Arming Slaves in Early Modern Maritime Asia

Abstract

While there are large literatures on both Islamic slave soldiers and the phenomenon of “arming slaves” in the Atlantic world, military slavery in early modern Asia is still poorly understood. Using a variety of Chinese, Latin, Spanish, and Portuguese sources, this article will argue that enslaved labour was frequently directed towards violence across early modern Asia, colonial or otherwise. At the same time, the phenomenon was far from uniform in the vast expanse of land and sea between East Africa and Japan. Rather, it is better to speak of a series of analogous regimes of bondage that interacted with each other across large distances, with the line between enslaved soldiers, mercenaries, and run-of-the-mill trader-raiders being vanishingly thin at times. Finally, all this existed within the context of military infrastructure in the broadest sense of the word that included fortresses, factories, and even war elephants.

Introduction

In his vast unpublished manuscript treatise on slavery in Asia, the Jesuit missionary Gomes Vaz (ca. 1542–1610) devoted several pages to the characteristics of certain slaves:

Not all [slaves] are the same. Rather, some excel others and can be said to be superior and more deserving than these others. For this reason, Cicero said in his Stoic Paradoxes that “steward slaves” were the most valuable. However, in our Indies (not that I am talking about anything else), the foremost, most outstanding and spirited in war and most disciplined in battle are the Japanese and Koreans. If you make an armed force of them and mix them with locals, they will no doubt be similar to the soldiers that Caesar Augustus chose in the Illyrian and German wars and called “volunteers” [i.e., citizen soldiers, not auxiliaries], and those that the Roman people, driven by extreme necessity and taking unusual measures, sent against Hannibal when he was laying waste to Italy. Indeed, this was shown well enough by the Japanese at the siege of Malacca when under André Furtado de Mendonça the Portuguese stronghold was completely surrounded by enemy forces. Japanese merchants happened to be there, having put ashore to do business, and although they were few in number (less than one hundred) the commander had at his disposal such power to assail the enemy that when the Japanese attacked their forces, they routed them and killed many in just one charge. I believe Abyssinians [i.e., Ethiopians] to be their equals except that Abyssinians are more faithful and serviceable to their masters than all other slaves, and you can safely entrust household affairs to their women who are not inferior to Portuguese women in this regard.¹

While elsewhere in his treatise Vaz states that slaves were purchased, gifted, and exchanged in Asia to undertake domestic, agricultural, and manufacturing-related labour, this passage suggests that the use of slaves for military purposes was also common. There are even surviving accounts of slaves or former slaves engaging in violence on behalf of their owners and employers, both European and non-European, and for self-defence. Of these, the most famous are probably the unnamed black slaves who were crucial to the defence of Macau against the Dutch in 1622.²

Of course, military slavery in general has long been known to historians, and has been closely studied in the case of Islamic military slavery (often referred to as the mamlūk system), especially in the case of the famous Janissaries who served as praetorian guards to the Ottoman and other Islamic emperors.³ This system extended into the Indian Ocean world with tens of thousands of slaves from northern India being forcefully integrated into the armies of the Ghaznavid Empire and other polities.⁴ East African slave soldiers also fought for the Muslim sultanates of the Deccan, and some even rose to positions of considerable prominence, most notably in the case of the Siddi prime minister of the Ahmadnagar Sultanate, Malik Ambar (1548–1626).⁵ As a result, Christopher Brown, Philip Morgan, and their collaborators concluded in an influential volume entitled Arming Slaves that military slavery was a widespread phenomenon in the Islamic world, West Africa, and the early modern Atlantic.⁶ Slaves fought in interimperial wars in the Caribbean and were frequently (although illegally) armed for the purposes of both defence and expansion in Brazil.⁷ In Europe, there is widespread evidence of slaves being used to protect and violently advance the agendas of their masters in urban contexts, and of slaves being used in naval warfare, most notably in the Mediterranean galleys.⁸ In his essay in the Arming Slaves volume, John Thornton underlined the number of soldiers in the West African kingdom of Ngondo who had a low status equivalent to slaves.⁹ All these examples are in addition to closely related phenomena, such as the militias of manumitted and free blacks in eighteenth-century New Spain, and the many examples of slaves engaging in armed resistance against those with ultimate control over them in the Atlantic world and beyond.¹⁰

Nonetheless, military slavery in Asia remains very poorly understood. In the particular case of Iberian Asia, the only focused treatment is a suggestive article by Timothy Walker, who has argued that the fundamental difference between Portuguese and Islamic military slavery in Western India was that the former used slaves for defence against external aggression while the latter had professional slave troops who were involved in a variety of military engagements.¹¹ Yet, it remains unclear how common enslaved soldiers actually were in Portuguese or other European forces in Asia.¹² Furthermore, since military historians have argued that there was relative technological parity and constant mutual borrowing between European and Asian merchants and militaries, at least until the mid-eighteenth century, the (as-yet-unanswered) question naturally arises of whether military slavery in the Portuguese Empire and Ming and early Qing China were likewise linked.¹³ In other words, vociferous it remains to be investigated whether there was interplay between conceptions of coerced labour among the polities and peoples of early modern Asia (e.g., Christian European, Chinese, Islamic) who we know certainly exchanged goods, ideas, and technology.

