TRADERS AND PORT-CITIES IN THE WESTERN INDIAN OCEAN IN THE SIXTEENTH AND SEVENTEENTH CENTURIES

In this essay I shall attempt to outline a general characteristic of the trading emporia of the Western Indian Ocean. I shall call this area the 'Arabian Seas', understood as an ancient network of commercial, cultural and social relations between India, the Red Sea, the Persian Gulf and Eastern Africa. The reason for regarding this area as a separate part of the Indian Ocean is, of course, fairly simple: one could sail from a Middle-Eastern port to the West coast of India in one trading-season, while, for pursuing one's journey further to the East one required another season. This network of ports provided points for reshipping the goods or loading them on camels and donkeyback. The ports were, to use a useful concept from the Levant-trade, steps or échelles between one trading-route and the next. While it has become increasingly normal to speak about the Bay of Bengal as a separate area of trade, I would consider the Arabian Seas (in the plural, as I include the interior of the Red Sea and the Gulf as well) to be another interlocking area. I would be very hesitant to designate this area as a 'world-economy', a term which is, however, currently en vogue. 1

In undertaking such an essay we may refer to two important theoretical lines of work. Firstly, we have the work of K. Polanyi, more particularly his essay "Ports of trade in early society". Polanyi interprets the 'port of trade' as a focal point of exchange between two different economic systems. Trade in this kind of port is regulated by the market, whereas in the economic systems which meet there, exchange or the appropriation of surplus may rather take place through 'redistribution' or 'reciprocity'. 2

Secondly, we have 'frontier' history, mainly exemplified by the work of O. Lattimore. 3 The latter put forward the concept of the 'frontier' as an extensive zone, where two very different economic systems meet, namely the 'pastoral economy' of Central Asia and the 'centralized agrarian system' of China. A plethora of intermediate forms between the two develop; pastoralism merges with extensive agriculture. Recently, this concept has been applied to China's maritime zone as well, where the agricultural economy of the hinterland met the maritime and fishing communities. To pacify this 'maritime frontier', a threat to the established order, the Chinese government engaged either in a policy to 'close' it by removing the peasants from the shore, thus offering them an opportunity to escape from the closed agricultural economy of the interior, or in a policy to incorporate the maritime areas through the use of military and naval force. 4

These two concepts can be merged if we regard the western Indian Ocean as the maritime frontier of Persia and Arabia on the one hand, and India on the other. Trade between the maritime zones intersected with that of the interior in a number of port-cities. These differed from the ones in the interior in having a special institutional regime.

The view of the Indian Ocean as a 'world-economy' has recently started to gain wide currency among the scholarly community. The Ocean is regarded as a wide trading zone, unifying the Asian economies. More seldomly, it has been seen as a frontier - an area where the pastoral/agricultural economies of the interior intersected with the fishing/trading economies of the coast. 5

CITY-HINTERLAND RELATIONS

It may well, however, be perceived as follows. Today's traveller, voyaging from Bombay or from Bombay to Surat may gather the impression that this is a very densely populated region; sprawling with towns, hamlets and cornfields. But up to the middle of the nineteenth century, and, in some cases till the construction of railroads between Ahmadabad and Bombay, this was not so. Southern Gujarat, from Daman up to Bulsar, was a thinly populated, dry region covered with shrubs, intersected by forested mountains. There were little roads, which, during the monsoon, were almost impassable. 6 Again, the Ratnagiri region, today one of the most densely populated in India, was during the seventeenth century largely covered by jungle, intersected by salt-pans and isolated hamlets. It was almost impossible to get from Chaúl to Goa by land, as the roads were interrupted by hillocks and forests. South of Goa though, were some densely populated tracts. Yet even those were intersected by wasteland such as the region around Karwar which was still covered with bush. 7 Kerala being an exception to the rule, one may well state that the West coast of India was a thinly populated region, apart from a tiny number of densely populated areas, intersected by wasteland.

What has been said for the West coast of India may be applied to the Gulf and the Yemen as well. Let us take a look at the Yemen: even though some of the valleys along the coastal ridge of the Hadramawt were densely settled, most of the Hadramawt coast was wasteland. The extensive Tihama plain was deserted, except for areas immediately along the main caravan-routes leading from al-Mocha to Tais and thence to Sa'na. The extent of cultivation in the Yemen, however, was closely related to the demand from cities. In the Yemen, towns created agriculture, not the other way round. 8

The coastal region was sparsely populated, and in many cases 'underpopulated'. Take Arabistan/Khuzestan in the Gulf, for instance, where, after the collapse of the dams and irrigation-works erected around Ahvaz, both population and cultivation declined dramatically, leaving most of the region to waste by 1700. 9

We can not venture an estimate of the population of the coastal regions of western India in the sixteenth and seventeenth centuries, for the source-material is woefully inadequate. Nor can we do so (for the time being, as sources are available but have not yet been studied) for the Yemen, al-Hasha and Iraq10 but we may make two comments.

First, the demographic balance of the western Indian Ocean was, in the sixteenth and seventeenth centuries, even more to the favour of India than nowadays, and this imbalance was increasing. There is some evidence that the population of those Middle Eastern regions bordering on the sea, like lower Iraq and the Hadramawt was decreasing, as was that of the Persian Gulf, especially in the region around today's Bushire. But the population of some Indian regions may have increased, more particularly Saurastra and Kathiawar. Sind though, seems to have participated in the Middle-Eastern trend towards a decrease.

The cause was often malaria, which tends to spread once irrigation systems are not kept in proper order. This was the case in lower Iraq during the wars in the thirties and the sixties of the seventeenth century for instance, and Khuzestan in the nineties of the sixteenth century. The ecological balance of the coast was, however, always precarious. To block the course of water for irrigation may spell disaster as well. This is demonstrated by 'old Goa' which was ravaged by epidemics emanating from the irrigation, and drinking water lake, to such an extent that it finally had to be abandoned. 11

Second, it seems that the region was getting drier. This may have been due partly to climatic changes (connected, perhaps, to the so-called 'little ice-age'), partly to deforestation. It resulted in the appearance of wide expanses of desert in Iran, the Yemen, Tigre and Eritrea. We read that in the early seventeenth century the Eritrean coast was still densely forested, whereas it is now an arid region, or about extensive agricultural lands around Bushire, today desert. In the early sixteenth century Socotra Island was covered by dense forest, which were wiped out in the next century. Part of these woods may have burned down, part destroyed by herds of goats and sheep imported from Hadramawt.

As regards India (where the evidence is more substantial), suffice to state that the West coast of India was suffering from similiar problems of deforestation. The capacity of the Indian eco-system to recuperate, however, was greater than that of the fragile system of the Red Sea and the Gulf shore. The latter region seems to have reached the limits of the population it could support in the sixteenth century. There is dramatic evidence extant from a survey of the coast of Oman in 1672. The decline of population was striking there: daddle-groves were abandoned everywhere; hamlets were depopulated; cities and towns had collapsed or were devastated; the desert was advancing throughout the region. 12

This demographic imbalance was the most fundamental factor in the history of the Arabian Seas: both the Gulf and the Yemen were only able to support a small population, whereas India still had extensive opportunities to expand. These were very much limited by the lack of capital to invest in agriculture, though. Indian agriculture in the seventeenth century was to a large extent 'extensive agriculture carried out with intensive methods'.

But because of the limited resources of manpower available in the Middle-East, often a labour nexus would occur. Indians might be sent as slaves, or contract-labourers to Mocha or Basra, to toil there, or they would enlist in the service of Middle Eastern-fleets, the Muscati-fleets recruited in Surat and Kathiawar. Bottlenecks forced up wages in the Gulf, so, on average, wages were much higher in Bandar Abbas than in Gujarat. The latter still had a pool of labour much more in the vicinity of the 'tribal regions'.13

What may also be confidently stated is that population and cultivation along the shores of the Arabian Seas were concentrated in a number of population-enclaves. Some were extensive, like the Saurastra coastal strip, some within a circumference of a few hours walking, such as the cultivated strip of land around Aden, or the grainlands on Qeshm Island wedged in by mountains and wasteland on all sides. In India we may mention the small cultivated area around the desolate island of Diu, closed off by the Girnar hills and the dense forests from the hinterland. 14 These enclaves served as 'gardens' for the ports. Take Essin, near Bandar Abbas. The inhabitants of the latter town had financed the excavation of irrigation canals there, and possessed daddle-groves. Another case may be found around Aden, where the citizens invested in the most profitable cultivation of gum trees. 15

It is often supposed that the cities in the Arabian Seas were mainly trading-centres with no connections with the agriculture in the surrounding areas. This impression would be faulty, for most inhabitants of the towns were to some extent interested in it. Though the landscape along the shore was that of a town, trade was not the main source of income. Agriculture was, closely followed by fishing. Though trading-centres might perish, often followed by a decline in cultivation too, (a very strenuous pursuit particulary in the salty soil of the Gulf region and the Tihama) fishing hamlets would subsist.

