Two main challenges need to be overcome in order to successfully define slavery. These are (i) formulating a general definition that reflects crucial differences in how slavery has been—and continues to be—practised at different times and places; and (ii) developing a general definition that clearly distinguishes slavery from related forms of human bondage, such as serfdom or pawnship.1 Most efforts to address these challenges have revolved around the decisive contribution of individual ownership, violent dominion, property rights, and extreme exploitation. These themes have not only been reflected in key legal instruments, such as the 1926 Slavery Convention, they have also long dominated popular understandings of slavery, serving as a series of benchmarks against which various examples of bondage have tended to be conceptualized and classified. The most recent example of this dynamic concerns ongoing debate over the defining features of ‘contemporary forms of slavery’, such as human trafficking, bonded labour, wartime enslavement, and the severe abuse of migrant workers.
In order to decide where slavery begins and ends, modern human rights activists and other actors have returned to these key themes of ownership and exploitation in order to help to determine which types of case and circumstance are sufficiently similar to historical slave systems that they deserve to be legitimately classified as instances of slavery. With slavery now ostensibly abolished across the globe, this has proved to be both an analytically difficult and politically contentious exercise. At an analytical level, the main problem has been determining what constitutes slavery in many illicit and irregular settings. In the absence of a formal legal signifier, it has proved difficult to formulate a series of criteria that can distinguish slave from non-slave in situations involving complex and often idiosyncratic variations in levels of consent, coercion, compensation, and working conditions. At a political level, the main problem has been a tendency to invoke slavery as a rhetorical device in order to prioritize many different causes, such as rape as ‘slavery’, or economic injustice as ‘slavery’.2 Taken to a logic conclusion, this rhetorical inflation reduces slavery to little more than a hollowed out placeholder that covers virtually any form of exploitation or abuse.
In this ambiguous and politically contentious environment, contemporary human rights activists have tended to favour their own definitions of slavery, rather than relying upon established legal instruments.3 Since the mid-1990s a number of new definitions of slavery have been formulated and disseminated, with the most popular model coming from Kevin Bales, who in 1999 defined slavery as ‘the total control of one person over another for the purpose of economic exploitation’.4 Whilst this definition is by no means without merit, it nonetheless suffers from several shortcomings. Firstly, and most obviously, it lacks a clear legal foundation, and is therefore open to charges that this is simply one person’s opinion as to what slavery might look like. Secondly, and more significantly, it unduly prioritizes economic exploitation as a defining feature of slavery as a general category. While economic exploitation was undoubtedly a key feature of transatlantic slavery, it does not necessarily follow that slavery per se should be condensed to economic goals or economic exploitation. It is here, I would argue, that efforts to define slavery in a contemporary setting ultimately require an historical canvass which places the history of slavery in Africa, Asia, and the Middle East alongside more familiar images of plantations in the Americas.
In order to credibly determine what ‘counts’ as slavery today, we need to take into account the diverse forms that slavery has taken historically. As numerous historians have demonstrated, slavery has always been a tremendously diverse institution, both in terms of variations within individual slave systems and variations between slave systems. Over the centuries, slaves have been forced into service in many different capacities. The most common starting point here is the slave plantation, which reached an historical zenith in the colonial Americas. In addition to field hands, slaves have also been routinely used as artisans, bureaucrats, concubines, miners, servants, sailors, soldiers, and sacrifices.5 In some circumstances, slaves have even been purchased to serve as (potential) wives or heirs.6 As these examples begin to make clear, not all of these roles have been chieflydefined by economic considerations or commercial calculations. While all forms of slavery have economic dimensions, these have regularly taken the form of expenses incurred pursuing other goals, such as prestige, consumption, warfare, or reproduction, rather than commercial enrichment. Moreover, we also have numerous examples of slave owners behaving in ways which are incompatible with their commercial interests, such as the frequent abuse of slaves for sadistic—rather than strategic—purposes, which would in turn compromise their productive potential and overall value. These non-commercial dimensions are pertinent to ongoing efforts to define slavery today. This argument is particularly relevant in relation to cases of wartime enslavement and forced military service, but it also has applications to various other practices, such as forced marriage. While the issues involved may not necessarily align with inherited impressions of (transatlantic) slavery, this chapter aims to demonstrate that there are nonetheless credible legal and historical grounds for rethinking prevailing models.
