The Definition of Slavery in Eighteenth-Century Thinking: Not the True Roman Slavery, John W. Cairns
4
The Definition of Slavery in Eighteenth-Century Thinking
Not the True Roman Slavery
John W. Cairns*
A. Roman, Early Modern, and Modern Definitions
The Roman jurist Iavolenus Priscus (born mid-first century CE) famously remarked that all definitions in law were hazardous, for there were few that could not be overturned.1 Two millennia later his point remains valid. It is certainly apposite when we come to consider the failure of the Roman jurists to define—or perhaps to describe—slavery in any kind of apparently satisfactory way. Roughly a century later than Iavolenus, Florentinus explained slavery by contrasting it with freedom:
Freedom is one’s natural power of doing what one pleases, save insofar as it is ruled out either by coercion or law. Slavery is a product of the ius gentium, whereby someone against nature is made subject to the ownership of another. Slaves (servi) are so-called because generals are accustomed not to kill but to sell their captives and thereby to save (servare) them.2
In the early third century CE, Tryphoninus stated as an aside in a discussion of natural obligations and the condictio indebiti that freedom was found in the natural law and ownership of human beings was introduced by the ius gentium.3 By ius gentium, the Roman jurists meant laws common to all or most peoples. It is worth pointing out that slavery provided the only instance in which the jurists recognized a conflict between ius gentium and ius naturale.4
That the Roman jurists’ explanation of slavery may appear so unsatisfactory to the modern eye may be the result of their inductive approach to law. They did not consider definitions to be conclusive or determinative but instead descriptive: this indeed is the force of Iavolenus’s remark.5 For the Romans slaves were simply a fact. Some human beings were unfree and subject to ownership; the consequences of their status were best understood by comparison with the free person’s powers (facultates) to act autonomously. But these discussions and the looseness of the legal usage of terms such as natural law and ius gentium gave considerable scope for later developments in interpretation.
The nature of slavery and its definition in the middle ages is outwith the scope of this discussion.6 But the views of the late Baroque age, an age of significant European slave trading as European powers developed colonial empires, can be traced in the work of, for example, Ulric Huber (1636–1694), one of the most important and influential of the Dutch jurists of the late seventeenth century.7 Huber broadly defined a slave as a man who owes unending labour in return for necessaries.8 He stated that slavery was established by the law of nations and was the subjection of one to the ownership of another contrary to nature. He pointed out that slavery had many justae causae. Huber added that ‘contrary to nature’ here meant contrary to the original state of men; it did not mean ‘contrary to the law of nature or the dictate of right reason’.9 This analysis was traditional, and indeed commonplace in the eighteenth century. Thus, in 1751, the Scots jurist Andrew McDouall (1685–1760) wrote:
The parents being slaves, the children behoved to be of the same condition.10
This explanation reflected the discussions of the modern natural lawyers, as well as serving as an interpretation of the Roman law.11 Huber’s account of slavery—though acknowledging it could arise in many ways—has a curiously contractual tone. The account of the German author and professor Johann Gottleib Heineccius (1681–1741) is rather different in focus. Heineccius wrote some of the most popular legal textbooks in the eighteenth century, using a supposedly rational, deductive method.12 In that on the Institutes, the German author explained that homo and persona differed completely in law: the term homo dealt with a man in whatever human form he was; the second, persona, meant the civil status of a man. He later deduced that while a slave was a homo, he did not have a civil persona but was a res, a thing, lacking family, citizenship, and freedom. This meant that slaves were property.13
This type of discussion provides the background to the important views of two Scots, John Millar (1735–1801) and his teacher Adam Smith (1723–1790), both of whom were familiar with the works of Heineccius and similar literature.14 Millar, Regius Professor of Civil Law in the University of Glasgow from 1761–1801, discussed slavery and other forms of service in his classes.15 In one set of student notes from his course on Justinian’s Institutes in the 1770s, he is recorded as telling his class:
It is impossible to define the precise degree of subjection which constitutes what is called Slavery the idea of which is different in different countries—We shall therefore consider the state of persons at Rome, called slaves without any general definition of slavery.16
Though Millar’s teacher Smith also paid some attention to slavery both in his published works and in his Lectures on Jurisprudence, he also took slavery as a self-evident category, not requiring precise definition. This meant that Millar was willing to consider the Roman coloni as slaves;17 Smith even classed serfdom as slavery, when he pointed out that slavery was found over most of the contemporary globe, other than a small part of Europe.18 Failure to define entailed reliance on descriptions of the incidents of slavery to identify it.
Smith and Millar considered slavery in the context of the relations of subordination in families, a traditional approach linking it to the law of persons. But there can be no doubt that what they saw as distinguishing a slave from a regular servant was that slavery was a status that rendered one unfree and subject to ownership, whereas a regular servant was engaged under a contract of service for a specified time for specific wages. Both still saw such servants as dependants of a master, but, as Smith put it, such servants ‘have almost the same priviledges with their master, liberty, wages, etca’.19 In other words, slaves had none of these: slavery lacked these attributes of freedom.
