GREEK AND ROMAN SLAVERY

Introduction

When one thinks of the Greco-Roman world, one probably thinks of Alexander the Great or Julius Caesar. It was also, however, the world of Spartacus, perhaps the most famous slave in history, and tens of millions of its other inhabitants were also slaves. There is robust disagreement as to how to reconstruct their lives. This might seem surprising. Greek and Roman slavery have been studied for over 150 years and comparatively little new evidence has been discovered in that time. Historians have an inbuilt ability to disagree with one another, but one might still have expected increasing consensus. The continuing debate is, however, largely the product of three factors that lend the study of ancient slavery its distinctiveness and help it to make a very individual contribution to the general study of slaves.

First, “Greco-Roman” actually covers a number of very different slave systems. Our evidence allows us only snapshots of some of these, such as Athens between about 450 and 300 BC, and Roman Italy between about 100 BC and AD 200. These in turn were subject to considerable change over time. In addition, slaves fulfilled a great range of different roles within Athens and Rome. All these differences make generalisations difficult, but also offer interesting comparisons.

Second, historians of Greek and Roman slavery lack the kind of bureaucratic records used in the study of more modern slave societies. They are forced to glean information from a wide range of material, including drama and poetry. Their especial sensitivity to both the possibilities and the problems of such material is perhaps their most distinctive contribution to the overall study of slavery. The great variety of potential “readings” of texts, however, provides a second explanation for continued debate.

Modern geography, curiously, provides the final reason (see McKeown, 2007, especially ch. 2–4). While there are no “national” schools of thought on ancient slavery, different areas have developed somewhat different emphases in their work. English-speaking scholars have tended to emphasise conflict between master and slave. French and Italian scholars have often taken a similar line, though with a greater willingness to apply Marxist ideas of class conflict. On the other hand, some German-speaking scholars, while recognising the inhumanity of slavery, have also asked why slavery was able to function successfully as long as it did, and have examined how slaves and ex-slaves were assimilated into the wider society.

If these differences help to explain why the past 150 years have not produced much agreement, they also explain why historians of  provide provocative studies where one can find most of one’s assumptions about slavery challenged, as well as a variety of very different, though equally sophisticated, ways of reading evidence. Antiquity, moreover, offers us relatively well documented slave societies where, unlike many modern examples, neither capitalism nor race was crucial, even if racism existed and slave owners expected to profit from their slaves.

Early Greek slaveries and slavery in Athens

Homer’s Iliad and Odyssey provide us with a (fictionalised) portrait of early Greek society of around 700 BC. Chattel slaves (outsiders bought for cash) and more numerous “semi-free” dependents served local “lords” and were apparently treated similarly. It has been suggested that Homeric servitude was a quasi-contractual relationship, with an expectation of rewards for the “good” servant. While it is certainly significant that the text emphasised the idea of the “good servant” and “good lord”, the servant’s rewards were precarious, and slavery was a fate one wanted to avoid. For example, after Odysseus bloodily regained control of his palace after 20 years away, he hanged 12 of his maids, apparently for sleeping with other noblemen (Garlan, 1988, 29–37).1

The comparative unimportance of chattel slavery in Homer may represent the norm in archaic Greece (700–500 BC). For some states, however, it remained the pattern into the “classical” period (500–300 BC). The Spartans, for example, lived off the labour of helots (Cartledge, 1979, ch. 10). Most historians believe these were the original inhabitants of Messenia and Laconia, which were conquered by the Spartans, though they may have been the losers in a struggle within Spartan society. They heavily outnumbered the Spartans, though the traditional ratio of 7:1 may be exaggerated. Some contemporaries referred to helots as slaves, but modern scholars usually place them between slave and free: “communal slaves” or even “serfs”. Helots were owned by the state rather than by the individuals whose farms they worked. They lived in family units and were allowed to keep part of their crop. Recent archaeological work suggests that there may have been sizeable, socially differentiated settlements of helots in Messenia (Hodkinson, 2008).

Unfortunately, most of our evidence about helots comes from non-Spartans writing long after the events they describe and who had a tendency to sensationalise. Ancient writers clearly believed, however, that helots were harshly treated, humiliated and sometimes murdered. Helots responded with rebellions, particularly in Messenia. They had a common language and identity, and lived separately from their masters in areas that provided favourable territory for flight and resistance. There have been recent attempts to downplay the tensions between helot and Spartan. The weight of the evidence, however, suggests serious conflict. The Messenian helots were eventually able to assert their independence with the support of Sparta’s Greek enemies after 370/69 BC, and helotry seems to have disappeared in Laconia some time after the Roman conquest of the second century BC.

Some states, notably Athens, chose to use chattel slaves. Little evidence survives to explain the history of the process that culminated in perhaps a third of the population of classical Athens being slaves (Finley, 1998, ch. 2; Morris, 2002). Moses Finley argued that slavery on such a scale was historically unusual because of the problems in controlling slaves and the costs in acquiring and maintaining them. He believed that the exploitation of internal labour sources (as in Sparta) was the usual choice for most elites in history. Finley suggested that three preconditions were needed before the rise of mass chattel slavery. (1) Large, privately controlled farms – if a ruler could simply appropriate a farmer’s surplus, or if farmers only produced at subsistence level, there would be little incentive to buy an extra labourer. (2) The development of market exchange – this helped to ensure a constant supply of slaves. It also created opportunities to sell any surpluses they created and made it easier to provision them. (3) Crucially, a lack of internal sources of labour – Finley believed that democratic structures helped protect ordinary Athenians from exploitation, most notably when debt-bondage was abolished in 594 BC. This forced larger landowners to import external labour to fulfil their manpower needs.

