References to protection appear widely in the historical records of imperial interactions in the early modern world. Empires touted their capacity to protect not only their own subjects as they travelled and settled in distant territories but also, under certain circumstances, other subjects and sojourners. Mechanisms for proffering protection ranged from ad hoc interventions in jurisdictional disputes to more systematic arrangements, such as the sale of safe conduct passes or the negotiation of treaty obligations. The discourse of protection – what we call ‘protection talk’ – represented a basic currency of interpolity relations, pervading the language of treaties, diplomatic correspondence, tribute ceremonies, appeals for military assistance, voyage narratives, and petitions from individual subjects to sovereigns.
And yet, for all its ubiquity, discourse about protection retained a quality of imprecision that makes it difficult to pin down precise legal statuses and responsibilities. It was often unclear who was protecting whom or what was the exact nature of the relationship between the protector and the protected individual, group, vessel, or polity. Within the confines of a particular encounter, meanings of protection could swing between weightily substantial and airily vague. At times, claims to protection seemed to be empty rhetoric, little more than an irrelevant promise observed by neither side; at other times, they signalled potent, and seemingly irrevocable, ties between protected groups and protectors, with significant implications, such as commitments to go to war.
How then to understand protection in the early modern world? A first step is to query protection’s supposed functions. It seems obvious that protection had a dual character. It could be offered internally to subjects and also extended beyond the realms of the polity to outside groups. In the ‘inside’ register, sovereigns provided protection to their subjects in return for allegiance. A ruler’s legitimacy depended in part on the ability to back up claims about the capacity to produce domestic order and thereby protect subjects from arbitrary violence. In an ‘outside’ register, protection found expression in security arrangements between political communities, particularly in the promises of shelter from enemies in exchange for tribute, or in pledges of mutual aid in defence against shared enemies.Footnote 1
While this distinction seems clear on paper, the boundaries between the two varieties of protection were far less evident in practice. Over several centuries, composite polities with uncertain territorial reach – some describing themselves as empires, some not – jockeyed for control over trade routes, territories, free and coerced labour, and resources. In a period when fluid and overlapping spheres of influence were the rule rather than the exception, political agents routinely invoked protection to describe arrangements of power across polities and used such notions to structure varied interactions. The lines separating domestic and foreign protection were frequently blurred, with many protection claims creatively combining ‘inside’ and ‘outside’ protection. The exercise of protection over subjects by imperial powers (‘inside’ protection) sometimes served to thrust imperial influence and power into new territories and realms (‘outside’ protection). In other contexts, security arrangements between two polities (‘outside’ protection) transitioned into claims of political dominance and the extension of rule (‘inside’ protection).
This blurring of lines meant that a certain cultivated ambiguity ran through protection transactions. Rather than undermining communications or destabilizing interactions, this basic ambiguity framed interpolity relations and guaranteed protection’s pervasive appeal. Political agents or rulers might invoke protection in describing a pact between equals, in characterizing a tributary relationship with a clear overlord and subordinates, or even in labelling a domestic arrangement in which the less powerful party represented subject populations. Equally strikingly, these interpretations could co-exist (often for extended periods of time) within the same basic framework. This capacity to sustain multiple understandings of power inside cross-polity relations helps explain the long life of protection. Protection talk was, in other words, ubiquitous between the fifteenth and nineteenth centuries precisely because it was so flexible.
We are not the first to note the importance of protection in structuring cross-polity relations. But the systematic study of protection has mainly been confined to historians of empire and international law in the nineteenth and twentieth centuries. One group has probed the role of empires – especially the British empire in the early nineteenth century – in promoting protection as a framework for imperial consolidation and expansion, while other scholars have debated the origins of the emergence of the ‘responsibility to protect’ doctrine and have traced the entrance of protected states and protectorates into the lexicon of international law terms.Footnote 2 The story of protection in the early modern world deserves more study – and not merely to fill in the pre-history of these later trends. Recognizing the intentionally capacious and ambiguous uses of protection in early modern empires illuminates a distinctive, long phase of interpolity relations in world history. In contrast to the consolidating campaigns of early nineteenth-century global empires, the debates about quasi-sovereignty in the late nineteenth century, or the international project to establish protected states and protectorates in the twentieth century, networks of protection in the early modern world underpinned the creation and persistence of overlapping zones of political and legal power.Footnote 3
The view of empires as webs of protection opens a new window on the emergence and persistence of inter-imperial (or interpolity) zones of overlapping and shifting political and legal power. In focusing attention on these zones, our perspective builds on two bodies of recent scholarship. The first emphasizes the weblike nature of early modern empires. Protection arrangements were one of several phenomena that gave early modern empires the form of overlapping networks. Historians have pointed especially to the generative power of commercial relations; migratory circuits, including forced migration; and systems for the circulation of imperial officials.Footnote 4 Claims about protection touched all these types of networks. As we will show by analysing examples of imperial strategies, agents of empire proffered protection in order to cultivate trading partnerships and political alliances. The resulting webs of protection did not just influence empires; in a very real sense, such relationships constituted empires.