According to the jurisprudence of Smith and Millar a freeman enjoyed privileges deriving from his being protected against injury as a man, as a member of a family, and as a member of a state. (The gender-specific language is theirs.) Some of these privileges were natural; others were civil or acquired.20 Enslaved servants lacked these privileges. Smith listed their disadvantages: in early times, masters had an arbitrary power over the lives of such slaves and might put them to death; masters controlled the liberty of their slaves; having the sole right to the work and labour of their slaves, masters could transfer them to another master; freedom was at the will and caprice of the master; slaves were incapable of having any property and the fruits of their labour went to their master; whatever a slave acquired belonged to the master; slaves could only contract with the consent of the master, to whom any profit from the contract belonged; there was no restraint, at least in early times, on the master’s cruelty. As well as these disadvantages, Smith pointed out that slaves could not marry; they could live together in contubernium, but this was not a recognized marriage and vulnerable to the actions of the master. At ancient Rome, according to Smith, slaves were largely excluded from the consolations of (pre-Christian) religion.21
Smith had based his discussion of slavery largely on Roman law; but he argued that, except for religion, ‘our legall [sic] slaves … are indeed … in much the same condition’ as those of the Romans.22 Thus, while he considered that villeins were slaves, he stressed that ‘their slavery was of a milder kind than that known among the antient Greeks and Romans, or even in our West Indian colonies’.23 Roman slavery provided a touchstone by which other forms of slavery could be judged. Contemporary slavery in the colonies was considered as comparable to that of Rome. This was a common view found, for example, in some contemporary editions of Heineccius’s textbook.24
The significance of Roman slavery as a standard by which to judge and to understand slavery more generally persisted into the next century. Thomas Cobb (1823–1862), the learned antebellum apologist for racially-based slavery in the southern states of the United States of America, drew a contrast between what he described as ‘pure slavery’ and lesser forms of ‘involuntary servitude’:
Absolute or Pure Slavery is the condition of that individual, over whose life, liberty, and property another has the unlimited control. The former is termed a slave; the latter is termed the master. Slavery, in its more usual and limited signification, is applied to all involuntary servitude, which is not inflicted as a punishment for crime. The former exists at this day in none of the civilized nations of the world; the latter has, at some time, been incorporated into the social system of every nation whose history has been deemed worthy of record. In the former condition the slave loses all personality, and is viewed merely as property; in the latter, while treated under the general class of things, he possesses various rights as a person, and is treated as such by the law.25
Although he did not state this explicitly, Cobb appears to have identified slavery at Rome with ‘absolute slavery’. Explicitly relying on Heineccius, he wrote:
Among the Romans, the slave was classed as a thing: Res. He was ‘Homo sed non persona.’ He was considered ‘pro nullo et mortuo, quia nec statu familiae nec civitatis nec libertatis gaudet’ [‘as no one and dead, because he enjoys neither family, nor citizenship, nor freedom’].26
Earlier he had stated that ‘[t]he status of the slave, in the Roman law, was literally as a thing and not as a person…. His general status was “pro quadrupedibus”’. Cobb emphasized that the paterfamilias ruled his household from which derived ‘the power to kill the son and the slave with impunity; a power recognized, as to the latter, until the days of Antoninus, when it was abolished’. He emphasized that the power of the master was ‘unlimited’.27 For Cobb, as for others, Roman slavery was paradigmatic of absolute—or near-absolute—slavery.
The Roman jurists did not explicitly list the powers enjoyed by the master or the corresponding limitations on the slave. While the slave codes of the New World focused on the policing of slaves and slavery, the Roman legal sources, other than a few general texts, tend to be most concerned with the way the slaves operated or affected outcomes in the general legal system. Slavery is rarely addressed directly in the legal texts as an institution; discussion is more commonly scattered through treatment of different areas of the law. It is, of course, possible to infer the attributes of ownership, the authority of masters, and the limitations on the autonomy of slaves from the Roman texts, and modern scholars of Roman law have attempted to set them out. Obviously, there was a right to sell; a right to anything acquired by the slave; a right to kill (if moderated during the Empire into, finally, a right of chastisement); a right to take the children of a slave-mother as slaves; a right to give in noxal surrender, if the slave had committed a delict; and a right to free (though restricted from time to time by the law). Slaves suffered from obvious legal disabilities. They could not sue or be sued. If they gave evidence, it had to be under torture. They could not make a contract binding at civil law; they could incur, however, what was described as a naturalis obligatio, which could in some circumstances have a legal effect. They could not contract a marriage recognized by civil law.28 The casuistic nature of Roman law can, however, leave some doubts in the mind of the reader over some of the finer points.
The evidence thus suggests that Roman slavery has been taken as providing an almost ideal type of slavery, by which other forms of enforced or unfree labour might be judged. Roman law defined slavery by reference to ownership and in opposition to freedom; while the extent and nature of the rights over and legal disabilities of the slave that followed were never tightly defined, they may be listed. In the eighteenth and nineteenth centuries, the list of these incidents and attributes of ownership of human beings remained substantially the same.