Finley may have exaggerated the importance of the abolition of debt-bondage. The development of democracy and of slavery were more probably processes lasting several hundred years into the fifth century. Indeed, some historians argue that rather than democracy helping to create slavery, slavery helped the development of democracy by allowing farmers time to engage in politics. There was also clearly some demand for chattel slaves even before 594 BC, and it has been suggested that Finley underestimated the positive economic attractions of slave labour (such as its flexibility). The infrequency with which we see slave systems on the Athenian scale, however, suggests that more than economic factors are required to explain its rise. For the moment, therefore, Finley’s position still represents the orthodoxy.

By the classical period Athens was the urban centre of a triangle of territory (“Attica”) approximately 40 miles on each side. At its height, between about 450 and 300 BC, modern estimates suggest a free population of 150,000–250,000 with 50,000–100,000 slaves. The figure for slaves is, however, largely an educated guess. Our sources imply that most slaves came from the Balkans and from Turkey, though some were from other Greek cities. The market was the main mechanism of acquiring slaves, with warfare, kidnapping and possibly child exposure the likeliest ultimate sources of those traded. There is little surviving evidence of slave breeding (or slave families), though it may have been significant.

The debate about the spread of slave ownership in Athens indicates the problems and choices faced by historians of ancient slavery. The traditional picture suggested that the richest few thousand Athenians used slaves to produce the bulk of their wealth. Another 10,000 middling farmers, the hoplites, allegedly owned one or two slaves as well. It was suggested that a growing population and attendant land division forced hoplites to intensify production with extra labour. Hired labour was despised and undependable, and only slave labour allowed farmers the free time required to engage in democratic life. The remaining 10,000–15,000 poorer citizens also aspired to own slaves, but may not have been able to afford them.

Some historians have, however, argued that slave ownership was restricted to the rich (Jameson, 2002). Texts implying that hoplites owned slaves may reflect their wishes rather than reality. Even if they needed extra labour (and the evidence for a land shortage and the supposed intensification it necessitated has been questioned), it has been suggested that extra labour could have been found from family or neighbours or hired workers (ordinary Athenians may have been more willing to be hired out than aristocratic contemporaries suggested). Computer models of probable farm production also suggest that it is doubtful whether hoplites could have afforded to buy and keep slaves. Lastly, it has been argued that the democracy improved the position of smaller farmers, directly through cash payments and indirectly by preventing exploitation by the rich. This, in turn, allegedly removed part of the need to intensify production, the supposed reason for buying a slave.

This position, however, implies a peasantry so poor that it would have been vulnerable both to periodic famine and to control by wealthy patrons, even given the benefits of democracy. Neither phenomenon is very significant in our sources. In addition, the computer modelling of hoplite farm production has such a wide margin of error that it cannot rule out the possibility of slave ownership. There are, then, problems with the argument that slave ownership was restricted to the rich, especially as it also requires actively arguing away all the literary evidence pointing to hoplite ownership of slaves. Relatively widespread slave ownership seems likelier.

Up to a third of slaves in Athens may have worked in her silver mines. We know little about their lives. It has been suggested that expensive inscriptions left behind by some, and the skilled work seen in the mines, indicate that their life may have been relatively tolerable. This seems doubtful. Mining in antiquity was often a punishment, and slave rental agreements involving the replacement of dead miners also hint at low life expectancy. The surviving inscriptions may well be the product of slave supervisors, not ordinary slaves, and the high standard of the mining work may simply show effective control.

Most slaves in Athens probably were involved in mixed agricultural and domestic work, living alongside their masters. One author described slave barracks and an overseer on the estate of a very wealthy landowner, but concentrations of more than a dozen slaves were probably quite unusual, and ownership of just one or two slaves the norm. In addition, the property of the wealthy was probably generally dotted about the countryside rather than in single units.

The remaining slaves did mostly artisan work. Some lived separately from their masters, paying a form of rent. A small number of slaves helped run businesses, including primitive banks, or acted as commercial agents. There were, finally, some state-owned slaves serving as record keepers, “civil servants”, street cleaners and even policemen. A few of these may have led quite independent lives.

Apart from their job, the temperament of their master was probably the other crucial element in a slave’s life. Comic sources suggest casual violence against slaves, though the interpretation of such material is problematic, as we shall see. Athenians, however, unlike the Romans, spent little time discussing abuses committed against slaves. This could be interpreted optimistically, suggesting the absence of severe abuse, but Athenian sources generally either ignore slaves altogether or adopt a cynical attitude towards them. Slaves certainly had little legal protection. A right of asylum existed, but the fate of slaves who used it is unknown. Killing one’s own slave was an offence against the gods, but any religious stigma could be removed with purification rites. Killing someone else’s slave was treated legally as damage to property. To the surprise of contemporaries, slaves were technically protected alongside the free from hubris, assault designed to humiliate. Even if it dated from a time when Athenians could be enslaved for debt, the law was certainly applied in the classical period. It may, however, have protected the owner’s honour rather than the slave’s, or been used to discourage misbehaviour towards slaves that might later be applied to citizens.

