I like a plantation in a pure soil; that is, where people are not displanted, to the end, to plant in others. For else it is rather an extirpation, than a plantation.
Francis BaconFootnote 1Introduction
In his influential edited volume Slavery, bondage and dependency in Southeast Asia, Anthony Reid suggests that long-term slave-based systems of production were absent from agriculture in Southeast Asia, and had an ambiguous presence at best in other areas of economic activity. The argument he presents suggests that indigenous slavery in the region merged into a ‘kind of serfdom or household membership’, a situation that continued after the arrival of Europeans whose slave-holding practices were profoundly shaped by the local traditions they encountered: ‘slavery in the European colonies owed more to the Southeast Asian environment than to European legal ideas’.Footnote 2 Reid's analysis is insightful and his conclusions persuasive. But he also notes a single exception to this general picture: ‘the Dutch perkenier system for producing nutmeg in Banda with hundreds of slave labourers on large estates’.Footnote 3 The nutmeg estates of the Banda Islands, in eastern Indonesia, provide a rare unequivocal example of a slave mode of production in Southeast Asia, and its sole instance in an agricultural context. The islands have a similar status within established accounts of slavery in Asia more generally. While some degree of geographic and historical variation is usually acknowledged, European slavery practices in Asia are regarded as distinct from colonial slavery in the New World, where European systems were imported wholesale.Footnote 4 Against this conclusion, the perkenier system in the Banda Islands has been described as a form of exploitation ‘unheard of in Asia’, one that represented a ‘Caribbean cuckoo in an Asian nest’.Footnote 5 In other words, Dutch nutmeg cultivation in the Bandas constituted a New World style system of slavery operating in an Asian context.
This paper questions these depictions of slavery in the Banda Islands. I contend instead that slavery in the nutmeg estates was consistent with the general picture described by Reid, where ‘the Southeast Asian character of slavery always asserted itself’.Footnote 6 Key traits include the prominence of domestic households in economic and social terms and the existence of spheres of slave autonomy, along with diverse opportunities for manumission. The perkenier system (Dutch perkeniersstelsel) for producing nutmeg and mace was certainly one of very few historical situations where Asian slaves worked on European-owned farms or plantations.Footnote 7 Established by the Dutch East India Company (Verenigde Oostindische Compagnie or VOC) following their final conquest of the Banda Islands in 1621, the perkeniersstelsel represented an early-modern vision of a totalising administrative order encompassing labour and the environment (including both slaves and estate-holders). It is appropriate to conceive of the perkeniersstelsel in terms of a nascent application of the rationalising instrumentality of modernity in the service of agricultural production, and it clearly forms a precursor of later colonial plantation regimes. I would argue, however, that such an ambitious project was not fully realised. One critical shortcoming, of most interest here, was that the Dutch administration failed to isolate slavery in the nutmeg estates from broader practices existing elsewhere in the Indonesian archipelago. Several factors are explored as contributing to this situation, which also serve to distinguish the nutmeg estates of the Bandas from slave-based plantations in New World locations.
My fundamental concern here is not historical revision, but rather (as an anthropologist) to understand key dimensions of social life in the Banda Islands today. Established views suggest the islands were transformed absolutely by the VOC conquest, in which much of the pre-conquest population was killed or driven from the islands to be replaced by imported slaves labouring in nutmeg estates.Footnote 8 Bruno Lasker, an important early scholar of slavery in Southeast Asia, refers to the Bandas as the site of an ‘iniquitous colonial experiment’ whose ‘unforeseen legacy’ involved the creation of a uniquely out-of-place population, one he characterises as the ‘social residue’ of colonialism.Footnote 9 Yet the bulk of the contemporary population today declares itself to be meaningfully Bandanese, a perspective enacted in part through a range of ceremonial and ritual activities viewed as inherited from the pre-conquest era.Footnote 10 My assertion is that slavery in the post-conquest Banda Islands played a crucial role in engendering a transformative reiteration of Bandanese identity. This was able to occur in large part because the perkenier system and the slave labourers it imported were not isolated from the wider cultural sphere.
The result was a dynamic process of culture-building that drew together a remnant pre-conquest population and imported slaves, and ultimately also involved the estate-holders and other incomers. It is this historical process that underpins the robust assertions of local identity found in the Banda Islands today from a population that simultaneously acknowledges heterogeneous ancestral origins. Moreover, local identity in the islands remains remarkably processual, where the possibility of accessing local identity – that is, of becoming Bandanese – remains for more recent in-migrants and their descendants. It is my view that this constitutes the truly unforeseen legacy of slavery in the Bandas, the cultural outcome of slavery practices that eluded the administrative vision of the perkeniersstelsel and represent instead a variation consistent with the general situation of slavery in the archipelago. By the time the perkeniersstelsel entered its final phase of unfree labour in the latter part of the nineteenth century, importing numbers of indentured workers or ‘contract coolies’, this historical experience had established a ready route for their potential acceptance and incorporation into the local populace. The continuing presence of a strongly asserted form of local identity in the islands is testament to the elusiveness of totalising schemes of governance, whether in the realm of economic production or colonial administration.
The perkeniersstelsel as plantation
The Banda Islands are the original source of the spices nutmeg and mace, both products of a single tree (Myristica fragrans) native to this small island group. The perkeniersstelsel derives its name from a series of estates established in the Bandas by the Dutch as territorial units of silvicultural management and spice production. This followed the final brutal conquest of the islands in 1621 by the quasi-sovereign VOC, which sponsored a major military expedition under Jan Pieterszoon Coen to seize possession of the still extensive areas not yet under its control as a result of earlier conflicts. The expedition succeeded, after decimating the local population and razing their fortified settlements. Prior to Dutch conquest, the population of the Bandas is generally estimated as between 13,000–15,000 people. Readings of historical sources suggest around 1,000 Bandanese survived in the islands.Footnote 11 The Company re-established and intensified production of the valuable spices through establishing a series of estates known as perken (singular: perk). This term evokes notions of a garden-bed, but in the context of the Banda Islands it has been translated variously as garden concessions, orchards, groves and even parks. Perkenier then, refers to farmers, planters or ‘park-keepers’, although colonial-era Dutch dictionaries more often simply provide: ‘a nutmeg planter’. Hence perkeniersstelsel can be glossed as the ‘nutmeg-planter system’, and it was unique to the Banda Islands.