A second important strand of literature that provides context for our efforts focuses on imperial borderlands and is concerned with the interpenetration of political power, legal authority, and markets in areas where empires met and where they interacted with multiple polities.Footnote 5 Attention to protection talk helps us to understand the influence of ascendant imperial power in such zones. On the seas, references to protection provided an effective mechanism for projecting imperial power selectively into areas where political and military control was thin. In borderlands areas, mutually intelligible ideas about protection spanned very different political and social systems. The discourse of mutual protection framed alliances and gave them some stability, at the same time embedding within them the possibility of conquest, a context in which only one power could claim the capacity to protect. As empires consolidated power, protection claims shifted away from languages of alliance to discourses of subjecthood, in the process casting shadows of sovereignty that anticipated imperial annexation.Footnote 6
Protection was central to European expansion across the early modern world, and it should be acknowledged that Europeans were the overall beneficiaries of widespread notions of protection that could be introduced, often without provoking immediate resistance or dissent, and subsequently bent to serve particular interests. But there was nothing inevitable about this process, and we should be careful of assuming an upward curve of European influence running steadily through protection transactions across the early modern period. Rather, political actors operated with the expectation of the stability of interpenetrating power rather than the inevitability of imperial consolidation. In other words, the projection of the power to protect did not always lead to imperial consolidation and expansion. Power grabs that intimated conquest sometimes reverted to relations of fragile alliance. The juxtaposition of different meanings of protection framed the fluid reconfiguration of imperial power and interpolity relations that was so central to regional and global regimes of the early modern world.
Our analysis commences with the seas. Sprawling maritime operations under the loose command of individual imperial powers utilized protection to establish their influence and channel their interests. Rather than ruling the seas, vessels and captains representing even the most powerful empires had to navigate overlapping networks of protection. We then apply some of these insights to territorial arrangements in which polities invoked protection to similar ends and effects. In analysing the ways in which empires used protection to describe relationships of alliance, we show how talk about protection could signal both the stability of political coexistence and the possibility that alliances prefigured conquest. The focus throughout is on the blurred lines between ‘inside’ and ‘outside’ registers of protection and the patterns of interpolity relations produced when networks of protection intersected and overlapped.
Extortion and regulation on the seas
The maritime licence has long been regarded as an indispensable tool of European expansion, with historians’ commentary highlighting its use by the Portuguese in the Indian Ocean as well as by various successors and rivals.Footnote 7 The mechanics of the Portuguese cartaz system are well known. Capitalizing on the military advantage provided by the light cannons their ships carried, the Portuguese sold cartazes, or shipping licences, to Indian Ocean shippers to force them to call at Portuguese-controlled ports. The system generated revenue and prizes for the Portuguese, who collected licence fees and also profited when they attacked vessels not carrying cartazes. There were Mediterranean precedents for such a system, and, in the seventeenth century, other European powers, including the Dutch and the English, developed similar pass systems in the Indian Ocean.Footnote 8 But scholars have challenged the view that such systems represented European innovation. The Indian Ocean was not a completely non-militarized space before European entry in 1500. The Mamluk sultans of Egypt, for example, developed a maritime passport apparatus that looked remarkably similar to the Portuguese cartaz, and local rulers on the Malabar coast implemented their own systems of ‘enforced tribute’.Footnote 9
While valuable in undercutting assumptions about European exceptionalism, such analysis proceeds from a relatively narrow understanding of the cartaz system as a mafia-style protection racket in which a militarized sea power sold shelter from its own violence, extracting payments from merchants in exchange for ships’ safe passage.Footnote 10 Even when an empire proffered protection as a means of extending and consolidating its influence, guarantees of safe passage worked as a flexible tool rather than the centrepiece of a singular strategy. And when other powers were in play – as they almost always were on the polyglot seas – many agents referred to protection, in the process generating overlapping networks. The result was not so much a regulatory order as a framework for desultory violence.Footnote 11 Looking more closely at the actual workings of the Portuguese cartaz system showcases the fluidity of such systems.
The guarantee of ‘safe conduct’ by the Portuguese in the Indian Ocean took on many forms. Portuguese agents sold cartazes to Indian Ocean shippers, taxed cartaz-carrying ships in Portuguese-controlled ports, and regarded any ships without passes as fair game for capture.Footnote 12 The system generated some revenue but was mainly valuable for its effects in shoring up Portuguese monopolies for some commodities along particular trade routes and in identifying a stream of legal prizes for continued raiding. Particularly during the first decades of Portuguese maritime presence in the Indian Ocean, we observe something more complex still: an array of overlapping strategies to apply safe conduct as an instrument to consolidate and sustain Portuguese power.