Slaves of both sexes were subject to sexual abuse from their master. There were few taboos against a male citizen seeking sex outside marriage, and slave prostitution was also an acceptable part of daily life. The only hint of a limitation on sexual use was an apparent distaste of the practice of castrating (Greek) slaves.2

Slaves were barred from pleading in courts, except in some commercial cases in the fourth century BC. When slaves testified, they typically did so under torture (Mirhady, 2000). Curiously, we have little evidence of this from surviving court speeches, and many examples of litigants refusing to surrender their slaves for torture. Some believe that torture was a legal fiction and never took place. Others have suggested that torture was common, and regarded as so effective in establishing the truth that it was unnecessary afterwards to continue with the court case, explaining the absence of references to it in court speeches. The continued demands to hand slaves over for judicial examination do indeed make little sense if torture was not a real possibility. In any case, the alleged incapacity of the slave to speak the truth without physical compulsion, and their physical vulnerability in law, helped to distinguish slave from free on an ideological level.

Athenian drama is potentially one of the most fruitful sources of evidence for general attitudes towards slaves (DuBois, 2003, 137–52). At first sight it suggests a degree of sympathy. Euripides’ plays Hecuba, The Trojan Women and Andromache have as their heroines slaves who were captive royal women from the time of the (mythical) Trojan War. It could, however, be questioned how far Athenians associated these mythical royal characters with contemporary slaves. Indeed, it has been argued that the plays chiefly allowed Athenian male citizens to express their own fears of enslavement. When Euripides deals with ordinary non-Greek slaves, he is quite capable of portraying them as fawning cowards. Even statements that clearly appear positive towards slaves need to be examined in their dramatic context. For example, a character in Euripides’ Ion claims there was no difference between a slave and a free man.3 This is often quoted as an example of an author questioning the justice of slavery. The character speaking is, however, a slave who is conspiring with his mistress to murder a fellow servant who they feel is rising above his place. The passage is, therefore, rather ironic. Even where slaves are clearly depicted showing an unselfish loyalty to their owners, for example in Euripides’ Alcestis, this may be done more to throw into relief the actions of a particularly selfish free character than to praise the slaves.

The comic plays of Aristophanes are, unlike tragedies, at least set in the contemporary world of the audience rather than a mythologised past (DuBois, 2003, 119–25). Slave characters such as Xanthias in The Frogs and Carion in Wealth are near equals of their masters. Both characters, however, largely disappear from the action as the plays reach their climax. Carion is also less noble than his master. Xanthias’ character in The Frogs can only be understood against the political backdrop of the play. Athenians sometimes conscripted slaves into their navy and had recently offered some of them citizenship. Aristophanes appears to have disapproved, and Xanthias’ ability to swap places with his cowardly master in The Frogs may be a bad-tempered commentary on Athens’ decision.4

Aristophanes’ comedies are, as noted, full of jokes about casual brutality towards slaves. Together with the material from the law courts and philosophers, they reinforce the association of slavery and physical vulnerability. In a grimly funny scene in 'The Frogs', Xanthias engages in a rather peculiar contest with his master: they should both be whipped until one of them shows pain. His master may be a god, but Xanthias reckons that he has more than enough experience and will hold out longer.5 There is, admittedly, certainly trust between master and slave in Aristophanes’ comedies (as well as in Menander’s comedies of about 300 BC). This trust mirrors some of our evidence from non-comic sources: for example, in the use of slaves as commercial managers, mentioned above. Efforts by some historians to stress the positive elements in drama may be over-stated, however. Slaves remained secondary characters and their status was regarded as degraded and vulnerable.

Integration, resistance and anti-slavery thought

Writing around 400 BC, the “Old Oligarch” complained that Athenian slaves led relatively privileged lives, refusing to step aside in the street and indistinguishable from the free, who could not strike them.6 Considerable independence was indeed given to a few slaves, especially those involved in commercial activities. In addition, slaves sometimes allegedly claimed to be citizens. Athenian orators even occasionally claimed their opponents had slave parentage, and Athens felt forced to make several wholesale checks on the identity of its citizens. Some ex-slaves were even granted citizenship: they included the bankers Pasio and Phormio, amongst the richest Athenians known. It has been suggested recently that the dividing line between slave and free may, therefore, have sometimes been fluid (Vlassopoulos, 2009).

There are, however, reasons to be doubtful. Firstly, the testimony of the Old Oligarch is dubious. The booklet is a vicious attack on Athenian democracy and the “rule of the poor”. Its views on the privileged lives of slaves should perhaps be taken with a pinch of salt. In one law court speech, we hear of a citizen boy sent into a garden to pull up plants. The speaker’s enemies allegedly hoped that the garden owner would man-handle the boy assuming him to be a slave, thereby committing assault.7 Contrary to the Old Oligarch, this implies that one could normally differentiate slave from free (and that a slave could be beaten). Secondly, periodic scares about citizenship lists actually suggest a continuing Athenian desire to maintain barriers dividing citizen from slave or free immigrant. While isolated allegations of slaves impersonating citizens indicate that such cases were conceivable, they do not tell us how often this may have happened: defendants may have been telling the truth when they claimed they were free. Thirdly, Pasio and Phormio were probably highly atypical in gaining citizenship, which required a vote of at least 6000 citizens.

Athenians certainly did free slaves, though we cannot tell how often, or why. Freedom might repay loyalty, and sometimes acquaintances provided a loan for a slave to buy freedom. Semi-independent artisan slaves were probably more able to buy their freedom than mining or agricultural slaves, but we can see nothing like Roman peculium, slave pocket-money. It is possible that ex-masters had some rights to the work of their freed slaves, but the evidence for this in Athens is poor. Ex-slaves who remained in Athens became metics alongside free immigrants (Whitehead, 1977). They lacked political rights as well as the crucial privilege to own land. Athenian xenophobia would not have helped integration, nor would alleged contempt for the manual trades pursued by many slaves and ex-slaves. Athenian drama suggests that metics were best advised to remain inconspicuous. 8 When metics fought to help restore the democracy after a particularly bloody right-wing coup in 404/03 BC, Athenian citizens were ultimately very grudging in their thanks (Whitehead, 1977, 154–59). The social mobility of non-citizens, and particularly non-Greek slaves, should not be exaggerated.

