CHAPTER 6
TRANS-SAHARAN COMMUNICATION,
EARLY ARAB  CROSSINGS AND IMPRESSIONS


Before Islam

In pre-historic times the Sahara was well watered and settled with people, whose traces can still be found. About 5000 BC the climate began to become dry, and by 3000 BC much of the area was desert, with only patches of inhabitable land or oases left.  Movement around the Sahara was then more difficult, but still possible.  Yet for earlier times there is no indication of widespread commercial or political networks.  Pastoral and food-gathering people simply roamed or settled where they pleased, content with a self-sufficient clan organization that did not need wide contacts.  Study of human remains shows that the Negroid population reached far up into the Sahara, although the north of the continent was occupied by Berbers.

Around the year 1000 BC some long-distance communication routes began to develop.  Evidence is found in chariot drawings on rocks in the Sahara in sites strung out along two main recognizable routes, one from Morocco to the Niger bend, and the other from around Tripoli through the Fazzān and on to Gao, with a branch to Aļr and continuing, as we will argue, into Nigeria.  Horses had come into Egypt around 1600 BC, but were of a small pony type which could not be mounted but only pull a chariot. They had come to the Maghrib with the Phoenicians from 11000 BC and soon afterwards were adopted by Berbers of the desert.  The chariots could only carry one or two men and could not go far without repair, and it is likely they went in relays over longer distances.  Heavier loads still had to be transported by donkeys or oxen, while chariots seem to have been prestige possessions of the wealthy or the chiefs, and used only for sport racing or hunting or pleasure.  Nevertheless they are an indication of long-distance contact over the Sahara in the first millennium BC.  The exploratory trip of the Nassamonians (a Libyan tribe) to the Niger river in the 5th century BC, mentioned by Herodotus, followed an established chariot route.  It is probable that the Phoenicians joined in some of the chariot excursions across the Sahara.

An indication that the Phoenicians probably crossed the desert is the fact that the technology of iron smelting, which they introduced to North Africa, spread to Nok in Nigeria by the 5th century BC.  Nok is right in line with the Aļr branch of the eastern chariot route.1  The Phoenician origin of this technique is arguable also from the fact that iron smelting is a complicated art which would hardly be developed spontaneously or independently; for example the old empires of Mexico and Peru never had it.2  On the other hand, on the basis of Phoenician secrecy about their trade routes (Strabo reports) and presumably also about their technology, and on the basis of the short time (100-200 years) between the appearance of iron technology at Carthage and at Nok, it is also arguable that Nok iron technology is autochthonous.3  Sub-Saharan Africa had the market, resources and the fuel to carry on iron working, which did not exist in the desert to the same extent.  Little is known, but some sort of political and trading network must have grown up around the Nok industrial centres.  Iron production is known to have spread to Yelwa by the Year 100 AD, but to Daima in Borno only by around 810 AD.

A revolution in trans-Saharan contact came about with the introduction of the camel, an animal perfectly suited to the desert and capable of carrying heavy loads.  It was introduced into Egypt before 1000 BC, but became common in the Roman Maghrib only in the 1st century AD, when Roman rule also came to North Africa.  The camel displaced the horse in the desert both for the Romans and the Berbers, but the horse, whether chariot pulling ponies or newly bred mountable varieties, continued to be used within the Roman settled territory.  The Romans did not fully appreciate the strategical advantage of the camel and its possibilities for desert penetration and defence. More importantly, they had no interest in crossing the desert, because they knew of no incentive to do so.  The habitable land they occupied in the North of Africa had long been part of the Mediterranean world, and the security of the Roman Empire only intensified its links with all lands bordering on the Mediterranean, to the exclusion of distant and hardly known West Africa. Nevertheless, by the end of the 3rd century AD gold, notably from Jenne-Jeno, began to flow to Carthage.4

By conquering North Africa, the Arabs became heirs to this gold trade.  Without the trans-Mediterranean links that the Romans and Byzantines had, the Arabs developed the more difficult trans-Saharan contact with West Africa by means of the camel.   At the same time the Arabs and their Berber allies came to appreciate the value of the camel as an ideal vehicle for raids, and initiated military expeditions into the desert. This resulted in the development of the slave trade, especially with Kanem.

