Empires of the Word - Nicholas Ostler [55]
And this indeed was to be the pattern with all the further spreads of Islam that occurred in the second millennium, notably from North Africa south of the Sahara, from Egypt and Arabia down the coast to East Africa and Madagascar, from Baghdad and Bokhara into Siberia and central Asia, from Afghanistan into India, from India into South-East Asia: Arabic was accepted as a sacred language, but had no tendency to spread as a vernacular, or even as a lingua franca for contacts among the new Muslim populations. Except for the Hausa speakers of West Africa, none of the converted communities spoke Afro-Asiatic languages; so this conforms to the linguistic constraint.*
Before leaving the subject of the spread of Arabic and its limits, it is right to consider one other way in which Arabic might have been expected to spread, but in fact did not. At least from the beginning of the first century AD to the advent of European adventurers in the fifteenth, it is known that Arab sailors, with perhaps some Persian competition, undertook most of the marine trade between the Near East and the coasts of Africa and India.
The first testimony dates from the first century AD, in the Greek guide for sailors Períplous Thalássēs Eruthraías, ‘Voyage Round the Indian Ocean’.
(§16) Two days’ sail beyond there lies the very last market-town of the continent of Azania [East Africa], which is called Rhapta; which has its name from the sewed boats [rháptōn ploiaríōn] already mentioned; in which there is ivory in great quantity, and tortoise-shell. Along this coast live men of piratical habits, very tall, and under separate chiefs for each place. The Mapharitic chief governs it under some ancient right that subjects it to the sovereignty of the state that is become first in Arabia. And the people of Muza now hold it under his authority, and send thither many large ships; using Arab captains and agents, who are familiar with the natives and intermarry with them, and who know the whole coast and understand the language …
(§21) Beyond these places in a bay at the foot of the left side of the gulf, there is a place by the shore called Muza, a market-town established by law, distant altogether from Berenice [Ras Banas] for those sailing southward, about 12,000 stadia. And the whole place is crowded with Arab shipowners and seafaring men, and is busy with affairs of commerce; for they carry on a trade with the far-side coast and with Barygaza [Broach, in western India], sending their own ships there.77
Wherever Rhapta (Dar es Salaam?), Muza (al Mukha?) and Mapharitis (Ma’afir?) were, it is clear from this that Arab trade involvement with both sides of the Indian Ocean goes back for well over six hundred years before Muhammad. It is also a known feature of Arab ships, up until 1500, that their hulls were stitched together, not nailed or pegged.78 The 1001 Nights’ stories of Sindbad the Sailor (in fact, more a maritime merchant than a sailor) had a strong basis in Arab fact.*
This means that Arabic would have been heard in all the ports along the shores of the Indian Ocean from Mozambique to Malabar and Coromandel in southern India. Surely this might have had a linguistic effect, at least in the creation of a trade jargon? There is, after all, ample precedent, both, as we have seen, in the way that Phoenician was spread round the Mediterranean, and in more recent centuries as European powers have brought their languages to the parts of the world where they went to trade. Trade is usually accounted the first factor that set English on the road to becoming a world language.*
In fact, the only vestige of such influence from Arabic is found in East Africa, where Swahili, the major Bantu language, shows heavy signs of Arabic influence. Its very name is derived from Arabic sawāil, ‘coasts’. Counting up to ten, the