We have no examples of slave rebellions from Athens, and few from the rest of the Greek world. They may have been under-reported, but Greek writers typically associated rebellion with helots rather than slaves. In truth, considerable difficulties faced would-be slave rebels in Athens. Slaves mostly lived with or near their masters and in small households. Differences in origins and occupation would also have undermined potential solidarity. Geographical conditions did not provide places of easy refuge. “Slave contentment” therefore need not explain the lack of rebellion, as has been suggested. Similarly, the willingness of Athenians and others to conscript slaves (generally as rowers rather than as infantry) probably indicates desperation in times of crisis rather than an expectation of devotion (Hunt, 1998). There are one or two stories of slaves murdering masters, but it is difficult to tell how major a problem this was perceived to be. The willingness to execute a whole household when a master was murdered betokens a horror of the possibility, though it need not imply much about its actual frequency. Some literary passages have been used to suggest that Athenian slave owners feared their slaves. Plato wondered, for example, about the fate awaiting an owner marooned with his slaves.9 Plato’s point was, however, that slave owners in Athens were not on the equivalent of a desert island.

Regardless of the fear of violence, owners still clearly faced other forms of resistance. While there is no suggestion that Athenians chained the bulk of their slaves, there were certainly concerns that slaves might run away, as allegedly more than 20,000 did in wartime after 413 BC. Greek writers also mention fears of slaves pilfering and gossiping, the latter of real concern in an honour-based society such as Athens. References in Aristophanes’ comedies to the tortoise-like back of the slave, hardened by repeated whippings, may also indicate poor work as a form of resistance. Talking back was an option allowed fictional comic slaves, though legal and other sources defined slaves as people who could not say what they wanted to. Ultimately, while some recent authors have tried valiantly to rescue the story of slave resistance from the silence of our sources, slaves in Athens were notable chiefly by their incapacity to defend themselves.

Slavery was largely accepted by the free as a fact of life. The philosopher Aristotle produced a short justification of slavery (Garnsey 1996, ch. 8), suggesting that at least some criticism of the institution existed, though this was probably directed more at Spartan helotry than Athenian chattel slavery. Aristotle’s defence rested on four key assumptions. (1) Racism: Aristotle believed that easterners were intelligent but soft, northerners being strong but stupid. (2) The belief that some people had to be controlled for their own good. (3) The perception that slaves looked and behaved “slavishly”, probably a warped reflection of the roles they were forced to perform. (4) Someone had to do the work. Aristotle knew that some slaves, notably Greek war captives, did not fit his criteria of natural slaves, and his justification of slavery has been condemned for its inconsistencies. Contemporary Athenians, however, may not have noticed them as they shared his core assumptions. The possibility of one’s own enslavement in times of war may also further have eased any pangs of conscience. Some fourth-century writers did, however, begin to question the morality of enslaving fellow Greeks. It is noteworthy that the mistreatment of a woman originally captured from the Greek town of Olynthus by Philip of Macedon in the 340s provoked scandal in Athens, a very unusual example of outrage over the abuse of a slave.10

Roman slavery

The Roman Empire had approximately 50–100 million inhabitants at its height, 7–14 million of them in Italy. It was, therefore, a vastly larger society than classical Athens. It has been argued that 30–40 per cent of the population of Italy were slaves, and 10–15 per cent of the whole empire. These estimates are speculative, however, and the current trend is to revise them downwards. In general, one would expect Italy, home to the imperial elite, to have had the highest proportion of slaves in the empire, followed by wealthy areas in the east, and then the more Romanised areas in the west, especially northern Africa, southern Spain and southern France. Slavery may have been relatively restricted in other areas such as northern France.

Slaves came from the children of existing slaves, free children left exposed to die by their parents and raised as slaves, or individuals purchased or captured outside the empire. It has been argued that at least 75 per cent of Rome’s slaves must have been bred, otherwise the sources outside the empire would have been denuded and improbably high numbers of exposed children would have been required from within the empire (Scheidel, 2005). Whilst criticised, this remains the most convincing picture of Roman slave sources, if 10 per cent or more of the empire’s inhabitants were indeed slaves. It implies that few female slaves of child-bearing age were freed.

The orthodox explanation for the massive development of chattel slavery, particularly in Italy, comes from the work of Moses Finley (1998, ch. 2) and Keith Hopkins (1978, ch. 1–2). Both stressed the unavailability of indigenous labour. Roman peasants were vital to the army and had important political rights that discouraged their direct economic exploitation. Hopkins argued that the profits from the expansion of the empire flooded into the hands of the elite, allowing them to buy peasant farms which had been undermined by their owners’ long absences on military service abroad. The elite also used their new wealth to buy slaves to work this land. The army and cities – whose populations grew as peasants left the land – also provided markets for the new slave-run estates.

Increasing doubts have been raised about this model (De Ligt, 2006). (1) Literary evidence and the results of archaeological surveys suggest that small farms may have survived on a much greater scale than Hopkins believed. (2) Rather than undermine small farmers, military service may have helped them by removing surplus labourers at points in the life cycle of farms. (3) Long-term urban growth in imperial Italy implies a continuously replenished supply of rural emigrants and therefore a continuing population of small farmers. (4) Finally, it has been argued that slavery was not well suited to grain production, and that only a few hundred thousand slaves (not the two or three million claimed) would have been required to produce other key Roman crops (notably olives and grapes).