Albeit, little is known about fishing. This was the pursuit of a large part of the population, and history can often be explained by looking to populational conditions. For instance, fishermen from Oman were supposed by the Dutch East-India Company (VOC) to be sturdy sailors, whereas mariners, recruited along the Konkani coast, were only fit for coastal navigation. This is explained by pointing to the fact that fishing along the Konkani coast is confined to coastal waters. Fishermen from Oman had to venture far afield for a plentiful catch. Political history can impinge upon fishing too: the conflicts between the Omani imamate and the city-states along the Hadramawti shore in the seventeenth century, and between Oman and Mogadishu in my view, by and large, involved fishing-rights and dues upon this. 16

There was little overland communication between these populated enclaves. The clearest cases are found in the Gulf: not a single road existed between Bandar Abbas and the neighbouring port of Kung/Lingeh. Communications between Musqat and Shur were extremely bad. This was the case not just in the Gulf: the links between the various Hadramawti ports were tedious as well, and the overland connections between Jiddah and Mocha were a difficult venture due to the 'stony tracks' and to thieves.

The West Indian Ocean, detail from a map by João Teixeira I, 1643

Ports were linked overseas. Some were so heavily dependent upon this that they had virtually no communication with the hinterland. Portuguese Musqat for instance, was linked to the Omani interior by four mountain passes which were all considered extremely dangerous because of robbers, and very difficult to pass through. Suhar enjoyed better links with the hinterland, but these too were tedious. 17

Such places as Aden, Musqat, Qatif, Saif, or Suakin may be perceived as virtual islets floating in wasteland and supplied by the sea.

These 'emporia' were exceptions. Generally, the road network converging upon the harbours was turned towards their communications with the hinterland. They often exercised but little influence upon their immediate surroundings. This is clearly perceived along the Swahili coast. The history of many of the ports can be seen in the light of their communications with important cities in the hinterland.

We could cite several cases. To mention one: the Bahmani port of Goa derived its importance mainly from its status as a 'royal port' from whence horses were brought to Bidar and Gulbargâ. Goa was one of the few harbours along the Konkani coast, which were not given out in fief, or controlled by the Sawantwadi raja. 18 Another, Barcelore (Basrur) derived its status in the early sixteenth century mainly from being a transitional port for Hampi/Vijayanagar. After the fall of Vijayanagar, however, its trade was given a new span of life through the expansion of the rice trade in the sixties of the sixteenth century. 19

Communications were concentrated on the interior rather than with neighbouring ports regions. This is the case of the Banadir coast: ports like Mogadishu and Zeila, both situated on the verge of the desert or pastures, had to obtain everything from the trading caravans connecting then to Zanj and Tigre. A striking case too was the grand emporium of Suakin which had to be supplied completely from Ethiopia, or Egypt, as the site was altogether sterile. As regards Hormuz, the saying was that Hormuz had to obtain "its water from the mainland, its wood from the sea".20

The cities in the Arabian Seas clearly formed a frontier society. They were the nodal points between the densely populated and regulated agricultural tracts in the interior, and the sparsely populated coastal fringes.

THE HARSHNESS OF LIFE

The coastal lands were mostly unable to feed the cities. They were heavily dependant upon the import of foodstuffs from the interior. It was a frontier society in that it was also highly sensitive to the slightest freak of nature that might cause a disaster. For instance, a grasshopper-plague in 1687 in Rig caused starvation in Basra, similiary in Mucha in 1688. 21

We are not well informed about the diet in these maritime cities. Seemingly it was one-sided and poor at best. Salaries were low; where they were higher, as in Bandar Abbas, prices were too. This city was one of the most expensive places in the Indian Ocean. It was, however, exceptionally well-endowed with supplies as large amounts of cattle from the Makran were sold there. So too was Surat, which had good pastures nearby in Souali. These high-priced enclaves tended to attract foodstuffs from throughout the Indian Ocean. Hormuz and Bandar Abbas were supplied largely from the Hadramawt and from the Canara coast. Surat was provided for from Mewar and from the Canara coast. The resources of the towns in the Konkan, which were not supplied from abroad, though, were rapidly exhausted. To buy lifestock for the VOC's fleets blockading Goa, one had to obtain this from far away Navsari. 22

Provided Bandar Abbas was provisioned from the sea, we find no mention of starvation there, while mention of famine in the Yemen, the Konkan and Gujarat are legion. Even with a bountiful harvest, most of the poor in the towns of Gujarat seemed to have lingered on the verge of starvation. Saurastra was at least twice visited by famine in the seventeenth century: in 1632 (the infamous satyagraho kol) and in 1660. In Kerala, the Dutch thought that the population was 'content' with extremely little food. South-Kerala was plagued by epidemics in 1670, which caused many deaths. 1632, and the eighties of the seventeenth century were years of disaster for the Konkani coast. The population was then decimated by famine and plague. Portuguese texts still have to be studied for what they contain about this subject in the sixteenth century, but I may assume that the situation was hardly any better. They record, for instance, a famine in Vijayanagar in 1554, when in many places people were forced to feed from corpses. 23

Because of the congestion of people caused by rack-renting, the massing of manufacturing around the shipyards and the settlements of merchants, as well as the malnutrition or at least monotonous diet, the townfolk were prone to epidemics. The history of Surat in the 1680's seems a tedious succession of plagues and scarcity, with the city nonetheless enjoying its peak of commercial prosperity! 24

An account could be written on the history of the plague in the Arabian Seas. Apparently the area had two sources of contagious diseases: the cities of the Ottoman empire and particularly the huge conurbation of people from all over the world, rats and lice in Istanbul. Of course, the origins of epidemics are habitually traced to neighbours, so the Ottomans habitually blamed Persia for the plague. However, Istanbul seems to have been the hub from which the great plague of 1681 spread eastward to Anatolia, and thence to Isfahan and upward to Marw, Basra, Bandar Abbas and India. Again, the Euphratus delta was a notably unhealthy area, repeatedly ravaged by epidemics. 25

Secondly, we have Eastern Africa: Mozambique and the Zambesi valley in particular, where African diseases merged with plagues from Europe and Asia. The area was so unhealthy that even statistically one had but a one in four change to survive a stay at Mozambique. 26

But, although spots like the Gulf, Bombay, Mozambique and Diu were notorious 'fever-holes', all port cities were perilious. They were meeting-places of diseases from all over the globe. Even the mortality rate in such a renownedly pleasant station as Cochin was high. This applies not just to Europeans for, although they were particulary prone to disease, Asians and Africans succumbed too in graveyards such as Mozambique, Hormuz or Diu. The reluctance of Indian and Persian nobles to establish themselves on the seashore, which has often been commented on, may have been perfectly logical. For, apart from spices, the ships carried diseases as well.

POPULATION

The epidemics and the constant malnutriton caused a very rapid turnover of the population. This would have caused depopulation without a natural disaster, or populational migration. Take a way station such as Socotra for instance: though the religious history of this island is still shrouded in obscurity, it seems from 1500 to 1650 to have turned from a mainly Christian to a Muslim island. The Christian population gradually disappeared and was replenished with newcomers from the mainland, such as slaves from the Banadir coast who were mostly converted to Islam. One cannot ascribe this to massive flight of the population to elsewhere, so this seems to have been caused mainly by the gradual demise of the Christian population, due to epidemics. 27

As the maritime areas were heavily dependent upon immigration, this caused their population to be transitory, composed of people from all over the world. Although most pre-industrial cities had to be replenished by immigration, the cities of the Arabian Seas markedly displayed this characteristic. Apart from the 'normal' strains of urban life such as 'pollution' and venereal disease, the population of the port towns faced the additional hazards of famine and war.

These were a constant drain on the population of both the coast and the interior, serving as a magnet for wide areas. Traders and artisans were attracted to Surat from as far afield as the Coromandel coast or Hindustan. But they decimated them too: in Surat in the summer of 1686 it was estimated that as many as two hundred people succumbed daily to the plague. The plague emptied Berhampur, however, whereas Surat's population was on the increase. As regards the harshness of life in Bandar Abbas, we may quote VOC-commander Ysbrand Godske: "in the summer the temperature mounts there to unbearable heat, which causes sudden afflictions of fever, both among us (the Dutch) and among the indigenous people, so that one day one may be alive and well, the next ill, and a few hours later dead".28

This shortage of manpower gave rise to a fairly extensive slave-trade to the grave-yards on the Euphratus, to Bahrein, (where to diseases were added the perils of pearl-diving) and to the Tihama from Ethiopia and Eastern-Africa. It gave rise to a demand for trained personnel as well, particularly after epidemics, which often induced indigenous states to look for labourers among the Europeans established in the Arabian Seas. This happened among the Portuguese in particular, as they were often married to indigenous wives. 29

M. Pearson has made a case for seeing these 'fleeting' communities as a proof of the fundamental unity of the Indian Ocean. 30 One might argue, however, that the opposite was also true: they were part of 'trading diasporas' spread throughout the world. One may drop the first word though, as there was a diaspora of manpower, slavery and expertise as well. These diasporas were not confined to the Indian Ocean: the bania-trading network extended to Russia, the casado-network was linked to the trade of the Americas. 30

One may see these diasporas as an expansion of traders, or, more often, humble mercenaries and peddlers, constituting a veritable counter-expansion to European expansion. These communities followed the cash-flows from America, and the increasing flows of bulk-goods. They were being drawn in by the emigrational pull of the port-cities.

