Trading partners across the Indian Ocean: the making of maritime communities

HIMANSHU PRABHA RAY

The conventional framework of ancient Indian history emphasizes that trading activity proliferated during certain historical periods, and declined thereafter.

Implicit in the conventional pattern has been the assumption that maritime trade was dependent on demand from states and empires and was hence sporadic. In terms of the Indian Ocean trading network, the traditional view holds that local trading communities of the Indian subcontinent played a minimal role, especially in the western Indian Ocean, since it was the foreign - European and Arab - demand for luxuries that triggered it.

This view is not supported by the nature of commodities involved in the Indian Ocean network, however, which ranged from salt and cowries at one end of the scale to metals, medicinal plants, aromatics, agricultural produce, costly textiles, silverware and dancing girls, at the other. Such commodities appear in early texts. The Arthasdstra, a Sanskrit treatise on political economy¹ dated between the third century bce and third century ce provides an interesting array of possibilities for the acquisition of commodities. For example horses could be obtained as gifts, purchased, obtained in war, bred in stables, received in return for help or temporarily borrowed (ii.30.1).^{^[279]} ^{^[280]} Other early texts also mention trade. There are references both in early Buddhist literature and in inscriptions to the srenι or trade associations; the number eighteen is often mentioned, though nowhere are these eighteen enumerated. In the Pali canonical Jataka literature, there are frequent references to the chief of the guilds in the retinue of the king (Book ii: no. 154) and he is generally revered as a rich and powerful figure. These suggest that local demand played a more important role than the conventional narrative suggests.

Andre Wink has also stressed continuities, arguing against the decline of Roman trade in the Indian Ocean region in the early centuries of the Common Era because Greek-Byzantine traders again became active in the India trade from the fourth to the sixth centuries.^{^[281]} In the fifth-sixth centuries Persian commerce synchronized with the ascendancy of the Sasanian Empire, and with the coming of Islam there was an increase in trading networks in the Indian Ocean. Wink suggested that control of trade was a motivating factor in the Arab conquest of Makran, Sindh, Kathiawar and Kutch and that the tenth century marked the emergence of an integrated Muslim trading empire. In emphasizing the role of Islam in promoting trade, Wink points to the disappearance of Buddhism from India, and contends that, due to restrictions stipulated in the Law Books or Dharmasatras on maritime travel, the Hindu population turned to ‘agrarian pursuits and production, away from trade and maritime transport'.^{^[282]}

Wink's emphasis on continuity of trading activity in the western Indian Ocean is well taken, but there was also change, with shifts in coastal centres and the emergence of new settlements. For example, Bharuch, Sopara and Kalyan on the western Indian coast were important outlets for trade in the early centuries of the Common Era, but gave way to Valabhi in the Gulf of Cambay around the middle of the first millennium. Subsequently, Arab attacks on Valabhi around the latter half of the eighth century led to its abandonment.^{^[283]} There was the consequent rise of Stambhatirtha, or Cambay, at the head of the Gulf of Cambay. Another important coastal settlement was that of Somnath (modern Prabhas Patan) on the Gujarat coast, which continued to play an active role in maritime trade (see Map ιι.ι).

Trade was also shaped by the presence of Hindu, Buddhist and Jain religious places. Somnath, for example, was the location of an important Hindu temple and site of pilgrimage.

At Somnath, archaeological evidence of historical settlement dates to the fourth century bce, but religious structures, such as temples, emerge only in the fifth-sixth centuries ce. After the tenth century, shops and markets were located in the vicinity of temples and

Map ιι.i. Gujarat

contributed materials to the performance of rituals and festivities. For example, a thirteenth-century record, inscribed on a long slab of polished black stone from one of the temples at Somnath, refers to the purchase of shops by the benefactor and their donation to the temple.

Kanheri near the present city of Mumbai was the largest Buddhist monastic complex along the west coast of India dating from the first to the eleventh century. Merchants and traders were the major patrons of the monastic site. A six-line inscription in cave ιι, dated to 12 September 854, records the visit of a devout worshipper of the Buddha from Gauda or Bengal on the east coast and a permanent endowment made by him of one hundred drammas or silver coins for the construction of a meditation room and clothing for monks residing at the monastic complex. This grant also draws attention to the interconnectedness of the two seas, i.e. the Arabian Sea and the Bay of Bengal.

In addition to the varied religious landscape, maritime travel in the medieval period involved a changing array of watercraft. Cargo carriers formed the foundation of trading ventures. Merchants and traders in some cases certainly owned ships and watercraft, but they neither manned nor sailed these. A thirteenth-century inscription from western India (the Ana- vada inscription of 1291) refers to gifts to a temple by shipowners. More often, goods and cargoes were entrusted to the captain of the vessel, who was then responsible for their sale and profit. Thus maritime activity involved diverse groups, from the owners of watercraft to those who commanded them and still others who sailed them.

