1Ancient Indian Ocean trade, while avoiding many of the pitfalls of overland commerce at lower costs, was still a risky business.1 Even if mariners timed their ventures according to seasonal winds, adapted their vessels with new technologies, and possessed the celestial knowledge necessary to navigate the open ocean,2 ships were still vulnerable to a variety of hazards. Sudden shoals, attacks from brigands, or state-sponsored pirating efforts all threatened ocean-going vessels, their crews, and their merchandise – when left unaddressed, such factors dramatically increased the transaction costs of maritime ventures for all parties involved. When a ship went down, the investor(s) financing the operation often bore the burden of a cargo lost at sea, but traders onboard could well lose their lives.

2In an ancient world full of unknowns, with capital and life on the line, what steps were taken to mitigate the risks to ships on the Indian Ocean? This paper examines the potential dangers of two waterways connected to the western Indian Ocean – namely, the Red Sea and the Gulf of Khambhat (or Cambay) – together with some of the measures taken to reduce them in the first three centuries of the Common Era. After a brief introduction to the hazards posed by these contested waterways, this paper will investigate some of the top-down strategies taken by ancient states to secure them; then, it will turn to possible bottom-up tactics employed by trading ventures to ensure their safety. Where appropriate, this paper appeals to examples of “human factors” from later periods of Indian Ocean commerce, not only to fill the gaps left by the ancient evidence, but also in service of one of the major aims of the Red Sea 9 meeting in Lyon – namely, Indian Ocean connectivity in the “longue durée”.3 For the aid of the reader, geographical points of interest are displayed in fig. 1.

Fig. 1 – The ancient Indian Ocean world, ca 1‑300 CE, AWMC map rendered under CC BY 4.0 (J.A. Simmons).

Geography and piracy

3Despite their difference in size and orientation, the Red Sea and Gulf of Khambhat were connected in ancient conceptions of space as part of the larger “Erythraean Sea” (“Erythra Thalassa” or “Mare Rubrum”) of Greco-Roman geographies or the “Great Sea” (“Mahāsāgara/samudra”) of Indian sources. Both waterways have been exploited by humans for much of recorded history and remain crucial in modern shipping despite repeated warnings of danger in nautical traditions. Both bodies of water host a variety of natural hazards to ships, especially those without mechanized propulsion and real-time bathymetry. Due to the nature of its geological formation, the Red Sea has shallow shelves along both its African and Arabian coasts, which has encouraged the growth of numerous barrier reefs and atolls. Safe travel in Antiquity was limited to narrow corridors along either coast, or else a wider stretch in the center of the sea;4 cabotage within these narrow coastal bands could prove dangerous, especially in areas with irregular currents and sudden reefs. Atmospheric conditions, while providing the tricky wind regime necessary for seasonal travel, also imperiled sailors with violent sandstorms.

4The Gulf of Khambhat, a bay off the Arabian Sea surrounded by the modern Indian state of Gujarat, is defined by its numerous shoals. Several long, ribbon-like sandbanks run in a north-south orientation at its wider end, and alluvial deposits from multiple rivers define its estuaries. The high tidal variance and velocities when entering the gulf compound the risks that these sandy banks posed to vessels, as ships could run aground without much warning.5 These coastal hazards, described in “shipping manuals” as early as the first-century Periplus of the Erythraean Sea (allegedly composed by an anonymous Greco-Egyptian for Greek-speaking mariners on the “Erythra Thalassa”), could prove exceedingly dangerous to ships with navigators unfamiliar with the region.6 Despite these dangers, the pre-modern ports of this temperamental waterway, and in particular its principal port at Bharukaccha (modern Bharuch in Gujarat), known as Barygaza to the author of the Periplus, proved great lures for trade, as demonstrated by the many stone anchors found along the coast of ancient Saurashtra.7