While the great trading-diasporas, such as that of the banias, the khatris, or the Armenians have attracted most attention, large parts of this diaspora of manpower constituted a drifting 'sea-proletariate' as F. C. Lane called it, 31 spread throughout the Indian Ocean and beyond.

One example are the 'Ethiopian' slaves or sidhis which were found in large numbers along the Konkani coast so as to form separate communities in places like Rajabag, or Gheria. They constituted a large community in Surat as well where they worked as soldiers and upon the ships. African slaves are found working in the daddle-groves of the Tigris delta, in Hormuz and were even sold, apparently from Persia and the Ottoman-empire, to Russia. 32

While these sidhi-communities go back to Pharaonic times, the Luso-Indian communities date back to the sixteenth century. To plot the spread of the Luso-Indian lançado diaspora upon a map would be to track the geographical extent of trade carried on in the Indian Ocean. While work upon this theme is proceeding on the Bay of Bengal area33, this has not so far been undertaken for the Arabian Seas. The lançados were as widely spread there as they were in the Bay of Bengal. One may find Luso-Indian traders or mercenaries involved in the ivory and gold-trade or in running the affairs of state in Malawi where their trading-networks intersected with those stretching from Luanda. One might find them at Agra and Lahore, at Bijapur or at Golkunda/ Hyderabad - both of these communities including several hundred persons --, or at Damvia in Ethiopia in the early seventeenth century, or Laribandar on the Indus river. 34

We may dwell upon the crucial role of mercenaries from Baluchistan to the Portuguese and later Musqati maritime trade and warfare too. But this may suffice to state our case that immigration was always necessary to replete the otherwise rapidly decreasing reserves of manpower of the ports.

These migrants mainly came from 'poor' areas: the Indian mountain-ranges such as the Satmala range in Gujarat, the western Ghats along the Konkani coast in India, or the mountain terraces of Oman. The latter area was particulary involved in the export of man-power even before the Islamic period: Omani-communities may be found from the Maldives to Madagascar, from Malindi to Sumatra.

Often these were areas of staunch and ascetic religious beliefs and practices, suffusing 'new blood' in the expansion of Islam. One may mention the role of qazis from the Hadramawt in the jihad waged in the dar al-harb of such places as Atjeh, or Kerala. This migration assured the prosperity of the areas involved: the 'overseas fortunes' were invested in land or irrigation-works. But this would foster envy and rising expectations in the mountains. The 'hillbillies' wanted to grasp the riches and loot the coastal towns. Where the forces of mountains and desert outweighed the garden regions, like in Oman, this would ensue in a pattern of fitna and assabiya, to use Ibn Khaldun's terminology. The cyclical rise and fall of the power of the state, occasioned by civil strife. Maritime expansion was followed by unrest, contraction of the seaborne empire, foreign invasion, and a new assembly of the tribes, under the sway of the imam, to withstand the foreigners, which ensued in turn in a new wave of maritime expansion and imamate. 35

Elsewhere, the mountains, overlooking the plain, might be a cause of worry, but the tribes hardly wielded the power to reduce the coastal towns. The latter had to settle with the tribes in allowing them their shares of the profits gained from commerce. For instance, Baluchi-sultans were allowed a substantial benifit of the custom-dues of Hormuz. On their own accord, however, they hardly disposed of sufficient power to conquer this stronghold. If tribute was not paid, they could, and did, interrupt the landborne trade between Hormuz and Sind, which was still quite important in the sixteenth century. Hardly an unsurmountable problem, but harmful to Hormuz's trade, as Hormuz was largely supplied with Punjabi textiles overland. 36

Once newcomers from the mountains had arrived in the coastal towns they rapidly merged into the melting-pot. The port-towns were a 'frontier' society too, in forging a synthesis between the various cultures, cohabiting there. This stands out in considering the Swahili coast, where the often tense cohabition between African traders, artisans and workers, Arab sailors and merchants, Indian (slave-) traders and Persian 'ulama forged a distinct cultural synthesis. 37Several other cases, for instance, the forging of Islamic identity; a synthesis between Hadramawt and Hindu creeds, along the Malabar coast may be mentioned. 38 Why dwell upon such cases, though, while we have one at hand with Portuguese Goa, which will be more familiar to my readership?

"Goa is different" as Goan nationalists would insist, in some respects, but it can be considered as a typical port town of the Arabian Seas too.

First, it was typical in being not a large town: about 40,000 inhabitants at most by around 1580, a figure which can not be compared to the huge cities in the interior like Agra, Delhi, Baghdad, or Isfahan. Surat, which may have counted up to 200,000 inhabitants in 1700, was an exception. 39

Essentially the main constraint upon the size of the population was that these towns could not be provisioned. The urban authorities had to keep a watchful eye upon its supply, whether through the food-dues (collecta) at Goa, through forced deliveries of goods from Qeshm in Hormuz, or through a separate office to send grain from the Nile to the fortresses on the Red Sea, as was the Ottoman practice. 40

Second, it was typical in this respect too. While Goa could be partly supplied by the sea, this was not sufficient. It had to be provisioned from the Deccan, and depended upon the regular supply of grain from thence; it was estimated that the Adil-shah commanders on the Baghalaghat could starve the city within three months.

Third, it was altogether typical in being a 'graveyard'; one out of three Portuguese arrived there may have died within a year. Lest it be thought that this was a typically European concern, the Ottomans were struggling with manning their fortresses on the Red Sea, and the Safavids had the greatest of difficulties in manning the ones in the Strait of Hormuz. Most troops were pressed to serve there. These contingents of 'mulazim' or militia, were mostly in very bad condition, and apt to mutiny. Apparently, however, the empire was most reluctant to commit its precious qizilbas -troops to a lengthy stay there. 41

Fourth, because of this shortage of manpower, Goa depended heavily upon immigration. Workers on the building-sites and the shipyards were recruited by jobbers in the wide surroundings; many returned to their homesteads during the monsoon. Goa still had a sizeable Hindu population, even after the forced conversion of the population of Tiswari, Bardes and Salcete to Christianity. Many Hindus owned allegiance to the Punda temples, and visited it on pilgrimages. In the seventeenth century, the number of Hindus may still have surpassed that of the Christians. 42 What is stated here for Goa applies just as well to many other port towns such as Surat, or Basra, which was also a magnet for traders, herdsmen, or workers from al-Hasha. 43

Fifth, it was typical in being a blend of Hindu, Islamic and Christian beliefs and practices. While I would leave the study of this matter to art historians, this emerges clearly enough from any study of the iconography of the Goan churches.

Finally, nevertheless, the luscious golden luxury of its churches should not obscure the fact that Goa dourada was an extremely violent society. Apart from the public delight in executions, hangings and 'burning at the stake' parties which were as well frequented as in contemporary Europe, Goa's society was typified by the extent of corporal violence. Even during peaceful periods, slaves were not permitted to bear arms, so that they would not engage those of another lord. Hindus were not permitted to, and fidalgos only restrictedly. 44 And this is by no means atypical. The cities were also full of skirmishes, wars, or rumours of war. To hire an assassin was the easiest thing in the world in Surat. Due to the many different nations cohabiting, conflicts might resemble today's 'communalist strife'. In Goa crowds of Christians demanded the immediate expulsion of Hindu-traders, sucking the life-blood of the poor. If the viceroy would act against the banias, this was often prompted by such communal tensions. 45

While the various ethnic groupings tended to cluster in wards (mahal, barrio); borders between these were no way always identical to caste, religious, or 'ethnic' groupings ('ethnic' groupings such as, for instance, the 'black' wards found throughout the Arabian Seas). Wards were rather formed by a cluster of dwellings around the house of a powerful merchant, or priest. These were often ethnic or caste-groupings, as seems to have been the case in Goa, certainly in Hormuz, or Surat. But one should see the various groups involved in trade as 'nations', a descriptive group with separate judicial and economic privileges allotted by the state. Central to these was the allegiance to a 'saint', or 'sanctuary'. They were 'descriptive', as not the state but merchants would decide who was a member. 45