As they encountered Indian Ocean polities and traders, Portuguese agents adopted postures ranging from warfare to negotiation to withdrawal.Footnote 13 Portuguese arrangements of suzerainty with regard to coastal polities incorporated some of the same language and structures of maritime protection. Further, Portuguese officials resorted to similar proffers of protection when seeking to regulate the activities of often unruly and self-interested subjects. When viewed in the context of overlapping networks of Portuguese patronage and command sustained by raiding, profit, and corruption, the cartaz system appears as one of a set of practices utilized to fashion the empire as a sprawling web of protection arrangements.Footnote 14
The wide range of protection practices of the Portuguese is apparent in the commentary of Afonso de Albuquerque, who led Portuguese campaigns at Goa and Melaka and established the network of ports that formed the key components of empire in the east. Protection arrangements underlay the Portuguese foothold in those ports. For example, Albuquerque’s attack on Ormuz resulted in a truce that left the port in the hands of the kingdom of Khaja Attar and ‘ordained that the king should rule in the name of the king of Portugal’ and provide tribute to the Portuguese.Footnote 15 The insecurity of the Portuguese position was more norm than exception. Even in Goa and Melaka, where his forces prevailed and the Portuguese installed themselves as rulers, Albuquerque worried about the tenuousness of Portuguese command. Imploring the Portuguese king to send more arms and ships, he alerted the king that Moslem traders who had been forced to pay for Portuguese protection were ‘wait[ing] for the moment when they can take the noose from their necks and set to work against you’. He warned of the absurdity of trying to keep even seemingly peaceful clients under control ‘with fair words and offers of peace and protection’.Footnote 16 The dangers posed by the weak ties created through bonds of protection encompassed individual Portuguese subjects who ranged outside the narrow ambit of Portuguese jurisdiction.Footnote 17
This volatility of protection arrangements in the ports extended to the way in which the Portuguese approached safe conduct on the seas. When merchants from the Coromandel coast appealed to Albuquerque for a safe conduct pass for ships sailing to Melaka and for a ship ready to sail from Melaka with a shipment of cloth, Albuquerque reported the cases as examples of the profitability of exchanging safe conduct guarantees for ‘the king’s portion’.Footnote 18 Yet, in the same letter, he bemoaned the loss of revenue resulting from the mere regulation of peaceful trade: ‘Now that we have peace and friendship with everybody except the Sabaio and Calicut, there are no prizes or booty’ from which to pay troops and maintain the fleet.Footnote 19 The benefits of providing protection had to be balanced against the opportunity costs of diminished raiding opportunities.
Under Albuquerque, crown officials were already applying similar tactics to the control of unruly Portuguese subjects in the east. It was, after all, by virtue of a licence from the king that the viceroy held the right to seize the property of non-compliant Portuguese and reassign captaincies of individual ships.Footnote 20 For example, the letter issued to Lopo Soares as Captain Major and Governor of India in 1515 invested him ‘with the power of removing and displacing captains of the fortresses … and captains of the naos or vessels … and all other officials, even if they go from here by our order and warrant’. Opportunities for private profit drew individual Portuguese to act as independent agents and even to seek to establish themselves as semi-autonomous rulers: Diogo Veloso installed a king of Cambodia under his sway and constructed a fortress under his own command; Filipe de Brito e Nicote parlayed his service as a mercenary in Burma into a governorship there, established a feitoria (trading post), and collected revenue in Lower Burma. Both appealed to Goa for recognition by the Portuguese crown.Footnote 21 In such cases, the difference between corruption and service to the king was a matter of interpretation.
Albuquerque labelled Portuguese captains as criminals for flouting his directives, and he confiscated and redistributed ships, providing protection to favoured subordinates by allowing them to sail in convoys with the Portuguese fleet.Footnote 22 Taken together, the set of Portuguese practices involving the sale or provision of protection, on land or on sea, in the Indian Ocean comprised a flexible framework for regulating both foreign and Portuguese shipping, and for structuring relations along a wide continuum of arrangements, with local rulers and Portuguese subjects.Footnote 23
Other pass systems, including those introduced by non-European powers, also evolved in unexpected ways by expanding to incorporate new groups or by channelling violence toward foreign upstarts and potential rivals. This was the case with the famous vermilion-seal or shuinjō system, which was consolidated in the early seventeenth century by the newly established Tokugawa regime in Japan. The shuinjō system was created as a domestic regulatory system with what seemed, at least at first, to be clearly defined boundaries. The Tokugawa military government, or Bakufu, issued passes (shuinjō) authorizing the holders to undertake individual voyages from Japan to stated destinations. The system was designed to monitor, control, and restrict access to foreign trade. Only trusted merchants and lords received passes, thereby ensuring that the profits represented by long-distance trade did not spill out into unfriendly corners of Japan, especially to Kyushu-based rivals such as the powerful Shimazu clan of Satsuma. Since the Tokugawa regime maintained no navy (or, indeed, vessels of any kind), the system functioned at the port level rather than on the open seas. Only vessels carrying these passes were allowed to depart from Japanese ports, and the Tokugawa shogun wrote to rulers across East and Southeast Asia warning them that any vessel not carrying one of these documents should not be permitted to trade.Footnote 24
Although the system was created for domestic merchants, it began to bulge in unexpected directions once it became clear that the Tokugawa would intervene aggressively to guarantee the security of any vessel carrying one of these documents. In 1610, for example, the Portuguese watched in horror as one of their massive trading ships was sunk in Nagasaki harbour for a perceived infringement of a shuinjō.Footnote 25 In the aftermath of this episode, foreign merchants, including later the Portuguese captains placed in charge of the crucial Macao–Nagasaki route, jostled to secure access to Tokugawa-issued shipping passes.Footnote 26 The result was the development of a lucrative marketplace where shuinjō were exchanged or sold by European and Chinese traders in order to secure the best combination of destination and timing.Footnote 27