The ambiguity of the term perken may reflect the status of the Bandas as an early venture by the Dutch into tropical agriculture incorporating slave labour. This enterprise broadly paralleled earlier efforts by other European powers, in particular Spain and Portugal, in which islands and island populations also played a prominent part. In the mid-1450s the Spanish enslaved the entire indigenous population of the Canary Islands to labour in large-scale sugar cultivation established there.Footnote 12 Canary Islanders were later imported as slaves by the Portuguese to labour in their own Atlantic sugar-cane plantations on Madeira Island.Footnote 13 This pattern of indigenous enslavement was repeated when sugar-cane plantations began to be established in the Caribbean between 1500 and 1580, though by this time African slaves had begun to form the main source of labour.Footnote 14 By the 1550s São Tomé, the last of the Atlantic Islands, was predominantly worked by African slaves and was the closest among these islands to the plantation model that was to become a New World norm.Footnote 15
The perkeniersstelsel of the Banda Islands is a recognisable form of early tropical plantation. Judy Bieber defines the ‘plantation complex’ in the New World as consisting of ‘large estates, funded through European capital, using predominantly slave labour in the production of tropical export crops for European consumption’, adding that plantation societies depended on slavery as ‘a predominant, institutionalised form of labour’.Footnote 16 Alec Gordon characterises the late colonial ‘Asian plantation’ by its combination of foreign ownership in a conquered territory; a relatively large-size farming unit; an export orientation; and a dependence on cheap land and cheap labour (facilitated by a colonial administration).Footnote 17 The early modern nutmeg estates of the Bandas feature aspects of both depictions.
However, while the Bandas provide a Southeast Asian example of a remnant island population enslaved (in part) within plantation estates, in this case an exotic commodity was not introduced for cultivation; this immediately distinguishes the perken from both New World plantations and those in late colonial settings in Asia. The Bandanese enslaved in the islands following VOC conquest were cultivating a forest product they had themselves developed for trade many generations before, and doing so in their own lands, now under Dutch authority. The essential method remained much the same, though intensified: an understorey of nutmeg trees protected from strong winds and harsh sun by a semi-closed canopy of larger trees (notably Canarium spp., known locally as kenari). The newly enslaved Bandanese were not simply ‘labour’ in the perkeniersstelsel — they were the original silviculturalists and traders of the islands whose expertise was recognised and utilised by the VOC. Enslaved Bandanese were deliberately distributed about the islands to make use of their expertise in cultivation and spice production, with several hundred individuals initially exiled to Batavia being returned to the Bandas for this very purpose.Footnote 18 In an account of the post-conquest population in 1638, a VOC official recorded Bandanese as forming some 13 per cent of the total slave population in the islands of 2,199.Footnote 19 Perhaps it was this situation, as much as the violence of the conquest itself, which prompted Lasker to describe the perkeniersstelsel as ‘a particularly ruthless experiment in slavery by a colonial power’.Footnote 20
The destruction of Bandanese society was sought by the VOC as part of its ambitions to monopolise the European spice trade at its source, both for profit and to provide leverage for expanding Dutch interests in wider inter-Asiatic trading.Footnote 21 To this latter end, Coen emphasised the need for Dutch colonists to be established in the Indies archipelago as a burger class (that is, a non-VOC citizenry) ‘balancing’ the native population, and at least initially, envisaged as participants in regional trading networks.Footnote 22 For several centuries after the Dutch conquest the Banda Islands remained a regional trading entrepôt and redistribution centre rather than an isolated European plantation enclave. This marked a continuation of a longstanding status, shaped by the central location of the islands in terms of local physical geography and seasonal sailing conditions.Footnote 23 As Roy Ellen notes, ‘the political economy of Banda had been transformed, but with the aim of maintaining a pattern of production and export that had preceded it’.Footnote 24 In short, key elements of the pre-conquest life-world of the Banda Islands remained of vital importance under VOC rule and in fact formed the very focus of the Dutch presence. Such differences must be prominent in any comparative consideration of the nutmeg estates in the Bandas and plantations existing elsewhere, especially in the same period.
Another important point of contrast between the perkeniersstelsel and New World plantations involves the latter's dependence on the Atlantic trade in African slaves. Alongside a proportion of the remaining Bandanese, the VOC initially sourced slaves through their established trading presence outside the archipelago (e.g. on the Coromandel and Bengal coasts of India). But it was not long, however, before local markets and suppliers were emphasised, especially in the eastern archipelago, where slaves became available in increasing numbers.Footnote 25 The slave trade in the Maluku region – where the Banda Islands are found – was longstanding, consisting mainly of people seized in raids on enemy villages.Footnote 26 This trade intensified dramatically after the arrival of Europeans, most notably the Dutch, who signed contracts with local rulers in Maluku in the early seventeenth century for the provision of slaves, as well as acquiring them directly or by proxy. This demand for slaves actually exacerbated the incidence of inter-village warfare in the region.Footnote 27 In its wider regional form, the slave trade drew on upland and interior peoples from Seram, Timor, Borneo and Buru as well as Papuans, Balinese and people from South Sulawesi.Footnote 28 It flourished over a considerable period — Malacca, Makassar, Buton and Timor were all key centres of the slave trade at various times, while Java, Bali, Lombok, Sumbawa, Buton, Alor, Aru and Tanimbar were all involved in exporting slaves.Footnote 29
Vincent Loth's description of the perkeniersstelsel as ‘an early example of total colonisation by a western power’ relies in part on an argument that the estates constituted a ‘fully inclusive colonial system of agricultural production’ marked by the ‘extreme rationalisation of productional means’.Footnote 30 Drawing on VOC records, he suggests that the post-conquest islands can be characterised by ‘the central role of the VOC in the lay-out and implementation of the new agricultural system’ and ‘the firm grip the colonial government kept on the activities of the perkeniers’.Footnote 31 The estates, perkeniers and slaves are entirely subject to the (re) organisation of spice production via the perkeniersstelsel: ‘every aspect of natural and human life [was] tuned to nutmeg production’.Footnote 32 In this view, the administration seems to exercise a capacity and influence in excess of the most ambitious forms of modern governmentality: ‘nothing in Banda can be seen outside of the framework of VOC dominance’.Footnote 33 While the VOC documents Loth cites certainly suggest considerable formal attention was paid to diverse realms of regulation and supervision, other evidence would be needed to gauge the efficacy of such regimes, particularly in respect of emerging social and economic relations involving estate-holders and their slaves.
In practical terms, the Dutch administration in the Bandas seems to have been unable to isolate slavery within the perkeniersstelsel from wider related socio-cultural practices that were long established in the region. Perken boundaries quickly became porous economically, demographically and culturally, as Company-purchased slaves intended for the estates (the perkenslaven) blurred with the privately owned slaves of perkenier households. I suggest that it ultimately becomes difficult to separate the use of slave labour in support of spice production from that which served household economies. At times, the latter even appears to challenge the dominance of the former. In this sense the social practice of slavery in the Banda Islands can be said to have eluded VOC control, and exhibits key features that Reid argues were endemic throughout the archipelago and Southeast Asia. In this respect (as in many others), the ‘firm grip’ the VOC may have sought in the Bandas proved elusive here, as elsewhere.
