RESISTING SLAVERY AT ROME

Keith bradley

The orator Pupius Piso, wishing to avoid being unnecessarily disturbed, ordered his slaves to answer his questions but not add anything to their answers. He then wanted to give a welcome to Clodius, who was holding office, and gave instructions that he should be invited to dinner. He set up a splendid feast. The time came, the other guests arrived, Clodius was expected. Piso kept sending the slave who was responsible for invitations to see if he was coming. Evening came; Clodius was despaired of. ‘Did you invite him?’ Piso asked his slave. ‘Yes.’ ‘Then why didn’t he come?’ ‘Because he declined.’ ‘Then why didn’t you tell me?’ ‘Because you didn’t ask.’ Such is the way of the Roman slave!

This anecdote is recorded by Plutarch (Moralia 5iid—e) of M. Pupius Piso, the consul of 6i bc. It may not be literally true. But if it has any plausibility at all, which it must, it suggests that slave-owners in the Roman world of Plutarch’s era were well aware that their slaves could present challenges to their authority at any time and even place them, if only for a frustrating moment, in a position of powerlessness they normally expected their slaves alone to occupy. To express and circulate the idea that a slave could crushingly embarrass his master by obeying his instructions to the letter was to acknowledge that slaves were capable of resisting slavery.

Like chattel slaves in other periods of history, Roman slaves were deracinated, disempowered beings who enjoyed no personal or social identity other than that which derived from association with their owners. They were permitted no formal ties of kinship and, lacking all legal personality and rights, were forcibly held at their owners’ discretion in shameful, infan-tilising subjection. Varro (On the Latin Language 9.55, 59) said that their lack of a gentilicium rendered slaves virtually devoid of gender, Valerius Maximus (6.2.8) that they could never remove slavery’s smell, Augustine (City of God 19.18) that theirs was a condition of unimaginable ill. The relationship between master and slave was by definition asymmetrical, comparable to that between a tyrant and his subjects (Philostratus, Life of Apollonius 7.42). But as Plutarch’s story suggests, it was not altogether

One-sided. Slaves were human chattels, and human agency could manifest itself in the relationship from moment to moment. Unlike the animals to which they were often compared, slaves were not easily manipulable, but had to be managed with thought and discretion to make sure that they did what was required of them, through rewards (munerd) and punishments (poenae) in the vocabulary of the moralist Seneca (To Marcia 10.6). The relationship between master and slave was one therefore that on both sides involved constant adjustment, refinement and negotiation. It was a contest of wills, a psychological struggle for power in which the energies of both sides were constantly implicated.

Resistance is a category of analysis that historians of slavery commonly employ to refer to acts of defiance and protest by which slaves, not only at Rome but in all slave societies, contested the presumptive right of slave-owners to demand services from and impose claims upon them, the claim of labour included. It is often set against a contrasting category of accommodation, a term commonly used to refer to forms of behaviour showing how slaves accepted or acquiesced in their enslavement. Sometimes accommodation is understood to have incorporated subtle modes of subversive behaviour and to have constituted a form of resistance in its own right, clandestine, purposefully concealed activities that while less potentially threatening to the slaves who undertook them were motivated by the same spirit of opposition to enslavement visible in more open and dangerous acts ofdefiance and protest. In either case resistance is a concept that usefully structures answers to the question of how slaves responded to slavery.

Regrettably there is little evidence from Roman slaves themselves that allows direct views of their responses to slavery to be seen. Rome produced no Frederick Douglass or Harriet Jacobs to give detailed accounts of slaves’ experiences of life in slavery — or if it did, the accounts have not survived. The history of Roman slavery depends instead on sources of information that overwhelmingly represent the views of the slave-owning classes, and this means that any attempt to penetrate the mind of the slave — a necessary condition for explaining slave behaviour — has to be largely a matter of inference. In late antiquity Augustine (Confessions 9.8) recorded that his mother had once suffered an insulting reproach from a female slave. The incident occurred in the family wine cellar where a quarrel had broken out between the two women. Monica, who was young at the time, was a secret drinker and the slave chastised her for her misbehaviour, with words all the more bitter, Augustine said, because they came from someone who as a slave was more or less an enemy (the conjunction of ideas is notable). Monica was stung into abandoning her bad habit. But precisely why the slave taunted her to begin with was a mystery: either she was angry and wanted to hurt Monica, or else she was afraid of being punished if it were later revealed that she had not disclosed Monica’s guilty secret. The story as Augustine tells it must work to his mother’s credit, and he tells it carefully. But the fact that he offered two explanations of the slave’s behaviour shows plainly enough that he did not know, and never could have known, what was in the woman’s mind when the interchange took place; his explanations only project what he thought likely and must have depended on what his mother told him long after the event itself. The slave woman could have had an altogether different view of things, but there is no way of seeing it. The story serves as a paradigmatic illustration therefore that the intentions behind Roman slaves’ actions can only be inferred from the record of events, which means that a certain degree of speculation must always surround any reconstruction of slave activity. When a slave kept for the sole purpose of amusing an owner and his guests with smart verbal cracks was at his most impudent, and, it seemed to Seneca (On Constancy 11.3), most entertaining, who knows what was happening in the slave’s mind? Was the slave simply doing the job expected of him, or was he covertly expressing through his work opposition to the man who owned him? Who was exploiting whom? Or again, when the slaves who carried Fronto (Letters to M. Caesar 5.44) to the baths in his sedan chair and caused a painful injury to his knee, was it just a matter of chance, or had the slaves wilfully conspired to inflict harm on their master under the guise of an accident?