At the same time, foreign merchants with no capacity to claim these documents also clamoured for Tokugawa protection, and a string of aggrieved traders arrived in Edo to demand that the shogun intervene to secure their vessels.Footnote 28 The regime responded by tacking on additional provisions that significantly expanded the remit of its authority over the seas. In 1621 the Bakufu decreed that all merchants in Japanese coastal waters (very vaguely defined) were entitled to Bakufu protection if they were attacked, regardless of whether or not they carried a shuinjō. In 1659 the Bakufu went one step further by stipulating that Chinese merchants on their way to Japan with goods intended for the Japanese marketplace could claim protection even if they were thousands of miles away from the archipelago.Footnote 29 The result was a significant expansion of the boundaries of what had started off as a domestic regulatory system and the insertion of Tokugawa protection as a key determinant in maritime conflicts in the East China Sea. In 1665, for example, the Dutch East India Company (Verenigde Oostindische Compagnie or VOC) was forced to call off a maritime campaign directed against Zheng shipping based in Taiwan after Tokugawa officials made it clear that the shogun’s protection extended to Chinese vessels travelling between Taiwan and Japan, even if the company was in open war with their home state.Footnote 30
It was also the case that a single power could maintain very different – and sometimes fundamentally contradictory – protection systems. In 1641 the VOC implemented a defined pass system, the pascedullen system, after the capture of Melaka, in which all shipping was required to call at this port in order to purchase passes.Footnote 31 But prior to this it had operated overlapping protection systems, all functioning according to strikingly different principles. In places such as the Banda archipelago, where the Company was determined to maintain monopoly control over key licences, the VOC created a narrowly defined pass system in which any Bandanese ship ‘found without a passport’ was liable to be ‘taken by the Hollanders and declared as good prize with all their cargo’.Footnote 32 In Batavia, by contrast, VOC officials issued passes to Chinese merchants in order to provide security and develop the port as a regional hub. And in the Taiwan Straits, the Company simply took over pre-existing protection networks run by local pirates and targeting Fujianese fishermen who were required to provide a share of their catch in return for security.Footnote 33
One of the reasons why so many different pass systems coexisted was that they were not always engineered from the top down by imperial powers seeking to bring order to the sea-lanes. Instead, such systems often emerged and derived strength from the actions of vulnerable merchants who petitioned relentlessly for passes and flags. Although the English East India Company (EIC) had no formal pass system in Asian waters, English agents did supply documentation of protection. The head of the English factory in Japan, Richard Cocks, was kept busy issuing documents for Chinese merchants. On 27 February 1618, he handed over ‘2 lettrs of favour … w’th 3 flagges, two new and one ould’ to Chinese merchants.Footnote 34 Two days later, he gave a flag and a letter for a Chinese merchant headed to Taiwan.Footnote 35 On 5 March, he was back in action, handing over ‘an Eng. flag & a letter of favor’, and the following day he gave a merchant called Fingo Shiquan ‘a letter of favour & an English flag’, although he was not even sure where the vessel in question was headed or what the name of the captain was – a lapse clearly evident in the two blank spaces in the diary where this information should have appeared.Footnote 36 In the space of just over a week, Cocks had handed over five letters and six flags, each designed to assure particular merchants of English protection even in the absence of a wider officially sanctioned regime.Footnote 37 In these and other cases, entrepreneurial merchants sought to bolster their connection with a European sponsor by carrying documents, displaying multiple flags, and even taking on board European crewmembers.
If pass systems were essentially flexible, what legal identity or guarantees did carrying a pass actually confer? At times, the answer seems to have been very little. A single vessel could carry multiple documents issued by both European and non-European powers. In 1622, for example, a VOC ship eagerly looking for prizes to plunder sighted a strange junk in the waters near Taiwan. When questioned, the captain of the junk produced three documents: a letter from a Dutch merchant, a similar document issued by the head of the EIC factory in Japan, and the ‘Emperour’s pass’ (a Japanese shuinjō).Footnote 38 Such permutations were not limited to the first decades of European expansion but continued to characterize maritime encounters into the late eighteenth century. In 1780, for example, an English ship encountered the Istambul, which was owned by the wealthy Surat-based merchant Saleh Chalebi, on its way from Surat to Basra. Although he carried an English pass, the captain, when questioned, explained that he had no obligation to fly an English flag or claim English protection because he also carried both French and Dutch passes.Footnote 39 The proliferation of papers showing sponsorship in Indian Ocean shipping made the operation of pass systems at times very similar to maritime practice in the Atlantic, where sea raiders routinely carried papers documenting authorization by multiple, different sponsors, as well as multiple flags, and calculated which documentation to produce in maritime encounters and in prize courts.Footnote 40 Across the early modern world, mariners and merchants hoped to convert sometimes flimsy and even fraudulent paper claims into effective protection.
Some traders deployed ingenious systems to minimize costs. Gujarati merchants sometimes departed legally with a cartaz, then arranged for the same document to be sent back so that the next vessel could sail with it.Footnote 41 Sometimes the issuers of passes undercut the value of their own documents in order to secure advantage. In 1622, for example, Dutch warships attacked Indian vessels exiting Mocha harbour with passes issued by the VOC chief in that port, sparking a ferocious outcry from local merchants and rulers over what they termed ‘false passes’.Footnote 42 In other cases, privateers and pirates deliberately forged passes to evade capture.Footnote 43
Often a single document could be subject to very different interpretations. The multiple uses of protection by the Portuguese opened opportunities for conflicting interpretations of its legal significance by both Portuguese agents and those purchasing cartazes. Writing in 1727, one Portuguese viceroy, João Saldanha da Gama, argued that the purchase of cartazes transformed recipients into effective vassals of the Portuguese crown – an outcome hardly consistent with a supposedly pervasive transactional logic.Footnote 44 Asian rulers with ships sailing under the protection of these documents often interpreted the arrangement very differently. For the Mughal emperors, their receipt of free cartazes, together with the fact that some of the emperor’s ships were permitted to sail without any licence as long as they paid customs duties at the port of Daman, enabled the regime to avoid any suggestion that carrying a cartaz required the acknowledgement of Portuguese suzerainty over the bearer.Footnote 45
These examples show that pass systems operated with a recognized degree of ambiguity about the relation between the protecting sovereign and the ships and crews sailing with licences. Such ambiguities were consistent with a more systemic flexibility about the purpose of proffering protection. Issuing authorities guaranteed safe conduct on an ad hoc basis, sometimes to ships belonging to formally recognized allies, sometimes opportunistically to others in efforts to establish monopolies on the trade of certain commodities or to control certain trading routes and ports, and sometimes haphazardly in response to merchant demands. At the same time, very similar licensing arrangements aimed to regulate the commercial activities of diverse sets of subjects and subordinates.