Answering such questions is made particularly difficult by the thorny definitional issues that haunt all histories of slavery. In fact, it is often extraordinarily difficult to draw a clear line between slave soldiers and the large numbers of poorly documented mercenaries (Portuguese, Turkish, Japanese, Indian, etc.) who are known to have roamed maritime Asia, taking advantage of any opportunity that presented itself in the period's constant process of trading, raiding, and empire building.¹⁴ At the same time, it is not always obvious how to categorise enslaved camp followers, enslaved people involved in the construction of fortifications, and enslaved sailors on ships with guns, all of whom were essential to the large-scale violence that permeated life in early modern Asia, without being combatants in the strict sense of the word.¹⁵

To shed light on these issues, this article offers a view from early modern maritime Asia on the question of what Christopher Brown and Philip Morgan call “arming slaves.” In so doing, it will make two related arguments using a variety of published and unpublished sources in Chinese, Portuguese, Spanish, and Latin. First, it will show the widespread use of coerced violent labour beyond the familiar contexts of the Atlantic world and the Ottoman Empire, including in the South China Sea. Some of these applications were adaptations of Islamic mamlūk slavery. However, there were also autochthonous forms of coerced violent labour as well as the importation of Mediterranean customs, with constant interactions between roughly analogous elements of these various regimes of bondage. This conglomerate of forms was then directed to serve the interests of states, companies, and powerful individuals who all wanted to bend others to their will.

Second, the article will argue that military slavery in the narrow sense (i.e., combatants who were placed in a subordinate and alienable socio-legal category broadly associated with property) overlapped significantly with related phenomena, including mercenary armies, conscription, press-ganging, slave bodyguards, and allied troops. Combat would also not have been possible without the free and unfree labour that manned ships, built fortresses, and produced weapons, gunpowder, and food. All this should be considered part of the same larger process of building war machines in the age of war capitalism. As such, this article contributes to the ongoing project of understanding the myriad encounters between regimes of bondage that characterised the early modern world market in human beings. These larger arguments will emerge from the article's three case studies that will treat in turn “black” military slaves in maritime East Asia, the reach of Japanese military slavery, and the interactions in Southeast Asia between different models of bonded violent labour.

“Black” Military Slavery in China and the South China Sea

Early modern China and the South China Sea saw the mobilisation of an array of coerced violent labour, particularly during the unstable final years of the Ming state and the subsequent Manchu conquest. In the European colonial case, the most quoted example of this is probably the 1622 defence of Macau. This was necessitated by a large-scale Dutch attack from the sea, during which, various accounts suggest, black slaves (escravos negros) played an important role in holding back the eight hundred–strong landing force that came ashore at the foot of the Guia hill. These unnamed slaves were probably a mixture of East Africans, Malays, and Indians brought to Macau as part of wider patterns of Portuguese maritime trade from Mozambique, Malacca, and elsewhere, and perhaps included some of the Abyssinians mentioned by Vaz, whose martial skills were also emphasised by the unknown Indian artist who produced the Codex Casanatense (Figure 1).¹⁶ Indeed, the role of “black” fighters (with or without direct reference to their status) is mentioned in passing by a range of European- and Asian-language sources. For instance, the VOC historian François Valentijn (1666–1727) records the Dutch–Johor assault on Portuguese Malacca in 1606, when the city was defended by “400 Portuguese and black soldiers armed with muskets and pikes.”¹⁷ This echoes the Sinhalese poetic chronicle, the Rajasinha Hatana:

Thus, like the Full Moon in the midst of Stars,

King Rajasingha, with his royal brother and the courter train, did end the war.

Destroying the hosts of Portugal, but sparing the Padres,

With the blacks and the double-natured Thupassis, for mercy's sake.

And so, the proud enemy was crushed and the glory of our King blazoned forth.¹⁸

Figure 1. Abyssinians from the Codex Casanatense (Rome, Biblioteca Casanatense, cod. 1889, 3–4).

This passing reference, tucked in between luxurious tributes to Rajasingha II of Kandy (r. 1636–87) and attacks on the Portuguese and the Luso-Portuguese Thupassis (topazes), is a reminder that the events in Macau in 1622 were doubtless repeated countless times in sieges along the Portuguese trade route in the sixteenth and seventeenth centuries.