The geographical picture the Arabs found consisted of an inhabitable section of North Africa called the Maghrib, meaning “the West” (from the standpoint of the Arabs coming from the East), lying between the Mediterranean and the desert.  Since the desert was likened to an ocean and camels to ships, the Arabs thought of the Maghrib as an island lying between two seas; so they called it al-jazā’ir (Algeria), meaning “islands”.  To the south of the desert there was first a strip of land suitable for grazing, but not farming.  This the Arabs called the sāhil (sahel), meaning the “coast”, bordering upon the desert sea.  Beyond the Sahel lies the savanna lands suitable for farming, and further south is the forest.  This whole land south of the desert the Arabs called “Sudan”, meaning the “country of the Blacks”.

The dynamics of economic and political organization in West Africa

A general principle in this discussion is that wider political organizations depend on wider economic organizations, and if there is no wider economic interest a wider political entity will not emerge.   Before about the year 2000 BC the prevalent economy of West Africa was hunting and food-gathering; agriculture was not well developed and even pastoral economy was minimal.  On the political level the family or clan was normally the highest unit; its members were bound together by blood and looked after one another’s interests without any help from the outside. Their economic self-sufficiency meant that no wider political organization was needed beyond a clan patriarchy or matriarchy, such as is true of Pygmies in some parts of D.R. Congo today.

By the year 1000 BC agriculture was well established, and sedentary farmers, along with nomadic pastoral peoples, represented the normal way of life in the savanna, while hunting and food gathering continued in the forest zones.  Agriculture is a complicated economic enterprise because it demands farming and grinding tools, whether stone or iron, granaries and semi-permanent houses, wells for drinking water, ropes, pottery, gourd containers, tanned skins, some cloth etc.  All this required a variety of skilled artisans.  These many needs could not be supplied within the clan, but called for a wider economic and political organization: the village, comprising several independent families or clans under the leadership of a chief or king.

Depending on the locality, a village would need some military power to defend itself against outside raids. This was particularly true near the desert where the nomads often raided, and in fact nomadic raids drove the Black population very much to the south of its former boundaries.5

As people moved beyond the subsistent economy of the village and began to demand more diverse goods which were not always available in their home area, different local economies, often based on different climactic conditions, joined to complement one another in a kind of common market.  For traders to pass from one area to another, security was needed.  Also some management, control or organization of the trade network might be required.  A common market, therefore, naturally leads to a common political system. Thus the early states of West Africa grew up in response to the need to provide security and coordination for an economic network, and the kings who provided this were recompensed by control of the trade (as in Ghana the gold trade was a royal monopoly) or by tolls on traders who passed through their territory.  West African kingdoms were not like the kingdoms of Europe or elsewhere which were based on control of land, which was scarce, giving rise to a tenant or feudal system.  In Africa land was abundant, and political power was the management of people, not land. And the common project for which human management was necessary was not monumental buildings, like the pyramids of Egypt or the aqueducts of Rome, but the complex organization of long-distance trade.

Since empires based on a common market embraced different ethnic and language groups with no natural sense of nationhood, a political system had to be devised to give cohesion to the state. The government could not be a direct-rule unitary system as in a modern nation, because there were no effective communications or means of implementing decisions over a wide area.  The monarchical system of the ethnic homeland, modified by an aristocracy representing clan or business interests, was further modified outside the ethnic homeland by a vassalage system of indirect rule.  Lastly, outlying states were drawn into the political network by treaties or alliances which guaranteed freedom and security of movement for trading agents. A kind of empire thus emerged which had its sanctions not in traditional customs or a regular army, but in the prestige of the stronger power and occasional punitive raids on recalcitrant vassal kings, who might be replaced by men more cooperative with the imperial power.  In such circumstances traditional religion, bound to particular localities and clans, was prone to yield to a more universal religion which would become a unifying ideology for the empire.  The size of such an empire depended on how far its influence carried, rather than on any determined borders.