In order to develop this overall line of argument, the chapter is divided into two main sections. In the first, I consider the evolving status of slavery under international law, paying particular attention to the 1926 and 1956 Slavery Conventions, and the ambiguities which continue to surround the legal definition of slavery and related forms of bondage. The main argument here is that the concept of ‘slavery in all its forms’ can help us to successfully navigate between the unhelpful extremes of definitional minimalism (ie only legal slavery is ‘proper’ slavery) and rhetorical inflation The second section then goes on to consider a number of ways in which experiences of enslavement outside the Americas can help to inform efforts to conceptualize and define slavery today. Of particular importance here are parallels between the historical experiences of slave soldiers and the contemporary experiences of some child soldiers. While many practices could also have been considered in this context, I have focused on the nexus between war and slavery in order to highlight the non-commercial aspects of slavery.
Like other contributions to this collection, this chapter takes the 1926 definition of slavery as its primary point of departure. As I have already intimated, I am particularly interested here in the concept of ‘slavery in all its forms’, which chiefly acquires legal and analytical currency through its incorporation within 1926, but can also be found in other international instruments, such as the 1948 United Nations Declaration of Human Rights, which affirms that ‘[n]o one shall be held in slavery or servitude; slavery and the slave trade shall be prohibited in all their forms’. In more recent times, ‘slavery in all its forms’ has tended to be viewed as much the same as ‘contemporary forms of slavery’, which is most prominently associated with the United Nations Working Group (1975–2006) and Special Rapporteur on Contemporary Forms of Slavery (2008–present). There are undoubtedly similarities at work here, but it is nonetheless important to distinguish between these formulas. While ‘slavery in all its forms’ enjoys a clear legal foundation through the 1926 Slavery Convention and other related instruments, the more recent category of ‘contemporary forms of slavery’ can be best understood as political invention which has ultimately resulted in confusion, rather than clarity. In order to effectively determine what counts as slavery today, the operative threshold should not be whether or not various practices and institutions are broadly similar to slavery (ie contemporary forms of slavery), but instead whether specific cases are the de facto equivalent of historical experiences of slavery (ie slavery in all its forms).7
The drafting history and subsequent reception of the 1926 Slavery Convention has attracted considerable intellectual interest in recent times, following a long period of relative neglect. In the analysis which follows I am primarily concerned with two themes that have rarely featured prominently in discussions to date. First, we have the specific concept of ‘slavery in all its forms’, and its connection to various non-European slave systems which remained in operation during the first half of the twentieth century. When the 1926 Convention was being developed, the most pressing issue was not ‘New World’ slavery, which had already ended, but ongoing slave systems in Africa, Asia and the Middle East. Second, we have the way in which the 1926 definition was operationalized by European authorities concerned with various slave systems which continued under colonial jurisdiction. As Jean Allain has persuasively demonstrated, most international attention in this period was directed towards non-European governments, most notably Liberia and Ethiopia, rather than European tutelage of ‘backward peoples’. By alternatively focusing upon colonial responses I aim to (re)emphasize that the most common initial approach to 1926 was to circumscribe and minimize its application as much as possible. This offers an important counterpoint to the rhetorical inflation of the second half of the twentieth century.