THE MORPHOLOGY OF TOWNS

The history of the Indian Ocean is generally perceived as a history of the rise and demise of port towns. This is all well known; one factor is the complementarity between the Red Sea and the Persian Gulf, trade to the one area opening when the other closed; another the rivalry of the ports in South and North Mozambique; still another that between Cambay, Surat and the northern Konkan (ports such as Dabhol and later on Bombay): the primacy of these ports depended, up to the nineteenth century, upon the importance of the trading-routes in the interior. But we should guard against dramatizing. Surat, for instance, was already an important port in the sixteenth century. Cambay's trade may have declined somewhat, but it continued to be an important trading centre throughout the seventeenth. The history of the port towns is often not as dramatic as suggested by such simple notices as 'the rise of Surat and the decline of Cambay'. Yet, while port-cities display more urban continuity as is often thought, and the 'grand emporia' by no means monopolized the trade of a region, it is true they were often markedly different from European towns. Many towns did not display the urban continuity of the latter, and rose and fell quite abruptly. 47

What struck most European observers when visiting a port town in the Arabian seas was the delapidated state most were in. Buildings were in disrepair; sewage non-existent; if public buildings existed at all, they were badly constructed; and if cities were walled these often were in disrepair. While some of this may apply to most Islamic cities, the rather shabby town of Bandar Abbas contrasted to the splendour of Isfahan; the shambles of Mocha to Sa'na or Zabid. 48

This is in a way a mirage of today's underdevelopment. Though the urban squalor may have increased with the 'population-explosion', seventeenth century Surat was as unwholesome a place as today's Bombay. But a more appropriate comparision might be with the 'boom-towns' on the Brasilian or American 'frontier', which pass from foundation to decay without ever having known a flowering. 49

What they have in common was their provisional character. The shores of the Gulf were strewn with abandoned sites, like a forlorn mining-town in Minas Gerais. Take, for instance, Jask, formerly an important harbour, which was visited by VOC -sailors in 1645:

They witnessed some abandoned holes, wooden crosses, whereupon plackats were nailed; a few pieces of wood from which they saw graves, and a few goats, which is everything they saw. 50

A real ghost-town this! Such sites grew virtually overnight from tiny hamlets to large centres. Urban institutions and services had no time to develop. Once the trading routes, upon which they depended, declined, they would suddenly decay. One might mention the fate of Bandar Kung, which consisted of only twenty houses in 1628, grew to one of the largest cities in the Gulf round 1670, and disappeared in the eighteenth century. 51 In the Gulf, cities rose and fell dramatically: great emporia like Shiraf, Hormuz, or Kung went through such a cycle within a span of a hundred years. Cities like Hudeida, Mocha, or Luheia have a fairly recent history too, though, and rather resembled the 'boom-towns' of the Gulf. 52 Indian trading-centres of importance in the seventeenth century like Sidlmisr in the Konkan have disappeared, or, like the great inland emporium of Rajabagh been reduced to insignificant hamlets. Others, such as Bhaunagar in Gujarat can boast but a recent past. The fate of ports depended often upon the dynasty which founded it. One may not only think of the case of Bhaunagar or Purbandar in the eighteenth century, but of Diu as well, being the fief of Malik Ayaz or even of some trading-route to a major city in the hinterland, such as the Rajabad-Belgaum route in the seventeenth century, which made Rajabag's fortune. 53

Once trading routes shifted, ports might dwindle into oblivion. Take, apart from the classical case of Hormuz, Sindi-ports like Mandvir, an important port in the sixteenth century and still a substantial harbour in the 1670's, which thereafter disappeared. Again, many ports in the Konkan experienced such a brief period of prosperity followed by a sudden decline, such as Vengurla, which mainly prospered because of the trade conducted there by the VOC.

Yet these, 'disposable' towns are but one face of the cities in the Arabian Seas. Take, on the contrary, the ancient towns of the Banadir coast: Mugadishu and Zeila, city-states, surviving from the tenth to twentieth century. While their fortunes widely varied, these towns continued to attract trade throughout this span of time. 54

Though ruled by a sultan, the ports of the African littoral have more in common with the European city-states than often realized. There, like in Europe, the city was run by a powerful council of eldermen. To be a citizen of Malindi, or Kilwa was something to be proud of. They would hire foreigners to fight among themselves (be they Muscati or Portuguese hirelings). Often they would be split by a lengthy conflict between 'nobles' and mainly African 'commoners', strife which even extended to the language spoken: Arab or Swahili. But this did not impinge upon the feeling of urban solidarity. Though 'urban air' hardly made one free on the East-African littoral, slavery in these cities was milder than in the areas considered to be dependencies, for servants there largely worked in the household, while in the plantations they laboured in agriculture. 55

The African port cities may be termed city-states; Jidda, Cananore, Purakkad, Aden, or Cochin may be termed so too, however, although by the sixteenth century ruling over an extensive sweep of the shores of the Gulf, the Turanshabi-the Turanshah ruled, but only over the small stretch of land of the Biaban and Bandar Abbas-province; the remainder either consisted of isolated fortresses, such as as-Shur, or islands, like Bahrein. The Turanshah paid tribute to the indigenous rulers to ascertain its paramountcy. We might accordingly distinguish 'market', or 'port cities', which tended to rise and decline due to shifts in trade-routes, and had hardly any urban institutions, and 'city-states', which existed over prolonged stretches of time, mostly ruled by a class of urban 'notables' on the other. 56 They might experience prolonged periods of decline, like Kilwa in the sixteenth century and yet would survive, and often resume, a prominent place. The main port might have shifted to a near bay, or inlet: take the rivalry between Lamu, Patte and Kilwa. Their history was more stable than that of the' port cities', either because of their religious importance, a strong civic sense of its inhabitants, or, most importantly, because these ports were not on the fringes of large territorial agrarian empires, like Vijayanagar, the Deccan sultanates, Gujarat, or the Mughal and the Safavid-empire. While the rise of the Mughal, Safavid and Ottoman empires along the Arabian Seas in the sixteenth century way have well stimulated trade, this spelled disaster to many small city-states, a form of institutional organisation which was common in the fifteenth century. Many were subdued by empires, such as Aden, Massawa, or Basra, or by the Portuguese like Daman, or Bassein. They might be forced, too, to enter a precarious alliance with one or several of these states, as was the case with Basrur, Kilwa and Mombassa. 57 The seventeenth century was the high tide of 'imperial rule' over the older city-states. Most were weakly absorbed into the administrative apparatus. The exception is Mozambique where the city-state rose to prominence again, whereas the interior was disturbed by the decline of the Monomotapa-state. 58 In the eighteenth century city-states began to shake off the clutches of empires; some built maritime states (such as the Bani Ka'ab in Kuwait), or sought British support. To an extent Oman, under the sway of Ibn Saif, may be perceived as such. An exception is Kerala, where city-states like Purakkad or Quilon existed throughout the seventeenth century. Instead, the eighteenth witnessed the rise of a territorial state: Travancore, and of its 'market city' of Aleppey. 59 ^^ INSTITUTIONAL SET-UP Within this context of the relation between military-agricultural empires and their maritime fringes, Polanyi's concept of a 'port of trade' is of significance. These harbours connected the 'agricultural economies' in the interior to the trade along the coast-line. This necessitated a particular institutional arrangement by the interior states. That coastal commerce was significant is evident from recent research. The Indian economy was in the sixteenth and seventeenth century evincing something like a commercial revolution. Take the increasing taxation in cash, and the growth of market relations in the interior. 60 About Persia, let alone the Yemen, little is known. The impact of the booming coffee cultivation upon the agricultural economy of the Yemen, for instance, is a moot question. The very rise of this, however, would seemingly point to the capability of the peasantry to respond to demand on the world market. It would have worked an increase of cash-flows into the Yemeni-economy too, for the spice-trade touched only its fringes. As regards Armenia, the part of the Safavid empire best studied, some have not hesitated to speak of the rise of an Armenian capitalist class in the sixteenth century. 61 As states were confronted with a cash nexus to expand the revenue base, port towns were of paramount importance for sluicing cash into the economy. They were often no more than points for the distribution of spices though one should not slighten the importance of other goods imported such as horses (generally a royal monopoly), or luxury-goods, which were often important for the exchanges of gifts at the court such as carpets (exported in large quantities from Bandar Abbas and Basra) or swords (a speciality of Hormuz). These were generally royal monopolies. The state was hardly able to control the trade with precious metals, although the Persian empire often tried to. To cream off cash from the port cities, a distinct institutional set-up was needed, allowing for the free conduct of trade, while benefitting the state. Amsterdam, 1672 (by courtesy of Universiteitsbibliotheek, Amsterdam).

Easiest was a tribute-arrangement where merchants or rulers of ports were to offer presents to the lords of 'agricultural empires'. This was the case in remote spots in the Persian Gulf such as Bahrein or Qatif, and in the Swahili-cities Mombassa paid tribute to the Muzungolos, Querimba to the Yao 'empire' in the seventeenth century, for instance.