Nonetheless, the 1622 defence of Macau remains the best-documented case of enslaved blacks in action. For instance, a Dutch source claims that the slaves had been plied with alcohol and opium, while an observer writing in Spanish referred to them as “drunk,” which he used to explain their fearlessness in the face of such overwhelming odds, and noted that the Dutch got away lightly as the black defenders and other servants (gente de servicio) lingered to behead the wounded and despoil their bodies. The Jesuit chronicler Fernão Queiroz also talks about a female black slave who rushed at the enemy wielding a long-handed axe called a halberd.¹⁹ This and other acts of derring-do and the death of several slave defenders were acknowledged and rewarded not only by Portuguese observers, but also by the Chinese provincial admiral, who was reportedly so impressed that he donated two hundred piculs of rice to the slaves.²⁰

In addition to these intriguing details, there is much more to this story than is usually assumed. Whereas Boxer focused on the dyad of black slaves and Portuguese soldier-merchants involved in the defence of Macau, the reality is that the much-lauded Portuguese office holders and black slaves fought alongside various other combatants. This included not just what remained of the free Portuguese and Eurasian population, who had stayed behind during the period of the Canton Fair and following the Portuguese accession to a Ming request for troops to fight the Manchu, but also alongside unidentified Japanese defenders. This group was probably a combination of Japanese merchants, émigré Christians fleeing Hideyoshi's persecution, and slaves who had been purchased by Portuguese and other merchants in Nagasaki and transported to Macau. Importantly, in fighting alongside Japanese men of various statuses, Macau's defenders were not unlike their Dutch adversaries. On the way to Macau with their Malay and Bandanese crews and accompanied by two English ships that wanted to observe the events, the Dutch fleet encountered a Siamese junk. This was manned by a mixture of Siamese and Japanese seamen who sensed an opportunity and asked to join the armada that by then consisted of 1,300 men.²¹ What proportion of this total were slaves in any sense of the word is unclear. However, some proportion were almost certainly enslaved, as it is well known that Southeast Asian merchants purchased Japanese people in Nagasaki during the Sengoku period, while early modern Siam featured the widespread use of coerced domestic and agricultural labour that likely spilled out into maritime Asia.²² In other words, the much-discussed 1622 defence of Macau by its dark-skinned enslaved inhabitants was embedded in a larger system of coerced and transient labour that was regularly put to violent ends.

Furthermore, while it is true that Macau did not have a slave garrison per se, as the Portuguese did not use slaves as permanent professional soldiers, Macau's slave population was seen as an integral part of its military infrastructure, since the city usually had fewer than thirty soldiers sent from Goa in the period to defend it.²³ Most of these were low-status Eurasians and European- and Asian-born Portuguese, including some criminals who had been given a reprieve from penal slavery in the Mediterranean galleys (or in auxiliary naval work like rope-making) and exiled to Asia.²⁴ As a result, when noting the slave population of Macau (around six per casado household for a total of around five thousand in 1634), António Bocarro identified them specifically as “armed slaves.”²⁵ This did not mean that every single slave could be expected to wield a pike or a harquebus (especially as there was considerable diversity among the city's enslaved population), although many of those who already served as bodyguards to their masters were probably quite adept with such weapons. Rather, they formed a workforce that could contribute to fighting in a variety of ways. For instance, in Goa we know that slaves were put to work in the gunpowder factory, a task also commonly assigned to convicts, again showing the blurring of the distinction between enslaved combatants and other slaves, as well as between different types of coerced labour.²⁶ Indeed, both offensive and defensive fighting required not only men (and on some occasions women) to fight, but also weapons, gunpowder, food, functioning ships, sturdy fortresses, and in many cases long supply chains, all of which relied on a mixture of free and variously coerced labour depending on what could be mustered at that moment.

In considering the encounter between different regimes of bondage in early modern Asia, it is also important to note that some dark-skinned slaves of the sort documented in Macau and Malacca were either sold to Chinese buyers or fled their masters and found their way into China proper.²⁷ For instance, the Ming scholar-official Cai Ruxian (蔡汝賢) recorded in his 1587 Portraits of Eastern Barbarians (東夷圖説):

[The black slaves] are supremely strong—each of them can carry loads up to several hundred jin [i.e., over one hundred kilograms]. They will not flee even when facing the enemy, and can survive for one or even two days underwater. Some [Chinese] generals have purchased them as vanguard soldiers, and they are worth the expense. If they are matched with Chinese women, their children are also black. [The children] can learn our language after a long period of training, but [the parents] cannot. They are used by all foreigners just like slaves (nupu) in China. Some say they are braver than the white (白, bai) barbarians.²⁸