5Ancient sources indicate that pirates operated in both bodies of water as well.8 Ancient Greek authors, including Diodorus Siculus and the geographer Strabo, record that Nabataeans frequently raided Ptolemaic ships on the Red Sea,9 but the Periplus of the Erythraean Sea notes that the west coast of Arabia remained a hotbed of piracy in the first century CE.10 The Periplus also notes that an anchorage near Adulis, the entrepôt on the Eritrean coast with links to Aksum in the Ethiopian highlands, was susceptible to raids by unspecified barbarians.11 The Konkan coast, just south of the Gulf of Khambhat, had a strong association with piracy in Greco-Roman geographical traditions; in fact, in his geography of the second century, Ptolemy labels this region as a “Pirate Coast” of sorts (“andrôn peiratôn”).12 As we shall see, increased piracy in this region in the first and second centuries, as well as others throughout the wider Indian Ocean world, such as southern Arabia from the mid-second to mid-third century, may correlate with larger geopolitical conflicts.13 Nevertheless, it is difficult to know whether the piracy described in ancient sources was carried out by private criminal outfits, privateers, or formal military personnel. A contentious political situation along the Konkan or south Arabian coast may have given rise to aggressive “peer-polity” jockeying for access to incoming trade,14 which could qualify as “piracy” in the minds of its victims. In any case, it is clear that coastal areas along these bodies of water could prove dangerous.

Top-down strategies: state initiatives

6What could ancient states do to ensure the safety of mariners on dangerous waters? Any development of the coastal regions along these waterways reflects a conscious effort by states to gain access to the resources they offered, including those of maritime commerce. Infrastructure on the Red Sea coast, far from major population centers and difficult to sustain due to a dearth of littoral resources and natural harbors, required deliberate and consistent investment from states.15 A similar difficulty applies to the ports on the Gulf of Khambhat, which were very susceptible to silting due to the prodigious alluvial outputs of the gulf’s estuaries. However, ancient states had a major incentive to invest in these regions in the form of tax revenue. High rates of indirect taxation applied to Indian Ocean products are to be found on either end of the Arabian Sea, including the Roman tetarte tariff16 and those recommended in Indian shastric texts.17 We might extrapolate the potential gains for ancient states from the taxes to be paid in Egypt for the cargo preserved in the second-century “Muziris Papyrus”; even if this document reflects a particularly large cargo by contemporary standards (valued at approximately 7 million sestertii), the taxes from all Indian Ocean imports in a given year were likely a substantial source of revenue for the Roman Empire.18

7Infrastructure works undertaken by ancient states to link these ports to the interior – the fortified roads in the Eastern Desert of Egypt19 or through the steep passes onto the Deccan Plateau of India,20 warehouses for imported goods,21 and rest houses for traders22 – all provided opportunities for the deployment of armed personnel by the state to aid and protect those conducting commerce or, better yet, monitor their activities. These physical constructions – the literal paths on which traders depended for their safety – allowed states to manipulate the direction, speed, and volume of commerce; all the while, it provided merchants the certainty of a reliably safe environment for their commercial activities, thereby lowering transaction costs. Certain harbors made maritime trade even safer by virtue of their location according to regional wind regimes, as has been argued for the Red Sea port of Berenike.23 The maintenance of safe harbor environments was thus mutually beneficial for both state interests and commercial agents.

8In addition to infrastructure within their borders, ancient states could station military or other service personnel throughout dangerous waterways to advance their interests and provide aid to sailing vessels. Beyond the brief Roman military expeditions along the Red Sea during the reign of Augustus,24 the Romans deployed manpower throughout the Red Sea on a more regular basis, particularly in the second century CE. The Roman Red Sea defense network, as we might call it, had two major components, namely a designated Red Sea fleet25 and a military outpost with jurisdiction over the southern Red Sea based on the Farasan Islands.26 We might surmise some of the operational imperatives of these forces. A permanent fleet could deter piracy on the waterway, as well as provide aid to merchant vessels in distress. Perhaps it functioned like the earlier Ptolemaic “guardians of the Arabian Gulf” (“phylakon tou Arabiou mykhou”) reported in Strabo, which allegedly rescued a shipwrecked Indian sailor in the time of Ptolemy VIII Physcon (ca 120 BCE) – though the veracity of this episode has been rightly approached with caution.27