To refer to the theories relative to the maritime frontier in China, this may be called 'sealing-off' or 'excluding' the maritime frontier through buffer-states. Though this is probably less the case in Western than in Eastern Asia, it can not be ruled out that the state wanted to prevent its flock from merging with the 'fuzzy-wuzzies' along the seashore. For instance, in the case of the narrow straits of Hormuz, many a Persian farmer migrated to Oman, where the huge irrigation-canals (aflai) laid out there, are largely of Persian design. Persian schemes against Muscat, rampant both in the seventeenth and eighteenth centuries, may have been concerned with forestalling this migration. 62

This arrangement seems to have been normal in Vijayanagar. Again, the Safavid-empire, following the Il-khans, and the Timurids, long stuck to it in dealing with maritime buffer states like Bandar Rig, Hormuz, and in a way Lar (as Lar was mainly a centre for the reconveyance of maritime trade). 63

The other form was the 'inclusion' of the maritime areas within the agrarian empire. The requisite way was custom-farming. This exposes the leaseholder to the risk of an unsuccesful transaction, not the state, since he will not kill the chicken laying golden eggs. So, wherever we look, we find this, be it at Bandar Abbas, at Basra, at Mocha, Dabhol, or at Surat. In the latter place the customs were often leased by powerful grandees of the court. Even when officials were named, as was the governor of Surat, such a nomination would have to be underpinned with the bestowal of substantial 'gifts' to get a hold of such a position, and retain it. The 'institutionalized corruption' of the Estado da Índia, to use N. Steensgaard's term, well fits into a pre-established pattern, therefore. In fact, even the names of farms at Goa were derived from the Bahmani ones. The leaseholds of Hormuz, are unmistakably Turanshahi in origin. No doubt, the Portuguese were following their inclinations in giving out the state incomes too, but they were acting according to established practices. 64

"Jente onrada portuguesa da india" ("Members of portuguese gentry in India") (Cod. Casanatense, ibidem)

That this often had dire consequences is clear. As farmers but briefly enjoyed their leasehold, there was no use in investing in urban services. In the booming trading city of Surat in the seventeenth century, governors would rather construct splendid caravanserais than invest in amenities, such as improving drainage or sewage. The city exuded a stench of excrement, and its roads were little more than pools of filth during the monsoon. Though it was proposed to have some streets paved and illuminated, this was turned down by the merchants as being but an excuse for extortion. Even the wall of Surat, laid out in the seventies, was so badly constructed that it repeatedly collapsed and offered no resistance to any enemy whatsoever. 65 Nor did most other cities have a significantly better record in this regard. This sometimes might have disastrous consequences; entrances to the harbour were blocked by sandbanks and wrecks. This was, for instance, the case with Dabhol; the harbour of which had become inaccessible by 1639, as nobody cared to remove wrecks, nor had forestalled landslides, which had washed the hillside overlooking it into the bay. 66

Tax-farming, though often pernicious, was probably the best system at hand for an ancien régime state with limited resources at its disposal. The development of trade is therewith allowed to entrepreneurs, often with considerable resources at their disposal, who are allowed to develop trade. One may think of such powerful entrepreneurs in ships, money and manpower as Hodsia Athar, minister to the Turanshah in the early sixteenth century, or Sabibi Begam, Shah Jahan's beautiful and cunning daughter, who sought to monopolize Surat shipping in the middle of the seventeenth century. 67 Of course, alongside such clever folk one might find loafers or spendthrifts.

The greed of a custom-farmer might be forestalled, though, by the normal splitting of the administration in two, into the 'civil authority' and the 'military authority', customs and military, vali and shahbandar, or between fortaleza and alfândega. Whereupon 'military' and custom-authorities would counterbalance. Again, the Portuguese took over the established practice, though refining it by putting two distinct administrations which were not to impinge upon each other. This is hardly evinced in the Persian or Mughal administrative practices, for there, authorities generally overlapped. 68

The lay-out of cities would generally reflect this administrative division; thus the two main buildings would in general be the palace of the governor and the custom-house (or alfândega). The cities were often walled, but these were often badly constructed, or in disrepair. One could not find large edifices there. Most striking were in general the caravanserais, often the fulcrum of the entire town. Houses and shanties would be grouped around the mansions of opulent merchants, noblemen and nakhudas, with the intervening space between the mansions hemmed in by the slums. Ports were often sprawling, built on the outskirts of the town and interrupted by stretches of waste. Speculation would leave empty patches, wherein would move the poorest. The few stone buildings would be intersected by wooden shanties, exposing the towns to the everlasting danger of fire. This danger was increased by the many workshops intersecting the buildings, which used inflammable materials. Mocha was, for instance, burned to the ground completely in 1665 and Surat was always plagued by fires during the summer. 69

From the market-square, a tangled web of alleys extended, narrow and hemmed-in by low-storied houses. The cities were generally divided in 'quarters', of which a caravanserai was the centre.

But we should not generalise: the Kerala ports had a very different outlook; so too had the ancient city-states such like Kilwa, or Malindi, the 'bazaar' of Cananore, or Mugadhishu, towns surrounded by high walls within a labyrinthine web of houses often several storeys high. 70

"Caretas enque vão molheres per caminho no Reino de cambaya chamão-se sejualis" ("The carriages in which the women travel through the Kingdom of Cambay are called sejualis"), (Cod. Casanatense).

CONCLUSION

What all these towns had in common was their strong reliance upon the hinterland, both as regards trade and administration. While Surat in the first half of the eighteenth century increasingly evolved independent from the Mughal administration in Ahmedabad, the city still clung to a semblance of Mughal authority; the governor was seen as the nawab of the emperor. This fiction collapsed only in 1759, with Surat no longer sending something like tribute to Ahmedabad, but to Bombay instead. The Indian merchants were ruled over by a proper court of justice, and the nabab, too, clung to his majesty, while disposing of but a vague remembrance of power. Authority there was increasingly devolved to the British residents, and Bombay. Its trade (and this was the most basic development) increasingly turned towards the sea, and land-borne trade declined. Even though it lost part of its trade, particulary to the Gulf, the merchant-class of Surat still profitted from the supply of grain and foodstuffs to the Indian army, and from the supply of cotton and opium to the China-trade too. 71

"Jente parsia dorreyno de ormuz: mouros" ("Parsees from the Kingdom of Ormuz: Moors") (Cod. Casanatense).

Most European towns had taken over the administrative system of their Indian kindred, though the 'European' ward (generally the fortress-area) was of course ruled over by a European court of law. The 'native' or black town was administered by indigenous representatives, such as the guild-masters in Goa, or the 'headman' of the banias in Diu. Yet, with heavier European involvement in 'country-trade', and the rise of European empire in India, the 'fortress' became increasingly involved in planning the 'black town' and remodellng its judiciary.

Most importantly though, by the 1780's the port towns extended their power more and more to the hinterland. A new kind of town arose along the shores of the Arabian Seas: neither market-town, nor city-state, but administrative centre,wherein European civil servants cohabited with Indian rural powerholders. Cities, which mainly derived their trade from the cultivation of cashcrops in their dependencies. Generally they bore a strong military imprint, because of the European garrisons there. While Bombay is of course the prime case, this applies to Panjim as well, and probably to late Dutch Cochin.

Although this applies to India, however, the older pattern of port cities was hardly completely eclipsed throughout the Arabian Seas. In fact it bore its last and most precious flowers along the Swahili coast in the early nineteenth century. Everybody who has visited some Swahili town will have been impressed by the perseverance of the cultural and architectural tradition of the Arabian Seas there.

"nubis que abitão no estreito demeca da bãd. da ethiopia" ("Nobians women who live in the Straits of Mecca in Ethiopia") (Cod. Casanatense)

"Goa", in Nauwekeurige beschryvinge van Malabar en Coromandel (...); Philipus Baldeus, t'Amsterdam, by Johannes Janssonius van Waasberge, Johannes van Someren, Anno 1672. Colour engraving, Photograph from the Arquivo Histórico Ultramarino, Lisbon.

REFERENCES

1. cf. for instance A. Wink, "Al-hind India and Indonesia in the Islamic world economy ca. 700-1800", Itinerario 12 (1) (1988), pp. 33-70, and N. Steensgaard, "The Indian Ocean network and the emerging world-economy", in S. Chandra (ed.), The Indian Ocean: explorations in history, commerce and politics (London, 1987), pp. 125-150.

2. K. Polanyi, "Ports of trade in early society", The Journal of Economic History 23 (1) (1963), pp. 30-45.

3. Most notably in his Inner Asian frontiers of China (New York, 1940) cf. also his Studies in frontier history (Collected papers 1928-1958) (London, 1962) cf. for instance D. H. Miller and J. O. Steffen (ed.), The frontier: comparative studies (Oklahoma, 1977).

4. B. Wiethof, Chinas dritte Grenze, Der traditionelle chinesische Staat und der Küstennähe Seeraum (Wiesbaden, 1969), cf. also J. E. Wills, "Maritime China from Wang Chih to Shih Lang: themes in peripherial history", in J. E. Wills (ed.), From Ming to Ch'ing (New Haven,1979), pp.201-238.