The critics of the Finley/Hopkins model, however, themselves face difficulties. Firstly, even if we dramatically reduce our estimate of the number of slaves in Italy and accept that many small farms survived, slaves were still clearly crucial to key areas of the Roman rural and urban economy. For example, Roman agricultural writers assume large farms had permanent slave workforces. Secondly, the only alternative explanation proffered to the Finley/Hopkins model suggests that slavery developed because it offered a more flexible and profitable labour force than free workers or tenants. If this were true, one wonders, as with Athens, why mass chattel slavery has not been more common historically. Hopkins’ answer at least fits with the surviving literary evidence.

Who owned the slaves? We know very little about slave ownership among poorer Romans. It was probably a common aspiration, though slaves probably typically cost the equivalent of several years of food for a family. The ownership of hundreds of slaves does not seem unusual for the elite, consisting of the richest few thousand Romans, and ownership of thousands is known.

The sheer variety of slave jobs makes generalisation difficult. Most slaves were probably active in agriculture, domestic work or urban trades, with mining less significant than in Athens. A number of Roman authors, including Cato, Varro and Columella, discuss slaves in agriculture. Individual farms might have some dozens of slaves. They lived in barracks with a slave overseer, and masters were often absent for part or all of the year. Some slaves were chained, most probably not. We have little evidence of any family life for farm labourers except for overseers, though recent work has suggested that female slaves (and their children) may have been much more significant than previously believed. Apart from arable farming, slaves were also used in herding. Here we have a little more evidence of family life and indications of a degree of independence, perhaps one reason why herders are relatively prominent in the stories of slave rebellions discussed below.

Our evidence for domestic slavery is heavily biased towards the households of the very wealthy. The degree of specialisation in such households is striking: male and female slaves ministered to every physical and mental need of their owners. Romans were happy to use slaves in jobs requiring a high degree of training, including secretaries and teachers. They were proud of their educated, especially Greek-speaking, slaves. Domestic work was not, however, the only activity of city-based slaves. Epigraphic evidence indicates slaves engaging in an extraordinary range of trades, and legal evidence shows masters operating as silent partners in businesses run by slaves or ex-slaves. Slaves and ex-slaves are particularly prominent in the inscriptions put up by urban workers. It has been suggested that these indicate a pride in work consciously opposed to the anti-labour prejudices of the slaves’ masters. Little can be said of the organization of labour, apart from exceptional cases such as the factory production of pottery at Arezzo, though businesses were probably generally small-scale. It is clear, however, that many slaves and freedmen joined trade- and household-based social and funerary clubs known as collegia.

Slaves owned by the state, whether by local municipalities or by the emperor, form a final important group. As Rome expanded its empire in an ad hoc fashion, aristocrats were given short-term control of provinces and war zones, and relied on their slave households for secretarial support. The first emperors took over many of the functions of the Republican aristocracy and created something close to a professional bureaucracy staffed by slaves and ex-slaves. We shall see below that some of these could rise to positions of extraordinary influence. While many state-owned slaves had purely menial jobs, such as tending to aqueducts or making pipes, both municipal slaves and particularly the emperor’s slaves seem to have enjoyed advantages compared with ordinary slaves. Slaves owned by the municipalities were sometimes given pay and independent lodging. The imperial slaves appear to have had a higher chance than ordinary slaves of forming a partnership with a free woman which could become a legitimate marriage after manumission.

There is considerable debate over the treatment of Roman slaves, partly because of the difficulty in assessing the typicality of the surviving evidence (Bradley 1987 and 1994). The letters of Pliny the Younger and of Cicero, for example, as well as much of the surviving epigraphic material, suggest a close relationship between some slaves and their masters. These slaves were, however, predominantly urban rather than rural, and also often highly trained and especially useful to their masters. They may represent only a small minority, though it may still be significant that some owners wished to appear “good” masters to “good” slaves.

On the other hand, there are also shocking stories of abuse, as in Juvenal’s Satires or the philosophical work of Seneca. Some women, Juvenal wrote, paid an annual salary for someone to flog their slaves.11 There is, however, a problem with such anecdotal evidence. There were indeed professional slave-floggers and torturers. On the other hand, Juvenal’s text, in common with many other statements about Roman slavery, is part of a moralising discourse. Juvenal satirises the evil effects of female vanity and paranoia. Compare the story of how Vedius Pollio attempted to feed to his lampreys a slave who had broken an expensive goblet: the story is told to indicate Pollio’s meanness and viciousness.12 His guest, the first emperor Augustus, reacted with fury at his behaviour. The episode is often cited, but not always with its original moral.