Slaves accordingly were considered as belonging to the category of property, not that of persons, and thus were objects of the second not the first division of the law.29 This also appears to be the understanding of slavery in international law, where the Slavery Convention of 1926 defined it as follows in its first article:
Slavery is the status or condition of a person over whom any or all of the powers attaching to the right of ownership are exercised.30
In 1953, a Protocol of the United Nations adopted this, and in 1956 it was supplemented by a UN Convention on the Abolition of Slavery, the Slave Trade, and Institutions and Practices Similar to Slavery. The 1926 definition remains authoritative in international law.31 As with Roman law, it identifies slavery with dominium, ownership; but it further describes an enslaved person as one over whom ‘any or all of the powers attaching to the right of ownership are exercised’, and the emphasis on understanding slavery through the incidents of ownership is notable. This approach to the definition is quite compatible with Florentinus’s description of slavery in opposition to possession of freedom—the natural power (facultas) to do what one pleases, which necessarily means that free individuals are not subject to such powers in the hands of an owner; but enslaved ones are. Though Iavolenus may be correct that definitions in law are problematic, a definition that is workable remains necessary as enslavement is a crime in international law, and the application of that law requires a measure of precision so that the law is functional in a court and is not so broad and vague that it tends to the meaningless or conflicts with other principles.
Interpretation of the 1926 definition has raised problems. In 2005, the European Court of Human Rights understood it as referring to de iure slavery; it accordingly refused to apply article 1 to the situation of a female child who served for four years as an unpaid domestic worker in France, because ‘the evidence does not suggest that she was held in slavery in the proper sense, in other words that Mr and Mrs B. exercised a genuine right of legal ownership over her, thus reducing her to the status of an “object”’. The fact that slavery was illegal in France meant that though ‘the applicant was … clearly deprived of her personal autonomy’, her status did not correspond to ‘the “classic” meaning of slavery as it was practiced for centuries’ and embodied in the article.32 In contrast, in Kunarac et al, the Appeals Chamber for the International Criminal Tribunal for the Former Yugoslavia had earlier focused on a different part of the article. The Chamber noted that the article ‘speaks more guardedly “of a person over whom any or all of the powers attaching to the right of ownership are exercised”’. It commented that ‘[t]hat language is to be preferred’ and added:
The Appeals Chamber considers that the question whether a particular phenomenon is a form of enslavement will depend on the operation of the factors or indicia of enslavement identified by the Trial Chamber. These factors include ‘control of someone’s movement, control of physical environment, psychological control, measures taken to prevent or deter escape, force, threat of force or coercion, duration, assertion of exclusivity, subjection to cruel treatment and abuse, control of sexuality and forced labour’.33
In other words to be caught by the 1926 definition slavery did not need to be understood as referring to ‘a genuine right of legal ownership’; it encompassed de facto slavery. This approach has recently been endorsed by the High Court of Australia in The Queen v Tang. There the Court emphasized that what was significant was the type of powers exercised over an individual; if these were of the type that would have amounted to legal slavery when that had been possible, then article 1 of the 1926 Convention was applicable.34 The jurisprudence continues to develop and remains fluid; but it is unnecessary to discuss it further.35
B. Eighteenth-Century Cases and the Meaning of ‘Slavery’
Both use of the term ‘slavery in the proper sense’ by the European Court of Human Rights and the contrasting International Criminal Tribunal’s identification of slavery through various indicia or ‘powers’ echo much discussion in the eighteenth century, when courts and jurists in Europe struggled to understand and deal with colonial slavery. In particular, eighteenth-century courts were exercised by the position of enslaved individuals brought from the colonies to the European home countries of their masters and mistresses. The courts sought to understand whether slavery existed in law in their own systems, and if so, to consider its nature and how it related to other forms of labour and social relationships. This discussion was also much concerned with definitions of slavery (or the failure of them) and the incidents of slavery. Questions were raised as to whether an individual’s circumstances amounted to slavery with Roman law sometimes providing a test of ‘true’ slavery.
The most famous English-language case dealing with this type of issue is undoubtedly that of Somerset.36 Unfortunately, there is slight evidence of the arguments of counsel and even considerable dispute over the precise wording of the opinion of William Murray, Lord Mansfield (1705–1793).37 There are, however, four relevant Scottish civil cases:38 Sheddan v Montgomery (1757), a freedom case;39 Stewart Nicolson v Stewart Nicolson (1769–1771), a divorce case before the Commissaries of Edinburgh advocated to the Court of Session on, inter alia, the issue of the admissibility of evidence;40 Dalrymple v Spens (1770), another freedom case;41 and finally Knight v Wedderburn (1778), yet another freedom case.42 There is one criminal case, HMA v Bell or Belinda (1771), before a Circuit Court of Justiciary in which the court implicitly recognized the validity of the enslavement of a Bengali girl. On the bench were the Lord Justice-Clerk, Thomas Miller of Barskimming (1717–1789), and James Ferguson, Lord Pitfour (1701–1777).43