Hence there can be no simple caste attributions of the communities involved in trading activity. Moreover, the normative rules laid down in the Law Books need to be balanced against the narrative literature in Sanskrit, which provides glowing accounts of maritime travel, by merchants, craftspeople, musicians and others.

Trade also formed an important source of revenue for the state, which continued to be monarchical. From contemporary records in western India, it is evident that tax was collected for use of roads, ferry-crossings and harbour facilities and differed from tax charged on the sale and purchase of commodities in the marketplace. Inscriptions from the fifth century onwards refer to several officials engaged in collecting revenues from trading caravans. The city formed the nucleus of settlement and it is at the city gates that goods were taxed and the king obtained revenue in return for providing protection to local trading communities.

In this chapter I focus on four issues: the communities who participated in the trading networks; the impact of maritime trade on a littoral region, such as that of Gujarat; the organization of trans-oceanic trade in the western Indian Ocean; and finally the interconnectedness of the two sectors of the Indian Ocean, viz. the Arabian Sea and the Bay of Bengal. For each of these issues, I highlight particular sources or examples that provide special insight into maritime trade and its organization in a world historical context.

Local trading communities

Recent research suggests that fishing and sailing communities formed a distinct group and were the crucial component of all sea travel in the Indian Ocean.^{^[284]} Fishing was the traditional occupation of coastal groups in several pockets from at least the fifth millennium bce.^{^[285]} These communities adopted numerous occupations associated with the sea: fishing and harvesting other marine resources, salt making, sailing, trade, shipbuilding and piracy.

Fishing and sailing communities formed the foundation of maritime activity in the Indian Ocean and provided a continuum to seafaring throughout history, though no doubt their fortunes fluctuated over time. A common feature of the sewn boats of the Indian Ocean, for example, was the use of coir-rope for stitching. Because coconut palm plantations were restricted to certain parts of the Indian Ocean littoral, coir-rope would have been one of the commodities in demand in the boatbuilding settlements of the coast.^{^[286]} Thus, the building of the dhow, or traditional watercraft, involved trade and transportation of wood for planking and coconut coir for stitching from different regions of the Indian Ocean, thereby creating networks of interaction and sustaining vibrant exchange across the seas.

The first-century ce Periplus Maris Erythraei written by an anonymous sailor in koine Greek^{^[287]} is unique in that it provides the first detailed description of local boats in the Indian Ocean in the early centuries of the Common Era. It refers to several types that were used from the East African coast to the west coast of India. In the region of the Barbaroi is a small port of trade called ‘Avalites, where rafts and small craft put in' (section 7). The island of Menuthias, identified with the island of Pemba on the coast of East Africa, has sewn boats and dugout canoes that are used for fishing and catching turtles (section 15). The name of the southernmost of the towns of Azania, Rhapta, is said to derive from the name of the sewn boats, or rhapton ploiarion, found there (section i6). Muza is described as a port teeming with Arabs - shipowners or charterers and sailors who trade across the water and with Barygaza on the west coast of India, using their own outfits (section 21).

Another invaluable literary source for an understanding of the Indian Ocean network is the Christian Topography written in the sixth century by an Egyptian monk, Cosmas Indicopleustes ('India-voyager^,).¹° Cosmas was in all probability of Greek parentage and a native of Alexandria.

In the early part of his life, he was a merchant and had travelled widely in the western Indian Ocean. Later he retired from secular life and moved to the cloister, where he devoted himself to the composition of works on geography, cosmography and Scriptural exegesis.

The Christian Topography (Book xι: 367-8) refers to a series of coastal centres on the west coast of India, many of which overlap with those mentioned in the Periplus. The Arthasastra, for example, makes a distinction between local trade transacted in fortified cities of the interior, identified as nagara or city, and that originating at distant places and exchanged at the pattana, located either on sea coasts or on river banks of the interior. The two had a different organization of trade and different administrative structures, with itinerant merchants' organizations being active only at coastal centres. The use of the term patana or pattana is significant - it is a term that continued to be used in the India letters of the Cairo Geniza dated to the eleventh-twelfth century, which we will discuss later. Thus clearly pattana was a market centre with a distinctive administrative structure for the collection of levies that continued from the early centuries ce into the second millennium.