Cai expresses no surprise at the Chinese acquisition of such slaves and their use in military exploits, since exchanges in military technology and manpower between the Portuguese and officials of the Ming state were common in this period: Ottoman, Portuguese, Dutch, and English cannons found their way into an increasingly beleaguered Ming China; Jesuit missionaries were called upon as military engineers; and Portuguese soldiers and gunners were to be found on both sides of the Ming and Qing conflict.²⁹ Similar dynamics are documented in the kingdom of Arakan in the Bay of Bengal.³⁰ Today, we might be tempted to call some of these soldiers “mercenaries.” However, the term implies a default situation in which “national” professional armies face off against each other in international conflicts. In the premodern world, such a scenario was simply unimaginable. Indeed, the premodern decentralised and profit-driven model of warfare may have been more efficient and contributed directly to the military revolution in Europe.³¹

While it is hard to find further evidence for Chinese generals purchasing slaves from the Portuguese, we do know that a generation or so later the pirate general Zheng Zhilong (鄭芝龍) had a battalion of dark-skinned soldiers in his retinue under the leadership of one Luís de Matos. How exactly these Christian troops—who are said to have advanced to the traditional Reconquista battle cry of “Santiago!”—entered Zheng's service is unclear. The bishop of Puebla in New Spain, Juan de Palafox y Mendoza (1600–59), called them runaways (fugitivos) and it is quite possible that they had fled the tiny city of Macau, in the same way as Mozambican slaves who had escaped from Goa fought for the Dutch in Vengurla (just north of Goa) in the 1670s.³² At the same time, given the amount of time that Zheng spent in Macau, where he was baptised, it is not impossible that he purchased (or maybe hired) them there, or during his myriad other interactions with Portuguese and Dutch merchants in Macau and Japan.³³

In any case, Zheng's bodyguards are described by various observers. For instance, Palafox recounts how the corps of more than three hundred “black slaves” was reduced to around two hundred in the course of fighting the Qing. After Zheng's capture in 1646, his former bodyguards entered the service of the new dynasty in Anhui, where in 1649 the Spanish Franciscan Antonio Caballero records the presence of “some blacks from Macau, Christians and [now] soldiers of the mandarin,” presumably Li Lütai (李率泰), the viceroy of Min-Zhe and later of Liangguang. Although little evidence remains of these soldiers’ subsequent exploits, Palafox notes that they remained good Christians, abstaining from meat during Lent even when it was offered to them as a reward for outstanding service against pirates. Whether this is more hagiography than history is unclear.³⁴ Zheng's son, Zheng Chenggong (鄭成功), better-known today as Koxinga, also had two black battalions who were said to be slaves that had once belonged to the Dutch. Clearly, then, the use of slaves or former slaves as soldiers, including as full-time praetorian guards, was not limited to the mamlūk slaves of the Indian Ocean, but was mirrored in the connected world of the South China Sea, where they frequently entered into personal relationships with the powerful of the sort recently described by the Islamic jurist Bernard Freamon, as rooted in “loyalty as a tool of governance.”³⁵

Further examples of the encounter between regimes of bondage can be found in Korean sources. For instance, in relation to the Ming naval reinforcements that arrived in May 1598 to help push back Hideyoshi's second invasion, the authors of the Veritable Records of the Joseon Dynasty, writing in classical Chinese, recounted this conversation between a Ming admiral and King Seonjo of Joseon (1552–1608):

There was a seaman whose eyes were yellow. The face, hands and legs were all black. His hair was curly, like the wool of a black sheep. Yet, the top of his head had no hair at all, which he covered with a tied-up yellow cloth resembling a flat peach. Moreover, he could dive into the water and stay there for several days. [Such things are] rarely seen even by the people of the Central Plain [i.e., the Chinese]. This [Korea] is just a small place far away from the centre, how is it we came to see such an incredible soldier? Because of you, I was able to do so and am grateful to witness such a miracle. It seemed that our victory was now imminent.³⁶

Unlike Ming accounts, which usually explicitly mention that these seamen were enslaved (nu, 奴), there is no suggestion of his status in the Korean source. It is also not guaranteed that he had served previously on a Portuguese vessel. Yet, the parallels with accounts from Macau and the lack of an obvious alternative origin are suggestive of a previous transoceanic journey, perhaps from East Africa to Macau, or another southern Chinese port city.

