CHAPTER 19

The Land of Sofala

Coasts of Present-Day Tanzania and Mozambique,
End of the Thirteenth to the Beginning of the Fourteenth Century

Ibn Saîd al-Maghribî, born near Granada (in present-day Spain) in 1213 and who died in Tunis (capital of today’s Tunisia) in 1286, made the pilgrimage to the holy places of Arabia at least twice. His travels took him to Iran. He was the continuator of a family summa, a sort of geographical anthology, which, unfortunately, is mostly lost. As for Abû l-Fidâ (1273–1331), he was a prince of the Ayyubid family and governor of Hama in Syria in the time of the Mamluk sultans. A learned connoisseur of geographic knowledge, he skillfully compiled the writings of past authors whose manuscripts he was able to procure. We owe large extracts from the works of Ibn Saîd to Abû l-Fidâ, who dutifully cited them in his own encyclopedia, but not before correcting, for better or worse, what he had judged to be his predecessor’s aberrations.

Ibn Saîd and Abû l-Fidâ: Whom should we praise? Whom should we incriminate? They have transmitted to us the knowledge of fine, learned men like themselves regarding the African coast of the Indian Ocean. But this knowledge, as it has come down to us, has remained almost incomprehensible. Let’s start with the closest feature—already quite remote for men from the central regions of the Islamic world—Bab-el-Mandeb at the southern end of the Red Sea. Here first is the land of Berbera, which extended from west to east, opposite the south Arabian coast, before turning off toward the south, marking the profile of the Horn of Africa (as we say today, using an image that was not used by medieval cartographers). We recognize the name Khafouny or Hafouny: it’s Cape Hafun, a promontory known by all the region’s sailors since antiquity. We recognize Maqdishû: it’s Mogadishu. Once we have passed the arid coast of modern Somalia, we find ourselves in a more pleasant region, the land of Zanj. The only city names given to us are familiar: Malindi and Mombasa, two ports on the southern coast of Kenya (34). Next comes a desert, and then we arrive in the last known segment of the eastern coast of Africa: the land of Sofala. But this time, in spite of the names, we recognize nothing: a city named either Batyna or Banya; the capital of the people of Sofala, which is called Seyouna; an important port, Leyrana, whose inhabitants are Muslims; and finally the last stopover, Daghouta.

It was from this land, where people dressed in leopard skins, that the Zanj merchants bought the gold that they resold farther north to Muslim traders. The region had changed since al-Masûdî’s day in the ninth century (2): the land of Sofala, once a mysterious horizon far beyond the Coast of Zanj, had become more familiar. Now in the late thirteenth and early fourteenth centuries, we find entire communities that had converted to Islam. It is even known to some Arab authors that gold came from a land in the interior where it was harvested in tunnels dug like those of ants.

It is frustrating to be reduced to conjecture to place these names on the map. Our authors, however, took care to aid their readers by providing each place’s geographical coordinates, the distances in degrees from one port to another. But in fact, by positioning the points on a grid, we can only draw a horizontal coast stretching from Malindi to Kenya, all in longitude, barely crossing the equator by twelve degrees and seeming to link up with Southeast Asia. This is not what Africa looks like; so it must be said that our geographers believed that Africa’s eastern extremity connected to China, that the Indian Ocean was a closed sea. How are we to trust information inscribed in geography so manifestly erroneous? Or rather, to put it differently: how are we to use such precise geographical information inscribed in cosmography so manifestly different from our own?

And yet, a quarter turn clockwise, the longitudes change to latitudes; the profile of the land of Sofala becomes familiar: when it is superimposed on the coastal profile of Tanzania and the northern half of modern Mozambique, they match to a very accurate degree. Nevertheless, it is not any easier to identify cities or, say, the stopovers in question over vast distances. But at least we can locate them in the coastal segments. The last segment, angled to the southeast, which Abû l-Fidâ imagined being the farthest east, and which for us is the farthest south, outlines the coast of Mozambique between the Island of Mozambique and Beira at the mouth of the Pungwe River. It was here two centuries later that the Portuguese coming from the south, eager to harness the flood of gold from the Zimbabwean plateau for their own profit, erected a fort that they named Nova Sofala.

BIBLIOGRAPHICAL NOTE

The Arabic edition of Abû l-Fidâ’s text was published by Joseph Toussaint Reinaud and William McGuckin de Slane, Géographie d’Aboulféda (Paris: Imprimerie royale, 1840). For English readers, the essential texts on the East African coast are brought together in G.S.P. Freeman-Grenville, The East African Coast (Oxford: Clarendon Press, 1962).