How did the coming of Islam redefine maritime space? Dionisius Agius has shown that with the spread of Islam, Arabic was enriched by borrowings from Aramaic, Persian, Greek, Sanskrit and other Indian languages. He also cautions against equating religious with ethnic identity, as Arabic sources do not make this distinction and thus label as Muslim all non-Arab foreigners who converted to Islam irrespective of their ethnic background.¹¹ There are nevertheless exceptions. For example, the historian and geographer al-Mas‘udi (d. 956-7) states that the Sirafis and Omanis were the leading seafarers of the time, thereby highlighting regional coastal identities. ^[288] ^{^[289]}

The Annals of the Tang period (618-907) in China, on the other hand, do not refer to religious affiliations, but make a distinction between Persian and Arab traders, the former termed Po-ssu and the latter Ta-chi or To-che. In contrast, Muslim travellers such as Ibn Jubayr (d. 1217) and Ibn Battuta (d. 1368-9 or 1377) apply the term ‘Muslim' in a religious context to distinguish some communities from others that were not Islamic, such as Christians, Jews, Hindus and Zoroastrians. This is an important issue, which has implications not only for understanding the terms used to label medieval trade and the religious or ethnic identities attributed to the seafarers, but more significantly for comprehending emerging religious identities in the Indian Ocean region and the ways in which these were formulated and described.

From the eighth-ninth century onwards, maritime space expanded to include East Asia, and Chinese wares were distributed through the Indian Ocean networks.^{^[290]} ^{^[291] ^[292] ^[293]} Excavations at Fustat (Old Cairo) have yielded Chinese ceramics dating from the ninth through the fifteenth century, with the largest concentration from the tenth to the fourteenth century. 1³ These need to be used with caution for demarcating the extent and nature of the network, however. For example, the excavations at Siraf in the Persian Gulf yielded a Chinese stoneware fragment bearing two Arabic names - Yusuf and Mansur, or Maymun - incised before glazing the vessel, belonging to a jar probably sent by a merchant resident in China. 1⁴ But who was the owner of this vessel and for what purpose was it used? More broadly: Can the finds of Chinese ceramics, such as Changsha or Yuezhou Celadon Wares, be used to define the nationality of watercraft used in the transportation of the pottery, such as whether these were Chinese ships or Arab ships? Can they be used to determine the direction of the contacts?

A search for Chinese ceramics along the Indian coasts has yielded mainly post-eleventh-century ceramics, a majority belonging to the thirteenth and fourteenth centuries along the Tamil and Malabar coasts.1⁵ The evidence from Siraf (period ιι), however, presents a different picture. At the beginning of the ninth century the site reached its maximum extent, covering an area of over no hectares between the seashore and the foot of the mountains. This was also the period of intensive building activity coinciding with the construction of the great Friday mosque. Chinese imports, however, represent less than one per cent of the total amount of excavated ceramics from the site and only one-third of these are from luxury porcelain or Celadon pieces, while two-thirds come from stoneware jars. In contrast to the limited quantities dated to the ninth-tenth century, imports increased in the thirteenth-fourteenth century. In his 1969 survey in southern Iran, A. Williamson picked up about 10,000 Chinese sherds from coastal sites, but no more than ninety from inland centres.^{^[294]}

This uneven distribution of Chinese ceramics both temporally and spatially across the western Indian Ocean raises issues of peaks and lows in trading networks in specific regions. Second, the clustering of finds in coastal areas suggests the prevalence of contacts along the coast.

Gujarat: trade and religion in a littoral community

How did this oceanic trade impact centres further inland? The area of Gujarat in western India provides an excellent example of a littoral community where trade and religion combined, as evidenced by a variety of sources. Archaeological investigation in western India has provided information on several long-lasting coastal settlements. For example, after an archaeological exploration conducted in the mid-nineteenth century, the site of Mandvi on the estuary of the Rukmavati River at the entrance to the Gulf of Kutch has been known as a port town with links to both Oman and the East African coast. Dwarka, located on the north coast of Saurashtra, is another contemporary site with a long period of settlement from the first century bce to almost the present. Its importance stems both from its coastal location as well as the religious significance of its temples and sacred association as a centre of pilgrimage. Other continuously settled sites along the Gujarat coast include Porbandar, Somnath and Valabhi at the head of the Gulf of Cambay as mentioned earlier. Further south, Chaul - referred to as Campavati or Revatiksetra in the Epics - is mentioned in inscriptions and literary sources from the early centuries ce to the seventeenth century. In a copper plate record of 1094, three major coastal centres are identified: Chaul, Thane and Sopara, all in western India. Epigraphs dating from the tenth-eleventh

Trading partners across the Indian Ocean centuries refer to these coastal settlements as linked to each other, as well as to settlements in the interior.