The vagaries of pass systems made it difficult to pin down precise legal statuses and responsibilities, even when parties had incentive to do so. Records of litigation over captures provide a window on strategies to invoke protection in support of competing interests. Such conflicts posed questions about who was entitled to protection and about what precisely protection meant – in particular, whether it conferred subjecthood on pass holders. Consider the case of the Hamedy, a Persian vessel seized by the Portuguese in 1725 after departing from the port of Bandar-Kung on the Straits of Hormuz. The ship’s owner, a wealthy Persian merchant called Aga Mehdy, had refused a cartaz, opting instead to sail under the protection of a French flag. The French bitterly protested the capture and dispatched a representative, Louis de Saint Paul, to extricate the ship from Portuguese control and claim restitution. Saint Paul proceeded to marshal a string of arguments to prove that the Hamedy was bound to the French company, claiming, among other things, that carrying a French commission conferred French nationality.Footnote 46 In much the same way, an East India Company pass could convey subjecthood: Company passes stipulated that the holder could expect to receive the same protection as if ‘actually in the said Right Honorable Company’s Service’.Footnote 47
A 1771 case involving the Portuguese and the EIC centred on the issue of subjecthood and provides a good illustration of the positioning of different powers over meanings of protection. In that year, the Portuguese captured a vessel belonging to Dhanjishah Manjishah, a Parsee merchant who had served as an English broker in Surat.Footnote 48 The ship had sailed from Surat in November 1770 with a pass, but an EIC official, Lieutenant John Thisilton, had confiscated the document because of non-payment of convoy duties.Footnote 49 After Dhanjishah’s vessel was captured, English and Portuguese officials clashed for months over differing interpretations of their rights.
The English argued that the case was entirely straightforward and that the vessel should be immediately restored to them with its full cargo. Anyone who held a pass, even if the pass was temporarily unavailable, was by definition a ‘subject of the Hon’ble Company’s as the Act of granting is a sufficient indication of it’.Footnote 50 The pass granted to Dhanjishah’s vessel ‘was of the usual Tenor and is such as Serves a Protection to all our other Subjects’.Footnote 51 On the other side, the Portuguese argued that, when the ship was seized, it possessed no ‘paper or Authentic pass that confirms her to belong to a person subject to the British government’.Footnote 52 When confronted with the pass itself, Portuguese agents ridiculed the notion that it meant that Dhanjishah was an English subject. According to the Governor of Goa, ‘if the Hon’ble Company treat that Parseeman as a stranger why do they intend that I should repute and consider him as a Native of England’.Footnote 53 If Dhanjishah was not a subject, then the English had no right to issue him with a pass, and by doing so English authorities had usurped ‘the rights of the Crown of Portugal’.Footnote 54 The offence was not a minor one. For the English to issue such a document was ‘the same as entering violently one of the Countries belonging to Goa and depriving its owner thereof’.Footnote 55
The case went back and forth because it called into question English rights to issue passes in the first place and the rights of holders of these documents to claim compensation based on their status as Company subjects. As a result, officials in Bombay determined that they must support their rights ‘of granting Passes to all Persons’ with all possible efforts.Footnote 56 When the Portuguese refused to back down, English officials determined to take action and, in late 1772, they convinced the Nawab of Surat to seize Portuguese property equal to the value of the captured ship and its cargo. In this way, both the EIC and the Governor of Goa sought to assert the authority not only to issue passes but also to determine their validity. Dhanjishah was unlucky, but he was not behaving illogically. His withholding of payment to the English recognized the imperfect nature of the protection that the English agents were offering. Yet, when the ship was captured, he made the obvious choice of asserting English protection anyway.
This example and others like it caution against asserting too strongly that a particular imperial pass system was organizing maritime relations in a particular place and time. Pass systems had ragged edges that shifted according to the strategies of rival powers, local authorities, and merchants. The picture emerging from these profiles is one of overlapping protection networks of fluctuating strength and shifting dimensions. Imperial agents offered guarantees of safe conduct in different circumstances and arenas to subjects, to allies, and to foreign traders without formal political affiliation. The fungible and flexible nature of promises of protection flowed both from mariners’ manoeuvres and from the strategies of empires and their agents seeking to preserve the legitimacy of violent interventions in order to prop up their interests on seas and in ports. These webs of protection are best understood in many places as interpenetrating parts of a single interpolity legal regime – an image far removed from the standard account of sequential maritime imperial domination in the Indian Ocean. No one power conquered the seas. Every militarized maritime power structured its imperial enterprise around a family of relationships of protection. We now turn our attention to land, where similar networks of protection underpinned interpolity relations.
