Walk through the ruins on Kilwa Kisiwani, a small island off southern Tanzania, and you notice that almost every great building faces the water. The palace, the mosques and the merchants' houses were placed to watch for sails. That is the first thing to grasp about the Swahili coast: its towns were African towns whose front doors opened onto an ocean, and their history makes sense only if you hold the land and the sea in view at once.
The shape of the coast
The Swahili world stretched from the Benadir coast of southern Somalia, through the Lamu archipelago and the Kenyan shore, past the islands of Pemba, Zanzibar and Mafia, down to Kilwa and on toward northern Mozambique. Its people also reached the Comoro Islands and the northwest of Madagascar. Few towns lay more than a short walk from the shore, and many sat on islands, which offered safe anchorages and some protection from raids. Behind them stretched a narrow fertile belt, and behind that, drier bush country where other African peoples farmed and herded.
Coral reefs run along much of this coast, broken by channels where rivers enter the sea. The reefs gave builders their stone and fishermen their catch, but they also made sailing dangerous for strangers. Local pilots who knew the gaps in the reef were always in demand.
The winds
The ocean itself kept time. Swahili sailors call the northeast monsoon the kaskazi and the southwest monsoon the kusi. The kaskazi blows from roughly December to March and carried ships from Arabia, the Persian Gulf and western India down to Africa. The kusi, strongest from about May to September, carried them home. Between the two came calmer months when ships were repaired and goods were sorted.
The monsoon also set a southern limit. Ships from Arabia or India could sail comfortably as far as the Kilwa region and still catch the returning wind in the same year, but voyages farther south ran out of season. Kilwa sat near the southern limit that a ship from Arabia could reach and still return within one monsoon year. Goods from farther south, above all gold from the port of Sofala, therefore had to be gathered at Kilwa and shipped on. Geography helped make Kilwa a broker.
What the outsiders wrote
The oldest written description of the coast is a Greek merchant's handbook of the first century AD. It describes a coast it calls Azania and a market town named Rhapta, where traders bought ivory and tortoiseshell. Rhapta is a market town whose site has never been found, though scholars have searched near the Rufiji and Pangani rivers.
Arabic writers came later. Arabic geographers called the coast's Black African inhabitants the Zanj, and they called the region the land of the Zanj. The name Zanzibar probably comes from the same root joined to a Persian word for coast. The traveler al-Masudi sailed to East Africa early in the tenth century and wrote that ivory from the land of the Zanj went to Oman and then on to India and China. Al-Idrisi, writing in Sicily in 1154, gathered reports of coastal towns, and Ibn Battuta visited Mogadishu, Mombasa and Kilwa in 1331.
What the coast wrote about itself
Swahili towns also kept their own histories. The Kilwa Chronicle is a history of Kilwa's sultans compiled in the early sixteenth century; it survives in an Arabic copy and in a Portuguese summary made by the historian João de Barros. Other towns, such as Pate and Lamu, had chronicles written down much later from older oral tradition. These texts are precious but tricky. They were composed to honor ruling families, and they compress centuries into tidy genealogies.
The fourth great source is the ground itself. Archaeology gives the best evidence for the coast before about 1000 AD, when written sources are thin. Pottery, coins, glass beads and house foundations show when towns grew, what people ate and where their goods came from.
Worked example: reading a chronicle with care
Suppose you read in the Kilwa Chronicle that a certain sultan ruled for exactly forty years and built a great mosque. How much should you trust it?
Start with who wrote it and why. The chronicle was compiled at the court, around the time the Portuguese arrived, by people who wanted the dynasty to look ancient and legitimate. A round number like forty years is a warning sign; it may be a way of saying "a long time." Next, look for independent checks. Coins bearing the names of Kilwa's rulers have been dug up at Kilwa and elsewhere on the coast, and their order and style can confirm or correct the chronicle's list. Excavated mosques can be dated by their building phases and by the imported pottery found in their foundations.
When the sources agree, confidence rises. When they disagree, the historian does not simply pick the written text because it is written. Here the chronicle and the coins broadly match on the names of rulers but differ on dates, and archaeologists have moved some reigns by a century or more. The result is a history built from several kinds of evidence, each correcting the others, which is how the whole story of this coast has to be told.