The problem of interpretation is also very clear with one of our largest sources of evidence on Roman slavery: legal texts (Watson, 1987). Roman law dealt with slaves almost exclusively as property rather than as people. We should not, however, exaggerate the significance of the lack of recognition of the humanity of the slave in this material. The vast bulk of surviving Roman law concerns the adjudication of property disputes between the free, particularly disputes concerning business and inheritances. The willingness of Romans to use slaves as agents and representatives meant that quarrels concerning the consequences of slaves’ actions could become very complicated. Roman law concentrates on such issues and is, therefore, a potentially rather one-sided picture of attitudes towards slaves. There are, however, passages mirroring the kinds of abuses seen in other literature: for example, the possibility that a master might starve his slaves to death. Of course, the reason we hear of this was a desire to intervene to prevent it. Roman law attempted, at least theoretically, to protect slaves from some aspects of abuse by their masters from the first century AD onwards. For example, the emperor Hadrian forbad the throwing of slaves to wild beasts in public entertainments without the permission of a magistrate. Antoninus Pius attempted to check “excessive brutality” against slaves. Killing a slave “without just cause” was to be punished as severely as killing someone else’s slave.13 The murder of a slave was, therefore, not taken as seriously as the murder of a free man. In addition, “excessive” and “just cause” were doubtless tricky concepts to define, especially as fellow slave owners made the determination, and death from “justified” whipping was never legally punishable. While we have every right to be sceptical about the reach of legislation into the lives of most slaves, we should, however, also recognise that it at least reflected either a desire to establish new norms of behaviour, or existing ideas of “the done thing”. The evidence of Roman law, however, clearly conforms to Orlando Patterson’s (1982) famous definition of slaves as vulnerable, and this is true also concerning their lack of kin and of honour. Roman lawyers did not recognise any kinship relations between slaves, except for some prohibitions on incest, though we know from inscriptions that de facto families did exist. There were occasional qualms expressed about breaking up family groups, but separations must have been inevitable given Roman inheritance patterns. Only in the fourth century AD do we see any legislation attempting to prevent it. As for honour, protection of slaves from the actions of third parties was, as in Greece, conceived as protection of the property of the master. There was, however, potential recourse available with the master’s support for serious assault, and here the honour of the slave could, at least potentially, be considered.14

There were no legal controls on what a master might do sexually with his slaves, apart from repeated and apparently unsuccessful attempts to prevent the trade in castrated slaves.15 There was little discussion of what might constitute the limits of acceptable sexual behaviour towards slaves before the Christian era and, it has been suggested, even then (Bradley, 1987, 116–18; Glancy, 2002). Slave prostitution was widespread and there were no taboos against masters sleeping with their female slaves. Any child produced was a slave. Slave women were sometimes freed on the condition that they marry their ex-owner (though the practice was prevented at the very highest levels of society), and widowers (even emperors) also sometimes took slave or ex-slave concubines: new offspring would not then threaten the inheritance of a dead wife’s children. Sexual relations between masters and male slaves were probably less common than with female slaves, though poets such as Martial and Statius used erotic language to describe “pets” who were usually adolescent and pre-adolescent male slaves. Relationships between free women and slaves were, typically, not socially acceptable, and could face legal sanctions, though it is difficult to know how far they affected the lowest levels of society.

The slave in literature, particularly in drama, is again the focus of some very interesting recent work (Fitzgerald, 2000). “Cunning slaves” are often at the centre of the comedies of Plautus, usually struggling to help their master’s son get the girl of his dreams by tricking other characters, including their master. Other slaves give speeches stressing their loyalty, and in one play a “noble” slave saves his master from captivity by switching places with him. On the other hand, slaves are also called “whipping posts”, and threats of violence towards them are common. It is not difficult to see how different historians have used this material to argue both a negative and a positive picture of slave life.

Almost all of the arguments used can be turned on their head. The cheeky but loyal slave may depict not reality but the upside-down carnival world of comedy, representing a mechanism by which the young and powerless could mock their elders and social superiors. Slaves who give speeches about their loyalty almost immediately prove to be fools incapable of doing what their master wants. The noble slave who saves his master was, we discover, originally a free man kidnapped as a child and, fortuitously, the son of the man currently holding him prisoner. Finally, as in Athens, we cannot determine the reality of violence against slaves simply from the names they are called in comedy. “Whipping post” may be shorthand for “bad”, and it clearly indicates the possibility of brutality and punishment, but it tells us little about its frequency. The threat of extreme violence in drama may be comic in itself, and some characters are criticised precisely because of their unreasonable violence towards slaves.

The nature of the differing types of evidence on Roman slavery, therefore, allows great latitude in general attempts to reconstruct slave life. Much is hidden from us. Our evidence is patchy and potentially unrepresentative of slaves as a whole. One should not lose sight, however, of the vulnerability of the slave and the clear desire of slaves to gain freedom. While Romans certainly discussed the point at which the acceptable treatment of slaves turned into socially reprehensible mistreatment, that could be seen in a negative light, suggesting higher levels of abuse: one made a son obey, but broke a slave.

Integration, resistance and anti-slavery thought

Securing one’s freedom in Rome seems to have been a result of an emotional bond with one’s master/mistress or else an economic exchange making use of the peculium (Andreav 1993). The peculium consisted of money or goods that an owner allowed a slave. While it always remained the property of the owner, Romans had good reason not to take it away arbitrarily. The peculium allowed a form of “limited liability” insurance when slaves operated independently of their masters in business. It was also a crucial mechanism in stimulating labour as slaves saved to buy luxuries, including other slaves, or freedom. We do not know how many slaves had a peculium, though it appears to have been widespread among urban slaves. Nor can we calculate the chances of a slave being freed. One throwaway remark implied that slavery might typically last as little as six years. Given the demographic debate on Roman slavery, this seems an exaggeration, particularly for female slaves. This remains true even if the first Roman emperor Augustus felt the need to regulate manumission more closely. After manumission, freedmen had a duty of respect towards their masters even if some Roman writers complained that this was not always fulfilled. Masters also had considerable power over the ex-slave’s right to bequeath property and over the rights of freedwomen to marry. The freed also often had to complete operae, days of work for their patrons. The state was prepared to protect ex-slaves from particularly demeaning or inconvenient operae, but it also defended the rights of the ex-master.