From protected allies to protected subjects
Protection talk frequently commenced with the possibility of an alliance in which two external powers promised to guarantee each other’s security. In other cases, imperial agents pledged to protect newly encountered groups without claiming political authority over them, and militarily weak groups found shelter under protective regimes without signalling their intention to cede sovereignty or abandon jurisdictional claims. In other words, accepting protection could mean no concomitant cession of sovereignty, but the same kind of protection talk could transition into a framework for claiming sovereignty and reclassifying allies as subjects. Protection talk framed, foreshadowed, and in part constituted the process of conquest.
This process could happen with striking speed as putative allies suddenly found themselves reconstituted as subjects. The 1605 treaty concluded between the newly established VOC and elders on the Banda islands, where much of the world’s nutmeg was grown, stipulated that local authorities would take Dutch representatives under their ‘care and protection’, while the company promised to defend its new-found allies against outside aggression.Footnote 57 But four years later, in 1609, the terms of the relationship shifted dramatically as VOC representatives demanded permission to build a fortress in order to protect the Bandanese properly against external foes such as the Portuguese and the Spanish. That the fort would turn them from allies into something far closer to VOC vassals was obvious, and the Bandanese responded by attacking and killing a Dutch admiral. The Company retaliated with overwhelming force, quickly bringing the Bandanese back to the negotiating table, where they were forced to sign a new treaty that dispensed with any notion of an alliance. By 1621, the Bandanese found themselves reconfigured as VOC subjects, compelled to sign a new treaty stipulating that they would recognize the Company and its officials as their sovereign without accepting any other ‘princes or potentates’.Footnote 58
In such cases, protection served as a conduit for transforming relations. At the same time, as on the seas, the framework’s appeal lay in its ability to support multiple interpretations at the same time. An alliance between equals could, when viewed from a different perspective, represent a tributary relationship with clear overlord and subordinate. In many cases, this kind of ambiguity was deliberately inserted from the beginning. When concluding what they regarded as vassalage agreements, the Portuguese in Goa made use of the term ‘friend’ to smooth the way and conceal a hierarchical relationship.Footnote 59
In the Kongo, the relationship between Afonso I, the ruler of that kingdom (r. 1509–42), and the Portuguese crown was infused with ambiguity. An incident in 1516 illustrates the way in which fragmented Portuguese authority in the Kongo combined with formal recognition of the suzerainty of the Portuguese crown to create a space for Afonso’s legitimate rule. Portuguese agents were feuding over claims to local power, and the Portuguese envoy Álvaro Lopes killed another Portuguese agent, Diogo Fernandes, who was claiming judicial authority. The accused murderer, Lopes, took sanctuary in the church, but Afonso extracted him and packed him off to Portugal for trial.Footnote 60 The episode tells us something about the limits of Portuguese royal authority in the Kongo, where the power to protect Portuguese subjects – from Afonso or from Portuguese rivals – was weak and unreliable.Footnote 61 More tellingly, the incident showcases Afonso’s strategy of representing the royal houses as equals and characterizing his relationship to Portugal as one of alliance. He referred to the Portuguese king as irmão (brother), and called on him to act to contain the violence of his own Portuguese subjects. Yet, as adept as Afonso, the Portuguese monarch responded with language emphasizing his superiority to the Kongo king. Here the proxy for protection was tutelage. Royal letters to the Kongo outlined for Afonso how diplomacy should be conducted, and addressed him as a subordinate.Footnote 62 Who was protecting whom? It depended on who was describing Portuguese–Kongo relations.
Similarly, in Spain’s conquest in the New World, protection represented a useful framework, one that characterized pragmatic alliance-building as a process preparing the way for Spanish political dominance. For example, protection talk narrated the fluid and shifting power in relations between Spaniards and Tlaxcalans in sixteenth-century New Spain. It is commonplace to observe that Hernando Cortés would not have succeeded in taking Tenochtitlán if not for the actions of the Tlaxcalans. Historians typically refer to the relationship as an alliance: Spaniards relied on the aid of Tlaxcalans in defeating the Aztecs, regarded as common enemies. Structures of composite political authority of both Spaniards and Tlaxcalans made sense of the idea of an alliance between groups of different political and military strength, allowing both sides to avoid talk of conquest in the context of their relationship.Footnote 63 But mutual understanding is not the point. Spaniards and Tlaxcalans differed sharply in their take on the nature of the alliance and how long it would last. If the alliance resulted from a pact – a security arrangement with mutual benefits – then it might dissolve (or evolve) once the object of war – the defeat of the Aztecs – was realized, but the relationship might also endure. Tlaxcalans could angle to preserve the alliance while recognizing that they were under the protection of the Spanish crown; Spaniards could argue that Tlaxcalans had assumed a subordinate place within the Spanish empire, even as they retained elements of self-rule.
The ambiguities meant that no clear interpretation prevailed quickly. In 1529, the Spanish crown granted Tlaxcalans an exemption from being assigned an encomienda, an act that affirmed their status as free vassals.Footnote 64 Tlaxcalan vassalage still provided space for autonomy. Tlaxcalan elites fought to retain recognition of their political community as a municipality, even travelling to Spain to make a case for their right to a direct link with the monarch rather than a relationship mediated through Spanish overlords to whom they paid tribute and services. The arrangement established a form of self-government utilizing Spanish institutions.Footnote 65 Here Tlaxcalans used the language of protection to manoeuvre for degrees of autonomy. The crown, in turn, used the framework of protection in efforts to shelter Tlaxcalans from the worst abuses of Spanish colonists.Footnote 66 Here we see protection’s range of meanings framing a political field of options that no-one considered mutually exclusive: the preservation of limited indigenous authority, the subordination of indigenous political communities, and the regulation of the power of imperial agents.