5. cf. J. C. Heesterman, "The Hindu 'frontier", Itinerario 7 (1) (1989), pp. 1-25, and from a very different point of view C. R. Boxer, "Some possible fields of research in the history of Portuguese India", in J. Correia Afonso (ed.), Indo-Portuguese history: sources and problems (Bombay, 1985), p. 186, yet Aden was habitually perceived as the frontier of Islam O. Löfgren, Arabische Texte zur kentnis der Stadt Aden in Mittelalter (Upsalla, 1883), p.1.

6. On the road network in Mughal India compare J. Deloche, La circulation en Inde avant la révolution des transports (Paris, 1980), 2 vol. An excellent work which is, however, marred by resorting insufficiently to Company archives; for the road from Daman to Bulsar cf. for instance: General State Archive The Hague (Algemeen Rijksarchief henceforward ARA), Sweers/Manis Collection 5: Description of kingdoms on the road from Golkunda to Aleppo overland, the journey to Aleppo taking about five months, and for a latter period cf. also for instance S. C. Ghosh, The peninsula of Gujarat in the early nineteenth century (New Delhi, 1977).

7. A. Hamilton, A new account of the East Indies (W. Forter, ed.), 2 vol. (London, 1930), i. p. 248; timber from Karwar was generally used in the Bombay dockyards, S. Campbell (ed.) Gazetteer of Bombay city and island, 3 vol. (Bombay, 1911-1916), ii, p. 188; the Dutch were very impressed by Kerala's dense population; attorney-general to the Company Pieter van Dam even put forward a figure of around 10,000,000 inhabitants for the coast as a whole, which would seem to be a credible figure: ARA, Archief Staten-Generaal (Estates-General archives), 5741, despatches relative to the VOC (liassen VOC), Report by Pieter van Dam, 21 December 1683.

8. Utrecht Archives, Huydecooper Collection, bundle 621, Instruction for the chief merchant which Jan Jozua Kettelaar left, on his leaving Mucha, to his deputy mr. Daniël Hagendoorn, 1 August 1709; C. Niebuhr, Reise nach Arabien (Copenhagen, 1772/Graz, 1969), p. 223 ff.

9. For Ahvaz (Lhasa) cf. for instance A. Reis Machado (ed.), Fr. Gaspar de S Bernardino, Itinerário da India por terra até à Ilha de Chipre (Lisbon, 1953), p. 65; S. A. Kasravi, Tarikh-i panjsad sal-i Huzistan (Teheran, 1312/1343), p. 85.

10.A. Grohmann, Südarabien als Wirtschaftgebiet, 2 vol. (Vienna and Brün, 1922-1933); A. Brouwer of Amsterdam University has a thesis on the Yemeni economic history in preparation, which will be the main modem authority on this topic; much material on al-Hasha is still untapped in the Istanbul Archives, though some sixteenth-century sources have been published in O Lüfti Barkan, XV ve XVI asirlarda Osmanli imparatorlugunda ziraî ekonomikin hukuki ve malî esaslari vol. I, Kanunlar (Istanbul, 1943), but cf. J. Mandaville, "The Ottoman province of al-Hasha in the sixteenth and seventeenth century", Journal of the American Oriental Society 90 (3) (1970).

11. On the sanitary problems in Goa cf. for instance Desembargadores da misericórdia to crown, 23 December 1703, Historical Archives of Goa (henceforward HAG), Livros das monções 67, f. 136.

12. On Socotra, cf. T. Bent, Southern Arabia (London, 1900); Frei Paulo da Trindade, Conquista espiritual do Oriente 3 vol. (Lisbon, 1967-1969), iii,. 20-23; ARA, VOC Archives (henceforward VOC) 113, Summary and trustworthy report relative to cruising for the Portuguese carraks from Lisbon, 27 November 1633, f.1132, Report of Pieter van den Broecke f. 1311; VOC 1259, Description of a voyage along the shores of Arabia.

13. On the slave-trade from India to the Middle East, cf. VOC 1234, Report by Pieter van Sandvliet regarding the present state of Vengurla f. 1657.

14. R. A. de Bulhão Pato (gen. ed.), Documentos remettidos da India ou livros das monções, vol. I (Lisbon, 1880), p. 54; P. P. S. Pissurlencar, Assentos do Conselho de Estado, 5 vol. (Goa/Bastorá, 1953), vol. i, p. 234; J. R. Quadros, Diu. Apontamentos para a sua história e chorographia (Nova Goa, 1899) and on Qeshm, cf. for instance A. Sadid as Saltanija, Bandar Abbas va Khalig-i Fars (Teheran, 1334), pp. 124-132, and Biblioteca Nacional de Lisboa, Fundo Geral, codex 580, description of Quesm, 1614.

15. VOC 1379, Shiraz to Batavia, 2 October 1681, f. 2622; P. Baratzes, "Tractato do sitio de Adem", in G. Beccari (ed.), Rerum Aethiopicarum scriptores occidentales inediti ad saeculo XVI-XVII (Rome, 1907-1913), iv, p. 336.

16. VOC 882, Uitgaand briefboek Batavia, (Batavia out-letter book) A, Warrant for Adriaan Roothaas; rear-admiral of the fleets bound to the coast of India, 6 August 1658 f. 348; S. B. Miles, The countries and tribes of the Persian Gulf (London, 1908/1966), Gazetteer of Bombay presidency, 23 vol. (Bombay, 1876-1896), vol. xiv, Thana, pp. 236-256; Pissurlencar, Assentos do Conselho, i, 165; V. L. Grotanelli, Pescatori dell'Oceano Indico (Rome, 1955); A. A. Hamid-Hiyed, Relations between the Yemen and South Arabia during the Zeydi imamate of Adb-al Qasim (Unpublished Ph. D. degree thesis, University of Edinburgh, 1973), pp. 118-121.

17. A. Bocarro, Livro das plantas de todas as fortelezas, cidades e povações do Estado da Índia Oriental (Bastorá, 1936), vol. i, pp. 60-61.

18. P. P. S. Pissurlencar, Goa pré-Portuguesa através dos escritores Lusitanos do século XVI-XVII (Bastorá, 1936), there are several good studies on the history of sawantwadi in Marathi such as P. Pingulkar, Savantvadi samsthñca itihas (Bombay, 1911), which are, however, blissfully ignorant of Portuguese sources, J. de Barros, Décadas da Asia, second decade, book V, chapter ii; the Konkani ports were generally given out in leaseholds by the Bahmanis cf. Sa'id Ali Tabata'bai, Burhan ima'athir (Hyderabad, 1300), p. 137

19. M. Longworth Dames (ed.), The book of Duarte Barbosa, 2 vol. (London, 1918, 1921), i, p. 193; S. Subramanyam, "The Portuguese, the port of Basrur and the rice-trade", in, IESHR 21 (4), 1985, pp. 473-510.

20. On Hormuz, cf. in particular W. F. Sinclair (ed.), The journey of Pedro Teixeira from India to Italy by land with his chronicles of the kings of Hormuz (London, 1901); on the pre-Portuguese period one might refer to Fiorani Piacenti, L'emporio ed il regno di Hormuz (VII fino XVI d. chr.) Vicende storiche, problemi ed aspetti di un civiltà costeira del Golfo Persico (Milan, 1975), there are, as yet, no thorough studies on the Portuguese period, but cf. A. Faroughy, Histoire du royaume de Hormuz (Brussels 1949), or more in general on the European 'interest' in the Persian Gulf, M. N. Tumanovic, Evropejkie derzavi v Persidskom Zalive v 16-19 vc. (Moscow, 1982); on Zeila's trade cf. M. Abir, "Southern Ethiopia", in J. Gray and D. Birmingham (ed.), Pre-colonial African trade. Essays on trade in central and Eastern Africa before 1900 (London, 1970), pp. 119-137; for Mogadishu the reader is referred to the classic work of E. Cerrulli, Somalia Scritti varii e inediti, 2 vol. (Rome, 1957).

21. VOC 1688, Mukka daily register, 19 August 1687, f. 534.

22. VOC 1114, Daily notices regarding occurences in the offensive fleet f. 142, 18 January 1642, on the costs of fitting out ships in the port of Kung cf. the important document, "Lista dos generos do condestavel da não Nossa Senhora da estrela que fica neste porto de Congo", in HAG, Monções 74, f. 315-325. I hope to return to prices in Persia elsewhere.

23. The best source on wages in Kerala in the seventeenth century are the accounts department (with statement of allowances in food) of ships built by the VOC in Cochin, as regards famine in the Mughal empire, cf. W. H. Moreland, From Akbar to Aurengzeb A study in Indian economic history (London, 1923), p. 153; G. Schurhammer, Die Zeitgenôssische Quellen Portugiesisch Asien und seine Nachbarstaaten zur zeit des hl. Franz Xaviers (1538-1552) (Rome, 1962/Freiburg, 1973), p. 64.