Another important caveat is that such long-distance links do not mean that enslaved or recently freed military labour was conceptualised in the same way throughout maritime Asia. For instance, while the Portuguese seem to have lacked a formal category equivalent to the Islamic mamlūk soldier, the general assumption seems nonetheless to have been that slave labour could be put to a variety of violent ends, especially (although not necessarily exclusively) within the context of defence. This therefore became an important consideration when acquiring slaves. In contrast, when these slaves either escaped or were sold into the Chinese military system, they presumably became “housemen” (家丁, jiading): private vanguard troops of generals. This was in contrast to hired soldiers and hereditary garrison troops. The latter had been the default arrangement for imperial defence since the Yuan, but was slowly falling out of favour due to the monetisation of the Ming economy in the wake of the flood of silver from the Americas, resulting in an increasingly large body of fighting troops who owed their primary loyalty to the local commander who hired them using imperial funds, rather than directly to the empire.³⁷

Here, the fact that the Christian black slaves (黑奴, heinu) were not Chinese (唐人, tangren) was not of itself that unusual, as the hiring of foreigners (Mongols, Jurchens, Japanese, etc.) as combatants was common in Ming armies. These foreign soldiers were prized for their specialist military skills such as archery, horse riding, and mountaineering. Most common along the northern border, but not absent elsewhere, they were assigned to specific battalions according to their origin and assigned a Chinese commanding officer in a way reminiscent of Britain's later Indian Army. There was also a cross-border market in military slaves in the north, with the Jurchens buying Ming soldiers and Ming buying Jurchens. This would be critical to the rise of the Manchu. The subsequent victory of the Qing also likely did little to change this situation, as their famous Banner System was rooted in Turko-Mongolian military slavery that drew strength from integrating defeated enemies in a subordinate status and regularly featured chattel divisions.³⁸ It was hardly much of a leap then to integrate another group of degraded or desperate foreign combatants into this pre-existing structure.

Defining Japanese Military Slavery

The various enslaved and formerly enslaved black combatants and seamen documented in Macau, Ceylon, Korea, and elsewhere show the porous nature of military slavery in maritime Asia. At the same time, the vagueness in many sources about their exact status and their apparent mobility between different regimes of bondage underlines a recurring problem in the history of slavery: the question of definition. Indeed, much hinges on whether particular groups are categorised as military slaves, mercenaries, conscripted soldiers, press-ganged sailors, or native allies. This very same definitional question also arises with another diasporic group: that of the numerous Japanese men observed fighting in South China and across Southeast Asia in the sixteenth and seventeenth centuries. Certainly, there is evidence of Japanese slaves being used for defence in Portuguese cities in Asia. When Philip III attempted to republish a 1570 law against Japanese slavery in Goa, the city council responded that they would be left effectively defenceless against the Dutch.³⁹ Furthermore, many of the antislavery edicts promulgated by both European and Asian powers in the period were actually related in one way or another to the issue of Japanese military slavery. For instance, in 1613 Chinese authorities prohibited the purchase of Japanese slaves in Macau in these terms:

Banning the rearing and keeping of Japanese slaves: if old and new barbarian traders still dare to rear and keep Japanese slaves and let them hitch a ride on foreign trading ships as before, whosoever experienced such in the past is allowed to come and report; and [those law-breaking traders] will be strictly punished according to the military laws; but if they do not come forward, they will be severely punished.⁴⁰

This edict was a reaction to the large number of Japanese slaves in the city, whom the Chinese authorities saw as a potential source of instability, given the perceived connection between the Japanese (倭寇, wokou) and piracy, with the Chinese viceroy famously arguing that in any case the Portuguese had no need of Japanese labour when they had black slaves.⁴¹ This said, despite the decree's primary focus on slavery, many of the potentially armed Japanese in Macau that so concerned the Ming officials were not slaves at all (as well as not even being Japanese in some cases, as there was frequent conflation of overseas Chinese involved illegally in the Japan trade and actual Japanese people). For instance, the incident that triggered the official intervention in Macau involved the attempted theft of a ship by a group of armed Japanese merchants who had arrived on a vessel belonging to the daimyo of Arima.⁴² Similarly, in Japan the 1621 partial reiteration of Hideyoshi's 1587 antislavery proclamation was directed at stopping both the hiring and the purchasing of Japanese by Dutch merchants who transported them to Southeast Asia, where they were put to work in various different roles.⁴³ The distinction between Japanese military slaves, mercenaries, and armed contract labourers was therefore not always particularly sharp for contemporary observers.

A further complication is that the arming of slaves in cities frequently occurred outside the contexts of both state-sanctioned warfare and urban defence against external enemies. In 1596, a Portuguese edict banned armed slaves from walking the streets of Cochin alone. Whereas protecting their master's person was clearly acceptable (provided they were actually present), unsupervised slaves were a liability. This was likely a response to the fact that throughout the Iberian world slaves were regularly accused of committing crimes on their masters’ behalf and were regularly described as sneaking out to rob, rape, and murder for their own purposes. The risks, as well as the benefits, of arming slaves were therefore of considerable concern to authorities, since this could result in uncontrollable private uses of force, as well as ultimately the most feared of all: slave revolts.⁴⁴