Slavery and the economic
As noted, Coen's original social vision for the Banda Islands was for an émigré settler-colony dominated by a class of independent Dutch burgers. But a far more diverse society actually emerged in the islands, by the beginning of the eighteenth century, its ‘chief inhabitants or burghers’ were listed as comprising Europeans, Batavians, Ambonese, Ternatens and Chinese, in addition to regular in-migrants including Tanimbarese, Balinese, Butonese and Buginese, and ‘many emancipated slaves’.Footnote 34 Freed slaves, or mardijkers as they came to be called by the Dutch, constituted a large section of the population of all major settlements in the archipelago.Footnote 35 Originally this term referred to manumitted slaves and their descendants associated with the earlier Portuguese presence. Usually Christian converts of Asian (particularly Indian) origin, the group was enlarged during the seventeenth century ‘by mestizos and/or migrants from other parts of Asia’, many of whom were freed slaves from within the archipelago.Footnote 36 Doubtless the mardijker population of the Bandas had a similar character. Slaves, however, remained the numerically dominant section of Banda's population over several centuries. In 1638, a total of 2,199 slaves constituted some two-thirds of the islands' total population of 3,843 persons.Footnote 37 By 1794, slave numbers had increased to 4,112 or around three-quarters of the islands' population at that time; importantly, more than half of these were privately owned.Footnote 38
Household versus market
The predominance of the household economy is a major element in Reid's depiction of European slavery practices in the archipelago as more influenced by existing regional patterns than able to impose European ones. The productive activities of slaves in Southeast Asia before the arrival of Europeans were integrated with the domestic or household realm of their owners, a situation that continued with the European presence. For Reid, this forms a clear contrast to the slave mode of production seen in the New World — again, with the exception of the perkeniersstelsel: ‘A slave mode of production did not exist in the sense of a significantly different production system from that of serfdom … the only definite exception to this point is the Dutch perkenier system for producing nutmeg in Banda.’Footnote 39
References to ‘serfdom’ and ‘slave mode of production’ suggest conceptual links to theories of colonialism and historical transitions to capitalism.Footnote 40 However, Reid bypasses these sometimes rarefied debates by defining a slave mode of production as simply involving ‘centralized control of a large body of workers who would produce on a scale different from the household-oriented economy’.Footnote 41 The juxtaposition of a limited, household scale of production versus a more extensive, market-oriented harnessing of productive labour is a key element in his general characterisation of slavery in the indigenous Southeast Asian environment.Footnote 42
Clearly, individual perken in the post-conquest Banda Islands were eventually able to attain levels of spice production far exceeding those of the pre-VOC period. However, the boundary separating slave labour centrally directed towards spice production versus decentralised household-oriented economic activity in the Bandas emerges as ambiguous, much like the status of the perkeniers themselves. From the very beginnings of the perkeniersstelsel, the position of the perkeniers was unclear. Were the post-conquest Banda Islands a VOC plantation venture, an agricultural settler colony or an early trading settlement? Were the perkeniers settler-farmers, perken managers or burger-traders? The question was never absolutely settled, in part because of disappointment with the ‘quality’ of the Dutch settlers attracted to the perkeniersstelsel scheme, combined with VOC reluctance to sanction unrestricted trading even by Dutch civilians.Footnote 43 The Company retained formal ownership of the perken and appears to have envisaged something of a feudal relationship with the perkeniers, who were seen as tenant-vassals (leenmannen); but the latter soon grew to regard the perken as their own property and themselves as a class of landowners.Footnote 44 Within decades perken began changing hands through processes of purchase and inheritance, generally unsanctioned by the Company, in which there was both splintering and amalgamation of estates. As Hanna notes, ‘in the course of a generation or two, there evolved a new social and economic system founded as much upon precedent as upon VOC prescription, the two of which often conflicted’.Footnote 45
One important area of conflict involved distinctions between publicly and privately owned slaves. Both were present in the Banda Islands, and the number associated with individual households appears at times as large as those attached to the estates. As noted previously, in 1794 over half of the 4,112 slaves in the islands were categorised as ‘private’ or house-slaves, as against perkenslaven. By 1823, private slaves totalled 1,437 compared to 1,727 slaves and convicts recorded as labouring in the perken.Footnote 46 Centrally managed and domestically integrated forms of slave-based productive efforts co-existed in the Bandas, and the extent to which these two forms of labour were consistently distinct is uncertain. Privately owned slaves offered a range of economic benefits to their owners by engaging in a variety of profitable activities, including public works. Though the Company maintained a small number of slaves for this purpose in the islands (as it did in Batavia), labour needs were such that private slaves could regularly be hired out for such work, as one observer noted in 1796:
the want of free inhabitants to labour and attend the different trades [means that] when works of any magnitude are carried on, it is necessary to hire at a very dear rate the few artisans who are willing to work and the private slaves of individuals, whose labour their masters turn to great advantage at particular times.Footnote 47
Other areas of ‘great advantage’ to which private slaves were directed included fishing and growing fruit and vegetables. As managers of the perken, it seems the perkeniers also regularly directed the labour of perken slaves towards similar ends. When a hurricane in the eighteenth century caused widespread damage to the nutmeg forests, the administration struggled to interest the perkeniers in schemes to rehabilitate their perken. They apparently ‘preferred to use their slaves to grow vegetables for consumption or sale, rather than to care for the nutmeg or even to harvest them properly’.Footnote 48 And an early-nineteenth-century observer noted that:
[the perkeniers] were in the habit of privately employing the labour of the public slaves, which ought to be exclusively directed to the care and cultivation of the parks and to due collection of their produce.Footnote 49
Alongside direct economic return, activities of this kind likely allowed the perkenier to supplement or even replace VOC rations, which could be in short supply and, in the case of rice, formed a readily saleable commodity.Footnote 50 At the same time, benefits also existed for the slaves involved in such efforts. In 1796, at a juncture when the Bandas were in severe want of rice, a British army officer then serving in the islands suggested:
the park-slaves were in general better fed and taken care of than any other, having the priviledge of cultivating all over the park [i.e. perk], where they found it convenient, plantains, jams [yams], sweet-potatoe, kanarynuts etc for their own use, and tho the Admiral empowered me, to make such addition to their allowance of rice as should appear necessary, they did not seem to stand in any need of it.Footnote 51
It appears that perken slaves exercised a degree of autonomy over their gardening activities, a situation that has been documented in contexts of colonial slavery elsewhere. Indeed, standard conceptions of a ‘slave mode of production’ in the New World are themselves sometimes problematised by historical evidence of slaves cultivating their own food plots and independently selling surplus food.Footnote 52 It is highly likely that perken slaves in the Bandas were involved in similar private market-oriented activities (almost certainly so in regard to kenari). In any case, the economic character of their labour cannot be wholly reduced to the production of nutmeg and mace for the VOC.
