Short History of Mafia


by Peter Byrne


Mafia lies across the historic trade routes of the ancient Azania Coast, running from Somaliland to Sofala, in Mozambique. The trade route yielded gold, ivory, slaves, live animals, animal furs and skins. Mafia’s location would have provided an excellent stop for refilling water containers and making repairs to ships, as it was a peaceful island that welcomed travellers. The more important port was Kilwa, where ships could be fully victualled and cargoes bartered.

Our historical research shows that Egyptian, Phoenician, Roman, Arab, Indian, Greek, Chinese, Portuguese, French, American and British ships have visited Mafia from at least 500 BC. Peoples of many different foreign origins have lived in Mafia: African (Madagascar, Mozambique), Indian, Yemeni, Baluchi, Omani, Shirazi, Chinese, Japanese and from several European countries. The racial potpourri can be seen today in the wide variety of skin colour and stature of the local inhabitants. The historical mix has also left a culture of great tolerance and self-sufficiency.

Mafia has at different times been the property of the Sultan of Kilwa, the Sultanate of Oman, Germany and Britain, before Tanzania became independent.

The history of Mafia has been a study of special interest for us and our library holds many papers, books and reports of Mafia down through the ages, its people and its trade and cultural links to the World.

The Periplus of the Erythraean Sea, (Indian Ocean) written in approximately AD50[1], describes a well established trade route, linking Arabia with Azania, as the east coast of Africa was known in the Graeco-Roman era. The principal port of trade was Mocha (or Merku and Mark’a) in present day Yemen, and the last port in Azania was Rhapta, lying some two courses (a sailing measure, possibly tacks) from the island of Menouthesias, itself 300 stadia – a measure of distance equivalent to about 50 km – from the coast. Menouthesias was “…a low island covered with trees in which are rivers…”. according to the Periplus. And Rhapta lay to the south “…beside and to the east of a cape with a river…” according to a separate source, Ptolemy[2], in his famous Geographia. The locations of both Menouthesias and Rhapta have confounded scholars since the Periplus was first translated in 1912. Some scholars argue that Zanzibar or Pemba may be the fabled Menouthesias with Rhapta somwehere between Bagamoyo and Dar es Salaam.

The trade links to Mocha indicate that the Sabaeans, ancestors of the Yemenis, claimed ancient right to overlordship of the Azania coast, although this is believed to have been an arrangement to reduce trade competition rather than the result of conquest. Rhapta and its hinterland was governed, undoubtedly tenuously, by these people, believed to be the Ma’afir, a tribe of Himyaritic stock. The control from the Ma’afir may explain the name of “Mafia”. [3]

It is not too fanciful to suggest that Mafia is the Menouthesias of the ancients and Rhapta was in the area of Kilwa. Ptolemy located Rhapta at 8o South (where the Delta lies) and “…near a big river…”; these geographical descriptions and the mention of many crocodiles in the old writings certainly support the possibility of the Rufiji Delta as Cape Rhapton. The respected archaeologist Neville Chittick believed this was possible and at his death in 1984 was investigating the Rufiji Delta for evidence of the lost metropolis of Rhapta.

A Persian family, apparently from the town of Shiraz in Persia, led by Ali ibn Sultan al Husayn ben Ali settled in Kilwa in AD975. The Kilwa Chronicle states that he purchased Kilwa from the ruling chief for a great quantity of cloth (sufficient to encircle the island !). Bashat, one of his seven sons, settled in Mafia to govern under this new Kilwa Sultanate and he is thought to have established the towns of Kua and possibly Kisimani Mafia. Chittick dated the ruins there to the early 11th Century.

Kilwa prospered from the gold and ivory trades, tariffs on cargoes, and as a source of pitch and resin as it is a convenient port for victualling and re-caulking ships. Here the literature is vague: Did Kilwa already exist as a significant port (Rhapta)? How did this new settler from Shiraz assume such economic and political power so quickly? It is hard to believe that Kilwa was not already prosperous and therefore attractive to the Persians, who stepped into or complemented an economic and power vacuum, possibly because the Yemenis were itinerant trader-sailors and did not settle[4]. It is certainly an exceptionally good port for a sailor, with an entrance that is easily negotiated and a superb protected anchorage in deep water.