As previously mentioned, not all armed Japanese in Portuguese cities, even those advancing the interests of Portuguese individuals or institutions, were slaves. Far from it. As Lúcio de Sousa has noted, when setting the punishments for Japanese found to be carrying arms in Macau, the Habsburg authorities made a distinction between what we might be tempted to call slave soldiers and employed mercenaries:

No Japanese of any rank and condition whatsoever shall reside in the said city of Macau or be held outside it, or may they, or any slave of any other nation, freed or captive, bring or be able to carry a katana, large or small, even in the company of their master, on pain of all those things that are found under it in the form of this law of mine: if they are a slave and have a master they will be sentenced to captivity in my galleys in India for life, and if they are free they are to be exiled for ten years to the same galleys.⁴⁵

While the document here presumed some sort of dependency (in the form of either ownership or employment) on a Portuguese or other Christian master over whom the Habsburgs might hope to claim some sort of authority, it is important to note that this was far from always the case. Many Japanese combatants in Iberian cities were entirely free agents, including the “merchants” that Vaz mentioned took up arms during the 1606 siege of Malacca.⁴⁶ In a recent study of Malacca in this period, Paulo Pinto concludes that the Japanese combatants who significantly swayed the outcome in favour of the defenders were “mercenaries,” in the sense of individuals who were not considered property and engaged in violence of various sorts to support themselves. This group later disbanded and was paid by the Portuguese factor before disappearing from the records, suggesting that they were an independent force of some sort who went off to seek opportunities for profit and plunder elsewhere. Similar dynamics are also observed in Dutch Batavia.⁴⁷ This sort of Japanese soldier of fortune may even have been the model for the striking image preserved in the Boxer Codex of a man wearing Japanese armour and carrying both a katana and a matchlock musket (Figure 2).⁴⁸ These were also likely the troops considered essential to the ultimately unrealised Habsburg plan to invade Ming China, and the sort that populated the Japanese militias found across maritime Asia, whose plundering of ships and cities led to irate letters being sent (including from Spanish Manila) to their indifferent sovereign in Edo.⁴⁹

Figure 2. Japanese soldier, (Boxer Codex, fol. 91v.,Lilly Library, Bloomington, Indiana). Used with permission.

Finally, even as the bulk of the evidence supports Timothy Walker's proposition that slaves in Asia were employed in a defensive capacity (especially in the case of the Portuguese), the offensive use of Japanese soldiers, some of whom are branded as slaves in contemporary accounts, nonetheless also crops up from time to time. For instance, there were around five hundred Japanese soldiers in the service of the kings of Ayutthaya in the late sixteenth and early seventeenth centuries, the most famous being Yamada Nagamasa (1590–1630).⁵⁰ While many of these were trader-raiders (among whom were numerous émigré Japanese Christians) who amassed considerable wealth and influence, others apparently occupied a status that was closer to that of the slaves owned by wealthier Japanese in Ayutthaya. For instance, the Dutch merchant Pieter Willemsz van Elbinck (also known as Peter Floris, d. 1615) clearly describes a group of Japanese “slaves” who took revenge on the court of King Si Saowaphak of Ayutthaya following the murder of their master.⁵¹ While evidence is very thin on the ground, some of these may have been purchased from better-off Japanese merchants, but it is important to remember that Siamese merchants are also documented buying people in Nagasaki, with the various prohibitions against selling Japanese slaves to foreigners targeting not only the Portuguese, but also the Siamese and Cambodians involved in the slave trade.⁵²

In sum, in this age of what Sven Beckert has dubbed “war capitalism,” it is wise to eschew reifying categories like “merchants,” “mercenaries,” and “slave soldiers,” and speak instead in the first instance of mobile populations of trader-raiders who engaged in commerce and piracy, to differing degrees according to their skills, resources, and current needs. While some came from notable merchant families in the Iberian Peninsula, many were runaway slaves, freedmen, or other people who had in one way or another fallen through the cracks of their societies and headed off to make their fortune by whatever means necessary. Many of these trader-raiders also owned what in European languages were frequently described as “slaves,” whose role in combat it is often convenient to call military slavery. At the same time, the division between those who in the historiography are often (although not always helpfully) referred to as slave soldiers, mercenaries, and formal allies is not always clear. Indeed, it is often difficult to know what to make of the many groups who fought alongside Iberian forces in Asia, whose origins, statuses, and intentions are often hard to reconstruct.⁵³ This is particularly true for the port cities of the Malay Archipelago, where different regimes of bondage collided, and which will be the subject of the final case study.