Private slaves and perken slaves seem to have engaged in very similar activities at times. In fact the distinction between the two categories appears to have been deeply and deliberately compromised, both by perkeniers and VOC personnel. Within a generation or two, a pattern developed whereby perkeniers would transfer frail private slaves to plantation labour pools, while withdrawing more able-bodied individuals as substitutes.Footnote 53 This practice was documented also among leading VOC officials in the Bandas (in the eighteenth century): ‘it has been a custom amongst the governors and other servants in authority to exchange from time to time their own private slaves, that were grown unfit for labour, with the best and youngest among those of the Company [i.e. the VOC]’.Footnote 54
Plainly, the labour of public slaves in the Banda Islands was regularly utilised in efforts serving private, household economies. Indeed, Hanna notes of the perkeniers: ‘they diverted the best of their [perken] slaves to work as domestics in their Neira households or as crews on their trading vessels’.Footnote 55 The involvement of slaves in inter-island trading is the key arena in a Banda setting where, as Jack Goody suggests, ‘more discriminating attention’ needs to be paid to the significance of slave labour.Footnote 56
Inter-island trade
Focus on the perkeniersstelsel encourages attention to terrestrial activities alone in the Banda Islands. However, a critical dimension of the historical character of economic activity in the islands involves their role as a maritime trading nexus, ultimately linking local spheres of trade to the Indian Ocean world (and beyond) in a pattern established over many centuries.Footnote 57 As noted previously, soon after their conquest of the Bandas the Dutch began seeking to reinstate local and regional trading networks which had previously ensured supplies of basic commodities to the islands. By 1650, the ‘regularities of local trade had re-established themselves with Banda Neira as the hub’.Footnote 58 And from the earliest period, perkeniers and other burgers in the Bandas – including VOC personnel – participated in this ‘extensive and perennial’ inter-island trade.Footnote 59 François Valentijn observed of the perkeniers in the latter part of the seventeenth century that ‘they have the stoutest and fastest boats which one might wish and voyage on them to distant islands’.Footnote 60 It seems most wealthy Europeans in the islands owned trading vessels, which they despatched regularly to regional markets in Macassar, Timor and Aru for slaves and a range of other goods.Footnote 61 Both legal and illicit trading – i.e. in violation of VOC trading restrictions – were widespread, and formed one of the most profitable areas of economic activity in the islands. Smuggling was endemic, including spices, and some Banda merchants at times reputedly even supplied arms and military stores to native groups on neighbouring islands who were in open conflict with the Dutch.Footnote 62
The intensity of this inter-island trade was such that it frequently threatened to eclipse the commitment of perkeniers to managing the nutmeg groves, with many absenting themselves from their estates in order to pursue trading profits.Footnote 63 As noted above, moreover, the perkeniers in the latter part of the seventeenth century diverted the best slaves away from plantation labour to crew their trading vessels. Clearly such activities were of concern to the Company, and various regulatory arrangements existed at different times to police the activities of perkeniers. Yet the picture that emerges hardly suggests an administration exercising complete control. In the eighteenth century, for example, one strategy of administrative surveillance involved officials a British observer refers to as ‘boscwagters’, charged with inspecting the perken to ascertain:
the planters employ the slaves, not for their own private purposes, but in attendance on the nutmeg trees, to take care that the mace and nutmegs are properly cared and to note the quantity, that none be smuggled off. They have certain walks allotted to them, and are to make constant reports to the Governor of the state of the different parks and the quantity likewise yielded to them.Footnote 64
The activities of these officials were supervised by the Governor or other ‘chief persons of the settlement’, who were in turn expected to visit every perken on a regular basis. Nevertheless, the same observer states that the system proved ineffective, with key members of the administration themselves neglecting their responsibilities in the estates, ‘attending only to the advantages to be derived from trade with neighbouring islands’.Footnote 65
Local trade involved nearby island groups such as Southeast Seram, where significant numbers of Bandanese refugees from the conquest had settled.Footnote 66 The considerable degree of traffic between islands provided opportunities not just for the continued circulation and exchange of traditional knowledge about the Bandas, but for perken slaves to abscond. Valentijn, in the latter part of the seventeenth century, describes significant numbers of perken slaves regularly fleeing the estates by sea: ‘no matter how close a watch they may keep, as many as 20 to 30 slaves may escape by boat at night and few ever return’.Footnote 67 The VOC soon sought agreements from the rulers of neighbouring islands concerning the repatriation of runaway slaves, but to little practical effect.Footnote 68 Gwyn Campbell and Edward Alpers note that in the larger city ports of the colonial Indies, lacking the kinds of centralised authorities needed to facilitate the recapture of runaway slaves, slave-owners tended to treat their domestic slaves particularly well ‘in order to retain them’.Footnote 69 A similar logic likely applied in the case of Banda perken. Hanna goes as far as to assert that slaves in the Bandas ‘seemed to consort on friendly terms with their masters only when both were engaged in contraband traffic in spices or other commodities’.Footnote 70
As with the cultivation of gardens, participation in inter-island trading offered direct benefits to slaves, including opportunities to escape from servitude: ‘[trading] afforded slaves the opportunity to become both affluent and mobile and a slave-sailor-trader soon earned his freedom or freed himself’.Footnote 71 Reid highlights the existence of diverse opportunities for manumission as another persistent feature of pre-colonial slavery. He notes that slaves ‘often succeeded in buying themselves out of slavery with the money they accumulated in their own time’.Footnote 72 That this applied in the Banda Islands once again points to continuities between the shape of slave-holding practices in these islands and those elsewhere in the region, notwithstanding the presence of the perkeniersstelsel.
Slaves, sociality and culture-building
It seems clear that slavery in the Bandas Islands had a far wider significance than supporting intensive production of nutmeg in the perkeniersstelsel. A dense web of relations came to link perkeniers and slaves across diverse realms of social life, inside and outside the perken. Perkenier households became enmeshed with private and perken slaves not just economically but socially, culturally and demographically, a situation that applied to European slave-holders throughout the archipelago. Indeed, Reid suggests that the most common function of European-owned slaves during the earliest period of the European presence involved various forms of ‘domestic service’, in particular ‘to display (and if necessary defend) the wealth and status of the owner’.Footnote 73 This provides further evidence of the European adoption of extant Southeast Asian patterns of slavery. Here again, parallels are evident in the Banda Islands.