9Similarly, the Farasan Islands provided an additional safe haven for sailors along the dangerous Arabian coastal route quite near to the pivotal Mandeb strait.28 In fact, Red Sea forces could be deployed to the Bab el‑Mandeb in response to threats that might hinder commerce: the Roman detachment stationed on Farasan was in a unique position to counteract piracy in the southern Red Sea up until the Bab el‑Mandeb;29 and a Roman attack on Aden at some point in the first century CE, an event briefly mentioned in the Periplus, may well represent an early iteration of military interventionism (though the historicity of this event has been debated).30 While we have no explicit testimony for the primary mission of Roman forces in the region, we might gain insight from a comparative perspective. Indeed, other regional powers used military force for the reasons outlined above: e.g. a third-century Aksumite king, who, in the so-called Monumentum Adulitanum II, boasts of military victories over various powers dwelling across the sea in Arabia (“peran de tês Erythrâs thalasses oikoûntas”) and commands that there be peace for those travelling by land or sea (“hodeusthai met’ eirenes kai pleesthai”).31

10The Western Kṣatrapas of Gujarat also stationed personnel on the Gulf of Khambhat and surrounding coastal areas in order to protect commercial vessels from dangers, as the Periplus notes on two separate occasions. For foreign vessels entering the gulf, the Kṣatrapas supposedly deployed royal fishermen and pilots with intimate knowledge of the dangerous shoals and currents of the waterway.32 In addition, the Periplus notes that foreign vessels that reached the neighboring Konkan coast would be escorted back to the port of Barygaza (Bharuch) under Kṣatrapa guard.33 The need for the Kṣatrapas to protect commercial vessels would make sense in light of the contemporary military campaigns launched by the neighboring Sātavāhana dynasty to reclaim territory on the Deccan Plateau lost a generation earlier, as well as the abundant examples of their counterstriking of Kṣatrapa silver coins. The passages in the Periplus, when read together with corroborating evidence of inland conflict and the known physical dangers of the Gulf of Khambhat, seem to record explicit actions undertaken by the Western Kṣatrapas to provide some recourse to mercantile vessels sailing through treacherous waterways.

11Diplomacy was another useful tool. Roman diplomatic involvement in Arabia spanned well beyond its arrangements with the Nabataeans and other Arabian tribes near its borders.34 The Periplus lists Charibael, a south Arabian king linked with the Sabaean Karib’îl Watâr Yuhan’im, as a friend to the emperors due to frequent embassies and gifts.35 Further south Arabian inscriptions of the second and third centuries record religious ceremonies and diplomatic missions between various Arabian polities, some of which involve Indian and Palmyrene envoys as well.36 Good relations with Arabian peoples governing important sea lanes allowed for ships to proceed through foreign waters without harassment. In terms of the Gulf of Khambhat and its environs, marriage alliances between the Sātavāhanas (based at Paithan, or Paithana in the Periplus) and regional dynasties including the Mahārathīs and Mahābojas,37 may well have aided them in their more rapacious tendencies along the Konkan coast.38 The Kṣatrapas (based at Ujjain, or Ozene in the Periplus) made their own alliances with other dynasties throughout the subcontinent, including the neighboring Ābhīras, the Licchavīs of Vaishali, and later the Ikṣvākus of Nagarjunakonda.39 We could view all of the initiatives of states, from regional diplomacy and skirmishes to fortification and infrastructure, as creating a monitoring system that guided the inflow of goods, and thus, increased tax revenue; but even if states prioritized security and diplomatic pageantry over policies directly aimed at commerce, their actions could result in incidental protections for commercial actors. Of course, those seeking to avoid taxes and state oversight could always try to circumvent them at more dangerous headings, but in doing so, they would eschew potential benefits offered by the latter.40