The Encounter Between Regimes of Bondage in Southeast Asia

In Malacca, following the conquest of the city by Alfonso de Albuquerque in 1511, work began on a new stone fortress complete with two wells and an imposing four-story keep. This impressive piece of military-commercial infrastructure, which far outstripped the previous wooden fort, was essential to the Portuguese control of the city that only ended in 1641 at the hands of a joint Dutch–Johor force. Importantly, its construction also made use of a very particular workforce. As the Portuguese chronicler João de Barros described, an elephant carried the heaviest objects and the former sultan's royal slaves did the rest. As regards the latter, he continued:

In the course of these works Afonso de Albuquerque took advantage of a subgroup of the Malaccan people called ambarages, which means “slaves of the king,” as indeed they belonged to the king who provided them with food rations; and if not, they earned it and took care of themselves, their wives, and children; the king had over three thousand such people. And because at the beginning of the project Afonso de Albuquerque knew who some of these slaves were, while others were still hiding in the forest, others in the plantations (dusuns), and still others in the city, their identity being unknown to him, he ordered that it be publicly proclaimed that every slave who belonged to King Muhammed, if he came to Albuquerque to ask for food, he would be spared and be granted the freedoms that he had before; and anyone who brought Albuquerque an escaped royal slave or if he presented someone who seemed to be a royal slave, then Albuquerque would give him as much. This proclamation was the cause of many free people becoming slaves, because men had an incentive to take poor free people from the plantations and forests and present them as slaves of the king. These people were then added to the list of royal slaves, thereby enslaving them, their wives, and children. And the worst thing was that many people did this to harm each other, denouncing someone as a slave with two witnesses, and there was no more miserable person than this, and the business of these ambarages was the cause of much harm, as will be seen.⁵⁴

Among other things, this passage reminds us that animals were also part of the military infrastructure of premodern Asia since they were able to move heavy objects for construction. This is in addition to their use in battle, which, as the Codex Casanatense makes clear, was known to the Portuguese (Figure 3).⁵⁵

Figure 3. Image of a war elephant (Rome, Biblioteca Casanatense, cod. 1889), 76–7.

Most of the passage, however, focuses on the unfortunate consequences of the encounter between the regimes of bondage of the Portuguese and the Malacca Sultanate in a moment of geopolitical transition. For some context, contemporary observers noted that in this part of the Malay Archipelago there was an array of terms to refer to subtly different types of public and private slaves that could sometimes also include Chinese, Dutch, and Portuguese seamen who had been captured and pressed into work, usually subsequently being circumcised and forced to convert to Islam. In the case of the hamba raja [ambarages] described by Barros, hamba referred specifically to “slaves” to higher powers, namely either to Allah, or, as here, to the king (raja). This class of slaves seems to have been present in various Malay sultanates, where they served as factors in trade, labourers in large-scale construction projects, and as bodyguards or executioners for their monarchs, gaining relative prestige, although not political power, in their societies.⁵⁶ Of the total slave population of Malacca reported by Portuguese observers in 1511, between 50 percent and 85 percent fell into this category.⁵⁷ While their status has been described as an extreme form of contemporary European vassalage or corvée labour, a more accurate analogy is arguably that they represent a Southeast Asian mirror of the mamlūk system resulting from the close commercial and diplomatic links between Malacca and the Ottoman Empire.⁵⁸

Admittedly, hamba raja was a more capacious category. It seems to have included both high status merchants and fairly low-status slaves who undertook backbreaking tasks for the sultan, and were frequently originally criminals who had preferred slavery to the sultan to capital punishment. They were also different from Ottoman Janissaries in that they were not a burden on the sultan's purse, since they were expected to devote half their time to supporting themselves. This has led to comparisons with vassals (vassali) or serfs (adscriptitii) in medieval Europe, in contrast to slaves (servi) per se, a fact also suggested by the patrilineal hereditability of the status. However, within their particular context they seem to have occupied a status different from even the most subordinated vassals in the region, such as the Moluccans whom António Galvão (d. 1557) reported were required to “besiege towns, to move temples, royal palaces, and other similar things” for their lord.⁵⁹ Furthermore, Vaz's treatise makes it clear that hamba raja could be bought and sold, with some at least ending up in the hands of Christian owners, since he discusses at some length the hereditability of the status on the basis of his personal experience in Malacca and Ceylon:

The royal slaves, of whom there is a great number among the princes of the South Seas, share the nature of serfdom, and are called Ambarages. If they are legitimate, they follow the status of their father, whatever status the mother may have; but if they are illegitimate, they belong to the mother's sort, not the father's, for we learned this once at Malacca when we were interpreting cases of conscience for the clergy, and also noted it from discussions on the island of Taprobane [i.e., Ceylon] with some of the chief vassals and diplomats of the impious King of Dachen [sic].⁶⁰

Concomitantly, the laws of Malacca (Undang-Undang Melaka) imposed a fine of fourteen times the value of the slave on those who stole a hamba raja who was later recovered, and in other cases the text demanded the enslavement of the thief. There were also fines of smaller multiples for stealing the slaves of other members of the royal family and royal officials. This implies hamba raja, who among other things contributed to the defence of the sultan, were essentially alienable and could be assigned a market value of some sort. As such, their status overlapped in significant ways with what is conventionally considered to be military slavery.⁶¹ In the construction of the fortress at Malacca, therefore, we see the combination and convergence of European, Islamic, and autochthonous regimes of bondage to support the construction of military infrastructure, which was just as essential to the large-scale outbursts of violence that punctuated life in maritime Asia as the marshalling of military labour of various statuses.