In the fifth century ce, Valabhi, located on the Bhavnagar creek in the Gulf of Cambay, emerged not merely as an outlet for maritime trade, but also as the capital of the ruling dynasty of the Maitrakas (493-776) and the core area for religious consolidation.^{^[295]} Contemporary Sanskrit literature such as the Dasakumdracaritam by Dandin (fl. c. 600) describes Valabhi as a prosperous trading centre. One of the stories in the text is set in the city of Valabhi in Saurashtra where the chief of sea traders (ndvika-pati) lived. According to the text, he possessed immense wealth, almost like Kubera, the Lord of Wealth.^{^[296]} References to Valabhi's wealthy resident and travelling communities of traders (vaniggrdma) are found in the copper plate donations from Toramana's reign dating to the late fifth-early sixth centuries. These record gifts made by the trading community of Vadrapalli to the temple of Jayas- wami or Narayana belonging to the queen mother. Maharaja Bhuta and Maharaja Matrdas also made donations of certain villages to the temple. The main commodities referred to in the plates are molasses, salt, cotton and grain and a detailed list of revenues transferred is also itemized These revenues were calculated on the basis of vessel-load, donkey-load, and cartload of the produce. Vadrapalli was probably located 8 kilometres to the west of Sanjeli and signatories to the donation included traders from Ujjain, Kannauj, Mathura and perhaps Mandasor. A goldsmith constructed a lake near the temple.^{^[297]}

The self-assurance of the trading community is evident from the charter of Visnusena of 592, issued from Lohata in the Kathiawar region, which reiterates the customary practices followed by the group and the acceptance of these by the ruler. This charter is addressed to a list of officials as is usual in Maitraka records. The charter assures protection to the community of merchants (vaniggrdma) established in the region and endorses their continued functioning. It provides a detailed list of seventy-two trade regulations or customary laws to be followed. Some of the regulations are of great interest to this discussion, as they indicate not only an active coastal trade,

but also a lower rate of tax for commodities for religious purposes. For example, it is specified that merchants staying away for a year were not required to pay an entrance fee on their return. Other clauses specify duties that were to be paid. A boat full of containers was charged twelve silver coins, but if the containers were for religious purposes, they were charged only one and a quarter silver coins. In the case of a boat carrying paddy it was half this amount. Other items frequently transported by boat included dried ginger sticks, bamboo, wine, leather, buffaloes, camels and bulls.^{^[298]}

Indigo or nιla is mentioned as one of the items exported from areas such as Gujarat and the charter of Visnusena mentions a tax on the pressing of the indigo dye. Indigo also figures as one of the items of trade from India westward to Egypt handled by the business house of Ibn ‘Awkal between 980 and 1030.^{^[299]} One of the three varieties mentioned includes Sindani indigo - Sindan/Sandan being identified with the Konkan coast of the subcontinent.

The variety of taxable commodities mentioned in the charter of Visnusena is an indication of the diverse nature of trade in the region. These included oil mills, sugarcane, wine, cumin seed, black mustard and coriander. The charter also refers to a tax on dyers of cloth, weavers, shoemakers and retailers hawking goods on foot. Others such as blacksmiths, carpenters, barbers, and potters could be recruited for forced labour under the supervision of officers.^{^[300]}

This consolidation of economic activity along with expansion of political authority and increasing involvement of brahmanas and Buddhist monastic complexes marks a departure from the trading patterns of earlier periods. Clearly the Maitrakas were attempting to widen their resource base. The inscriptions provide valuable information about the commodities involved in coastal traffic, customary practices adopted by the trading communities, and the differential tax levied on goods required or donated to religious establishments.

It is significant that while archaeological data and distribution networks of ceramics indicate local, regional and oceanic interaction starting from the third-second millennium bce onwards, inscriptions from Gujarat are largely silent about trading activities until well into the fifth-sixth centuries ce. From

this period onwards inscriptions reflect an increasing complexity in commodities exchanged and in the nature of transactions conducted.

The presence of mosques in several parts of the Indian Ocean littoral and the active role of Muslim merchants and shipowners in the maritime network raise issues regarding cross-cultural interaction especially in Gujarat. On the coast of Kutch in Gujarat, Bhadresvara is an early site, which was also the site of a twelfth-century Jaina temple. The local Muslim community of the Isma‘ili sect is said to have built a mosque here with the permission of the Jaina Council, but the most notable structure is the shrine of Ibrahim dated on the basis of an in situ inscription to 1159-60, i.e. nearly half a century earlier than the Islamic monuments at Delhi or Ajmer.

In Bhadresvara, and at several other sites in Gujarat such as Cambay, Somnath and Patan, there are a number of Muslim tombstones dated between the mid-twelfth and early thirteenth centuries. Many of the architectural features at Bhadresvara are also found in a mosque built by early settlers at Junagarh, where an inscription records that the chief merchant and shipowner AbuTQasim b. ‘Ali al-Idhaji built the mosque in 685 AH (1286-7 ce).^{^[301]} Junagarh is an ancient site, as evident by the location of a Mauryan inscription of the fourth-third century bce, and below the mosque is a Buddhist cave now regarded as a Muslim shrine.