Most modern discussions of slavery and international law begin with the pioneering work of the League of Nations, which established a foundation upon which most recent jurisprudence concerned with slavery has been built.8 In the aftermath of the First World War, the immediate catalyst for action was slavery in Africa in general, and Ethiopia in particular.9 This eventually translated into the establishment of the Temporary Slavery Commission, which was tasked with examining ‘slavery in all its forms’. Despite working with limited and/or unreliable information, this Commission nonetheless identified a number of territories where slavery remained an ongoing issue, including ‘raids properly so called’ on the ‘borders of the Saharan desert’ and in Ethiopia, the continuation of at least some slave trading in many parts of Africa, an ‘open slave trade’ in several Asian states, and the continuation of legal slavery in ‘Thibet and Nepal, and in most of the Mohammedan States of the East, such as Afghanistan, the Hedjaz and other Arabian States’.10 When it came to colonial rule, the Commission concluded that residual cases persisted—despite having been abolished in law—because ‘the civilising influence of Colonial Powers, however energetic and vigilant they may be, cannot change completely in a few years the habits ingrained by centuries’.11
The Commission was unable to settle upon a single definition of slavery. It instead divided its final Report into thematic chapters, which included material devoted to ‘Compulsory Labour’ and ‘Practices Restrictive On the Liberty of the Person’.12 In the case of the latter, it was recognized that a number of practices, such as the acquisition of girls by payment of dowry, the abuse of adoption, and the manipulation of debt-bondage could all degenerate into slavery, yet the precise threshold between slavery and other ‘restrictive’ practices was never clearly articulated. As part of their submission to the Commission, the premier non-governmental organization focusing upon slavery—the Anti-Slavery and Aboriginal Protection Society (founded in 1839)—highlighted four areas of particular concern: (i) contract labour, (ii) forced labour, (iii) ‘adoption’, ‘pledging’ and domestic slavery, and (iv) slavery. Emphasizing the ‘necessity of defining as a broad principle what practices may be said to-day to amount to modern slavery’, the Society boldly suggested that ‘any system in which force or fraud are exercised to secure control over the labourer for private ends partakes of slavery’.13 This expansive formula was politically problematic, since all colonial powers continued to make regular use of forced labour schemes involving the systematic exploitation and abuse of local ‘natives’.14
The question of definitions would be chiefly addressed through the 1926 Slavery Convention, which took the decisive step of defining slavery as ‘the status or condition of a person over whom any or all of the powers attaching to the right of ownership are exercised’, and committed parties both ‘to prevent and suppress the slave trade’ and ‘[t]o bring about, progressively and as soon as possible, the complete abolition of slavery in all its forms’. Frustratingly, the precise nature of these forms was never formally defined, so the burden chiefly fell upon signatories to the Convention to determine the scope of their new obligations. In the negotiations that preceded the final Convention, the expansive terrain considered by the Temporary Commission was circumscribed in a number of ways.
The preliminary draft of what eventually became the 1926 Convention contained a more open ended obligation to ‘bring about progressively and as soon as possible the disappearance of slavery in every form, notably in the case of domestic slavery and similar conditions’, but this was eventually discarded in favour of the shorter framing of ‘slavery in all its forms’.15 This revision was ostensibly justified on the grounds that ‘it was believed that such practices came within the definition of slavery’, thereby rendering further elaboration redundant. As part of this justification, specific reference was made to ‘domestic slavery’, ‘debt-slavery’, ‘the enslavement of persons disguised as the adoption of children’, and ‘the acquisition of girls by payment of dowry, etc’, but this list was immediately qualified by a further statement that ‘even if these last practices do not come under the definition of slavery … they must be combated’.16 As Allain has argued, this apparent ambiguity can be resolved by observing that the key point at issue here was not whether or not broad categories of human bondage fall within the terms of the 1926 definition, but instead whether or not individual cases of human bondage associated with these categories can be classified in terms of ‘powers attaching to the right of ownership’, and can therefore be legally defined as slavery. Allain maintains that this shift in language was designed to ‘make plain that there was but one type of slavery’.17 I favour a somewhat different position, which holds that the language and drafting history of the Convention clearly envisages multiple forms of slavery, yet all of these forms can only be credibly classified as such in cases that satisfy the operative threshold of ‘powers attaching to the right of ownership’.
The drafting of the 1926 Slavery Convention was heavily influenced by colonial politics, with various European officials working to circumscribe the scope of their international obligations in relation to legal slavery and its aftermath (with progressive rather than immediate abolition), forced labour (which was widely held to be legitimate, albeit sometimes subject to abuse), and related forms of human bondage (which were to be distinguished from ‘true’ slavery).18 These considerations also played an ongoing role in the way in which the 1926 Convention was subsequently interpreted. Throughout the 1920s and 1930s there was a concerted diplomatic effort to restrict the practical scope of the definition of slavery (unless a non-European government was the subject of inquiry).