24. VOC 1425, Surat to Batavia, 6 May 1686 f. 235 verso, nor were the Dutch the only ones to remark on this, cf. A. Martineau (ed.), Mémoires de François Martin, 3 vol. (Paris, 1931-1934), ii, p. 420, so Batavia's grumbles that the VOC at Surat seemed but to write about disasters to cover up its dismal failure in the proper conduct of trade, are only partly warranted.

25. ARA, Collection Hoge Regering Batavia (High government of the Indies at Batavia) 877, thorough description of Persia by D. van Rheede, 18 November 1756, chapter 11; on epidemics in the Ottoman empire, cf. for instance D. Panzac, "La peste au Smyrne au XVIII^e siècle" Annales ESC 28 (1973), pp. 1071-1091, but we are still awaiting a larger work on the topic.

26. On the mortality of Portuguese settlers in the Zambesi-valley cf. for instance HAG Monções 44, statistics entitled "Os casados que vierão de Moçambique pello Sennd" and "o nome dos homens brancos neste governo de Senna"; F. Hoppe, Portugiesisch Ost-Afrika zur zeit des Marquês de Pombal (Berlin, 1965) pp. 70-75; C. R. Boxer, "Mozambique island and the carreira da Índia", in Océan Indien et méditterranée (Lisbon, 1962), p. 95-133; G. MacTheal, Records of South-Eastern Africa (London, 1898-1903), 9 vol., iii, p. 462.

27. G. Schurhammer, Franz Zaver Sein Leben und seine Zeit (Freiburg, i. b. 1955-1973), 2 parts in 4 vol., ii, p. 356-360; A. Brásio, As missões Portuguesas de Socotorá (Lisbon, 1943); J. P. da Costa, "Socotorá e o domínio Português no Oriente" Anais da Academia Portuguesa da história, second series, vol. XXIV, (II) (Coimbra, 1977), p. 325.

28. ARA, Archief Staten-Generaal, 5740, despatches relative to the VOC, Report by Ysbrand Goske; VOC 1611, "Piece report delivered to the Rt. Honorable Willem van Outshoorn, Governor-general, relative to the main things having occurred in Surat during my stay there" by Pieter Ketting f. 292; ARA, Hoge Regering Batavia 838, "Memoir for the honourable Johannes Peacock by Jan Schreuder", f. 52-61.

29. M. A. L. Cruz, "Exiles and renegades in early sixteenth century Portuguese India", The Indian economic and social history review 23 (3) (1986), pp. 249-262, cf. also G. Winius, "The shadow empire of Goa in the Bay of Bengal" Itinerario 7 (2) (1983), pp. 83-101 and my Koningen, Compagnieen en Kapers De Arabische zeeën 1640-1700 (Ph. D. Thesis Leiden, 1991), ch. I; on the 'Ethiopian' slave trade compare in particular J. B. Kelly, Britain and the Persian Gulf 1795-1880 (Oxford, 1968), pp. 410-419.

30. M. N. Pearson, "The subject", in M. N. Pearson and A. Das Gupta (ed.), India and the Indian Ocean (Calcutta, 1987), p. 18.

31. F. Cl. Lane, Venice: a maritime republic (Maryland, 1967), p. 166.

32. Most known sources about the Ethiopian role in India have been listed by R. Pankhurst, An introduction to the economic history of Ethiopia (London, 1956), appendix; a reference to kafirs being brought as a present by the Russian ambassador to the Shah in 1647 will be found in VOC 1162, daily register of Nicolaas Verburg, 4 April 1647.

33. Take, for instance, S. Subramanyam, "The Coromandel-Malacca trade in the 16th century: A study of its evolving structure" Moyen Orient et Océan Indien 3 (1986), pp. 55-80; the same, "Comércio e comerciantes na costa de Coromandel no século XVI" Revista de história económica e social; J. Flores, "The Straits of Ceylon and the maritime trade in early sixteenth century India" (Paper submitted to the XIth European Conference on modem South Asian Studies, July 1990 Amsterdam).

34. A few documents on the Agra community may be found in Archiefbewaarplaats Jezuїten in Nederland (depository of Jesuit archives in the Netherlands) Nijmegen cf on the Agra community also Documentaзгo ultramarina Portuguesa vol iv (transcripts from British Museum) (Lisbon, 1972) and in particular A. B. Amвnзio Graзias, Uma dona portuguesa na corte do Grгo Mogor (Nova Goa, 1903); on Malawi and the Zambesi valley, cf. M. Newitt, Portuguese settlement on the Zambesi (London, 1973) and J. Isaacman, Mozambique. The Africanisa-tion of a European institution. The Zambesi Praзos 1750-1902 (Madison, 1972); on Golkunda, cf. T. Bowrey, A geographical account of the countries round the Bay of Bengal (London 1905), p. 111. Already during the conquest of Goa by Albuquerque Portuguese free-booters were present among the Adilshahi host A. Baiгo (ed.), Comentбrios do grande Afonso de Albuquerque (Coimbra, 1921-1923), 2. vol., ii p. 15.

35. J. C. Wilkinson, The imamate tradition of Oman (Cambridge, 1987).

36. A. and J. Cortesгo (ed.), The suma oriental of Tomй Pires (London, 1944), p. 31; P. P. S. Pissurlencar, Regimentos das fortalezas da Нndia, (Bastorб, 1956), p. 178, ARA Collection Sweers/Manis 5: route.

37. cf. for instance R. L. Pouwels, "Islam and islamic leadership in the coastal communities of eastern Africa" (Unpublished Ph. D. Thesis, UCLA, 1979).

38. S. Dale, Islamic society on the South Asian frontier. The Mappillas of Malabar 1498-1932 (Oxford, 1980); G. Bouchon, "Les musulmans du Kerala а l'йpoque de la dйcouverte Portugais" in Mare Luso-Indicum ii (1974), pp. 3-59; S. Bayly, Saints, goddesses and kings. Muslims and Christians in South Indian society (Cambridge, 1989), p. 189 and on the tribal fishermen on the West coast of India, cf. V. B. Punekar, The son kolis of Bombay (Bombay, 1959).

39. M. N. Pearson, "Goa during the first century of Portuguese rule", Itinerario 8 (1) (1984), p. 39, cf. H. K. Naqvi, Urbanisation and urban centres under the great Mughal 1556-1706 An essay in interpretation (Simla, 1972).

40. T. R. de Souza, Medieval Goa (New Delhi, 1976), p. 259; S. J. Shaw. The financial and administrative organisation of Ottoman Egypt 1517-1598 (Princeton, 1962), pp. 261-263; the Portuguese at Hormuz often discussed whether to exchange this Ottoman grain with spices, Schurhammer, Zeitgenцssischev Quellen, p. 334; J. H. de Cunha Rivara, Archivo Portuguкs Oriental (Nova Goa, 1857-1867), 9 vol., ii, p. 109; J. Wicki (ed.), "Duas relaзхes sobre a situaзгo da Нndia portuguesa nos anos 1568 e 1569", Studia 8(1961), p. 153; J. Aubin (ed.), "Titulo das rendas que rende a villa de Ormuz", Mare Luso-Indicum II, p.220.

41. C. R. Boxer, The Portuguese seaborne empire (London, 1969), p. 234; C. Orhonlu, Osmanli imparatorlugunda gьney siyaseti. Habes eylatin (Ankara, 1974), p. 130; VOC 1243, Persia to Batavia, 31 May 1664, f. 1929 verso.

42. J. Wicki (ed.), O livro do pai dos cristгos (Lisbon, 1969), pp. 216-218; Governadores to king, 16 March, HAG, Monзхes 36; K. N. Pawar (ed.), Tarabai Kalina kagadpatre, (Kolhapur, 1967-1973), 3 vol. iii. doc. 97, T. R. de Souza, Medieval Goa, p. 65 ff.

43.VOC 1179, Description of Basra f. 983.

44. HAG 8784, livro azul, Alvarб, 23 April 1634; A. Baiгo, A inquisiзгo de Goa, 2 vol. (Lisbon, 1930-1945), ii. p. 513; ARA Collection Geleynssen de Jong, Goa-daily register 1 November 1642.

45. M. N. Pearson, "The crowd in Portuguese India" in Coastal western India Studies from the Portuguese record (New Delhi, 1985).

46. C. H. Alexandrowicz, An introduction to the history of the law of nations in the East Indies (Oxford, 1967).

47. Compare for a reference to Surat around 1580 as being the main port of Gujarat Couto, Da Бsia, dйc. IV, b. 6 ch. XI; on Cambay, cf. A. Almeida Calado (ed.), Livro que trata das cousas da India e do Japгo (Coimbra, 1957), p. 57-60.