How did this changing religious landscape translate into interactions between communities? Arabic inscriptions and also bilingual epigraphs are important sources for insights into communication between communities after the advent of Islam. The partnership between the Nakhuda (master of the ship) and local communities is best exemplified by the Somanath-Veraval inscription of 1264, consisting of two slabs of stone inscribed with a Sanskrit epigraph and, two months later, with an Arabic record. Located at Somanatha-Pattana on the Gujarat coast, the inscription records endowments for the maintenance of a mosque (dharmasthana in Sanskrit) and for providing it other services by Nakhuda Nur al-Dawla wa-l-Din Firuz, son of Khoja Nakhuda Abu Ibrahim of Hurmuja-desa or Hormuz at the mouth of the Persian Gulf, together with the local community leaders (thakura in Sanskrit). At that time the king or raja of the coastline (velakula) of Hormuja was amir sri rukana dina, that is, Sultan Rukn al-Din. The Sanskrit version of the inscription provides a detailed description of the land acquired as a part of the endowment. This land was located in the vicinity of the city and at least four local residents are listed as providing the land for the endowment. In addition Nur al-DTn acquired and donated the products of an oil mill and two shops or marketplaces (hattas). The surplus, if any would emanate from these, was to be sent to the holy places of Mecca and Medina.^{^[302]} This transaction was witnessed by all the jamdthas or groups of Somnath who were also responsible for the upkeep of the property. The groups named are: Ndkhudd-ndvika; gamchikas or oil men along with their preacher; chunakdras or whitewashers; and also mushalamdnas or Muslims of the town.

It is important that the Arabic version of the inscription contains just twenty-eight lines as compared to the Sanskrit record of forty-three lines. While it agrees in several significant details with the Sanskrit version, it lists the imam and the mu'azzin of the mosque as beneficiaries. The Arabic version credits Abu IbrahTm ibn Muhammad al-‘ Iraqi, the father of Nur al-DTn Firuz, with the foundation of the mosque and one of the titles that it attaches to the latter is that of malik muluk al-tujjdr, or leader of business-leaders. The Arabic text does not contain details of the land transfer, nor does it mention that the surplus was to be sent to Mecca and Medina. It concludes by quoting from the Quran against tampering with the endowment. It has been suggested that while public inscriptions engraved in Sanskrit were a part of the administrative procedure of the subcontinent, this was not valid for the central Islamic lands, where paper documents were the norm and foundation inscriptions containing basic information about the founder were monumentalized in ornate epigraphs on the buildings. In keeping with the legal requirements of Islamic law the mosque was definable as a waqf (i.e. a charitable endowment), and trustees and beneficiaries had to be appointed. The Arabic version is thus crucial to an understanding of the establishment of the institution of waqf in western India. The Somanath-Veraval inscription indicates that there were ‘two separate registers encoding the interactions between Muslims and their non-Muslim business partners. One was the prescribed register, constituted by Islamic laws defining what was allowable and forbidden in the commercial dealings of its followers. The other was the practical register, evidenced by historical documents.'^{^[303]}

This overview of the early history of Gujarat has presented insights into dynamic interactions between merchants, shippers and, most importantly, religious institutions in the coastal regions of the Indian subcontinent. These littoral communities participated in the larger maritime network of the western Indian Ocean.

Trading centres on the Red Sea and trans-oceanic travel

Trading centres on the Red Sea were closely involved in trade with the west coast of India. Interesting evidence for this involvement is provided by more than five hundred paper fragments, written in the Arabic script, recovered during excavations at Quseir al-Qadim on the Red Sea coast about 500 kilometres south of Suez. This evidence is perhaps unique in the Indian Ocean world as, unlike the Cairo Geniza and other contemporary documents, it can be located within an archaeological context. Material remains from the excavations of the shaykh's house at Quseir al-Qadim help provide a cultural milieu to the texts. It is significant that there is no mention of Quseir in the Geniza documents, though they refer to ‘Aydhab and Qus frequently.^{^[304]} This divergence alerts us to the need to analyse a range of sources rather than a single data base in order to arrive at a holistic picture.

A survey of the fragments from the central area of the town confirms that the majority were addressed to or written on behalf of a certain Shaykh Abu Mufarrij and his son Shaykh Ibrahim, and to a lesser extent to other people associated with the family business in the first half of the thirteenth century. Inside the house, several hundred paper fragments, mainly personal and business correspondence written in Arabic, were found along with other small objects. These documents have been studied and translated by Li Guo and the present discussion is largely based on his work. The general pattern was for the letter to be written on the recto, while the verso contained the recipient's address.^{^[305]}

Shaykh Abu Mufarrij was the owner of a prosperous transit warehouse in the port of Quseir. His elder son Ibrahim worked as a merchant and broker rising to the rank of rayyis, or head of trade, and also a khatιb, or one who gave the sermon at the local mosque. Thus it is evident that the shaykh and his son were merchants who did not travel themselves, but sent shipments through others. The overlapping responsibilities of being a head of trade and also a khatιb in the local mosque underscores the close relationship between economic and religious activities. Pilgrims to Mecca via the Egyptian Red Sea southern route formed an important category of clients. One of the few complete letters requesting supplies written from a ship stuck in Yemen would suggest ‘some kind of regular transport of correspondence, supplies, merchandise, and even cash through a kind of maritime shuttle service between Quseir and Yemen'.^{^[306]} The coastal centre of al-Qasr al-Yamani, where the ship was stuck, cannot be identified, but it was most likely on the Yemeni side of the Red Sea and the ship was travelling along the Yemeni coast for Hijaz.