In such empires the trading network was often in the hands of one tribe whose agents spread outside their homeland to every corner of the common market area.  The Malians (also known as Wangara, Dyula, Malinke or Mandingo) were particularly noted for this activity.

In point of fact, West African empires arose in response to two types of international trade: the trans-Saharan trade initiated by the Arabs once their rule over North Africa was consolidated in the 8th century, and the Atlantic trade with the Europeans which became important from the 18th century onwards.  The Arabs came to the Niger bend in search of gold, and to the lake Chad region in search of slaves.  Ivory and skins were secondary items.  The Arabs, or their Berber partners, bought these with salt from the Sahara and luxury goods such as wheat, dates, raisons, glassware, lamps, swords and other metal products, and even some re-exports from Europe, such as paper.  Slaves were often exchanged for horses, not the local ponies, but the large cavalry mounts of North Africa which made slave hunting in the savanna all the more effective and nourished the vicious circle of more intense slave hunting to pay for more horses.  A transfer of technology made the production of iron more common in West Africa, both for tools and weapons.

At this point, we can refute the West African legends which attribute foreign white ancestry to their founding dynasties.6  Islamic influence may be responsible for the manufacture of such legends, but even where nomadic white rulers conquered a Sudanic people, as may have occurred in Kanem, this is no explanation for the evolution of states in West Africa.  The nomads had primitive political traditions based on shifting alliances of clans, whereas the Sudanese kingdoms had far more elaborate political institutions.7

The Arab writers on West Africa

A fair number of Arab writers have left us descriptions of their travels or reports they heard concerning West Africa. Two major studies with translations of the most relevant passages are Joseph Cuoq, Recueil des sources arabes concernant l’Afrique occidentale du VIIIe au XVIe sičcle (Paris, 1975) and N. Levtzion and J.F.P. Hopkins, Corpus of early Arabic sources for West African history (London, 1981), which I refer to as “LH”. For the Arabic there is the edition of L.E. Kubbel and V.V. Matreev, !D"$F84, 3FH@R>484 (2 vols., Moscow, 1960 & 65), referred to as “KL”. This edition is not complete; so for some writers I use other Arabic editions.  The following are the Arab authors and writings I make use of:

There are other writers not used here, and I do not reproduce every passage, since some later writings are merely copying from earlier ones.  Also I omit some stories and details which are not so important.  The translations are my own, although relying much on the interpretations of LH and Cuoq.

cUqba ibn-Nāfic

The most famous early Arab to penetrate the desert was the conqueror cUqba ibn-Nāfic.  The story of his exploits across Libya towards Kanem was told in Chapter 3.

[Ibn-cAbdalhakam, 363:] cUbaydallāh [ibn al-Habāb, 116-123/734-741] sent Habīb ibn-abī-cUbayda al-Fihrī [grandson of cUqba] on a raid to as-Sūs [southern Morocco] and the land of the Blacks.  He achieved a victory over them such as was never seen and got all the gold he wanted. He also captured a girl or two from the people the Berbers call Ijjān.  Their women have only one breast.

Comment: The story of cUqba’s exploits was taken up by West African legend, and he is made out to be the ancestor of the western Saharan Kunta tribe and of the Fulani.

Arab legends about West Africa

Some early Arab accounts paint a frightening picture of West Africans.  This may have been motivated by racial and religious prejudice to justify slave raiding, or it may be a device of gold traders to scare off would-be competitors from venturing across the Sahara.

One of the legends repeated by several Arab geographers is the story of the “curse of Hām”, a tradition developed from an indirect reference in Qur’ān 11:42-47 and a forced interpretation of Genesis 9:18-28.   Ibn-Qutayba (Kitāb al-macārif, 26), quoting Wahb ibn-Munabbih (d.728-32), says, “Hām son of Nūh was a white man with a beautiful face and shape, but God changed his colour and that of his descendants because of his father’s curse... God multiplied them and increased their number; they are the Blacks.”