12In the end, the amount of change that states could effect was limited. Investment by ancient states along these waterways was essentially targeted and did not cover hundreds of kilometers of coastline. Despite the normative precepts regarding state responsibilities to curb piracy preserved in ancient sources,41 actual deployments of security personnel surely fluctuated depending on the dynamics of political economy – we might cautiously draw inspiration from current events, in which the collective benefits to global commerce provided by maritime security forces gives way to nationalist posturing in political rhetoric.42 Moreover, states could incentivize the pirating (or even the destruction) of vessels rather than protecting them as a way to gain revenue or strike a devastating blow to their enemies. The aforementioned Nabataean raids on Ptolemaic ships serve as early examples of this; in a more dramatic instance, a south Arabian inscription describes a Sabaean attack on the Ḥaḍrami port of Qani’ (Kane of the Periplus) ca 225 CE, in which the besieging forces occupied and set fire to ships in the harbor.43

13Diplomacy was also far more effective on a regional scale, but we should be cautious not to ascribe too much consequence to such arrangements. In fact, the Western Kṣatrapa and Sātavāhana royal families themselves eventually became intertwined through marriage in the second century, as exemplified by inscriptions in which the “mahākṣatrapa” Rudradāman I mentions his “closeness of relation” to the Sātavāhana royal Vāsiṭṭhiputta Sātakaṇṇi;44 nevertheless, their relationship did not prevent warfare between the two polities over the course of the century. Diplomatic ventures between Arabian polities must be read in light of the periodic conflicts between Saba’, Ḥaḍramawt, and Ḥimyar, together with the rise of Aksumite hegemony beginning in the third century.45 Moreover, it is hard to find any tangible policy outcomes of the oft-discussed Roman diplomatic encounters with Indians, Scythians, and Bactrians, let alone guaranteed protections for “Roman” traders across the Arabian Sea.46 Even if the alliance forged under Charibael mentioned in the Periplus lasted for centuries – a far from certain prospect – it is hard to imagine how any amicitia between Rome and the Sabaeans would have protected merchants during the sack of Qani’ in the early third century or the Aksumite incursions of subsequent decades. Mercantile networks, as opposed to diplomatic meetings, often proved more useful for the rapid dissemination of information about dangerous conditions – in fact, the “informal networks” established and maintained by the former often proved vital for the success of the latter.47

Bottom-up tactics: private initiatives

14What could those actually sailing on these bodies of water do to mitigate the associated risks? The principal resource utilized by private commercial agents was information. Maritime ventures could contract local coastal pilots with intimate knowledge and experience of dangerous waterways, harnessing information to their advantage.48 However, when not embodied in the physical form of the anonymous pilot, the collective experience of maritime ventures could also be disseminated in written form, surviving for us most explicitly in the Periplus of the Erythraean Sea. This synthetic document, a distant relative of portolans and pilot books so essential in pre‑modern shipping, was oriented toward individuals setting out from ports in Roman Egypt, as well as their financiers based in Alexandria or elsewhere.49 It not only lists the specific winds and timing for sailors attempting the transoceanic crossing,50 but also notes particular coastal hazards throughout the western Indian Ocean: e.g. where shoals were particularly dangerous; where piracy was to be expected; and where vessels ought to make use of knowledgeable pilots.51 Particularly dangerous waterways could receive extensive treatment in such texts – for instance, while the Periplus usually offers a sentence or two about the specific dangers of a port of call, it devotes six full sections to the merciless Gulf of Khambhat.52 Such “colloquial” or “aspirational” works, as well as more formal genres (e.g. geography), benefited from the practical wisdom of trading ventures;53 however, gaps in the knowledge or a dearth of contemporary sources upon which to draw are also reflected in these texts, such as the spotty coverage of the Persian Gulf and much of the Bay of Bengal in the Periplus.54