Did the warehouse of Shaykh Abu Mufarrij also function as a caravanserai where people stayed overnight? Given the multiple capacities in which the shaykh and his son worked, ranging from merchant, agent and manager to tax farmer, head of trade and market inspector, in addition to the social services the warehouse provided, such as funeral arrangements, medical care and magic practices, the answer should perhaps be positive. It appears also that the shaykh's warehouse functioned as a government agency in charge of grain distribution on the Red Sea and Indian Ocean trade routes, especially for the pilgrims to Mecca.

The Arabic documents from Quseir supplement not only the archaeological data, but also those available in the Geniza papers. These papers, Judeo-Arabic documents preserved in the Cairo Geniza written in Arabic in Hebrew characters and dating to the eleventh and twelfth centuries, also provide insights into the working of trade in the Red Sea and beyond. The 459 documents, comprising 523 shelf-marks related to the India trade, constitute a significant component of the Geniza records. Yemen on the south Arabian coast formed the fulcrum of a network that extended from centres on the Red Sea to the west coast of India, generally designated by the term diyar al-Yaman wal-Hind, or the lands of Yemen and Hind.^{^[307]} Documents in the Cairo Geniza may be categorized into letters, documents of legal character, and memos accompanying shipments, such as accounts and wage payments of a brass factory in western India.

The second category of documents or deeds of ‘partnership according to Muslim law' drawn up before a merchant embarked on a journey are of particular relevance to the present discussion. Invariably a merchant did business not only for himself, but also for others, or acted as an agent for one or several investors. On his return, the merchant was required to make a statement to the local rabbinical or Muslim court and the partners would write out a release indicating that the deed had been concluded to their satisfaction. It should be stressed that whenever possible, merchants preferred sending goods rather than gold, and silk is at times mentioned as a means of payment.

Most large-scale transactions were carried out through partnerships, and merchants preferred to spread out their funds among several partnerships to reduce risks, rather than invest all their funds in a single business venture. Three types of partnerships are evident in the India trade: those between Jewish merchants and their co-religionists; between merchants and shippers; and between Jewish and non-Jewish trading groups. Several prominent Jewish business families are represented in the documents. Joseph b. David Lebdi of Tripoli undertook a voyage from 1094/5 to 1097, which started at al-Mahdiya, Tunisia, and was destined for Nahrwara (Anahilavada) in Gujarat, the country of lac, textiles, steel and beads.

Lebdi arrived at ‘ Aydhab with eighty bales of lac packed in Nahrwara, Gujarat, and fifty bales of pepper. The overseas venture was in partnership with two other merchants, Abu Nasr and Farah, and was for an amount of 800 dinars. While Lebdi returned to Fustat, after receiving ‘governmental prescripts' allowing foreigners and non-Muslims safe passage, he left his two partners at ‘ Aydhab to send the lac in installments to Fustat. Unfortunately the two men were murdered, which led to complicated lawsuits.^{^[308]} Once the lawsuit was settled, Joseph Lebdi undertook a second voyage in 1099-1101. Lebdi also had dealings with Hasan b. Bundar, representative of merchants in Aden. His son was Madmun b. Hasan b. Bundar, a shipowner and representative of merchants in Aden and Nagid or leader of the Jewish community in the land of Yemen.

In contrast to the Lebdi family who travelled to India, Abraham Ben Yiju was resident largely at Mangalore on the south Kanara coast of India. Most of the documents relating to Ben Yiju indicate that his permanent residence in India was at Mangalore (Manjarur). However he also dispatched shipments to Aden from Fandarayna (Pantalayini) where he must have lived for some time and also from Dahbattan (Valarapattanam) and engaged in business in Faknur - all cities on the Malabar coast. On one occasion we are told that his family was in Jurbatan.