The same legend is also found in Yacqūbī (Ta’rīkh, 1:15).  Akhbār az-zamān (LH 34) adds: “Noah cursed Hām, praying that his face should become ugly and black, and that his descendants should become slaves to the progeny of Sām... They are the various peoples of the Sūdān.”

[Ad-Dīnawārī, Al-akhbār at-tiwāl, p. 12:]  Abraha.. penetrated the land of the Sūdān, who submitted to him.  He travelled on and traversed their territory until he reached a nation of mankind whose eyes and mouths are on their breasts.  It is said that they are a nation descended from Noah who incurred the wrath of God so that he changed their form...

[p. 33:] Alexander travelled until he reached the land of the Sūdān.  There he saw people like crows, naked and barefooted, who wandered in the jungles living on fruit and if they had a year of drought and famine ate one another.

[Ad-Dimashqī, Nukhbat ad-dahr fī ācjā’ib al-barr wa-l-bahr, p. 15:]  The equator is inhabited by Sūdānese tribes that can be reckoned as wild animals or cattle.  Their skin and hair are burnt and they are morally and physically perverse.  Their brains nearly boil from the intensity of the sun’s heat.

[17:] Man created there is ignorant, extremely black skinned, with burned hair, a violent character, perverse temperament, in morals more like wild beasts.  They cannot live in the [more northerly] second clime, and much less in the third or fourth or sixth, any more than someone from the sixth clime can live in the first or on the equator because of the different compostion of the air and the heat of the sun. [See also al-Hamdānī, Kitāb ifa jazīra al-cArab, p. 40; Ibn-Khaldūn, al-cIbar, I, p. 93.]

If the previous passages are highly embellished travellers’ tales, the following description by seems to be in part a factual report:

[Akhbār az-zamān, 87-88:] Many nations descend from Sūdān, son of Kancān... If they do not go naked, some of them wear skins or grass, and some fix animal horns on their heads. They eat white mice, which they call “manna from heaven”.

A man may marry up to ten women, and sleep each night with two of them.  He may have intercourse with them as he likes, but if does not, the king gives them a divorce after the third night.

They often suffer from drought. To make rain, they gather bones into a pile, set them on fire and walk around, with their hands lifted to the sky, and pronounce some formulae.  Then it rains, and they have water.

When one of them gets married, they smear his face with something like ink, and sit him on a mound, while they sit on another mound.  They place the woman before him, and build a reed hut over them and cover it with grass. They stay around it for three days, drinking beer made from sorghum and enjoying themselves.  Then they depart, and the husband takes his wife  home...

They have a huge tree, and make a feast every year in its honour.  They dance around it until its leaves fall upon them, which they consider a blessing. They adorn their women with copper rings, and put cowrie shells in their hair.

«— Chapter 5

The Spread of Islam
Through North to West Africa

Chapter 7 —»


1See the map in R. Mauny, “Trans-Saharan contacts and the iron age in West Africa”, Cambridge History of Africa, vol 2, p. 282.

2Cf. Mauny, op. cit., pp. 331-335.

3Cf. John A. Rustad, “The emergence of iron technology in West Africa”, in B.K. Swartz & R.A. Dumett, West African culture dynamics (The Hague, 1980), 227-246.

4Cf. Timothy Garrard, “Myth and mythology: the early trans-Saharan gold trade”, Journal of African History, 23 (1982), 443 ff.; Susan and Roderick McIntosh, Prehistoric investigations at Jenne, Mali (Cambridge U.P., 1980), “Finding West Africa’s oldest city”, National Geographic, 162:3 (Sept. 1982), 396-418.

5Cf. R. Mauny, op. cit., pp. 335 ff.

6These were upheld by early European authors such as Y. Urvoy, in his Histoire de l’empire du Bornou (Paris, 1949), p. 21.

7Cf. N. Levtzion, Ancient Ghana and Mali, p. 9.