Work being done by Professor Felix Chami and other archaelogists, anthropologits and linguists strongly supports the school of thought that between AD700 and AD1250 an identifiable Swahili culture emerged that was predominantly Bantu with influences from the Indian Ocean and Arabia, rather than an imported culture with minor aspects of Bantu living; the picture that the Bantu were merely Stone Age slaves is unacceptable. Much evidence from very old settlements and from the writings of travellers such as Idrisi and Masud suggest a well organised society that was able to raise an army that could defeat an Arab crusading force prselytizing Islam, led by an Omani Prince, in the 7th Century.

Evidence for the period AD300-700 is still unclear except that there was an apparent rise in native military and political power, undoubtedly based on economic strength as a mutual trading partner.

The ‘Swahili Period’ is now identified as AD1200-1500 when relations between the Middle East and Azania appear to have been very good and peaceful with much prosperity ensuing, the latter indicated by the proliferation of trade goods dating from this period. The Swahili culture was predominantly Muslim[5] although the southward spread of Islam was slow between the 10th and 12th Centuries, and there was for that period a clear divide at Malindi (Kenya) between the Islamic north (Swahili) and the land of Zinj[6], to the south. Interestingly there is no mention of slavery and a slave trade in the writings of this period.

Construction of buildings using limestone dates from the 10th Century and the use of beautiful Persian monochrome and Chinese bowls to decorate mosques and homes began in the 12th century (and is still used in hotels and private homes in Zanzibar as an attractive decorative feature).

Trade and exploration must have been extremely active for, by the 12th Century, Arab and Chinese explorers knew even of great areas of the interior, including the great lakes that were the focus of European exploration 600 years later.

There is growing evidence that the emergence of the Swahili culture may have originated in the central Tanzanian coast, in the area of the present Rufiji delta, where archaeologists have found sites of continuous occupation from the last centuries BC to the 12th Century. This area is now believed to be the crucible of the ancient (First Millenium) civilization, the earliest being the Triangular Iron Ware. The proximity of great herds of elephant for ivory, hippo for fat (used for boats) and mangrove poles for export to the Gulf could have been economic bases. Later sea level changes, sedimentation of the river mouth (leading to extensive mangrove development), salt incursions into groundwater, the spread of malaria, filaria and leprosy could have contributed to the decline of these towns.

Kilwa became the most prominent port in the trade of the Indian Ocean by the 14th Century, although it was in decline by the arrival of the first Europeans. The rise of Kilwa and its great influence from the 11th through 13th Centuries coincided with the early spread of Islam to eastern Africa and very active trade, especially after the 10th Century. During this process Mafia played an important albeit supporting economic role.

There are Early Iron Working sites on Mafia from BC300-AD300 that have produced Graeco-Roman beads, glassware and pottery from the Mediterranean World. It is now believed that there was contact between Azania and Arabia from well before Christ, with Phoenicians, Egyptians (a BC600 expedition sent by Pharaoh Necho) and Ethiopians. The Indonesians reached Madagascar 1500 years ago.

There is clear evidence in Mafia of contacts in the Graeco-Roman period from about BC200. Early writers (e.g. Pliny and the Periplus) describe cave-dwellers and our exciting finds on Juani Island in the last two years may be a link with this period. All of the trade goods from that period are represented in the sites and we now have skeletons that can be dated and ethnic origin determined from bone samples. Rome controlled the Red Sea and the Azania trade at this time, but like the Portuguese conquerors much later, their role was tenuous and could not compete with the Bantu-Arab links already in place, including inter-marriage.

The inhabitants of Mafia are recorded in the Kilwa Chronicle as the Mwera, who were ruled from SongoSongo Island by Muslim settlers prior to the arrival of Bashat. Who these rulers were we do not know, and what of the first 1,000 years AD ? Who were the people and what was the government of Mafia ? Dr. Chami of the University of Dar es Salaam believes that the first settlers were Early Iron Working and farming Bantu people[7], who crossed from the mainland and settled the islands, probably for their marine and forest resources. This culture was followed by a later native culture recognised as Triangular Iron Ware, which was extant with the settlement and domination from Arabia.

The Portuguese arrived in April 1498, when Vasco da Gama first sighted Mafia off to starboard on his way northwards, but the first Viceroy did not arrive to establish control and depose Arab rule until 1505. Portugal formally annexed the east coast of Africa in 1515 after the Papal bull of 1514 divided the known World between the warring Portugal and Spain. Portuguese control in Azania was always erratic and brutal and tenuous, at best, in Mafia.