Such examples may suggest that protection functioned simply as a cover for imperial aggression. But the interpolity dimensions of protection were sometimes relatively stable over long periods and vast spaces. Less powerful polities often helped to perpetuate competing networks of protection by shifting loyalties between overlords. In Cambodia, a weak state marooned between two more powerful polities, the lines were so blurred that local rulers were able to perform the apparently impossible balancing act of claiming political independence while also seeking shelter from two separate protectors – Siam as ‘Father and the Vietnamese [emperor as] … Mother’.Footnote 67 The result was a string of contradictions concealed within a basic imprecision about the nature of protection, an arrangement that enabled Cambodia to function, in the words of one contemporary, as an ‘independent country that is a slave of two’.Footnote 68 Part of the reason why a state such as Siam could tolerate such imprecision in its relations with less powerful neighbours was that it made use of a similar ambiguity in dealings with its putative overlord, Ming China, to which it offered tribute in return for protection. This ambiguity found perhaps its clearest expression in a single term, jingong, a Chinese word meaning ‘tribute’ and the term used to describe the relationship between Siam and China. While Siamese rulers referred to jingong in documents dispatched to China, they avoided the Thai equivalent in Siamese royal letters and opted instead for the transliterated word chim kong, which lacked a precise meaning or clear associations with subordination. This sleight-of-hand allowed the Siamese kings to present themselves within Siam as brother monarchs of an independent country, rather than as subservient vassals.Footnote 69
Imperial centres also recognized the ambiguous status of semi-subordinate polities over which they claimed protection. The Tokugawa regime in early modern Japan cultivated a system in which the ‘roles of inferiors … could simultaneously express both “autonomy” and “complete subservience”’.Footnote 70 The contradiction was so pronounced that some domains in the archipelago could appear as utterly subordinate political units in one moment and as virtually independent countries in the next. While this incongruity has flummoxed generations of scholars, who have veered between radically different assessments of the power of the Tokugawa regime, the ability to tolerate these contradictions appears to stem from a crucial but largely neglected distinction between two key concepts: uchi, the (often) hidden inside, and omote, the outside façade presented to the world.Footnote 71 The Tokugawa regime’s governing dynamic relied on keeping these two categories separate, allowing the domains considerable autonomy within their own spaces of operation as long as they preserved the outward illusion of perfect subservience in their dealings with Edo. What this arrangement meant in practice was that domains made use of two separate languages, one for internal consumption within the domain and a second for interactions with the central regime.
Protection talk reflected – and preserved for a long time – a similarly useful ambiguity in relations between the EIC and Indian rulers. The Company had inserted itself into the subcontinent by accepting the protection of the Mughal emperor. This relationship was solidified by the Mughal emperor’s appointment of the EIC as Diwani of Bengal, Behar, and Orissa in 1765, an arrangement that established the Company’s right to collect taxes in those territories. Along with the rights to revenue came the obligation to protect, and the Company was charged with maintaining ‘a large Army for the protection of the Provinces of Bengal’.Footnote 72 The EIC also entered into an influential pact with the Nawab of Awadh in which, in exchange for the Company’s pledge to defend Awadh, the Nawab agreed to pay for the military services of the EIC. The treaty inaugurated a ‘new pattern’ in which Indian rulers found themselves ‘struggling to maintain their authority and to meet the imperious demands of the East India Company for revenues to pay for their armies’.Footnote 73 The British empire in India was stitched together in a combination of militarism and diplomacy, as a series of treaties with Indian polities that combined the cession of control over external affairs by Indian states with an obligation for them to pay for the Company’s military presence.Footnote 74
The language of protection continued to permeate Anglo-Indian politics. By 1803, with power relations shifting, the EIC could assert that it was placing the ‘person, family and nominal authority’ of the Mughal emperor ‘under the protection of the British Government’.Footnote 75 What precisely this meant was still unclear, and English officials described the relationship in different terms depending on their audience.Footnote 76 The logic of protection even framed the system’s rupture. Following the Indian Rebellion of 1857, the Company placed the emperor, Bahadur Shah II, on trial for treason as a ‘subject of the British government of India’.Footnote 77 His subjecthood was declared to have arisen when ‘his grandfather, Shah Alam, after having been kept in rigorous confinement by the Mahrattas, on their defeat by the English in 1803, applied to the British Government for protection’. After that moment, ‘the titular kings of Delhi became pensioned subjects of the British’.Footnote 78 This reading of the history removed any trace of ambiguity about protection’s meaning as alliance rather than subjugation.
Although protection was an indispensable tool for European expansion, we should be wary of assuming that it always functioned as a vehicle for growing European power. If protective networks organized European imperial ambitions across the Americas, increasing European authority within such webs was hardly certain or predictable. It is true that Europeans often boasted to metropolitan, settler, and native audiences of their rights and responsibilities as overlords. British and New York officials brought Mohawk ‘kings’ to London in 1710 as embodiments of the British suzerainty in New York more than a decade before signing the 1722 Treaty of Albany.Footnote 79 Across eastern North America, Europeans encouraged American Indians to adopt the fictional kinship title of ‘father’ or the native category of ‘greatest lord’ in referring to Europeans. Yet the political language of alliance persisted, with many recorded references by North American Indian groups to Europeans as ‘brethren’, a clear strategy to point to Europeans’ and Indians’ equality of status as allies and to support Indian claims for European gifts required to solidify pacts of mutual security.Footnote 80 In some circumstances, such gifts operated as tribute and recognized the greater military power of Indians, casting Europeans as purchasing protection.