Ex-slaves exercised a degree in influence in at least three ways. (1) They were given citizenship, though there were political battles about how much weight their vote should carry. (2) Some were proverbially wealthy, helped by their experience as business agents, and possibly also by financial support and bequests from their patrons. They also had considerable success in fields such as medicine and literature. (3) Some wielded great influence in the imperial bureaucracy, particularly in the first century AD: the aristocrat Tacitus, for example, supposed that they effectively decided who would succeed the emperor Claudius.16 It has been suggested that these imperial bureaucrats were still slave-like in that they were kinless and without honour, though this may over-emphasise elite contempt of these men. It is probably better to see such slaves and ex-slaves as examples of status dissonance: scoring high in some status indicators, low in others.

There is some evidence to suggest that at the lowest social levels, slave and free and freed socialised easily. The stigma of being an ex-slave – or even being descended from a slave – could, however, be strong in some quarters. Petronius’ satire the Satyricon describes the fabulously wealthy fictional freedman Trimalchio. His crudity shows us how not to behave in society; he is the “parvenu who never arrived”. A number of children of freedmen did gain high political office in Rome, but social advancement was apparently easier outside the city of Rome: the descendants of ex-slaves were a vital element in the social elite of Italian municipalities such as Ostia. We should not simply accept the Roman aristocratic view of such men as tasteless upstarts, but is still clear that ex-slaves faced prejudice and discrimination. They and their descendants may nonetheless have constituted a large proportion of the upwardly mobile section of Roman society.

Rebellions involving tens of thousands of slaves occurred in Sicily in the late 130s BC, again in 104–01 BC, and finally in Italy in 73–71 BC (Urbainczyk 2008). They were focused on the countryside, and initially defeated full Roman armies before ultimate failure. The Sicilian rebellions allegedly even manifested a level of proto-state organisation. Perhaps surprisingly, some of the key ancient sources on these events manifest at least a limited degree of sympathy with the plight of the slaves and qualified admiration for some of the slave leaders (especially Spartacus).

A few minor rebellions occurred at other times in Roman history, but the concentration of the major rebellions within a 60-year period requires explanation. The empire had taken massive numbers of slaves, from specific areas such as Cilicia and Syria, who cooperated more easily than slaves of more diverse origin. Rome’s attention was also distracted by major foreign wars. Finally, contemporaries believed that imperial growth had brought a flood of cheap slaves and easy wealth, which encouraged some owners to treat their slaves particularly inhumanely, provoking insurrection. Some modern writers have downplayed the significance of the rebellions to the history of slavery, stressing problems with the surviving narratives, possible religious or nationalist causes, and the participation of free people in the revolts. While some of the details of the events are indeed questionable, the overall picture probably should not be rejected. There is little evidence that the religiosity of the slaves differed significantly from that of the Romans themselves, and Roman writers clearly believed that slaves, not disgruntled peasants, formed the bulk of the rebels. There is little evidence, however, that the slave rebellions prompted major change.

Some Roman politicians claimed that their opponents involved slaves in violent political struggles between citizens, using the collegia clubs where slaves, freed and free mixed. Such claims may simply be malicious, even if the Roman state was keen to regulate the collegia. Slave resistance was typically less dramatic than organised violence. Roman writers expressed horror at the killing of individual masters by slaves. Roman law declared that a slave household that failed to help a murdered master should be tortured and executed. The notion, however, that the Romans lived in fear of their slaves is highly debatable. It is difficult to tell whether murders were anything other than very exceptional. Most of our evidence implies that the obedience of slaves was expected and their loyalty hoped for. Romans did not lock their bedroom doors when they went to bed at night.

Slaves could, of course, resist their owners in non-violent ways. It is impossible to determine to what degree, though the strong association of slavery and discipline via the whip is suggestive. Roman law indicates many ways in which a slave could defraud or cause loss to a master, but such texts again inform us more about the possibility of such actions than their commonness. Concerns about runaways appear often. Those who wished to disappear were helped by the sheer size of the empire and the looseness of bureaucratic control, the relative independence of some slaves, and the difficulties of differentiating slave and free. Laws from the fourth century AD indicate that free men might help fugitives, possibly in order to gain labourers. Overall, however, effective opposition, including flight, would have been very dangerous and may have been less of a problem for slave owners than is sometimes implied.

Unlike Greece, Rome saw significant discussion of the morality of slavery (Bradley, 1994, ch. 7; Garnsey, 1996, ch. 9–14). The Stoic Seneca complained about the inhuman treatment of slaves, most famously in his Moral Letters 47. Stoics argued that men should be judged by their morality, not their legal status; masters addicted to luxury might be the “true” slaves, not those who served them. Stoics taught, however, that the troubles of the world such as slavery were to be endured, not resisted: the test of a man was how he dealt with his lot. Some early Christian thought was very similar. St Paul argued that in the eyes of God there was neither Jew nor Gentile, slave nor free, male nor female.17 He was not thereby implying that slavery be abolished (any more than gender), though he did suggest that masters should not treat their slaves harshly. Both the early Christians and the Stoics have been criticised for reinforcing slavery by undermining resistance among slaves, especially as Christians quickly moved to tell slaves that they should obey their masters. The Christian empire of the fourth century onwards certainly fully accepted slavery. Stoicism and Christianity may nonetheless have had some positive impact. The legislation protecting slaves in the early empire cannot be traced directly to Stoic thought, but it is possible that it was prompted by the kind of moralising debates we saw earlier in Roman literature, and they almost certainly were affected by such thinking. The legislation was limited, but the social values expressed were nonetheless significant. Christianity may have brought a little change regarding sexual behaviour, and perhaps a more positive attitude towards manual work. And if Christians told slaves to obey their masters, there were at least some injunctions for masters to behave properly too. Overall, however, Christianity quickly adapted itself to slave society and may have had less influence on slave life than Stoic-influenced pagan moralising.