In 1588 Kilwa was sacked by an African army of cannibals referred to as “Zimba” or “Muzimbe”, believed to be from central Africa (other literature sources suggest somewhere in southern Africa). This put an end to the remainder of Kilwa’s declining supremacy as a trading port and to its control over Mafia, for the cannibals literally devoured the inhabitants[8]. From this time Zanzibar became the epicentre of trade in Azania, especially with the rise of power of the Omani Arabs.

Control of Mafia changed hands frequently in the 16th and 17th Centuries, as Portugal’s fortunes declined, Oman’s interest waxed and waned, and the influence of other world powers played their part, The defeat of the Portuguese by Oman in Mombasa in 1698 ended what had been a troubled and cruel Portuguese rule and gave the Sultan of Oman control of the coast from Lamu to Kilwa.

In 1829 the town of Kua on Juani Island was destroyed by Sakalava cannibals from Madagascar and in 1872 the remarkable town of Kisimani Mafia was lost in a cyclone. By then the seat of power had moved to Chole Island, a more convenient and productive location. The arrival of the Sakalava prompted the Sultan of Zanzibar to send a punitive expedition that included some of his personal Baluchi regiment. Descendents of these Pakistani people are still to be found settled mainly in the area of Kitoni near Kismani Mafia. There is also evidence for minor settlement of Mafia by Madagascans, Chinese, Malay and Indonesian peoples (who first settled in Madagascar about 1,500 years ago). Pottery and coins indicate trade took place from, at latest, the 8th Century and our new finds may help to prove Dr Chami’s theories of a much earlier active trade.



The majority of Mafia inhabitants live in small villages scattered along the coast and interior of Mafia Island itself and on the other islands nearby (Chole, Jibondo, Juani and Banja). Typically villagers have a number of sources of income and subsistence: fishing amongst the members of a family, including crop farming, permanent tree crops and skilled work (carpenters, weavers, sailmakers, etc).

The farms are smallholdings, growing plots of cassava, rice, pigeon pea, pineapples, pawpaws and beans; it is typical to also find cashew, coconut and mango trees on each household’s land. Farmed areas are usually surrounded by woodland, grassland or coconut plantations. Low-lying areas (mbuga) are adapted to upland rice (a variety that is not “paddied” under water, as in Asia).

Large areas of the island – especially the south and northern sides – are planted to coconuts, mostly by pre-World War II German settlers and descendents of Omani Arab, Shatri, Shirazi and Baluchi pioneers. These provide work for many inhabitants and a source of cash income, especially during the dry season when most coconuts are harvested and home-grown food is in short supply.

The north-east of the island is covered by a dense coral rag forest, most now protected as the Mlola Forest Reserve. The north-central area is an undulating plateau that has spectacular baobabs, doum palms and Euphorbia.

Guests of Kinasi can make an excursion to any of the island villlages, where the people are welcoming and friendly. The people of the villages of the islands of Chole, Juani and Jibondo are engaged in a variety of activities from boat building to animal husbandry. Small scale farming (rice, cassava, fruits) is practiced on fertile soils and fishing and fish drying contributes to their income.


[1] The book was probably written in Alexandria by a Greek author in the First Cebntury, and is a guide to the ports and trade of Arabia, East Africa, India and the connecting route to China and comprises the first eye-witness written account of the coast of Azania.

[2] Claudius Ptolemy was also an Alexandrine Greek and composed Geographia in approximately AD150. The text however, is either heavily edited by other authors or has been added-to and is regarded as a compendium of all known and written information for sailors of that period.

[3] Portuguese maps and manuscripts from their earliest voyages (circa 1492-8) mark the Mafia archipelago as Monfia, Morfiya and Monfiyeh but these names are derivatives of the Arabic for “archipelago”.

[4] Settlement on the Azania coast was stimulated by an exodus from the north in the wake of the aggressive proselytizing of Islam from 400-800 AD.

[5] The word Swahili may have been adopted to denote someone who was a Musim; there is much debate about the word and its vernacular use; it is clear that it is not simply used to refer to the people of the coast nor those of mixed race (the latter were almost sure to be Muslims as well).

[6] The word Zinj is no longer believed to denote “black” people; rather there is much linguistic evidence to suggest it is derived from zi/za (body of water, the sea) and nji/nchi (territory/place).

[7] The coast of Tanzania is known to have been originally settled by Late Stone Age recently as 200-400 AD.

[8] An excellent, graphic account can be read in Esmond Bradley Martin’s Cargoes of the East.