These broader trends are reflected in the work of a further series of expert committees tasked by the League to examine slavery during the 1930s.19 Much like earlier League inquiries into slavery, these committees primarily focused upon territories where slavery remained legal, and upon territories in which slavery had only recently been officially abolished. In the case of the former, reports from this period habitually surveyed the handful of governments which still sanctioned slavery, or which had recently passed laws against slavery, but in most of the territories concerned reliable information proved to be in short supply. This was chiefly because the League was forced to rely primarily upon government submissions, which routinely proclaimed that slavery was no longer a problem. Many submissions took the form of blanket denials, but some colonial powers such as Britain and France sometimes favoured a more qualified stance, which involved cautiously admitting that residual cases continued in some jurisdictions. There were two main variants to this overall approach. The first saw various qualifiers being placed alongside slavery. Examples taken from reports from 1935 and 1937 include references to ‘natives living in … a servile state bordering on slavery’, ‘quasi-slaves’ ‘mild forms for slavery’, ‘semi-slaves’, ‘household captives’, ‘domestic serfs’, ‘voluntary slaves’, or ‘so-called slaves’.20 Much like earlier slave-holders in many parts of the Americas in the nineteenth century, European colonial officials consistently sought to portray local slave systems as ‘benign’, or ‘mild’ in an effort to minimize potential complications and challenges associated with anti-slavery.
In the second variant, this underlying logic was taken one step further with the suggestion that those involved ‘were not really slaves at all’.21 An illustrative example of this widespread theme comes from the following description of the aftermath of legal abolition:
It may be said that, in general, in all the territories where slavery no longer exists in law … the native population is daily become more and more aware of the regime of liberty assured to it by law. The reason why certain natives do not demand their freedom is that is appears to be, or is actually, in their interest to remain with their former masters. Moreover, the latter are afraid that their servants will leave them, and this induces them to make their lot easier. As a matter of fact, the position of ex-slaves, even when they do not claim their freedom, is approximately the same as that of free servants.22
This statement—and others along similar lines—was inherently disingenuous. While historians of slavery generally agree that the overall prevalence of slavery in Africa, Asia and the Middle East declined over the first half of the twentieth century, there were still significant numbers of slaves in many regions who had yet to experience much in the way of change during the 1930s. In theory, slaves were now free to leave, but this did not always work in practice due to ongoing surveillance, threats of retribution, socialization, and official complicity. In theory, ex-slaves were much the same as ‘free servants’, but in practice many continued to endure physical punishments, sexual abuse, economic exploitation, severe restrictions on movement, social discrimination, and fragile family relationships.23
League experts also found it necessary to expend considerable energies upholding their preferred distinction between slavery and other forms of human bondage. Despite a general agreement that other forms of bondage should be distinguished from ‘slavery in the true sense of the word’,24 each report nonetheless included a separate section concerned with ‘Practices Restrictive On the Liberty of the Person’ (1932, 1935) or ‘Other Institutions’ (1936, 1937, 1938). In 1936, to take but one example, this section included reference to ‘debt-slavery’ (amongst other things), which was divided into ‘household and domestic debt slaves’, and ‘agricultural debt slaves’. Despite the use of the term ‘slavery’ to describe the practices involved, this classification was also qualified by an observation that these practices were not slavery ‘unless any or all of the powers attaching the right of ownership are exercised by the master’.25 Other practices which also featured in these sections were serfdom, debt-bondage, marriage practices, domestic servitude by way of ‘adoption’, and pawnship. Despite a widespread recognition that these various practices shared features in common with slavery, there was nonetheless a general consensus that ‘proper’ slavery should continue to be prioritized. This concentration on legal slavery and legal abolition proved sufficient to minimize the 1926 definition until at least the mid-twentieth century.
This circumscribed approach to slavery gradually gave way to a more expansive vision over the course of the second half of the twentieth century. Under the auspices of the newly formed United Nations, the boundaries of slavery were reinterpreted in various ways, resulting in a political—but not necessarily legal—landscape that saw many different kinds of problems (re)defined as ‘contemporary forms of slavery’. This new phase in the history of organized anti-slavery was inaugurated by the 1956 Supplementary Convention on the Abolition of Slavery, the Slave Trade and Institutions and Practices Similar to Slavery. This did not replace the 1926 Convention, which was belatedly taken over by the United Nations in 1953, but was instead presented as an attempt to ‘augment’ and ‘intensify’ anti-slavery efforts. Of particular interest here is Article 1, which obligates parties to:
take all practicable and necessary legislative and other measures to bring about progressively and as soon as possible the complete abolition or abandonment of the following institutions and practices, where they still exist and whether or not they are covered by the definition of slavery contained in article 1 of the Slavery Convention signed at Geneva on 25 September 1926.26