15Abstract information about dangerous waters was anchored by fixed reference points. Visual cues helped indicate that ships reached particular stretches of coast, and, by extension, coastal hazards. The appearance of sea snakes from Sind to Malabar announced the approach to coastal India and, thus, the potential for shoals – in fact, swarms of sea snakes (along with sea fowl) continued to serve as visual markers for Indian Ocean sailors well into the modern period.55 Changes in the color of water56 or specific topographical features with descriptive names also helped to orient sailors with landmarks along their route. For instance, the so-called “Katakekaumene”, or “burnt” island, marked the end of dangerous terrain along the pirate-filled waters of the Arabian coast in the Periplus.57 While certain elements of maritime technology spread throughout the Indian Ocean world in Antiquity (e.g. Indian cotton sails in both Mediterranean and Indian weaving patterns, Mediterranean ship construction using Indian teak),58 it does not seem that tools for measuring depths, such as sounding weights,59 made the jump from the Mediterranean to the Red Sea to be used among the other tactics discussed here. Given the lack of their use in the “longue durée” of Red Sea and Indian Ocean sailing, perhaps experienced pilots did not need them.60 Perhaps future discoveries at the Egyptian ports will change our understanding.

16Beyond specialized texts, how could traders transmit knowledge of the dangers along the Red Sea or Gulf of Khambhat? Despite the importance of personal letters for communicating market conditions and circulating products within personal exchange networks,61 not many explicit mercantile letters, such as those of the medieval Geniza Archive, survive from the early centuries of the Common Era.62 Indeed, they may have proven less effective for transoceanic communication, as letters could only come and go with the seasonal winds, along with traders themselves. One way to get around the seasonal limitations of letters was to specify exact ports of call and return-by dates for trading ventures. Whether orally agreed upon or formalized in contracts, such leave-by dates gave some sense of surety for when cargos would arrive; they were especially relevant for the investors who poured large sums of capital into these ventures.63 It may be no surprise that the loan of the Muziris Papyrus specifies the exact route to be travelled by the cargo belonging to the creditor.64 The reliable schedule of the monsoon winds made fixing deadlines in Indian Ocean commerce a simple affair: ships could leave no later than July from the Mandeb Strait and could depart from India between December and January.65 In fact, a round-trip journey from Alexandria in Egypt to western India, there and back again, could be completed within a single calendar year. No shows at the end of the sailing season may have prompted members of their support networks, financial or otherwise, to fear the worst.

17Nevertheless, we can look to other means of communication that enabled the effective spread of information. Graffiti at key locations frequented by traders, such as roads in the Eastern Desert of Egypt,66 the Darb al‑Bakrah of northwestern Arabia,67 or the Hoq Cave on the island of Socotra,68 suggest both imitative and dialogic tendencies of their inscribers,69 and thus, that these writings were effective forms of communication. Word of mouth or rumor could often communicate information with a surprising level of accuracy, especially when the veracity of the information was of direct relevance to a financial operation.70 Other orally-transmitted sources of information, such as mnemonics, rhymes, and songs, were tools of choice by pre-modern Indian Ocean sailors to remember particular headings and wind regimes, and we ought not to dismiss these as effective repositories of knowledge.71 It has often been argued ex silentio that ancient cartographic traditions had little use for drawn representations of space so important in later periods,72 but even crude visual aids or sketches could help sailors delineate hazardous waterways.73

18Where was this information shared? Beyond individual conversations, word-of-mouth channels stemming from well-established “ports of trade” were essential for information sharing related to the situation across the sea during the previous sailing season.74 Aden was particularly useful for the waterways discussed here, hailed by the Periplus as a former meeting point just beyond the Bab el‑Mandeb;75 the island of Socotra, with its multicultural inhabitants known from ancient sources76 and extensive trader graffiti, may well have outlasted Aden in importance during Antiquity, especially for sailors hailing from ports on the Gulf of Khambhat. Physical structures for exclusive use by traders have been attested and could serve as places to access institutional memory. One such epigraphic example is a communal space for the association of Palmyrene Red Sea traders set up in Coptos by their colleague Zabdalas – one of many established for the benefit of the Palmyrene diaspora throughout the ancient world.77 Gatherings of shippers within these communal structures, as well as religious sites, which often boast graffiti written by traders of similar cultural backgrounds, disseminated information far more efficiently than chance conversations or written communications.78