The end of ancient slavery

At some point in the later Roman Empire, villa-based direct exploitation of agricultural slaves in Italy declined in favour of free- and slave-tenant farming. Slaves became increasingly like the rest of the rural poor as they were given a degree of autonomy and allowed to live in family units. There has been considerable debate as to why this change happened. It has been suggested that traditional slave exploitation simply became uneconomic at some point after 100 AD. Arguments that the end of Roman expansion at that time made captives more costly rest, however, on the questionable assumption that war was the main source of slaves and that breeding was uneconomic. Nor is there clear evidence, as some have claimed, that increasing supervision costs made slavery unprofitable. There is at least some archaeological support for the idea that the export markets of slave-run farms in Italy were undermined by the development of non-slave production centres elsewhere in the empire. Slave-run villas may also have been compromised by the decline of market production in favour of barter in the later empire. One feels slightly nervous, however, of such an economic explanation of the decline of Roman slavery when economic theories, as we saw, fail to clarify why slavery in Roman Italy reached such an unusual magnitude in the first place.

This, of course, brings us back to Moses Finley’s emphasis on the availability of exploitable indigenous “free” labour, his “default” historical system of exploitation (Finley, 1998, ch. 4). We saw that Italian peasants initially had important roles as voters and soldiers. They lost these as the empire first removed their voting rights and then recruited armies from outside Italy. Increasing economic pressure could therefore be placed upon them, making tenancy more profitable than directly supervised slaves: the remaining slaves could simply be converted into tenants. In addition, the decline in the market economy removed another of Finley’s preconditions for mass slavery.

A third position is possible, however. It could be argued that slavery never actually declined (Whittaker, 1993, ch. 5). Slaves appear often in our texts up to the end of the Western Empire in 476 AD, and also in sixth-century Byzantine law codes. Some medievalists argue for relatively high percentages of slaves in post-Roman societies such as Anglo-Saxon England, where slaves may have comprised ten per cent or more of the population. Historians of Roman slavery are, therefore, able to push the “decline” of the institution beyond the point they themselves study. It should be noted, however, that the move from direct agricultural exploitation of slaves to a tenant system requires explanation. Finley’s argument certainly appears the most attractive explanation in terms of its simplicity. It fails, however, to explain why Roman peasants had originally managed to gain such military and political influence in the first place and it is therefore difficult to prove or disprove. As such, it remains symbolic of much of the rest of the debate about Greek and Roman slavery.

Notes

1 Maidservants: Homer, Odyssey, 22.457ff, in A. T. Murray and G. E. Dimock (trans.), Homer, vol. 4. (Cambridge, MA: Harvard University Press, 1995).
2 Herodotus, Histories, 8: 105 in A. D. Godley, ed. and trans. Herodotus, vol. 4 (Cambridge, MA: Harvard University Press, 1925).
3 Euripides, Ion, 554–56 in David Kovacs, ed. and trans., Euripides, vol. 4 (Cambridge, MA: Harvard University Press, 1999).
4 Aristophanes, Frogs, 693ff in J. Henderson, ed. and trans., Aristophanes, vol. 4 (Cambridge, MA: Harvard University Press, 2002).
5 Ibid., 612–73.
6 Pseudo-Xenophon (“Old Oligarch”), Constitution of the Athenians, 1.10–11, in E. C. Marchant and Glen W. Bowersock, eds and trans, Xenophon, vol. 7 (Cambridge, MA: Harvard University Press, 1989).
7 Pseudo-Demosthenes, 53: 16 in A. T. Murray, ed. and trans., Demosthenes, vol. 6 (Cambridge, MA: Harvard University Press, 1939).
8 Euripides, Suppliant Women, 888–900 in David Kovacs, ed. and trans., Euripides, vol. 6 (Cambridge, MA: Harvard University Press, 1998).
9 Plato, Republic, in P. Shorey, ed. and trans., Plato, vol. 6 (Cambridge, MA: Harvard University Press, 1935), 578d–e.
10 Demosthenes 19.196–98, in C. A. Vince and J. H. Vince, eds and trans., Demosthenes, vol. 2 (Cambridge, MA: Harvard University Press, 1926).
11 Juvenal, Satires, in 6, 474ff S. M. Braund, ed. and trans., Juvenal and Persius (Cambridge, MA: Harvard University Press, 2004).
12 Seneca, On Anger, 3.40 in J. W. Basore, ed. and trans., Seneca, vol. 1 (Cambridge MA: Harvard University Press, 1989).
13 The Digest of Justinian, 48.8.11.2 in A. Watson, ed. and trans., The Digest of Justinian, vol. 2. (Philadelphia: University of Philadelphia Press, 1985), Justinian, Institutes, 1.8.2 in J. A. C. Thomas, ed. and trans., The Institutes of Justinian: Text, Translation and Commentary (Amsterdam: North Holland Publishing Co., 1975), 1.8.2.
14 Digest, op. cit., 47.10.15.44; 47.10.15.35.
15 Digest, op. cit., 48.8.3.4; 48.8.5–6.
16 Tacitus, Annals, 12.1–3, in J. Jackson, ed. and trans., Tacitus, vol. 4 (Cambridge, MA: Harvard University Press, 1937).
17 Galatians 3.28 in The Good News Bible (London: Collins, 1976).

Written by Niall McKeown in "The Routledge History of Slavery", Routledge, UK, 2011, edited by Gad Heuman and Trevor Burnard, excerpts pp. 19-33. Digitized, adapted and illustrated to be posted by Leopoldo Costa.