For decades after the perkeniersstelsel was established, the precise boundaries of individual perken were not fully demarcated, leading to episodes V.I. Van de Waal refers to euphemistically as ‘violent confusion’.Footnote 74 Hanna elaborates at this point: ‘little battles occurred between armies of slaves disputing prime lands on behalf of their masters’.Footnote 75 Slaves were also used in defence of trading interests. One Banda perkenier maintained a strong influence over commercial activities in the Aru Islands through a personal fleet of seven trading vessels. He reportedly used his slaves (‘who are very numerous’) to provide martial assistance to the small VOC garrison in that locale.Footnote 76
In addition to defending their owner's wealth, slaves in the Bandas were involved in unmistakable displays of status. In part this was direct and institutionalised: the highest officials in the islands received personal allotments of slaves from the Company as part of their conditions of rank.Footnote 77 An observer's description of the wives of the perkeniers and other ‘Dutch gentlemen’ in the Bandas in 1796, though, suggests a more pervasive symbolic dimension:
[their wives are] fond of dress and wear a number of jewels, they pride themselves particularly in the dress of their favourite slaves, who attend them in company, where they carry their mistresses betel-boxes, which are always an article of great fashion and expense, and sit on the floor, at their mistress' feet, unless otherwise employed or ordered.Footnote 78
The influence of regional culture on the very shape of perkenier households is unmistakable in this passage, in which slaves perform a role as a domestic entourage reflecting the eminence of their owners. That perkeniers' wives participated in such practices is unsurprising, given that estate-holders began to marry non-European women at a very early stage of their stay in the Bandas. VOC regulations demanded such wives be Christian; according to Hanna, this was often accomplished by training a favoured female slave in Christian catechism to allow for her baptism prior to seeking official approval for a marriage to take place.Footnote 79 The reference to ornate betel containers is particularly compelling — a longstanding symbol of status and wealth among many local populations in the region. The same writer goes on to note:
Most of the Dutch gentleman are married to native women of perhaps the tenth generation from Europe blood, and seem never to wish to quit the places they are in, to return to Europe; perhaps from reflection, that their wives, who are little more than the chief female slaves of their families, are unfit to be introduced into society there.Footnote 80
The VOC itself encouraged mixed marriages in 1633 after abandoning plans to import Dutch women to Batavia, in addition to turning a blind eye to ‘irregular unions’.Footnote 81 By the beginning of the eighteenth century Valentijn observed: ‘there is hardly a single Hollander of any consideration in Java that does not have a concubine’.Footnote 82 Reid notes that by the early nineteenth century, the trade in women throughout the region ‘came to resemble a large-scale “marriage market”’, which provided a large proportion of the female population of the British Straits Settlements.Footnote 83
Importantly, the exercise of some level of agency by the women involved should not be discounted. Campbell and Alpers argue that servile labour in both Asia and Indian Ocean Africa often sought to ameliorate their conditions and status by securing a position in the dominant society that could be improved over time. If skills in sailing and trading offered such strategic possibilities for male slaves of the perken, concubinage and marriage would have formed potential routes for social mobility open to women, particularly where their children were to be ‘born into important families’ — which in the Banda context meant the perken-holders.Footnote 84 Reid stresses continuities with pre-European slavery in this realm, observing that opportunities for upward mobility and an easier life appear to have been far greater for women than men.Footnote 85 Eric Jones agrees, linking several documented examples of female slaves absconding from their owners in the archipelago to early VOC slave statutes which attempted to remove plasticity in social relations and secure the permanent position of slaves. The result was friction and resistance:
Absent the reciprocal relationships in which slave devotion brought filial promotion and freedom, running away seemed the only avenue for upward advancement … if allowed to experience some degree of family intimacy, as a real part of their owner's household … many slaves were willing to tolerate their position.Footnote 86
Indeed, the rarity of slave revolts in Asia (and Indian Ocean Africa) has been attributed to the fact that most slaves were women ‘often involved in intimate relationships with their owners, and frequently offered greater opportunity to assimilate into the dominant society than male slaves, they were reluctant to take risks that might damage their children's interests’.Footnote 87
Both Hanna and Lasker provide a description of perkenier households in the Bandas which point to a unique level of ‘hybridity’, though both place emphasis on ideas of racial difference rather than socio-cultural practice. Lasker observes that ‘Many of the Dutch settlers [in the Bandas] married native women, and the population, with the arrival of new slaves from a wide area, became one of the most hybrid in the Western Pacific.’Footnote 88 Hanna makes a similar point: Perkeniers and their slaves began very soon to come to a mutual accommodation on the basis of miscegenation … the new perkenier households thus took on a composite racial aspect, as ex-slaves mothered Dutch children … extended households of the so-called ‘mixties’ – some slave, some free, some freed slaves – established themselves in the perken.Footnote 89
The Dutch expressions mixtie and mixtiezen are regularly glossed in English as ‘mestizo’. In discussing seventeenth-century Ambon, Gerrit Knaap for example points to a chronic shortage of European women suggests this meant that the demographic category of European citizenry (vrijburgerij, ‘free citizenry’) ‘was liable to “mestizofication”’. He maintains that the female partners of European burgers and VOC personnel were ‘to a large extent bearers of a mestizo culture, defined as a mixed culture with several European and Asian elements’.Footnote 90 Anne Stoler has argued that the contours of colonial society during the first two centuries of the Netherlands Indies were shaped by ‘Indische mestizo culture’, referring to the practices of Indies-born people of mixed European and local descent.Footnote 91 Such terms can be problematic in their potential entanglement with racial discourse concerning the consequences of mixing ‘blood’ (whether viewed positively or negatively), effectively blurring the social and the biological.Footnote 92 Nevertheless, they serve to emphasise the point that socio-cultural transformations in the Banda Islands following conquest had much in common with wider dynamics in the region, notwithstanding the presence of the perkeniersstelsel. Importantly, moreover, any consideration of the role and significance of slavery in the islands must consider the emergence of cultural practices common to slaves, perkeniers and burgers (European and otherwise).
Whether this phenomenon should be conceived in terms of mestizo or hybrid culture remains moot. The notion of hybridity has its own problems, as Emiko Ohnuki-Tierney observes: ‘hybridisation is the sine qua non of all cultures. All cultures are products of hybridity … culture is hybrid by its very nature’.Footnote 93 In a related critique, Marshall Sahlins dismisses the concept as being ‘too abstract and indeterminate’ and urges attention instead to the careful historicising of all cultural forms.Footnote 94 What is notable about the Banda Islands is not that a process of mutual cultural accommodation emerged which ultimately affected the character of perkenier families, but the way in which this facilitated the emergence of a reiterated Bandanese identity among the descendants of the perkeniers as well as those of slaves and other residents of the islands. Such a process was assuredly complex, but an important shared feature involved language.
Malay and Bandanese identity
Within the strategies for advancement pursued by slaves, acculturation generally played a critical role. Among the most important forms of slave acculturation highlighted by Campbell and Alpers was linguistic change: ‘for most [slave] owners it was vital that the slave speak the local language … in such circumstances, the native languages of slaves quickly ceded to that of the dominant society’.Footnote 95 Reid's work on slavery however, attests to the fact that ‘acculturation’ need not be unidirectional. And in the Banda Islands, the language that soon dominated social life both in and outside the perken was not that of the governing authorities, but one with a longstanding presence in the islands and the region: Malay. The emergence of Malay as the main everyday language of the Bandas was without doubt a critical factor in the dynamics of local culture-building in the islands, and one in which the perken slaves would have played a prominent, though unrecorded role.
Within the newly heterogeneous population of the Bandas following VOC conquest, and especially among the diversity of regionally sourced slaves, a pidgin form of Malay would almost certainly have served as an immediate means of everyday communication. Malay existed as a trade-linked lingua franca in the archipelago for centuries before the arrival of Europeans.Footnote 96 One of the earliest written examples of a Malay word-list was collected in the sixteenth-century Moluccas by the Portuguese, where it was described as being used ‘like Latin in Europe’.Footnote 97 In all likelihood this pidgin Malay would have gained first-language speakers within one generation, a situation some linguists refer to as the emergence of an ‘abrupt’ creole. The typical elements giving rise to rapidly emerging creoles of this kind have been described as follows:
members from various linguistic communities are uprooted from their native habitat and, for basic survival, are forced to develop a contact language which becomes their primary language almost immediately and the first language of any of their children born subsequent to the uprooting.Footnote 98
This is an apt description of the situation in the Bandas following the establishment of the perkeniersstelsel, and slavery is often considered an archetypical context for an ‘early creolised’ language of this kind.Footnote 99 The development of a locally distinctive form of Malay would in all likelihood have occurred relatively quickly; certainly Malay shrugged off early efforts to establish Dutch as the main language of the islands.
As early as 1624, three years after the VOC conquest, schools existed in the islands for the children of Europeans, Christianised former slaves and pensioned members of local militia. The official language of instruction was Dutch, but a year later aspects of Christian religious education were permitted in Malay for ‘native’ children.Footnote 100 By the middle of the seventeenth century, just decades after the conquest, Malay had become the main language of the islands, even in Christian preaching and VOC directives, while its widespread utility saw it rapidly become the language of general instruction in Banda's schools.Footnote 101 The situation was formalised in 1665 when the administration in Batavia decided the use of Dutch would no longer be encouraged, as it was deemed too expensive to provide education on the scale required and too difficult a language to enable effective widespread communication.Footnote 102 Dutch remained the official language of government, though mainly in written form. A 1712 account of an annual ceremony of allegiance to the VOC in the Bandas called Veroverings Dag (Conquest Day) describes Dutch officials conducting a public address in Malay to a diverse crowd that included local burgers and ‘village folk’.Footnote 103 This attests to its established status as the language of greatest intelligibility among the islands' general population by this time.
Dutch retained a presence in the islands as an ethnic marker and symbol of status, particularly among ‘Europeans’ (a category that included the descendants of European-local intermarriage).Footnote 104 Even among the latter, however, it is unlikely to have provided the main medium for everyday interaction. A tract describing life in the Banda perken in the nineteenth century refers to the prevalence of ‘Moluccan Malay’ (Moluksch Maleisch) among perkenier families, especially women.Footnote 105 And the author of a Bandanese childhood memoir notes of the 1930s:
Weakness in the Dutch language was the main reason why so many Bandanese failed at school and became fishermen. Unlike the Ambonese, we Bandanese didn't like to speak Dutch. Even the majority of the Bandanese burgers preferred to speak Bandanese-Malay, rather than the language of their forefathers.Footnote 106
One of my own early informants in the Bandas confirms this view. An aged member of a colonial era perk-holding family which claimed direct descent from one of the first Dutch perkeniers in the islands, he was also the owner of the sole remaining privately held nutmeg estate in the Bandas.Footnote 107 As a result of access to colonial schooling and a period in merchant shipping, he was a confident Dutch speaker; nevertheless the first language of his childhood home in the islands was Malay, and this was true of his own household. He observed that a similar situation applied to all the older perkenier families of the Bandas, and had been the situation for as long as anyone could recall.
By 1876, the early Dutch linguist De Clercq had listed the Banda Islands as one of several locations in the archipelago where a unique regional dialect of Malay could be found, one he labelled ‘Banda Malay’.Footnote 108 The local population of the islands today asserts the distinctiveness of this form of Malay and refer to it as ‘Bandanese’ (Bahasa Banda).Footnote 109 One feature of Bahasa Banda is its numerous lexical borrowings from Dutch, altered but recognisable. Linguistic creolisation is often used to stand paradigmatically for historical processes of cultural and demographic change, particularly in settings of colonial slavery where, as noted, this was widespread.Footnote 110 However, Sidney Mintz cautions that ‘language is not culture, only a part of culture … the linguistic model of creolisation is a model for languages, not a homology with culture itself’.Footnote 111 Nevertheless, while culture is clearly not reducible to language, language can play a central part in the reproduction and transmission of important cultural ideas. The rapidly established and longstanding dominance of Malay in the Banda Islands would have facilitated the retention and transmission of prominent ideational motifs common both among regional populations and at least to some extent throughout the broader Malay-speaking world. Prominent among these is a profound ontological and cosmological significance being given to origin narratives concerning significant sites and places in the local landscape, alongside special respect and status associated with those bearing such knowledge.Footnote 112 In this regard, the remnant population of pre-conquest Bandanese still present in the islands during the early period of the perkeniersstelsel could be expected to have exercised an influence disproportionate to their number. The local knowledge they were able to provide would have been sought after and valued — in a practical sense by the early Dutch administration and perkeniers (as we have seen), and in cosmological terms by others, notably the perken slaves.
Contemporary Bandanese envisage local identity in terms of a necessary obligation for newcomers to adapt to the customary practices associated with the islands as a locale, rather than those linked to places of ancestral origin elsewhere. This is seen as especially imperative for those born in the islands. A Christian Bandanese woman in the late colonial period explained her adherence to local traditions in the following terms: ‘we go to church because we are Christians, but we are also Bandanese. Born on Banda's soil we should follow the Bandanese customs inherited from the tete moyang (ancestors)’.Footnote 113 This rationale echoes a view still asserted among Bandanese today, underpinning commitment to a series of ritual activities linked to significant sites in the landscape venerated by both Muslims and Christians, in effect elevating identification with place beyond religious affiliation.Footnote 114 A series of chanted verses reserved exclusively for ritual occasions are understood as preserved from the pre-colonial past, and have been acknowledged by linguists as preserving examples of the indigenous language of the islands.Footnote 115
It is here that some valuable historical comparisons might well be drawn between the Banda Islands and the Caribbean, not as evidence for a ‘non-Asiatic mode of production’ in Southeast Asia but rather to consider the locally specific dynamics of social and cultural transformations associated with practices of slavery in different locales. In the New World, processes of this kind have often been glossed as ‘creolisation’.Footnote 116 However, Sidney Mintz has argued strongly that this term should be reserved to signify a specific geographical and historical process involving slaves in colonial Caribbean plantations, rather than referring to generalised cultural ‘mixing’:
[Creolisation] stood for centuries of culture-building, rather than culture mixing or culture blending, by those who became Caribbean people … they were creating forms by which to live … they had to devise new forms of communication, new norms of interaction, new ways to maintain life's meaning, and to do so using their hearts, heads and hands – as well as those materials they found lying around them.Footnote 117
Creolisation emerges in Mintz's work as a form of ‘embattled creativity’ involving people in chains constrained to deal with living meaningfully in a new location. Elsewhere, Mintz provides additional details:
Culturally, they were becoming their own persons. This involved the refashioning of cultural materials from more than one source … what typified creolisation was not the fragmentation of culture and the destruction of the very concept, but the creation and construction of culture out of fragmented, violent and disjunct pasts.Footnote 118
I suggest that the form of cultural creativity which emerged in the Banda Islands among perken slaves, and which ultimately caught up members of perkenier and other burger families, occurred in a far less disjunctive or fragmented setting than was the case in the Caribbean. This was a result of the communicative opportunities, shared social idioms, and cultural continuities available through the post-conquest dominance of the Malay language. In this sense, the slave population of the Bandas could be considered less geographically alienated, less culturally disparate, and less socially isolated than those in Caribbean plantations. As a consequence, the project of culture-building that took place was likely to have been far less ‘embattled’. It would still have incorporated elements of an emerging Bandanese population ‘becoming their own persons’, in Mintz' sense of moving ‘beyond the cultural and conceptual confines of their migrant parents’.Footnote 119 At the same time, however, the process would also have involved engaging a broad suite of common orientations and understandings that were able to find expression in a local setting as a post-conquest reiteration of Bandanese culture.
Bandanese themselves now use the term campur (‘mix / blend’) in describing their distinctive Malay as a ‘mixed language’ (bahasa campur) – linked to lexical borrowings from Dutch and regional languages – but also as an expression of self-description in the phrase orang campur (‘mixed/blended person’). The term entails an acknowledgement of heterogeneous ancestry as a widely shared and distinctive condition in the islands, in turn enmeshed with state-promulgated representations of Indonesia as a nation of distinct (and harmonious) ethnic groups (bangsa suku). Few Bandanese feel able to claim unequivocal affiliation with any single group framed in these terms, a situation that does not detract from their firm assertion of a fully emplaced local identity.Footnote 120 To be Bandanese (Orang Banda) today is also to be an orang campur, a juxtaposition that troubles national narratives of static, regionalised ethnic identities.
In my view, the Bandanese notion of being ‘mixed’ involves an implicit recognition of the part played by historical processes in constituting the population of the contemporary islands as they are, rather than indexing a significant concern with prior purity (ethnic or otherwise). Being both ‘mixed’ and Bandanese emphasises post-colonial subjectivity in the islands as an effect of power and temporality, where the valid assertion of being local relies upon practices of cultural reiteration, deliberately enacted. Under such conditions, continuous scope exists for the incorporation of further newcomers. This process has occurred in living memory, notably involving the final wave of unfree labour to arrive in the perken during the late colonial period.
The final wave: ‘contract people’
By the time slavery was formally abolished in the Netherlands Indies in 1863, the transition from slave labour in the Banda perken was already under way. In 1860, more than 60 per cent of the total perkeniersstelsel workforce fell into the category ‘contract coolies’ or indentured workers, a total of some 967 people, mainly male (805 men and 162 women).Footnote 121 This is the form of imported labour strongly associated with the social character of the perken by Bandanese today, still recalled from the late colonial period by the oldest residents in the islands. Despite its historical prominence, slavery does not appear in local narratives of the perkeniersstelsel era, known as the ‘time of the enterprise’ (waktu onderneming). Nonetheless, the perken slaves left their mark in the culture-building process I have suggested originated with them. Indentured labourers in the Banda Islands perken did not find themselves part of isolated social and cultural enclaves. Rather, they constituted the final wave of unfree labour to fill the perken, a well-worn path existed for themselves and their descendants to move from outsider status to that of community member and to becoming Bandanese.Footnote 122 This was a route followed by many thousands of slaves and other in-migrants over several centuries of cultural transformation in the Banda Islands.
The option of becoming local was open to the indentured labourers despite the fact that they held the lowest status within a pervasive and complex social hierarchy existing in the islands during the late colonial period (variously involving occupation, ethnic and religious identity and gender). The low regard in which they were held by village residents has a number of dimensions, which cannot be addressed in detail here, but a critical aspect appears to be the character of the labour contract itself. The local expression ‘contract people’ (orang kontrak) emphasises perken labour recruited under a penal sanction, with negative implications in local conceptions of power and agency. A succession of Coolie Ordinances from 1880 promulgated in the Netherlands Indies included a penal sanction, making it a criminal offence for coolie labourers not to comply with the obligations of their contract.Footnote 123 They were obliged to remain on the premises of the enterprise and to serve out their contract, which normally ran for three years. In effect, the labour contract bonded coolie labourers to their employer.Footnote 124 After 1911, more diverse conditions of indentured employment were created with the emergence of different categories of coolie labour. ‘Free coolies’ were contracted without penal clause; ‘contract coolies’ remained under a penal clause.Footnote 125
Maluku as a whole was a relatively minor colonial destination for indentured labour and consisted to an exceptionally high degree of ‘free coolies’. In 1924 for example, two out of every three indentured labourers in Maluku were free, reflecting regional variations in recruitment (island labourers throughout eastern Indonesia were ‘less inclined to accept the penal clause’).Footnote 126 For the coolie labour that was present in Maluku though, perken in the Banda Islands formed a key destination, and the largest proportion of contract coolies were also likely concentrated here.Footnote 127 Scholars also note that the proportion of women in contracts with penal clauses rose over time, suggesting that women were in a weaker position than men in negotiating the ‘better’ contract of a ‘free coolie’.Footnote 128 Memories of indentured perken labourers in the late colonial period point to a large proportion (even a majority) of women, recruited mainly from Java. Certainly the phrases ‘selling one's soul’ (jual jiwa) and ‘selling oneself’ (jual diri) are mentioned in the Bandas in relation to the contract people, and these expressions have been linked elsewhere to the penal sanctions.Footnote 129
One suggestion in local terms is that labour contracts of this sort were only entered into by people of little social standing — not just impoverished, but lacking recourse to any kin who could assist them. This view sits alongside highly gendered perspectives of labour mobility valorising the exercise of a spiritually potent masculine agency in actively pursuing knowledge (ilmu) and good fortune (rejeki) through self-directed travel. By contrast, tales abound of Dutch labour recruiters and their local assistants using dark powers (ilmu hitam) to coerce or trick the contract people into leaving their communities of origin. These recruiters are said to have targeted women in particular, in the street or marketplace, magically robbing them of their normal level of awareness (often using a secret touch). Full consciousness would be restored upon reaching the Banda Islands, at which time their signed contract would be brandished and their situation inescapable.Footnote 130 The marked gender imbalance among contract labour is often represented as a clever Dutch ploy to attract the arrival of free male labour migrants. A local aphorism invariably cited in this context suggests ‘where there's sugar, there're ants’ (ada gula, ada mir).Footnote 131 While such accounts appear rather dismissive of feminine agency, they are also somewhat recuperative of the status of female ancestors, who some suggest may well have been individuals of some status before becoming ensorcelled.
If plantation workforces elsewhere were usually ‘forced into a way of life that was at complete variance with that of the local population’, this was far less the case in the Bandas.Footnote 132 Shared religious affiliation was regarded as an important factor in bridging the differences (and sometimes, points of tension) between contract people and village residents. Dutch literacy was a condition of managerial-level appointments, essentially limiting such positions to Christians, who could access superior levels of colonial education. As a consequence, those who had most contact with each other on a daily basis at the lowest levels of the perken hierarchy were invariably Muslims, as were most of the villagers living adjacent to the nutmeg estates. Village-dwellers, particularly single men, would sometimes work on a seasonal basis in perken. One result is that intermarriage between perken labourers and local people was a feature of life in the late colonial era — having a female antecedent who was a contract person is not uncommon in the islands.
Crucially, such individuals were expected to actively pursue membership of the village community. Informants maintain that this involved outsiders in demonstrating the sincerity of their intent to abandon competing cultural affiliations, as evidenced not just by engaging in village-based religious and social activities but also by attending, assisting and ultimately participating in the ceremonial and ritual action marking most village communities in the Bandas as pre-European polities. As noted previously, these occasions are vital to the symbolic construction of local community, so much so that Bandanese Christians long embraced a tradition of participation alongside their Muslim neighbours in a set of activities concerned in large part with marking the arrival and dissemination of Islam in the islands.Footnote 133 Co-operation of this kind was testimony not just to the importance of post-conquest culture-building in the Banda Islands, but also to its ability to cut across axes of social distinction within the islands' population. As noted, Christians had access to superior forms of colonial education, allowing them to largely monopolise administrative and managerial positions in the late colonial perken. Nonetheless, all had a part to play in these collective ceremonial events, including members of perkenier families who still valorised their descent links to the Netherlands in the late colonial period, and were locally referred to as ‘Dutch’ (Orang Belanda), particularly those who retained a Dutch family name.
Many of these families are recalled as active supporters of important ritual events occurring in local villages, whether as participants, donors, sponsors or hosts. Elderly informants with memories of the late colonial era recall enjoying lavish food and refreshments prepared at the homes of Orang Belanda before and after ceremonial performances. Elders of one village remember receiving substantial financial assistance from Dutch perken-holders for a major village ceremony with links to surrounding estates. The specific ceremony, known as ‘washing the sacred heirloom well’ (cuci parigi pusaka) occurred only once each decade or so and attracted large crowds. The core of the event involves the elaborate ritual cleansing of a well, which is understood as sanctifying the water, ensuring its continued abundance, and generally providing the village and its population with prosperity and good fortune. As one informant stated:
The Dutch enjoyed the [well-washing] ritual; really enjoyed it, it ensured the well would be clean, the water of good quality, and plentiful. They needed to use the well to water nutmeg trees during hot weather. So when we held the village rituals the Dutch would help senior people with the cost of the ceremony. The perek [perken] needed that well.
Indeed, many locally resident descendants of Dutch perkenier families were regarded as Bandanese (alongside the Orang Belanda ethnic label), and their enthusiastic involvement in such activity, sometimes in a prominent leadership role, is often cited. The permeable character of local identity in the Bandas is the more remarkable in a region where relations to place and to place-linked forms of identity are frequently expressed in cultural idioms of ancestral domains and the performative tracing of genealogical origins, whether of individuals or social groups (such as clans).Footnote 134 By contrast, the cultural terms of emplaced identity in the Banda Islands demote concerns with ultimate origins, emphasising instead the intent of individuals who actively seek to become local.
Conclusion
Jack Goody makes the general observation that arguments from economic principles tend to neglect slavery's location in specific social systems.Footnote 135 Mintz suggests, moreover, that to try to address the nature of plantation slavery in general terms is a risky and unprofitable undertaking:
Not only was slavery different in the colonies of one power to another, but even within one imperial system, there were often significant differences in the slavery institution from colony to colony. Moreover, time and circumstance deeply affected the way slavery worked in particular milieu. Demography mattered, as did the prevailing work at which slaves were employed … these, and many other factors, much influenced what slavery was and how it was experienced.Footnote 136
In the book Slavery, bondage and dependency in Southeast Asia it is evident that Reid agrees with both perspectives. At the outset, he endorses a principle that research into slavery must begin ‘by seeing exactly what each institution means to the people within that system’.Footnote 137 In other words, careful exploration of social context is a prerequisite to understanding the position of ‘slave’. A major strength of the volume as a whole is the way in which its contributors avoid a priori conceptions of what slavery entails. Against this approach, reference to the nutmeg estates of the Banda Islands as likely representing the sole unequivocal example of a ‘slave mode of production’ in Southeast Asia seems incongruous, harking back to reductionist and synthetic economic perspectives.
Slavery in the post-conquest Banda Islands shared critical features identified by Reid as distinguishing European slave-holding practices in the region. To represent the Bandas as the regional exception, or as an enclave of New World-style slavery, evokes a view of slavery as completely contained within the perkeniersstelsel and as entirely determined by the Dutch administration. In both cases this involves depicting perken slaves as isolated from the broader texture of economic and social life in the islands and their region. I have argued that such a view is not sustainable. The categories of private and perken slave in the Bandas were porous, as were the boundaries separating perken labour from the domestic economies of perkenier households. Neither did the perkeniersstelsel monopolise economic action in the Banda Islands; a great deal of market-linked activity – both legal and illegal – occurred outside the system, relatively unmediated by the local administration. Inter-island trading was especially prominent. For individual perkeniers, household-focused productive activities could be as important as the more centralised production of the nutmeg estate; the inter-penetration of both realms was a feature of life in the Bandas soon after the VOC conquest.
Yet the perken did form a critical context for culture-building by the slave population imported to the Banda Islands, processes within which perkenier households were themselves entangled. It is here that depictions of the perkeniersstelsel as an exercise in ‘total colonisation’ are most unsatisfying. Rather than representing a uniquely ‘un-Asian’ system of agricultural exploitation in a Southeast Asian setting, the significance of the perkeniersstelsel is that it provided a specific context for the expression of a local variant of the intercultural transformations that characterised the European presence in the archipelago. The ultimate result was the emergence of a processual form of emplaced identity that allowed outsiders (and their descendants) to achieve local status as a result of actively demonstrating their intent to do so. This is the truly unforeseen legacy of slavery in the Banda Islands.