19Knowledge is all well and good, but sometimes brawn was needed – after all, in an age before rapid telecommunication, information became stale rather quickly, and proved essentially useless when a ship fell under attack from pirates. Supplying ships with weapons or armed personnel (e.g. archers) was one way to repel attacks from other vessels on the high seas. Arming crewmen on merchant ships helped to fill the gaps in protection left by security forces at ports, whether private contractors or those associated with the tax-collecting apparatus,79 and the small fleets deployed by ancient states, which could not enforce every stretch of dangerous coast. The Konkan coast is a good example: despite the state-sponsored efforts to protect merchant ships there, Greco-Roman textual sources recommend that trading vessels arm themselves with archers just in case.80 The later fictional work of the Indian author Daṇḍin contains another example, in which a foreign merchant ship manned with archers successfully fends off an attack from another vessel on the Bay of Bengal.81 Though well beyond the Gulf of Khambhat and Konkan coast, the find of a single spearhead from the fragmentary cargo of the Godawaya wreck off Sri Lanka, dating roughly to the period covered in this paper, may prove more suggestive than definitive when it comes to the use of weapons by such personnel.82

20Ultimately, even if crews prepared themselves with the necessary information and hired muscle, they could not mitigate all the risks posed by these dangerous waterways, and some ships were inevitably lost. The Godawaya wreck is one such example, a ship which sank in an area of coast defined by strong currents. Another commercial vessel went down in the Eliza Shoals along the Arabian coast of the Red Sea at some point in the third or fourth century.83 Several wrecks have been discovered near ancient ports along the African side as well, from Myos Hormos to Adulis.84 We should also consider that, as much as commercial vessels could try to avoid areas prone to piracy, pirating forces no doubt used similar strategies of information sharing in order to effectively pillage and elude authorities, as has been argued for Mediterranean-based piracy in the “longue durée”.85 Thus, the tactics used by crews to safeguard their cargoes were met by opposing tactics of criminal enterprise in a perpetual tango.

Conclusions

21Top-down strategies and bottom-up tactics both attempted to mitigate the risks of maritime trade in the dangerous waters of the Red Sea and Gulf of Khambhat. States invested in infrastructure and stationed manpower in otherwise hostile environments to encourage trade and generate revenue. Reliable information proved to be the most important resource for commercial ventures to mitigate these risks – perhaps an even more valuable commodity than the numerous products available for purchase throughout the Indian Ocean – but armed personnel protected cargo and crewmen alike from human violence.

22Traders were separated for long periods of time from their support networks, whether familial or commercial. The slowness of communication often meant that rumor outpaced official reports and those delayed by the forces of nature; in any case, the anxiety of loved ones can be felt on either end of the trade. A first-century papyrus letter recovered from Berenike, written by a mother Hikane to her harbor-man son Isidoros, asks why he has not written her while abroad in Arabia (along with other members of his family).86 Similarly, the worries of family members are palpable in Sanskrit literary texts based on the Bṛhatkathā. In the Bṛhatkathāślokasaṃgraha of Budhasvāmin, Sanudasa, a survivor of a shipwreck, is asked by the owner of a rest house whether he had encountered a clever merchant by his own name. Evidently, the heart-stopping news of the shipwreck had reached his uncle Gangattata; there was no way for the owner of the rest house to know that he was speaking to the missing individual.87 In another story of the later Prakrit Vasudevahiṇḍi, the fear becomes very real – a maritime merchant, less fortunate than Sanudasa, dies at sea, leaving his wife widowed and children orphaned.88

23A complete sense of safety for ancient mariners in such dire straits may have ultimately remained at the discretion of “force majeure”. Not much could be done to change the patterns of the monsoon winds on which these ventures relied, and advances in climatological studies have begun to isolate periods of volatility in hydroclimates throughout the Indian Ocean world.89 The religious graffiti on the island of Socotra, akin to the talismanic elements added to early modern Indian Ocean ships, may reflect a profoundly human strategy of propitiating the divine to protect their voyages.90 But beyond moving heaven and earth through thoughts and prayers, human agents had their ways of minimizing risk, lowering transaction costs, and making maritime commerce as safe and expedient as possible.