Such exchanges reflected a general pattern in the region, where people of various origins, as well as different markets and distinctive forms of military slavery, overlapped and interacted. This complex encounter is also visible in the crews who manned the many ships that crisscrossed Southeast Asia. Whether they were built in a Portuguese or Chinese shipyard, or flew a Dutch or Siamese flag, the vessels that engaged in the period's constant process of trading, raiding, and empire building all featured highly diverse crews of multiple statuses and origins, including enslaved and free Africans.⁶² This complexity is reflected in Malacca's maritime laws (Undang-Undang Laut Melaka), which stated that captains could expect to have free men, full slaves, and debt slaves (orang berhutang) among their crew.⁶³ The last of these groups was particularly common in Southeast Asia (although also documented in parts of Africa and imperial China), where rice-farming was precarious and the elites effectively became loan sharks, gaining access to their borrowers’ labour. Interestingly, the Undang-Undang Melaka also suggests that debt bondage could be inflicted on those found guilty of “slashing” (menetak) a free man. While both royal courts and the Inquisition could send criminals and heretics to the galleys (from where they might be sent to Asia), in parts of maritime Asia too, penal slavery had the potential to influence the composition of the crews manning the region's many ships that were constantly alternating between commerce and warfare, as the situation demanded.⁶⁴

This close relationship between debt bondage and combat on both land and sea was also found at the other end of the Malay Archipelago in the Philippines. There, indigenous elites (principales) allied to the Spanish eagerly embraced the Reconquista-tinted mixture of offence and defence repackaged as “just-war slavery” to engage Muslim trader-raiders in an ongoing series of conflicts usually called the Moro Wars.⁶⁵ Of those who actually did the fighting, many were debt slaves of Catholic principales who became the instruments of a distant empire. Others took part as a preferable alternative to undertaking mandated labour for their local lord or the Spanish in Manila. This undesirable form of coerced labour involved both the forced sale of foodstuffs at a particular price (bandala) and corvée labour (repartimiento): primarily shipbuilding, which in turn was an intrinsic part of the military infrastructure that allowed the expansion of Spanish frontiers and their defence against the ever-present Dutch. In principle, those subject to both the bandala and the repartimiento were supposed to be compensated, although the Spanish treasury in Manila was rarely able to do so.⁶⁶ This was just another example of the myriad interactions between different regimes of bondage in Asia that undergirded with force (or the threat of force) both European interests and those of Asian elites, allied or otherwise to the far-travelled trader-raiders.

Conclusion

As in the Atlantic world, the arming of slaves and others of low status was common in early modern Asia, whether those bearing arms were fighting under the flag of a European or non-European monarch—or even no monarch at all. This is perhaps best exemplified by the case of the black slaves and former slaves who helped defend Macau from Anglo-Siamese-Dutch invaders. Members of this little-known corner of the African diaspora (that may have also included very dark-skinned Indians or Southeast Asians whom Chinese sources did not differentiate from Africans) would later spill out into imperial China and Korea, where they could be found protecting the Zheng clan during the Ming-Qing transition and aboard a Chinese ship off the coast of Korea. Similarly, Japanese slaves and semi-independent trader-raiders could be found bearing arms along the maritime route between the Indian Ocean and the South China Sea. These various groups of ethnically diverse combatants frequently fought alongside convicts and other types of bonded people, whom the sources struggle to categorise. Thus, military slavery is a historical category that cannot be decisively separated from mercenary forces, corvée labour, slave bodyguards, and allied troops.

A final important insight is that like the military history of maritime Asia more broadly, military slavery in the region was characterised by the encounter and mixing of peoples and customs from a range of slave regimes and military traditions. This is perhaps best underscored by the labour that went into the construction of the new fortress at Malacca in 1511. There, in the newly conquered city, Ottoman ideas about royal slaves, Malay kingship, Portuguese fortress design, neo-Roman European concepts of slavery and freedom, and elephantine labour came together as Alfonso de Albuquerque attempted to cement his control of a key bottleneck along the maritime route between the Indian Ocean and the South China Sea. In fact, it is hard to imagine that, without such encounters between labour regimes that frequently relied on coercion and commodification, ships could have been manned, fortresses built, weapons forged, gunpowder manufactured, and soldiers fed in the highly violent world of early modern maritime Asia.