Ben Yiju was not only an agent for his partner merchant in Aden, but more importantly he owned a brass factory in India. The Geniza papers provide valuable insights into craft production in the eleventh and twelfth centuries. Customers in Aden provided the materials, i.e. copper and tin as scrap, and paid the artisans per piece according to weight. The Arabic terms nahds and sufr are used interchangeably to refer to a copper and tin alloy, while ‘yellow copper' is often used to designate a higher percentage of tin. The workers included slaves and Jews from Yemen, one of them termed Abram and another called ‘Iyar, who was charged with assaying measures and weights. The proprietor provided the workplace and the tools and also procured customers, while the artisans paid for the fuel used, which consisted of rice husks. The artisan was paid according to the weight of the object produced - a system also in use in Yemen.^{^[309]}

One further issue regarding Ben Yiju's sojourn in India remains to be addressed. How were goods transported between Mangalore, where Ben Yiju was based, and Aden where his customers lived? Over twenty places on India's west coast are mentioned in the Geniza records, and each ship or convoy had its pre-allocated landing place and was labelled accordingly, such as ‘one bound for Broach', or for Tana or Kulam.^{^[310]} Indian shipowners were designated as PTN SWMY corresponding to pattana sami in one inscription.^{^[311]} ^{^[312]} The term is translated as lord of the mart, or in Arabic shaykh al-suq.3⁴ Dahbattan or Valarapattanam is mentioned on the Kanara coast where Ben Yiju had his brass factory. An Indian shipowner mentioned several times is PDYAR who possessed several ships, one of which was commanded by a Muslim.

We have highlighted a range of trading practices in the western Indian Ocean. On the one hand were traders and merchants who did not travel themselves, such as Shaykh Abu Mufarrij and his son Shaykh Ibrahim, but who were nevertheless involved in trading networks on account of their ownership of prosperous warehouses at the Red Sea port of Quseir. In addition the father and son duo functioned in various capacities ranging from merchant, agent, and manager, to tax farmer; from head of trade and market inspector, to a deliverer of sermons. In contrast is the evidence from the Geniza documents, which trace histories of several families, such as Madmun b. Hasan b. Bundar who represented foreign merchants in local law suits in Aden, the Lebdi family who travelled to India, and Abraham Ben Yiju who was resident largely at Mangalore on the south Kanara coast of India.

While business partnerships and other relationships emerge from these documents, completely absent is any interaction with local rulers or political elites in the Indian subcontinent. This suggests that perhaps coastal centres functioned in a somewhat autonomous fashion and interacted with other littoral settlements, rather than with centres further inland.

Another significant change that emerges from the present discussion relates to legal frameworks for economic activities that were institutionalized in Gujarat from the ninth to the beginning of the fourteenth century. The Sanskrit-written Lekhapaddhati is a unique document, dated from 744-5 to 1475-6, that contains a collection of specimen letters and provides rules for drafting a variety of documents, such as rules for drawing up land grants, treaties between rulers, rules of administration and drafts of private letters. It refers to five departments of the state relating to trade and commerce: vydpdra karana (trade and commerce), veldkula karana (department of harbours), jalapatha karana (department of waterways), mandapikd karana (department of customs) and tanka sdld karana or mint.^{^[313]} What makes the Lekhapaddhati invaluable is its close correspondence with the contemporary inscriptions of the Chaulukya kings of Gujarat who ruled from 996 to 1241.

The integration of communities across religious affiliations is further substantiated by information on slave agents and the slave trade. One of the Geniza letters (II, 48) refers to the writer's unsuccessful attempts at purchasing a slave (wasιf) from a batch of new slaves brought to Aden, on the occasion of the Jewish festival of Sukkot, from bildd al-Zanj or East Africa. There are other references to the purchase of slaves who worked as household servants (II, 56-7). Unlike slaves from Zanj, Ben Yiju had a household retainer - an Indian slave named Bama who helped him in the business and was treated with some respect in the letters. An inscription dated 15 June 1126, discovered at a village about two hundred miles southeast of Mangalore, refers to a servant of warriors named Masaleya Bamma, while another from the same area records the name of setti or trader Bamma from a merchant family.^{^[314]}

The earliest document in Ben Yiju's archive, written soon after his arrival in India, is a deed of manumission, done in Mangalore, granting freedom to a slave girl whom he had purchased. Through manumission she became a Jewish convert and was given the name Berakha (Hebrew for blessing). The manumission deed contains several other phrases that have been identified only in Yemenite Jewish documents. It has been suggested on the basis of a court record written on cloth, but much effaced, that Jewish communities in Broach and Mangalore had some kind of autonomous juridical authority, though they were under the overarching jurisdiction of the rabbinical court of Aden.^{^[315]}

After spending eighteen years in India, Ben Yiju returned to Yemen with his family around 1140. He had two sons and one daughter from his wife Ashu, though one son had died as a young child in India (ιιι, 41, line 14).^{^[316]} It would seem that Ben Yiju's return to Yemen was not taken kindly by the local Jewish community, especially his marriage to his freed slave. This raised a further question about the legality of this liaison and the inheritance passing on to his son from this marriage. Four letters in the archives in Ben Yiju's handwriting relate to the legal position in the case of a Jewish man having illicit relations with a slave girl before her emancipation.

