considerations are implicit in Rubin, A., ‘Bronzes of the middle Benu’, W. Afr. J. Archaeol., iii (1973), 221–31Google Scholar. A case could be made for a greater degree of cultural continuity through the Benue region before the Tiv expansion which has separated the Doma zone from that of Jukun/Kwararafa. On Doma, and information on it reaching the Mediterranean between the fifteenth and seventeenth centuries, see Person, Y., ‘Dauma et Danhomé’ J. Afr. Hist., xv (1974), 547–61CrossRefGoogle Scholar. One is tempted to speculate on its position in the trade with the south and possibly in tin. Whether or not one presses the Jukun connexion of Kwararafa, I am assuming that records of attacks on Kano and even Katsina occurring as early as the thirteenth and fourteenth centuries are either anachronisms or relate to some different peoples and events, possibly particularist rural movements within Hausaland.

Hausa external trade and cultural influences are treated by Adamu, Mahdi, ‘Spread of Hausa culture’; and Lovejoy, P. E. in J. Hist. Soc. Nigeria, v (1971), 537–47Google Scholar. See the latter's further articles, ‘The Kambarin Beriberi’, J. Afr. Hist., xiv (1973), 633–51Google Scholar; and ‘Inter-regional monetary flows’, J Afr. Hist., xv (1974), 568–85Google Scholar, dealing mainly with the nineteenth century.