48. J. Lockyer, An account of trade in India (London, 1711), p. 123; G. La Roque, Voyage а l'Arabie heureuse (Amsterdam, 1711) p. 95; Beccari, Rerum Aethiopicarum x, p. 320 ff.

49. C. Lйvi-Strauss, Het triest der tropen. Reisverslag van een antropoloog (Utrecht, 1962), p. 84.

50. ARA Collection Geleynssen Journal or daily register kept by Claes Jansen aboard the "Zeemeeuw", 28 October 1645; on the previous importance of Jask, cf. Schurhammer, Die zeitgenцssische Quellen, p. 88.

51. On Kung, cf. J. Aubin, L'ambassade de Gregуrio Pereira а la court du chah Soltan Husain (Lisbon 1972), p. 10, and N. Steensgaard, The Asian trade revolution of the seventeenth century (Chicago, 1974), p. 358.

52. J. Baldry, "The early history of the Jemeni-port of al-Hudeida" Journal of Arabian Studies 7 (1885), p. 35-53; F. Macro, Bibliography of the Jemen and notes on Mocha (Coral Gables, 1960).

53. S. C. Misra (ed.), The mirat-i sikandiri (Baroda, 1961), p. 141, 162, J. Aubin, "Albuquerque et les nйgotations du Cambaie" in Mare Luso-Indicum I (1973), p. 5-95; the Portuguese relations with Gujarat have (deservedly) been well researched cf. M. N. Pearson, Merchants and rulers in Gujarat (Berkeley, 1976) and K. N. Mathews, Portuguese and the kingdom of Gujarat (New Delhi, 1986).

54. For descriptions of Zeila, cf. E. Celabi, Seyahatnamesi, 6 vol. (Istanbul, 1975), vi, p. 674-678, for Mogadishu apart from Cerulli, op. cit. and I. M. Lewis, The modern history of Somaliland (London, 1965) see also the rather peculiar text: "A description of Mocadixu" included in D. Defoe alias Captain Johnson, A general history of the robberies and murders of the most notorious pyrates of the English nation (M. Schlumhorn ed.) (Columbia, 1972), which, though apparently based largely upon hearsay, is one of the few descriptions of Mogadishu extant from the seventeenth century and has escaped the attention of scholars engaged in Somali studies.

55. J. Kirkman, Men and monuments on the East African-coast (London, 1964); the same and N. Chittik, Kilwa an Islamic tradнng city on the East African coast, 2 vol. (Nairobi, 1970); F. Cooper, Plantation slavery on the East Coast of Africa (New Haven, 1977).

56. cf. N. R. Bennett, "The Arab impact" in, B. A. Ogot (ed.) Zamani (Nairobi, 1974), pp. 210-228 and L. Harries, "The Arabs and Swahili culture" Africa 34 (3), pp,. 224-239.

57. This interpretation of the sixteenth century leans, as regards to the Ottoman aspect, heavily upon W. E. D. Allen, Problems of Turkish power in the sixteenth century (London, 1963), and the various work of J. Baquй-Gramont cf. also Sцzbaran, "Osmanli imparatorlogunda ve Hindustan yolu", Tьrk tarih dergisi (1977), p. 40-65; for the tribute arrangement prevailing at Hormuz, compare A. Ramos Coelho (ed.), Alguns documentos do arquivo nacional da Torre do Tombo acerca das navegaзхes e conquistas dos Portugueses (Lisbon, 1892), pp. 477-479.

58. J. Kirkman. "The Muzungulos of Mombasa" International journal of African historical studies 16 (1) (1933), p. 71; N. J. Hafkin, "Society, trade and politics in Northern Mozambique 1750-1902" (Unpublished Ph. D. Thesis, 1973), p. 22; compare too E. A. Alper, Ivory and slaves in East-central Africa (London, 1975).

59. A. M. Abdu Hakima, History of Eastern Arabia (Beirut, 1967); A. A. Amini, British interest in the Persian Gulf (Leiden, 1967); A. Das Gupta, Malabar in Asian trade (Cambridge, 1968).

60. cf. for instance F. Perlin, "Proto-industralisation and pre-colonial South Asia" Past and present 98 (1983), pp. 30-95.

61. P. T. Arutunian, Osvobodsdenije dvizenije Armjanskogo naroda v pervoi cetverti XVIII-veka (Moscow, 1954), p. 26.

62. J. C. Wilkinson, Water and tribal settlement in south-east Arabia a study of the aflaj of Oman (London, 1977); as regards Persian designs against Muscat, cf. L. Lockhart, Nadir Shah A critical study mainly based upon contemporary sources (London, 1938), p. 212-222, for the eighteenth century and my forthcoming study Silk, zambuccas and Silver. The Dutch and the Persian Gulf 1623-1703 for the seventeenth century.

63. R. Savory (transl.), Iskandar Beg. The history of shah 'Abbas (Boulder, Colorado, 1978), 2 vol., ii, p. 1202; J. Aubin, "Rйferences pour Lar mйdiйval" Journal Asiatique 1955, pp. 491-505.

64. On this point, cf. in particular Barros, Da Asia, dйcada II, livro X, chapter vii; Steensgaard, The Asian trade revolution, p. 93; B. W. Diffie and G. Winius, Foundations of the Portuguese empire (Minnesota, 1977), p. 333, and for sources, compare for Bandar Abbas for instance ARA Collection Aanwinsten (newly acquired documents) 1903 xix, schedule of customs levied at Bandar Abbas on goods received, or for Goa the papers of Afonso Mexia in J. H. da Cunha Rivara, Arquivo Portuguкs Oriental, V. doc. 5 and 53.

65. A. Das Gupta, Indian merchants and the decline of Surat (Wiesbaden 1976), pp. 9-26; J. Ovington, A voyage to Surat in the year 1689, p. 106; VOC 1345, Surat to Gentleman XVII 4 April 1672 f. 2354; ARA Hoppe Collection, considerations by Jan Schreuder chapter 2 & 96.

66. VOC 1122, Report by chief-merchant Jan van Twist fo. 478; on the port of Dabhol, cf. in particular (the rivalry between the Bijapuri port of Dabhol and Portuguese/Malikshahi Chaul would well repay research) A. Das Gupta, "Indian merchants and the western Indian Ocean, the early seventeenth century", Modern Asian Studies 19 (3) (1985), p. 484; VOC 1147, Van Twist to Governor-general, 16 March 1637 f. 156.

67. On Sahibi Begam's doings in Surat compare in particular ARA Collection aanwinsten, 1936 vii letters between Delhi court and Surat in the years 1653 up to 1656; H. van Santen (De Verenigde Oost-Indische Compagnie in Gujarat en Hindustan 1620-1660, Ph. D. Thesis, Leiden, 1982, p. 74-78), has not paid sufficient attention to Sahibi Begam's involvement in the 'royal shipping' carried out from Surat, which causes him to ascribe to Shah Jahan a mercantilist policy, whereas such was really only a private investment of Sahibi Begam. On Hodsia Athar, cf. for instance L. de Albuquerque (ed.), Crуnica do descobrimento e primeiras conquistas da Нndia pelos Portugueses (Lisbon, 1986), p. 331.

68. See for Surat, ARA Hoge Regering 838, Memoir Schreuder f. 62; J. C. Heesterman's "Littoral et intйrieur de l'Inde", in L. Blussй and H. Wesseling (ed.), History and underdevelopment (Leiden, 1980), p. 904, Heesterman's suggestion that 'continental' custom-dues in India were higher than coastal ones, is however by-and-large untenable.

69. VOC 1529, Surat daily register, April 1683.

70. For a description of Kilwa, cf. for instance Barros, Da Asia, dйc. i, book 8, ch. iv, probably the best evocation of such a closely knit city state is Van Leur's description of Mukalla in his Eenige opmerkingen betreffende de oude Aziatischen handel (Middelburg, 1936), p. 90, which is, however, linked to a very different unit by him: to wit Mukka.

71. Although this is mainly based upon the (original) Surat factory records from the late eighteenth century preserved in ARA, Collection voormalige Nederlandse bezittingen in Voor-Indiл (Former Dutch possesions in India), or ARA Hoge regering 842, Memoir by Johan Pecock to his successor, 16 December 1753; cf. also M. Torri, "In the deep blue sea, Surat and its merchant-class during the dyarchic era", Indian economic and social history review, vol. 19 (series 3) (1982), pp. 267-299.

* Rene Jan Barendse was appointed researcher in the Kern Institute, Leiden University and the Centre for Non-Western Studies at Leiden University in 1961 and has also taught Indian and Economic History at Leiden University and the Centre for Asia Studies, Amsterdam. He is the author of numerous books and articles connected principally to his fields of interest: medieval and modern Indian history and the history of the Dutch and Portuguese expansion and is a member of the executive committee of the foundation of South-Asian Research of the Dutch Academy of Science. Barendse has travelled on research trips to Portugal, India and within the Netherlands with grants from the Leiden University Research Foundation.