In contrast to Jewish law and the position of female slaves, the next issue relates to the status of slaves in western India. References to slaves involved in a variety of domestic chores occur in a range of literary sources and the terms dasa and dasι are used for them. The pillar edicts of the Mauryan ruler Asoka (269-232 bce), inscriptions written on stone pillars that proclaim his adherence to Buddhism, refer to correct behaviour towards brahmana, sramana, slaves and servants, while the Jatakas mention four kinds of slaves: those born slaves; those bought; those who become slaves of their own free will; and those driven by fear. There are scattered references to the import of slaves from India in the Mediterranean world, while the Periplus specifically mentions the import of singing boys and female concubines at Barygaza or Bharuch on the west coast. There are references to the exchange of slaves on the island of Socotra, where shippers came from the west coast of India.³⁹

In this regard the Lekhapaddhati is of relevance. The deed of the sale of a female slave dated 30 April 1230 / 19 April 1231 records that King Rana Pratap Singh brought a sixteen-year-old girl named Panuti after attacking another kingdom. She was sold as a female slave at the crossroads after the five leading men of the town were informed, and the merchant Asadhara bought her at a price of 504 Visalapriya drammas for doing household work in his house. The deed was signed by residents of the town as witnesses.^{^[317]} In another case, the deed refers to the sale of a female who voluntarily chose to become a slave to escape severe famine and harassment by mlecchas, that is, those who do not follow the norms ofVedic society. Her duties specified that she was to be employed for household work.

Interconnectedness of the Indian Ocean

From the ninth to the mid-fourteenth centuries, two of the merchant associations that dominated economic transactions in South India were the Manigramam and the Ayyavole. Associated with these were communities of craftsmen such as weavers, basket-makers, potters, leather-workers and so on. As late as the seventeenth century, the Ayyavole seemed to be concentrated in the cotton producing areas of Andhra Pradesh on the east coast of India. Though these two originated independently of each other, from the mid-thirteenth century onward, the Ayyavole association became so powerful that the Manigramam functioned in a subordinate capacity to it. Not only did these merchant associations develop powerful economic networks, but they also employed private armies.^{^[318]} The range of their operations extended well beyond the boundaries of the Indian subcontinent into Southeast Asia.

Several clusters of Tamil inscriptions have been found on the eastern fringes of the Indian Ocean from Burma to Sumatra. Of the seven midninth to late thirteenth century inscriptions written in the Tamil or part- Tamil language found so far in Southeast Asia, one has been found near Pagan in Burma, two just south of the Isthmus of Kra in the Malay peninsula, and four in north and west Java. Perhaps the easternmost record is the bilingual Tamil and Chinese language inscription associated with the remains of one of the two Siva temples at Quanzhou in South China. These inscriptions connect merchant associations operating out of South India with the founding or the endowing of temples or other structures for the use of a resident Indian merchant community. In the 1060s and 1070s Chinese records report that the imperial court received missions from the Chola kingdom of South India, as well as from the ruler of Srivijaya located on the island of Sumatra, but by the last decades of the eleventh century, the Chinese court had begun to encourage Chinese traders to venture out to sea.

Perhaps the most relevant example is the Buddhist monastery at Nagapattinam, which was a major landmark on the Tamil coast from the seventh to the nineteenth centuries. A Buddhist temple was erected at Nagapattinam specifically for Chinese Buddhists at the insistence of a Chinese ruler during the reign of the Pallava king Narasimhavarman II (c. 695-722). One of the later Srivijayan kings, Maravijayottunga-varman, is known to have provided for its construction, and the Chola king Rajaraja I granted revenues of a large village, Anaimangalam, for its upkeep in 1006.

In the tenth century, local versions of these merchant guilds named banigrdma appeared in the north coast ports of both Java and Bali, especially at Julah on the Balinese coast. There are seven Javanese inscriptions dating from 902 to 1053 that refer to merchant associations called banigrdma and to the various tax concessions granted to them. While some foreign merchants may have been included in these groups, these appear largely as indigenous organizations associated with the local economic networks as tax farmers.^{^[319]}

At this stage, certain conclusions may be drawn. First, information on seafaring communities from historical and archaeological sources indicates that their composition cut across ethnic boundaries. Second, the role of religions, such as Buddhism, Hinduism or Islam, in motivating and supporting seafaring activity needs to be recognized and accepted. Third, local demand played an important and steady role, and coastal centres frequently interacted with other littoral settlements. Finally, for an appreciation of cultural interchanges across the seas, it is crucial to highlight the diverse channels of communication, which also included oral transmission by priests and pilgrims, traders, wandering storytellers and entertainers. It is only then that a holistic understanding of cultural interaction across the maritime world will emerge.

Trading partners across the Indian Ocean: the making of maritime communities

More on the topic Trading partners across the Indian Ocean: the making of maritime communities: