Coerced labour and colonial governance in nineteenth and twentieth century Indonesia

ABSTRACT

This article discusses the different uses of mandatory labour services, or corvée labour, in nineteenth and twentieth century Indonesia. While slave labour, coerced cultivation and other forms of involuntary colonial labour have been elaborately studied, these much more pervasive and diverse types of mandatory services have not yet received the attention they deserve. The article argues that in Indonesia corvée labour became foundational to the exercise of modern colonial governance and the organization of the colonial state and its fiscal capacity, so foundational, in fact, that it impeded much of the aspirations of colonial civil servants to replace coerced services with monetary taxes. Studying corvée, it shows, elucidates otherwise hidden strategies and practices of the organization of colonial statecraft and governance. As such, corvée labour played a pivotal role in how colonialism unfolded and was experienced.

KEYWORDS:

• corvée labour
• Indonesia
• colonialism
• taxation

Introduction

In 1929, the Dutch envoy to the Forced Labour Convention in Geneva, organized by the International Labour Organization (ILO) to discuss the worldwide banishment of forced labour in colonial states, faced the almost impossible task of defending the millions of annual labour services still performed in the Dutch colony in Indonesia (Haga, 1930, p. 213; Frankema & Van Waijenburg 2014, p. 372). To make matters worse for him, present as well at the Convention was the prominent Indonesian nationalist Haji Agus Salim (1884–1954), who joined as advisor to the Dutch Confederation of Trade Unions. Salim delivered a passionate plea, eagerly unveiling the excessive (ab)use of corvée labour in the late Dutch colonial state.¹

Coerced labour characterized Dutch colonialism in both the Atlantic world and Asia. While Dutch use of coerced cultivation, slave, coolie and convict labour has deservedly received much attention (Breman, 1989; Elson, 1994; Cribb, 2017; Hoefte, 2017; Van Rossum, 2018; Hofmeister; Brandon & Bosma, 2021), corvée labour has remained somewhat overlooked. In colonial context, interpretations of corvée range from a method of cheap labour supply contextualized in terms of serfdom, indentured labour in substitution for slavery or convict labour to a continuation of semi-traditional indigenous customary services recycled by way of in-kind taxation to expand the state’s fiscal capacity (Reid, 1993; Hoefte, 2017; Van der Linden, 2017; Van Rossum, 2018; Van Waijenburg, 2018 Booth, 2019; Tomczak, 2021). Recently, Van Rossum and Tosun (2021) pungently resisted against the idea of corvée regimes as ‘regressive leftovers of Asian or (transplanted) European feudal pasts or revenue extraction by weak colonial authorities or instruments of light in-kind taxation.’ They argue such regimes were ‘refined’, ‘strategic’ and ‘modern […] methods of colonial exploitation that provided colonial actors more direct access to and control over the production of commercially interesting global commodities’ (Van Rossum & Tosun, 2021, p. 2).

While this may certainly have been the case concerning the era of global trading companies, in the later nineteenth and twentieth century, I argue, corvée labour also played an important role in the physical and ideological construction of the colonial state itself. In Indonesia, only indigenous populations performed corvée services, which, as such, became an important marker of racial identity that retained the socio-ethnic boundaries among certain population groups within the colonial state. The services being levied, however, knew many regional varieties and covered a wide range of different ‘feudalist’ types of semi-coerced communal labour services, mutualism and reciprocity. Occupied by the colonizer and melded into uniform models of forced corvée services, these were henceforth used to support the economic projects of the state. This was very much part of the colonization process; it helped tying indigenous people to the state by forcing them to work in service of that state, under the justification that colonial labour was some sort of continuation of precolonial labour duties. Hereto, the state actively integrated the ‘traditional’ patterns based on which these duties were levied into its own structures. The use of coerced labour thus mattered to the political-legal construction of colonial empire as much as to its economic demands.

Simultaneously, especially closer to the twentieth century, corvée begun playing an important role in the state’s self-legitimizing narrative of development, as an important governmental instrument to instil work ethics and new principles of modern colonial economic subjection, productivity and political organization. Its increasingly systematized administration reduced individuals to statistical numbers, anonymizing them in the collectivized schemes of colonial archives, charts and tables. It is this uniformization and centralization that obfuscates our understanding of corvée in (pre)colonial societies as it mixed new with old forms of free and coerced labour. Reducing corvée to only a powerful method of exploitation to furnish the production of profitable commodities would violate this complex reality of corvée in the nineteenth and twentieth century.

Furthermore, by the turn of the twentieth century, coerced labour was supposed to be a relic of the past; from the 1860s onward, the so-called ‘cultivation system’ in Java, the most notorious example of Dutch coerced labour schemes in Asia, was gradually dismantled, while slavery in the Dutch empire was formally abolished in 1863. Political support for coerced labour as an economic model seemingly further deteriorated under influence of political debates in the Netherlands and developments in agricultural industry (Bosma, 2011). Dutch officials begun envisioning a modern welfare state that relied on rational bureaucracy, wage labour and monetary taxes to integrally support the ‘development’ of colonized economies and societies to new levels of economic modernity. Already in 1854, a new ‘Government Regulation’ had called for the ‘gradual abolition’ of coerced services and their full replacement with monetary head-taxes.² The adoption of an ‘ethical policy’ around 1900, the Dutch take on the idea of colonialism as a supposedly benevolent, moralizing and civilizational endeavour, further discarded coerced labour as incompatible with the forces of modernization and progress.

Hence, reducing use of corvée labour and replacing it with monetary taxes became a vital ambition of Dutch colonial policy (Manse, 2022b). In fact, all over Southeast Asia, fiscal systems were supposedly transformed and systems of revenue farming and corvée labour replaced with more modern tax systems (Booth 2019, pp. 36–37). Yet, by the time of the Geneva Forced Labour Convention in 1929, more than 11 million people in Indonesia, or about a quarter of Indonesia’s indigenous population, still performed corvée services of some kind.³ Simultaneously, the Dutch ultimately disagreed with the ILO Forced Labour Convention’s final resolution (C029), which they refused to sign until three years after the Convention, when a clause was included that exempted specific ‘civil obligations’ – including the supposedly ‘traditional’ services performed by Indonesians to their chiefs – which facilitated use of coerced labour in not only the Dutch but also the French and British colonial empires (Frankema & Van Waijenburg, 2014, p. 391; Van Waijenburg, 2018).⁴ This seems directly opposed to the promises of Dutch ‘ethical’ colonialism. So why was this the case?

This article argues that ‘legitimized’ use of corvée services lingered on not only because of their attractiveness as an attractive source to fulfil the augmenting labour demands of the ambitious state and the capitalist export-economy (Breman, 2015; Van Rossum & Tosun, 2021), but also because of their deeply rooted importance to the socio-political fabric of nineteenth century colonial ideologies of governance and state formation, and because Dutch officials experienced many difficulties in implementing alternative forms of monetary taxation due to local political and practical constraints. While exploitative forms of corvée services may seem in tension with ambitions of governance, in late nineteenth century colonial argumentation their supposed ‘reformatory’ capacity was squared with moralizing ambitions and attempts at governance. In the twentieth century, monetary taxes were supposed to take over these functions, but by that time corvée labour occupied such a central place in Dutch colonialism that abolishing it required a full, risky intervention in the operation of the colonial state itself.

The article examines three cases of the implementation, use and attempted abolition of colonial corvée in three parts in Indonesia: Ambon, Java and West Sumatra. Each of these had its own unique socio-cultural constellations, but got subjected to similar systems of exploitation, trade-monopolization, coerced production and corvée under Dutch occupation. Local practices were merged and slowly uniformized into an encompassing regime of coerced labour. Each of these regions has been studied extensively (Elson, 1994; Kahn, 1993; Knaap, 2004), but few historians have ventured to study them in comparison or use coerced services as a binding lens to understand the relations between Dutch colonial governance and corvée. Doing so will elucidate otherwise hidden strategies and practices of the organization of coerced labour, which, in terms of labour use, was on par with slavery and coolie labour, but, contrary to these, also integrally connected to processes and practices of supposedly modern colonial statecraft and governance.

Colonial corvée labour and governance in Indonesia

Corvée services existed across the globe as obligations of tenant farmers to feudal landlords, monarchs and states (Bloch 1961, pp. 73, 254–265; Kiser & Levi, 2015, pp. 563–565; Webber & Wildavsky, 1986, pp. 149–227). In the Netherlands, for instance, they were levied as incidental, community-based conscription services to improve infrastructure up until the nineteenth century. In Asia, for instance in Thailand, Korea and Japan, corvée services were used as a condition for land tenancy or by way of conscription, but in direct relationship with taxation; by the nineteenth century, corvée was replaced with wage labour and monetary taxes in all these states (Mathias, 2011; Terwiel 2011, pp. 512–513; Booth 2019, p. 48; Miller, 2007).

All colonial states in Africa and Asia used corvée, often in relationship with taxes, to suppress the costs of governance and contribute to processes of state building, both literally, by using services for the construction of infrastructure and government buildings, and metaphorically, to fiscally integrate indigenous populations into empires and instil economic values and shape ‘governable persons’ (Bush & Maltby, 2004; Gardner, 2012, p. 58; Frankema & Van Waijenburg, 2014, pp. 390–391; Havik et al., 2015; Van Waijenburg, 2018, p. 49). The Dutch colonial state was no exception when it comes to this sort of argumentation. According to the foremost Dutch architects of coerced cultivation (see Manse, 2021a, pp. 38–41, 129–132, 222–225), coerced labour instilled obedience, discipline and ‘productive virtues,’ which would help morally reconstituting supposedly ‘lazy and irrational’ natives into model citizens and economic men responsive to market forces (Alatas, 1979; Noor, 2016; Schrauwers, 2021). Though such focus on native labour, the indigenous population was linked to larger global economic systems of progress and capitalism. Labour became the central force by which questions of welfare could be answered, while at the same time people could be ‘civilized’, and their physical and moral development improved to furthering the economic prosperity of the country, suggesting a ‘happy unity between welfare, on the one hand, and productivity and economic efficiency, on the other’ (Anghie, 2007, p. 166).

While the British had (unsuccessfully) experimented with coerced pepper production in Bengkulu already in 1695 (Kathirithamby-Wells, 1977, pp. 22–44, 75–79), and the Dutch had monopolized and implemented forced planting regimes in the Moluccas in the seventeenth and in Java in the eighteenth century (see below), encompassing corvée labour regimes in Indonesia emerged only in the nineteenth century. Still, the Dutch used all sorts of coerced labour in Indonesia from the moment they were able to. By monopolizing markets and putting local, existing principles of mandatory and social or community labour to its own economic use in Ambon and Java, the Dutch East India Company (VOC) had sketched out what would ultimately become the notorious cultivation systems of coerced cash crop production in Java (since 1830) and West Sumatra (since 1848). In all these cases, systematized coerced cultivation followed colonial warfare as an ad-hoc solution to repair, reorganize and profitably govern the chaotic debris of conquest. As such, the Dutch gained over 300 years of experience in using coerced labour. In the late nineteenth century, they arrived at a refined methodology of inventing and legitimizing all sorts of corvée services. Nowhere in Southeast Asia do we find such an elaborate, extensive, overused and (supposedly) uniform system of corvée labour as in Indonesia around 1900.

Coerced services were anchored in Dutch interpretations of existing labour obligations, in many cases supposedly rooted in conditions of social hierarchy and rights to land. In practice, during the nineteenth century, this crystallized into a collection of intermixed and overlapping complexes of coerced cultivation and various mandatory services. Corvée labour became the umbrella term for these services that covered all sorts of differentiated forms of community duties, reciprocal household obligations and mandatory services performed in lieu of taxation to support landlords, village chiefs, aristocracies and the colonial state.

In many village societies in Java and Sumatra, for instance, people performed specific community or household services of mutual assistance. These were performed based on moral obligation and generosity (for instance after natural disaster), or on reciprocity and labour exchange in return for access or entitlements to usufruct rights to land, protection and purveyance, to build collective security (of mutual subsistence) and economically resilient communities. These services, in many parts of Indonesia known as ‘gotong royong’, were hence ‘coerced’ not because peasants were bound to its performance by rulers or states, but rather because without performing them they were not entitled to these provisions. In colonial jargon, they were (sometimes) called desa (village) services, although it might be possible that these also included services that locally were not considered gotong royong; the colonizer often mixed-up various types of services and generalization of specific, local terminology often violates the local specifics of the terms under which services were delivered. Nonetheless, desa services and gotong royong begun to overlap with other labour obligations and as such were interpreted and used by the colonial (and the subsequent Indonesian national) state as yet another ‘traditional’ form of mandatory labour but now also in service of the state (Bowen, 1986; Van der Linden, 2017, pp. 491–492; Suwignyo, 2019.

These complexities caused quite some confusion within colonial officialdom which echoes on in current day literature. This has remained somewhat neglected in historiography, especially in comparison to other forms of unfree labour, even though various corvée services often overlapped and conflated with coerced cultivation and originated in the same economic motivations (Bosma, 2014). Services were no longer merely used in agricultural village communities to sustain themselves or as an obligation to landowning peasants or the aristocracy, but slowly also in support for production of commodities and cash crops. They contributed much to maintaining colonial export economies, but in themselves also formed a major ‘hidden’ fiscal contribution to state revenue by helping to reduce the cost of managing and performing state activities (Waijenburg, 2018), including, in Indonesia’s case, constructing and maintaining infrastructure and irrigation networks, performing postal services, manning fire brigades and many other public activities (Manse 2021a, pp. 202–203). The pretence of imitating or continuing existing practices covered a reality in which the Dutch actually designed new institutionalized labour regimes that reproduced the services and the terms under which they were levied. Corvée itself was no colonial invention, but the way it was wielded as a pluralistic instrument of governance in the modern era was.In

Indonesia, corvée also played a key role in the design of the organization of fiscal-political relationships. In Java, for instance, many services were paid to enable peasants to pay other monetary taxes to the government. Moreover, monetary head and income taxes were intended as a replacement of coerced cultivation and corvée services (Hup, 2021; Manse, 2021a). Additionally, both services and taxes were based on an acclaimed right to overlordship, while simultaneously levying corvée and taxes was often accomplished only by co-opting indigenous institutions (Hup, 2021) and local customary law or adat. This was formalized in the 1854 Government Regulation, which apart from ordering structural reduction in corvée services also stated that the nature in which services were performed and the conditions under which they were claimed, had to be constructed ‘in accordance to existing customs, establishments and needs.’⁵ This required a complex reconstitution – if not reinvention – of the many customs and ‘traditions’ of rulership and land-right holding, making corvée labour as much an economic-fiscal as a political problem.

Against the backdrop of these developments, coerced labour in Indonesia was, at least on paper, increasingly restricted and replaced with monetary head-taxes. However, many local officials noticed that, much to their own frustration, they were unable to enforce these restrictions and levy monetary taxes as the desired alternative. The last cultivation services in sugar and coffee were, for instance, only abolished in 1917/1919 (Houben & Seibert, 2013, p. 181). In fact, use of corvée increased following the general expansion of the Dutch empire in Indonesia and its ever increasing labour demand. In many newly conquered territories – even in those in which no socio-legal foothold for corvée seemed to exist – corvée services were introduced as a method to subject and domesticate populations in name of the moralizing ideas of ‘ethical colonialism’ (Manse, 2021b; cf. Booth, 2016, pp. 290–301). As much as they would have liked to, the Dutch could never do without corvée labour. Especially where economies remained largely uncapitalized (despite the advances of Dutch industry and its gigantic profits) the Dutch kept relying on service performance. This was, for example, the case in Palembang, the part of Southern Sumatra exemplified by Haji Agus Salim, where people were made to perform more than three times as many services as legally allowed, the majority of which were levied as ‘municipal services in accordance to local adat.’⁶

In order to fully comprehend the extra-economic role of corvée labour in (Dutch) colonialism, it is necessary to closely investigate how they were designed and used on the spot. Ambon, West Sumatra and Java, though culturally diverse and different, had in common that they were all characterized by similar systems of communal and individual rights to land. Dutch interpretations of these systems shaped new orders and hierarchies at village level which enabled levying of collective forms of coerced services but later on impeded the introduction of individual monetary taxes.

Corvée and coerced labour in Ambon

After forcefully monopolizing the Moluccan spice trade in the seventeenth century, the Dutch used Ambon’s most fundamental, indigenous political units, called nagari, as starting point for organizing the system of coerced production and governance. These nagari consisted of a number of kinship-related clans unified under and represented in the nagari council by a clan chief (Cooley, 1962b, p. 7, 1969; Van Fraassen, 1987 I, pp. 140–141, 173–264). One of these clan chiefs also headed this council as nagari chief, representing the entire nagari to the outside world. While clans had a dualistic social-religious function to guarantee spiritual and social wellbeing, the division of resources among villagers was organized through different lineage-groups called dati, which existed independently from the clans and coordinated usufruct-rights to land. Members of dati collectively held these rights and together cultivated sago (the principle Moluccan staple food) on these lands (Van Wouden, 1968, pp. 76, 148).

Participating in dati incurred the obligation to protect, develop and contribute to it by performing services (Van Hoëvell, 1875, 202–218; Adatrechtbundel Vol. 21, pp. 28–30, 33; Cooley, 1962a, pp. 57–58; Knaap, 2004, pp. 207–208). The Dutch used these services as base to demand forceful cultivation of predetermined amounts of cloves on dati lands by dati groups which were transformed into concentrated territorially defined units, all under a strict monopoly (Andaya, 1993, pp. 41–44, 151–156, 174–175; Knaap, 2004, pp. 23–25). Further, under company rule all dati were expected to deliver one man to participate in military expeditions called hongi, deployed annually to enforce the monopoly (Knaap, 2004, p. 39). Allocation of land was increasingly used as reward or compensation for services performance and tax payment (Knaap, 2004, pp. 178–185). The VOC drafted ‘dati-registers’ to administer the coerced cultivation services people had to perform and number of clove trees they had to maintain (Van Fraassen, 2018, p. 277), but also to ensure no new dati were established without Dutch permission (Adatrechtbundel 21, p. 28). Villagers were prohibited to live outside of or leave their nagari without their chiefs’ permission (Knaap, 2004, pp. 169–170, 218–219). The collected chiefs were made responsible for enforcing profitable clove production, to supply the VOC with fixed amounts of cloves in return for various compensations, salaries and privileges, which enhanced their status and dignity, while alternative political alliances and structures were debilitated (Cooley, 1962a, pp. 59, 142; 1969, 149–151, 158; Bartels 1977, pp. 14; Fraassen, 1987 II, pp. 460–502; Van Fraassen, 2018, pp. 208–209).

This way, the Dutch transformed a relatively mobile and dynamic society into territorially bound communities of agricultural `production, and its elected representatives into authoritarian rulers. New mandatory services were invented to support the Dutch administration under the name ‘court services’ (hofdiensten). Additionally, people performed ‘nagari services’ for their local community, for instance to construct and maintain village property such as council meeting houses or churches and mosques, and ‘kwarto services’, originally military services levied by nagari chiefs, but under the Dutch also levied for the construction of warehouses, fortresses, infrastructure and ships. Kwarto services’ were levied from one to five subjects per dati unit per month (Knaap, 2004, 228–259; Van Fraassen, 2018, pp. 140–147, 347–348). Its performance depended on proximity of villages to specific projects; the closer villages were located to fortresses or roads, the likelier its inhabitants had to maintain them in forced services (Knaap, 2004, pp. 181–183).

Essentially, these supposedly ‘traditional’ services were the earliest forms of registered and legalized unpaid corvée labour in the Dutch colonial empire. They were legitimized by referring to the alleged lack of proper working ethics among the Ambonese due to the supposed constant availability of sago as staple food, which supposedly prevented the rise of large-scale rice cultivation and its accordant state institutes and fiscal regime (Van Hoëvell, 1875, pp. 52–55). To the colonial mind, this explained the ‘careless’ attitude to life of the Ambonese and their inability to independently engage in more complex forms of political and economic organization beyond subsistence. Hereto, it was theorized, they required European supervision and coerced labour, to discipline them and make them into productive, organized state-subjects (Van Hoëvell, 1875, p. 75). This way, coerced services were claimed to be to the benefit of the people who performed them, and as such became a principle instrument to build and explain state-subject relationships.

In the late the eighteenth century, the Dutch spice-monopoly was breached, and Ambon’s economy slowly collapsed. Enlightened governors in the early nineteenth century attempted to dismantle the coerced production system, but found that 150 years of coerced labour policy, grafted on decades of structural reform of local social institutes, was not that simply eradicated (Van Fraassen, 2018, pp. 145–152). Though the hongi expeditions were abolished, the monopoly itself was continued because of the generally poor condition of colonial finances (Van Fraassen, 2018, pp. 146–147). While large-scale cultivation systems were simultaneously being set up in Java and West-Sumatra, coerced spice cultivation in Ambon was continued.

In the 1860s, a sudden drop in the spice market aggravated economic decline and undermined popular disinterest in clove-cultivation, and the Dutch finally abrogated the (by then no longer profitable) monopoly and stopped buying cloves for a guaranteed fixed price in 1868. This had profound consequences for the chiefs, who had become the true linchpins of the colonial coerced production system. The Dutch presumed rendering the spice market free would cause production to decrease, stimulating a price increase and motivating the Ambonese to restart clove production and earn taxable incomes.⁷ Simultaneously, they introduced a ‘nagari tax’, which in fact collectively taxed not the nagari, but the dati groups, through the dati chiefs who were made personally responsible to pay the tax on behalf of their members, to the nagari chiefs, who surrendered revenue, after subtracting their due shares, to the Dutch.⁸ Since dati chiefs were lineage-representatives rather than landholding chiefs, this system politically made little sense. As the underfunded local government had no administrative means to calculate the participation of each dati member at an individual level, it chose to operate through the same old structures as provided by the spice-monopoly.⁹

Simultaneously, the majority of nagari services were formally abolished in 1869 and the burdensome mandatory cultivation services prohibited in 1881 (Van Fraassen, 2018, pp. 453–454).¹⁰ Consequently, the chiefs lost most of their direct shares in popular labour power. They were entitled to an eight percent ‘collectors wage’ of monetary tax revenue.¹¹ Such shares and salaries were deemed important to support the chiefs’ material welfare and prestige, on which colonial authority depended. In various nagari however, economic stagnation further depressed tax revenue which became insufficient to fund the chiefly salaries and cuts. The state’s parsimony, characteristic of colonial regimes around the world (e.g. Frankema, 2010), withheld officials from making the necessary investments – each region was supposed to fund itself. Consequently, the once prestigious and wealthy spice-lords became impoverished village-kings, clinging to a glorious but bygone past.

One contemporary official, G.W.W.C. baron van Hoëvell (Resident [governor] of Ambon in 1891–1896), proposed to simply reduce the costs of administration by decreasing the number of chiefs by merging nagari.¹² This encountered severe resistance from the ruling classes as it violated local succession rights and ruling traditions.¹³ Still, the chiefs faced a significant welfare decrease, which worsened when after a temporary recovery in 1874–1890 clove prices plunged again in the 1890s, never to fully recover. Many impoverished chiefs resorted to corruption, embezzling tax money or continuing use of (formally abolished) services, causing further political crisis.¹⁴

In 1891, the government implemented a new corvée regulation that stipulated personal liability to corvée performance, rather than the collective liability of the dati. This was followed by an individual head-tax, replacing the old nagari tax to target individual incomes, reach deeper into the nagari and accumulate more revenue.¹⁵ This head-tax was dogged by similar pragmatic difficulties and arrears, and its revenue remained insufficient to fully fund the chiefs’ costly lifestyles.¹⁶ Furthering impoverishment of the chiefs wreaked such havoc on their prestige that their political authority diminished. They had become underpaid government employees, whose rapidly diminishing prestige and improper compensation structurally hollowed out the sociolegal fabric of society (Chauvel, 1990, 11–14). The political and material privileges of being chief no longer outweighed its administrative and moral burden and enthusiasm to fulfil the office gradually reduced.¹⁷

In 1919, new monthly allowances were introduced but simultaneously a number of older compensations were abolished and the remaining kwarto services were replaced with a 5 guilder stipend per ‘able-bodied man’, which most Ambonese were unable to pay.¹⁸ Taxes were structurally exchanged for formally abolished labour services, now performed outside of the government’s scope. This caused distrust and dissatisfaction, to which emerging nationalist movements eagerly responded by vocalizing popular grievances against taxation and corvée leading to small resistance and avoidance (Chauvel, 1990, 116–119; Van Fraassen, 2018, 353).¹⁹

As such, the problem of transforming corvée services into monetary taxes became very much a problem of indirect rule. The imposition of monetary taxes and its limited success in generating revenue undermined the supposed political legitimacy of the state, and impaired the wealth, prestige and political authority of the ruling institutes upon which authority hinged. The orderly world of supposedly transformative and disciplinary taxes and services levied by self-funding indigenous elites imagined by colonial officials from the 1860s onward required abolishing the chiefly entitlements that had made indirect rule work. The assumed high costs of administration did not allow for anything else. These entitlements had become deeply interwoven with fundamental aspects of local social organization. Thus, path-dependent on these institutes and unwilling to invest in Ambon’s collapsing and increasingly under-prioritized economy, Dutch officials, ill-prepared to provide new answers to old questions of fiscal organization and governance, unintentionally continued the fundaments of coerced labour policy under the pretence of monetary tax and modern bureaucracy, to which the elites responded by continuing use of, from then on, ‘illegal’ corvée services.

Labour obligations in Java

Java provided fertile ground for Dutch schemes of taxation, exploitation and indirect rule. The supreme rulers of the island’s predominant rice-cultivating sultanates delegated fiscal administration to lower, apanage-holding aristocrats, who managed layers of tax-collecting middlemen and redistributed rights to land and the fruits of labour. Control over men overshadowed claims to territory; only by recruiting indispensable labour power could rulers harvest territorial resources. As chiefs in their own rights, local tax collectors and village notables held their own rights, based on office-holding, to collect taxes and services, channelling resources to, ultimately, the central courts (Carey, 1986, p. 65; Moertono, 1985, p. 61).

In the eighteenth century, the Dutch begun participating in this system. Following a series of wars and treaties with indigenous states, the Dutch claimed sovereignty over large chunks of Java and entitlements to apanage rights and local labour power (Carey, 1986). The duty to perform services was seen as connected to land tenure (Schoch, 1891; cf. Van Vleuten, 1872a; Van Vollenhoven, 1919). Thus, to understand colonial coerced labour in Java, we must fathom Dutch interpretations of land tenure and the labour obligations it allegedly brought.

Javanese peasants rarely owned land as property. Rather, they held specific rights of avail which brought specific fiscal obligations to rulers who offered protection. The peasants who held these rights (sikep) delegated the fiscal burden it brought downwards, by leasing out shares of their land-rights to, or accepting labour in exchange for access to land from ‘landless’, patronage-bound, sharecropping dependents, who then performed services on the sikep’s behalf.²⁰ Based on their territorial claims, the Dutch begun demanding forced delivery and labour. After the VOC’s demise in 1799, governors H.W. Daendels, T.S. Raffles and their successors initially attempted to curtail such ‘illiberal’ and ‘feudalistic’ practices. However, their anti-aristocratic reforms, combined with diminishing living standards as result of decades of political crisis, erupted into the Java War of 1825–1830 which caused such a severe crisis in Dutch public finances that the Dutch king ordered to adopt a full-fledged coerced cultivation system of cash crops (sugar, indigo, coffee and tea) traded under a state monopoly to turn colonial costs into profits.

As in Ambon, this cultivation system consolidated specific aspects of social organization into an essentialist model upon which coerced services and taxes were grafted. The Dutch, assuming a new role as ‘feudal landlord’, distilled from complex local duties a system of coerced cultivation and corvée services, performed for the state, local rulers and village communities and chiefs. As in Ambon, chiefs and rulers participated in profits and received rewards and compensations, enhancing their political authority, wealth and status. Landholding was stimulated to maximize the number of taxable persons and fulfil the growing labour demand for mandatory production of cash-crops by peasants on designated parts of their lands (Breman, 1983, Breman 1989, pp. 43–44, p. 15; Holleman, 1981, p. 157). In some areas, it became customary to make all cultivators co-landowners after three years of sharecropping, and thereby liable to perform services on a personal base (Van Vleuten, 1872a).

In the case of Java, corvée services developed in tandem with coerced cash-crop production, but were left largely unregulated. The terms according to which they were levied remained unclear and diversified (Schoch, 1891; Fokkens, 1901-1903 I, pp. 3, 107). From 1854 onward, the government attempted to regulate, rationalize and monetize Java’s hitherto chaotic corvée system through ‘quinquennial revisions.’ In 1864, the second of these revisions defined a ‘corvée service’ as one day of twelve hours of labour, at a maximum of 52 days per person annually, depending on landholding – peasants with ‘half’ a share in land performed 26 days of labour.²¹ By 1929, the standard amount of services was 26 (or 13) days of service, which was still much more than the 6 days of labour demanded by the British in Malaya (Lubis & Nasution 2003, p. 60), or the 16 days of services demanded by the French in Indochina (Van Waijenburg, 2018, p. 73; Booth 2019, pp. 48–49).

Services had to take place eight paal (ca. 12 km) within place of residence. If performed further away, an overnight stay and travel time was calculated as part of the service. A theoretical distinction, inherited from precolonial times, was adopted between ‘regular’ and ‘special’ services.²² ‘Regular services’ were paid and used for construction, repair and maintenance of interregional infrastructure and major irrigation works (Van den Haspel, 1985, pp. 7, 16). ‘special’ or ‘irregular’ services were unpaid, and had been used for emergency repairs after natural disasters. During the cultivation system, they begun including various other ad-hoc services, which previously might have been considered gotong royong, such as delivering mail, occupying guardhouses, patrolling villages or guarding and maintaining government buildings or even private sugar fields (Van Bowen, 1986, p. 548; Margana, 2007, pp. 105–106; Van Vleuten, 1872a, p. 217).

A stricter bureaucracy of registration of services in standardized forms was imposed in 1851, but administration of corvée was still delegated to local chiefs who themselves were usually exempted from service performance and had an interest in maximizing local labour.²³ This led to extortion and maladministration, which contributed to the unpredictability and uncertainty of peasant life and led to an unbalanced distribution of the fiscal burden (Pieren, 1884, p. 4; Elson, 1994, pp. 9–12, 145). As in Ambon, the Dutch had reorganized relatively dynamic village hierarchies into the ‘feudalized’, ‘corrupted’ and ‘despotic’ society they claimed to eradicate (Elson, 1994, pp. 34, 155–161).

Subject to increasing critique by for its burdensome effects and monopolist character, the coerced cultivation system was gradually phased out from the 1860s onwards. All services were supposedly replaced with taxes, allowing peasants to ‘buy-off’ services by paying up and arranging a substitute labourer.²⁴ This transformed into an institutionalized, permanent exchange of various services for monetary taxes, which allowed the Dutch representative at the ILO in Geneva to present coerced labour in Indonesia as an alternative to taxation, and argue that people were allowed to choose between both.²⁵ In reality, this ‘choice’ was predetermined by social and practical conditions informed by the local labour demand and level of capitalization as well as the political relations between the Dutch and local indigenous elites, and between these elites and their subjects. Monetary taxes were structurally levied to a lesser extent than they were supposed to (Fokkens, 1912, 1914; Hasselman, 1905; Meijer Ranneft & Huender, 1926).

As in Ambon, in Java’s case this had much to do with the predicaments of indirect rule. Provincial and district rulers participated in the fruits of coerced labour which contributed to their personal households and prestige. Therefore, abolition of specific personal aristocratic ‘pancen services’ was endlessly delayed out of fear to alienate Java’s indispensable intermediating aristocracy and instigate another revolt. Many rulers reportedly favoured services over shares in monetary taxes.²⁶ When finally replaced with a head-tax in 1882, an exception was maintained for personal services at village level, allowing Dutch administrators to turn a blind eye when local rulers claimed shares in these after all.²⁷ Consequently, on paper, various services were abolished and the maximum number of labour days was reduced to 42,²⁸ but in practice, these restrictions on the uses of corvée were often ignored (Fokkens, 1885, pp. 15, 1914, p. 8).

A series of surveys by colonial civil servant Fokko Fokkens on Java’s integrated principles for landholding and corvée liability defined, and thereby created, new forms of corvée liability depending on the number, rather than size of shares in land that peasants enjoyed. These investigations adopted an ideological bias in discerning individual patterns of land-, income- and property-holding, all to appoint as many individually liable taxpayers as possible, regardless of their actual wealth or total size of their land rights (Goh, 1998, p. 29). This had profound consequences. In Kedu (central Java) for instance, peasants with multiple small shares in land were subsequently assessed to a larger extent than peasants with one large plot of land, even if their accumulated total of land they had access to was smaller.²⁹ In Banten, West Java, the government made dependent share-croppers (rather than only sikep) individually liable for head-tax payment, which drastically increased the tax burden per household, causing a rebellion necessitating the measure to be revoked (Kartodirdjo, 1966).

Such recurring revolts were often motivated by a fusion of religious and political-economic tensions. They responded as much to the regressive effects of misguided Dutch fiscal policy as to the deep social consequences of other aspects of colonial rule (Kartodirdjo, 1973). Revolt and silent resistance were used as a tool to negotiate in matters of taxation and service performance, which itself became a platform through which the state could be encountered, deflected, and contested (Manse 2021a; Manse, 2022b). While not immediately forestalling Dutch strategies of reproducing local realities through investigation and knowledge production by selecting and overgeneralizing practices that suited colonial ambitions (Stoler, 2009, pp. 29–13), this did force the Dutch to temper some of their state-building ambitions. At the same time, where the state met its limits, indigenous chiefs kept finding new ways to retain and expand their claim on local labour power, predominantly by using the largely unregulated local ‘village services’, which were part of, or overlapped, with gotong royong, and took place outside of the government’s control but in the same space as corvée services. Differences between them thus might not have always been entirely clear to colonial officials, even though in practice peasants seemed to make clear distinctions between which services they performed for their lords, which for the government, and which for the village or each other (Bowen, 1986, p. 559). Still, every diminishment of corvée performed for the government and the local aristocracy seemed to be accompanied by an expansion in the use of village services. Over the course of the nineteenth century, the function of village services had drastically changed. Originally, they comprised a variety of regionally specific duties, performed depending on social position, landholding or relations to local chiefs. They were generally thought of as more democratic because desa administrations were elected, but not all officials were convinced of their value or traditional base (e.g. Van Vleuten, 1872a, pp. 218–221, 224–225). During the nineteenth century, these distinctions blurred (Van den Haspel, 1985, p. 7), and by 1890, they had become a replication of government corvée levied at village level (Fokkens, 1901-1903 II, pp. 4–5, 8). Lack and refusal of proper supervision – the Dutch government principally refrained from intervening in the supposed self-organization of village politics – basically gave village chiefs carte blanche to maximize use of these local services for their own interest (Fokkens, 1902). Further, buy-off of various local services generated an extra source of income for village chiefs (Fokkens, 1901-1903 II, pp. 1, 5).

A report on village services concluded that an informal labour economy of ‘illegally levied’ desa services had emerged since the government had begun actively diminishing corvée. These services, its author argued, were a proto-capitalist form of reciprocity within self-regulating communities, not to be meddled with by the state (Hasselman, 1905). Such argumentation, of course, reduced the state’s responsibility for the impact colonial policy had had, which allowed for the continuation of various forms of corvée as local village services. In 1901, an overwhelming estimated amount of 200 million days of desa services (versus 25 million days of corvée services) was still levied annually.³⁰ These were essential to the functioning of the village in the absence of a properly functioning tax state. Buying off village services remained unpopular (Fokkens, 1901-1903 I, pp. 15–16). Hence, village chiefs kept recruiting villagers to maintain roads, man the fire brigade and police and patrol the village at night, making Java of the 1900s still very much a ‘night-watchers state in the most literal sense’ (Breman 2010, p. 133).

Corvée labour in West Sumatra

Dutch colonization of West Sumatra resulted from intervention in local civil war, which broke out in the 1820s when Padri, Islamic reformists returning from Mecca, sought to violently align local society to Wahabi teachings (Dobbin, 1972, pp. 5–10; Hadler, 2008, pp. 972–979). The Dutch, fearing the Padri’s influence on Sumatra, sent troops, and ultimately managed to quench the war in 1837, consequently adding a new province to their empire (Dobbin, 1983; Hadler, 2008, pp. 985–987). The challenges the Dutch subsequently faced were of similar nature as those in Java and Ambon: how to govern newly acquired territories devastated by war and political division at the lowest possible costs?

As before, the answer was found in operating through local customs and elites, which on West Sumatra proved particularly complex. As in Ambon, the principle socio-political units of the Minangkabau, the main ethnicity of the West Sumatran highlands, were called nagari. In West Sumatra, these were village communities consisting of multiple clans (suku) of lineages and sub-lineages that were further sub-divided into households (Benda-Beckmann & Benda-Beckman, 2013, pp. 41–42), that supposedly descended from a common foremother. Household members communally held usufruct rights to their lands and property (harta pusaka), which was inherited in matrilinear line (Benda-Beckmann & Benda-Beckman, 2013, pp. 48, 69, 77). Men held no such rights. They contributed to their mothers’ and sisters’ households – even after marriage (Von Benda-Beckmann 1979, pp. 108–110; Graves, 1981, p. 6) – and were expected to reclaim new lands, property and wealth, to add to their lineages’ property, but could not lay personal claim to any of these (Van Fraassen, 1970; Von Benda-Beckmann 1979, pp. 149–151).

Each social entity – from household to village – was headed by a representative. For the suku, these were called penghulu. They administrated social occurrences (marriage, death, birth) and consulted with surrounding penghulu in village councils to maintain intra-nagari order. The lineages were represented to the outside world by mamak kepala waris, who bore responsibility for their families’ wellbeing and property, and mediated in matters of property and inheritance under supervision of the penghulu.³¹ Because of high degrees of diversity in adat, many such matters were constantly deliberated and resolved through consensus-seeking (Postel-Coster, 1985, pp. 7–18, 34–38; Von Benda-Beckmann 1979, pp. 81, 155–157, 164). Unlike Java, Minangkabau society knew no strong, service-demanding aristocracy, but was characterized by consultative governance and communal labour responsibility, coordinated through constant socio-legal arbitration (Von Benda-Beckmann 1979, pp. 139–215).

This changed dramatically under Dutch occupation. The Dutch appointed ‘head-penghulu’ to gain influence over local councils and maintain the appearance of consultative and participative governance. In reality, the role of adat chiefs, with whom the Dutch had cooperated during the Padri-war, transformed from elected representatives to territorial chiefs. This way, social organization was remodelled towards the repressive colonial-capitalist structures of coerced labour and indirect rule for coerced production of coffee, the principle cash-crop in West Sumatra’s highlands. Hereto, nagari were made the basic administrative, territorial and socio-political units and grouped into overarching sub-districts called laras (which originally were loosely structured, non-territorial nagari networks), under chieftainship of selected penghulu (tuanku laras) who were converted into salaried supervisors of coerced labour (Von Benda-Beckmann & Von Benda-Beckmann 2013, p. 42, 66; Hadler, 2008, p. 987).

In 1862, coffee production was made compulsory, and private production and export prohibited. As in Ambon, the Dutch created specific working groups, kerja rodi, for coerced cultivation and adjacent corvée services, supervised by more invented officials, penghulu rodi, who were awarded fixed shares and production-dependent bonuses (Van Hasselt, 1882, p. 188; Kahn, 1980, p. 123; Kielstra, 1888, pp. 1453–1455, 1480–1482). Rodi services were levied to construct secondary access road, but main roads for which, technically, government corvée services had to be used (Hamerster et al., 1928 I). This way, the Dutch created an increasingly powerful ‘false adat elite’ (Hadler, 2008, p. 990). Communal labour duties traditionally performed for the penghulu were transformed into corvée services for constructing government buildings, storehouses and infrastructure and transporting coffee downhill to Padang (Graves, 1981, 63, 79; Kahn, 1980, pp. 164–165; Kielstra, 1888). Infracting in the appointment of chiefs and the pusaka system by overthrowing specific rights to adat positions had, as elsewhere in Indonesia, a lasting institutional impact. Chiefs had to bridge a widening gap between their socio-religious responsibilities towards their local communities and their duties to the colonial government. This artificial combination of adat-leadership and governance in service of exploitation pretended a myth of political partnership that became increasingly untenable.

As in Ambon and Java, West Sumatra’s coerced cultivation system was deconstructed not for moral, but pragmatic, political and economic reasons. In the 1880s, coffee production stagnated due to soil exhaustion, leaf disease, and (consequent) local disinterest, causing economic collapse and deficits (Kahn, 1993, pp. 172, 177, 204–205; Young, 1994, p. 26). The coerced production system had generated quite some regional inequality; in 1895, 75% of the total tax burden in West Sumatra was calculated to be carried by coffee producing families in the highlands, assessed at around half of the population, while peasants in the lowlands remained largely untaxed. Coffee had become ‘the only source of welfare being taxed (Manse 2022a).’³²

To solve this problem, government advisors suggested introducing similar head-taxes as in Java and Ambon, which equally conflicted with pertinent aspects of altered indigenous society rendered by coerced cultivation.³³ For instance, men had been made responsible to perform coerced services in their place of birth, supposedly following the Minangkabau principle that men contributed to their mothers’ households, lineages and villages – even though they sometimes lived miles away.³⁴ Furthermore, while coerced labour systems were indifferent to the individual burden they imposed upon people as they collectively ‘taxed’ labourers aggregated in household economies via the requirement to deliver crops via local chiefs, head-taxes supposedly equitably targeted (often inexistant or unknown) individual incomes. Finally, the proposed head-tax was biased towards men, always seen as the proper ‘family heads’ and taxable subjects by colonial civil servants. In Minangkabau society, such heads did not exist; political power and authority over property and income were divided between men and women.

Consequently, officials erroneously presumed Minangkabau men lacked incentive to develop proper work ethics to accumulate capital and income required for personal taxation (Heckler, 1905-1906, pp. 64–82). They were considered ‘poorly taxable subjects’ who ‘miss stimulus to increase their possessions to provide their families with higher prosperity’, so that ‘while the women work at home […] the men wander around.’³⁵ This was not true; men were expected to contribute to their families’ property. They simply had no personal rights of avail or inheritance over it. Still, colonial officials puzzled for decades over how to tax men at an individual level without violating the property rights of women. While some argued to continue coerced cultivation,³⁶ others advised to impose head-taxes upon men individually, enforced by coerced labour in case of non-payment (Heckler, 1905-1906, pp. 88–91). Their dispute obstructed the transformation of coerced labour into monetary taxes for many years.

By 1907, no income tax had been introduced in the Minangkabau highlands. Pressured by the urgency to replace coerced labour with monetary taxation due to the introduction of taxes elsewhere in the archipelago that year, the colonial government ultimately decided to designate harta pusaka as taxable objects and mamak kepala waris as accountable for their lineage members’ payment, enforced by the threat of confiscation of harta pusaka in case of non-payment.³⁷ As warned by several local officials, this was an enormous violation of adat and posed an immediate threat to the livelihood of families (Heckler, 1905-1906, pp. 88–91, 92–105). The tax caused much resentment in Minangkabau society. When introduced in 1908, many penghulu refused to cooperate. As Dutch officials attempted to enforce payment using military assistance, the situation escalated, developing into a full-fledged anti-colonial revolt which was harshly suppressed by the military, sending shockwaves throughout the region (Amran, 1988; Oki, 1977; Young, 1994). The revolt’s causes and consequences were furiously debated within the Dutch administration, parliament and the media. No isolated event, the revolt should be seen as an important episode in the structural tensions that the introduction of monetary taxation caused across Indonesia (Manse, 2021a), in West Sumatra’s case also as a response to decades of infraction in political leadership, family life and social organization (Oki, 1977, pp. 80, 103).

The political climate in West Sumatra never recovered. Further violation of land rights and increase of the tax burden (Hamerster et al., 1928) prompted more resistance, which ultimately fused with political activism leading to a series of brutally suppressed communist uprisings of 1926–1927 (Kahin, 1999, pp. 35–36, 46–49,; Kahn, 1993, p. 210; Oki, 1977, pp. 91, 113).³⁸ Taxes were paid poorly, which was blamed by some contemporary administrators to the purported subversive and contentious nature of the Minangkabau (Liefrinck, 1917 II, p. 1). In reality, these socio-political tensions, as well as low wages and a putative ‘lack of coolies’ generated a climate that was pragmatically unfavourable to levying monetary taxes, further aggravated by the crisis of the 1930s. Local officials kept relying on burdensome corvée services until the end of the Dutch empire.

Conclusions

Corvée labour is impossible to ignore in the case of colonial Indonesia as an important political and fiscal device that pillared and expressed the colonial state in many ways. Massive amounts of services were levied in, more than in many other colonial empires, as a consequence of the strong grasp the Dutch claimed over local sovereignty and ‘traditional’ indigenous society and its different forms of communal labour duties and services. Coerced labour was no simple form of labour exploitation per se, but a necessity that could only have been engaged in through careful legitimization. Thus, colonial corvée services were presented and legitimized, both in the colony and on the international stage to legitimize, as a continuation of such ‘traditional’ services. But crucially, while historically often connotated with premodern forms of social organization, in Indonesia corvée services were not, or no longer, levied by some ‘feudalist’ lord or ruler, but by the modern (colonial) state.

In the case of Indonesia, but also in specific parts of the French and British empires in Africa, coerced labour became a central component in the political organization of the colonial state itself for multiple reasons. First, making people work on vital aspects of the state’s economy, such as its infrastructural projects, supposedly generated new bonds between people and the state. Second, through corvée regimes, subject populations were registered, standardized and prepared for a new role in supposedly uniform fiscal regimes and modern economies. Third, allocation of corvée labour was used to smoothen ties with indigenous elites to support the delegation of governance through indirect rule. Finally, services were used as a civilizational tool that provided an important edifying force to integrally discipline populations, govern socioeconomic behaviour and influence customary law and political organization in which corvée was embedded. By levying ‘traditional’ services, colonial officials believed they could transform ‘traditional’ indigenous society from within.

In many colonial empires, including in Indonesia, corvée became intimately tied to taxation, as either a ‘tax in labour’, or as a replacement of ancient types of service-levying, to supposedly prepare colonized populations for more advanced forms of subjecthood and economic organization under the banner of modern, civilizational or ‘ethical’ colonial policy. The Dutch considered the transition from coerced labour to monetary tax an essential part of this discursive modernization process. In 1854 already, the Dutch colonial government attempted to endorse the gradual but full abolition of corvée (and other forms of coerced) labour, by replacing services with monetary taxes. Eighty years later, little of this ambition remained, as it turned out levying of services abounded across the archipelago. Obstructions and practical constraints in levying monetary taxes promoted local chiefs and rulers to revert to levy services; the deep imprints corvée had left in the colonial organization of indigenous society impeded the accumulation of capital necessary to pay tax and rendered many fundamental changes to the fabric of social organization across Indonesia – often transcending questions of labour supply and demand or fiscal revenue. Because corvée was structurally ingrained in ‘traditional’, precolonial service-levying and patterns of rule and reciprocity during the era of coerced cultivation, it became vital to the organization of many of Indonesia’s developing political economies. In other words, decades, or in the case of Ambon, centuries of using coerced labour as a governing tool was not that easily eradicated, as it had become a pedestal to expressing colonial power relations, organizing the state and supporting procedures of governance.

As a result, while corvée services were slowly abolished on paper, and replaced with monetary taxes, in reality both Dutch officials and indigenous rulers held open their options to continue using services as they found that in the daily realities they governed, corvée often provided more convenient and cost-effective answers to the problems they faced than monetary tax, involving the lowest amount of risk. The implementation of individual taxes as an alternative to communal labour services could in itself be reason for resistance, be it because of outright misguided fiscal policy or because of the structural inability of colonial officialdom to understand society in its own terms, as demonstrated by the revolts caused in Banten and West Sumatra. Furthermore, the chiefs who were supposed to enact the transitionary policy from coerced labour to monetary tax of course prioritized their own interests over the modernizing ambitions of the state. Since many of them profited from the levying of services more than they could from monetary taxes, corvee services were, while being replaced with monetary taxes across sovereign states in Asia in the nineteenth century, tacitly continued in colonial state, in particular in Indonesia. Colonial fiscal policy had some success, and taxes were introduced in virtually all corners of the Indonesian archipelago, but the continued usage of corvée labour in various forms kept overshadowing these successes.

This forces us to rethink the importance of corvée in context of other forms of coerced labour in and beyond colonial situations. While certainly also exploitative, colonial capitalist tools used to gain a direct charge on indigenous labour power, corvée services were even more important than that. They were instruments to explore and change popular behaviour, alter and, supposedly, ‘improve’ the living conditions of local populations and communicate and execute governing strategies. They fulfilled fiscal, legal, administrative and political purposes, assisted in the discursive legitimization and reasons of empire to shape the colonial state-building process and determine fiscal state-capacity and occupied a central position in the relationship between the state and subjects, similar to taxes in modern societies. Without them, the ‘modern’ Dutch colonial empire in Indonesia in the nineteenth and twentieth century could not have existed.

Autobiographical sketch

Maarten Manse is lecturer in history at the Vrije Universiteit in Amsterdam and a postdoctoral researcher at Linnaeus University in Sweden. He specialises in the political, socioeconomic and cultural history of Indonesia under colonial rule.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Arsip Nasional Republik Indonesia [National Archives of Indonesia; hereafter ANRI], General Secretariat Archives [Algemeene Seceratie; AS], ‘Grote Bundel’ [GB] TGA 8368, herein: Director of Interior Administration [DirBB] to Resident of Palembang, 4-9-1929. He drew this example from an article published in the journal he edited, Fadjar Asia, written by a certain Hambali bin Haji Ahjat (6-7-1929).

2. RR 1854, art. 57, in National Archives of the Netherlands [hereafter NA], Archief Ministerie van Koloniën: Openbaar Verbaalarchief [Ministry of Colonies, public minutes; hereafter MinKol OV] 1850–1900, inv. nr. 3196, Verbaal [minute; hereafter Vb.] 13-5-1879 nr. 3.

3. ANRI Archives of the Department of Interior Administration [hereafter DepBB] nr. 586, herein: resolution DepBB, 18-1-1929: ‘Statistiekkaart.’ Around 1895, the number was around 20 million (Van Rossum, 2018, 71).

4. See for the original text of the resolution: International Labour Organization, C029 - Forced Labour Convention, 1930 (No. 29). https://www.ilo.org/dyn/normlex/en/f?p=NORMLEXPUB:12100:0:NO:P12100_ILO_CODE:C029.

5. RR 1854, art. 57, quoted from Schoch (1891), 1.

6. ANRI AS GB MGS 5052, herein, DirBB to Governor General [hereafter GG], 6-3-1928; ANRI AS GB TGA 8368, herein: Controleur of Ogan Ilir to the Assistant Resident of the ‘Palembangse Benedenlanden’, 15-12-1929, Controleur Maroe Doea to the Assistant Resident of Palembang, 4-4-1930: Bijlage III […] Afschrift Proces-verbaal, Hambali bin Hadji Ahjat, Resident of Palembang to DirBB, 23-4-1930, DirBB to GG, 5-6-1930 and DirBB to GG, 5-6-1930.

7. NA MinKol OV 1850–1900, 1389, Vb. 9-10-1863 n8; NA MinKol 1850–1900 1366, Vb. 11-8-1863 n22, herein: Nota Governor-General, 30-7-1863.

8. Staatsblad van Nederlands-Indië [State Gazette of the Dutch East Indies; hereafter Stbl.] 1863 nr. 169; NA MinKol 1850–1900 1626, Vb. 8-7-1865 n17, herein: Governor Moluccas to Governor-General, 28-5-1864.

9. ANRI AS GB MGS 4263, herein: Besl. 26-8-1864 n32, Governor Moluccas to Governor-General, 30-5-1864; Anonymous (1868) (‘De Hervorming der Molukken’), 139; Stbl. 1865 n42, art. 4–10; NA MinKol 1850–1900 1389, Vb. 9-10-1863 n8.

10. Stbl. 1869n91.

11. Stbl. 1864n169; ANRI AS GB MGS 4263, herein: Besl. 26-8-1864 n32: GovMol to GG, 30-5-1864; Besl. 15-3-1865 n7, GovMol to GG, 27-12-1864; NA MinKol 185–1900 1389 Vb. 9–10-1863n8, herein: Raad van Indië [Council of the Indies; hereafter RvI] to GG, 30-1-1863.

12. ANRI AS GB MGS 4263, herein: MGS 20-1-1896: Resident Ambon to Governor-General, 1-10-1895.

13. Ibid., herein: advice RvI 16-12-1895, DirBB to GG, 11-12-1895.

14. Ibid., herein: MGS 22-6-1903: ‘Nota […] Controleur J. van Lier’

15. Stbl. 1891 n45; NA MinKol OV 4359, Vb. 12-4-1890 n29, herein: GG to MinKol, 28-1-1890; ANRI AS GB Besl. 75, herein: MGS, 19-12-1891: Res. Ambon to DirBB, 30-9-1891, Besl. 3-3-1892: DirBB to GG, 14-1-1892.

16. ANRI AS Besl. 1906 4–11 n28, herein: besl. 23-7-1906 and NA MinKol Memories van Overgave [MvO] 314: W.D. van Drunen Littel, Amboina, 1918.

17. NA MinKol 1901–1953 Mailrapport 126, 1919 n2105, herein: Res, Ambon to GG, 30-8-1919; NA MinKol 1901–1953 Mailrapport 125, 1919 n1945, herein: Resident Ambon to Governor-General, 12-7-1919.

18. NA MinKol 1901–1953 Mailr. 125, 1919 n1945, herein: Res. Ambon to GG, 12-7-1919; NA MinKol 1901–1953 Mailr. 165, 1921 n1931, herein: ‘Regentenbond’ to Governor-General, 2-12-1920; MvO 315: C.C. Ouwerling, Amboina, 1930, 42.

19. NA MinKol 1901–1953 OV 2122, Vb. 5-5-1920 n88, herein: ‘Rapport Kopenis over zijn onderzoek naar de landbouw in het gewest Ambon’, 1-11-1919, pp. 78–79.

20. Sikep were presumed to be descendants of the original founders of the village. See Holleman 1981, pp. 156–157, 179–187.

21. Stbl. 1864 n14.

22. NA MinKol 1850–1900 3196, Vb. 13-5-1879 n3, herein: Vb.

23. NA MinKol 1850–1900 3196, Vb. 13-5-1879 n3; Fokkens (1885):315; Elson (1994):257, 258–259.

24. NA MinKol OV 1850–1900 3742, Vb. 10-4-1884 n59, herein: Vb., ‘Nota A2’. From 1905 onward, peasants were allowed to ‘buy-off’ collectively. Stbl. 1905 n520; Fokkens (1914):16; Ernst (1890).

25. ANRI AS GB TGA 9331, herein: BGS January 9, 1934/BGS December 18, 1933: ‘Notulen vergadering’ RvI November 25, 1933.

26. NA MinKol 1850–1900 3190, Vb. 18-4-1879 n31, herein: Vb., pp. 5–7, 15–17, DirBB to GG, 28-12-1874.

27. Stbl. 1882 n136 (abolishment of pancen services) and 137 (introduction of head-tax); NA MinKol 1850–1900 3190, Vb. 18-4-1879 n31; NA MinKol 1850–1900 OV 3196 Vb. 13-5-1879 n3; NA MinKol 1850–1900 OV 3575, Vb. 24-10-1882 n3; NA MinKol 1850–1900 3241, Vb. 16-10-1879 n2.

28. Stbl. 1885 n211.

29. ANRI AS GB MGS 4397, herein: Besl. 28-8-1889: ‘Onderzoek naar de verplichte diensten der inlandsche bevolking in de residentie Kedoe’; Fokkens 1901.

30. ANRI AS GB Besl. 826, herein: MGS 25-4-1901, DirBB to GG. 26-1-1901.

31. In Dutch archives mamak kepala waris are referred to as ‘family chiefs.’ In 1930 the government counted about 12,000 of these. They were usually the eldest men who could claim direct descent from the family’s oldest (female) ancestor, but occasionally women, in case they were considered the most competent and knowledgeable person concerning familial lands and rights (Postel-Coster, 1985, p. 35; Von Benda-Beckmann 1979, pp. 61–72).

32. ANRI AS GB MGS 4233, herein: MGS 24-9-1904: Governor Sumatra’s Westcoast A.M. Joekes (in office 1898–1902) to GG, 18-7-1901, p. 20.

33. ANRI AS GB Besl. 9, herein: MAS 18-7-1890 n1728: ‘Advies Raad van Indië (RvI) van Wijck’ [1st page missing, date unknown], p. 26.

34. ANRI AS MGS 4299, herein: MGS 24-7-1904, Governor Sumatra’s Westcoast to GG, 9-3-1904.

35. ANRI AS GB MGS 4233, herein: MGS 24-9-1904 n3773: DirBB, 31-12-1901.

36. Ibid., herein: Ass-Res Hengel to Gov SWK, 18-3-1905, pp. 7–8.

37. Stbl. 1908 n93; NA OV 441 Vb. 6-3-1907 no. 27; ANRI AS MGS 4477, herein: MGS 2-2-1908: DepFin to GG 2-1-1908, Besl. 17-2-1908; Heckler (1905-1906) II: pp. 19–24.

38. The colonial police signalled a clear link between noncompliant taxpayers and the communist party’s influence (Oki, 1977, pp. 94–95).