Alliances were notoriously fragile. They broke down not only when colonists failed to deliver sufficient or proper gifts but also when networks of alliances functioned to spread violence, much like conduits carrying an electrical charge.Footnote 81 The volatility of relations of protection is illustrated by the Anishinaabewaki of the pays d’en haut, whose agents travelled to Montreal to participate in the signing of a peace pact in 1701 that would establish the French as arbiters of conflicts among Indian groups to the west. A short time later, Anishinaabewaki recognition of French protection disappeared as it became very clear that the French could not produce order even in French-controlled enclaves, much less in the vast middle reaches of the continent.Footnote 82
Promises of protection could not by themselves convert unstable networks of fragile alliances into polities we might recognize as fully fledged empires. Nor did slippery imperial rhetoric about receiving and providing protection always translate into increased influence. In Japan, for example, such notions proved nothing more than a dead end. Beginning in the 1630s, Dutch officials, exactly as their English counterparts were doing in India, began claiming that they were nothing more than loyal vassals intent only on serving the Tokugawa regime. However, rather than providing a springboard for increased influence, this rhetoric came to trap the Dutch, who found themselves encased instead within a rigid requirement to actually deliver on that service by, among other things, dispatching a vessel to aid in the suppression of a Christian revolt at Shimabara in 1637–38.Footnote 83 In the Americas, protection featured prominently in the persistence of unevenly bounded spheres of influence. What we call ‘empires’ in thickly contested borderlands took the form of unstable webs of alliances, in some cases assuming more definite outlines in direct response to conflicts over jurisdiction and protection.Footnote 84 Protection claims cast shadows of sovereignty, but there was a great deal of movement in the shadows.
Conclusion
This article has pointed to the prevalence of protection in cross-polity interrelations in the early modern world. In different regions, protection established lasting, if elastic, frameworks within which relationships of political authority and jurisdiction combined and recombined. Rather than resolving into different types of protection systems – domestic licensing agreements or mafia-like protection rackets imposed on vulnerable outsiders on the seas, or relationships of alliance or conquest on land – empires activated a range of meanings of protection in responding to local circumstances and the political strategies of their interlocutors. Protection operated as a flexible framework for interpolity relations more than simply as a tool of imperial consolidation.
The key to protection’s function in structuring powerfully formative conflicts lay in the blurred divide between ‘inside’ and ‘outside’ protection. Parties to agreements about protection might describe them as voluntary arrangements and treaties between autonomous polities. At the same time, sophisticated actors on all sides knew that alliances did sometimes slide into configurations of unequal power. Both imperial agents and subordinate groups found reasons to associate protection with the recognition of subjecthood. A transition toward an ‘inside’ variant of protection in which it figured as a necessary good provided by a legitimate sovereign was not necessarily irreversible. Political communities absorbed into empires had reason to maintain the accoutrements of rule and their own claims to be providers of protection.
The resulting webs of protection structured both interpolity and intra-imperial relations in the early modern world. The institutional forms associated with protection arrangements changed, but the general framework remained stable over a long period. Empires took shape within these configurations as irregular and overlapping areas of influence, abutting and interweaving with spheres of power associated with rival empires and an array of less powerful polities. We look in vain for a doctrine of protection recognized across an early modern law of nations. Yet we do find a coherent and repeating set of practices and utterances about protection that spanned culturally and legally disparate parts of the world and that gave shape to overlapping spheres of imperial control.
Protection was not the only framing device in a repertoire of practices that we label as ‘interpolity law’.Footnote 85 Protection talk joined a widely shared focus on protocol for marking authority, pervasive jockeying over jurisdiction, and positioning with regard to possession to compose a recurring and recognized framework for organizing relations across very different polities in the early modern world. This is not to say that there was anything approaching a universal law of nations.Footnote 86 Rather, the existence of parallel systems of political organization meant that diverse political communities shared some basic characteristics that allowed legal actors to rely on analogies, rather than elusive cross-cultural understandings, as they engaged with other peoples and polities.Footnote 87
The meanings of protection were multiple but finite. The blurring of ‘inside’ and ‘outside’ registers of protection was widespread and often purposeful, the result of observable strategies. Yet early modern imperial agents and others invested protection talk with specific alternative meanings in arguments over property, subjecthood, and authority. The political configurations that emerged through the politics of protection held vast regions in thrall. In many ways the early modern world, on land and sea, was an age of protection.
Lauren Benton is Nelson Tyrone, Jr. Professor of History and Professor of Law at Vanderbilt University. Her publications include Law and colonial cultures: legal regimes in world history, 1400–1900 (2002), which received the World History Association Book Prize and the J. Willard Hurst Prize in Legal History; A search for sovereignty: law and geography in European empires, 1400–1900 (2010); and most recently, with Lisa Ford, Rage for order: the British empire and the origins of international law, 1800–1850 (2016).
Adam Clulow teaches at Monash University. His first book, The Company and the shogun: the Dutch encounter with Tokugawa Japan (2014), received the Jerry Bentley Prize in World History from the American Historical Association, the International Convention of Asia Scholars Humanities Book Prize, the Forum on European Expansion and Global Interaction Book Prize, and the W. K. Hancock Prize from the Australian Historical Association.