The main period of concern in this chapter is the roughly four and a half centuries of Roman history from the age of the elder Cato to the age of the Severans, when chattel slaves comprised perhaps a quarter to a third of the population of Rome and Italy, with vast but strictly indeterminate numbers engaged principally in agriculture, mining and domestic service. (There is no indication at any time in this period that slaves were in short supply.) Elsewhere in the Roman Mediterranean the proportion of slaves in the overall population was smaller, and numbers varied from region to region. Slaves were more visible in some quarters than others. The ideology and practice of slave-owning prevailed everywhere, however, and no area was immune to its influence or completely free from it. The passage of slaves through their cities, towns and countryside for disposal in the Roman heartland must always have been a common sight for many of Rome’s provincial subjects.

The most obvious way in which slave resistance presents itself in this period is in the record of slave revolts, the most famous, or infamous, of which was the revolt led by the gladiator Spartacus in 73 bc. Spartacus and a group of followers broke out of a gladiatorial training-school at Capua, attracted massive support from rural slaves in central Italy, and quickly converted a small act of rebellion into a major war against Rome that came to involve thousands of rebels. Or so the course of events was seen by the authors of the narratives of the uprising that still survive. Organising themselves along paramilitary lines, for two years the slaves held on to a tenuous freedom won by their initial acts of insurgence and defeated both the local and the metropolitan forces despatched against them. Their successes in battle were so great that at one point they threatened Rome directly. The appointment, however, of M. Licinius Crassus (cos. 70 bc) in 72 to lead the Roman war effort proved a turning point. With a major legionary force at his disposal, Crassus systematically wore down and finally defeated the slaves in southern Italy. Spartacus himself died in battle, and Crassus, notoriously, crucified 6,000 of his supporters along the Appian Way between Capua and Rome.

Spartacus’ aims can be perceived now only dimly. Notions made popular by Howard Fast’s romantic novel of 1951 and the Stanley Kubrick film that followed, that Spartacus wished to abolish slavery and create a utopian classless society in which equality, including equality between men and women, was the watchword, have little foundation in fact. Freedom from the brutality of enslavement was the initial motivation for revolt, and in all likelihood Spartacus’ intent was to maintain freedom for as many as possible as long as possible. (The powerfulness of the desire to be free should never be underestimated.) This objective, however, was rendered difficult by the great dimensions of the uprising, which as far as can be told were completely fortuitous and never foreseen or planned by the original insurgents at Capua. The volatile political and military conditions of first-century Italy probably facilitated escalation of the movement, but in the event Spartacus was unable to exercise full control over the astonishingly high numbers of slaves who joined him, and unanimity of purpose proved impossible to achieve.

Spartacus’ revolt was not the first large-scale uprising of slaves with which Rome had to contend. Two similar slave wars occurred in Sicily in the later second century bc, both of which involved huge numbers of slaves and both of which required a substantial Roman military response. Again, however, there is little sign that either revolt was well co-ordinated or planned from the outset as a general insurrection of all Sicilian slaves: each began as a localised uprising that developed almost spontaneously into a larger action according to the extant accounts. On both occasions the rebels were able to turn themselves into an efficient fighting force, and, adopting methods and strategies comparable to those of slave maroons in later history, held out for several years against the armies Rome sent to quash them. They concentrated their energies in such mountain fastnesses as Enna and Tauromenium, ideal features of the Sicilian landscape for maroon-styled tactics (as in another context were the Pyrenees [Caesar, Civil War 3.19]), but in both cases Roman military might ultimately and inevitably prevailed. The variegated character of the slave population, the prevalence of large numbers of first-generation slaves, and the harsh conditions resulting from intense exploitation of the local agrarian economy were factors that contributed to the scale of the insurrections. But again as far as can be told, the rebels’ intentions were only to extricate themselves from slavery, not to alter in any fundamental way existing social and economic structures; as in the case of Spartacus’ revolt, there were no calls for ending slavery.

Other episodes of revolt occurred through the course of Roman history, but on a much smaller scale. The murder of L. Minucius Basilus (praet. 45 bc), one of Caesar’s assassins, reported by Appian (Civil Wars 3.98), is one example, the murder of the praetorian senator Larcius Macedo in ad 108 described by the younger Pliny (Letters 3.14) is another. Both were assaults on slave-owners who were believed to have been excessively cruel, a motive commonly understood to produce a violent reaction in slaves even if assailants faced the risk of subsequent execution (on the cross, for instance [cf. Sen. On Clemency 1.26.1]). There were probably many other similar attacks to judge from the passage under Augustus of the senatus consultum Silanianum, a senatorial decree later modified under Nero and several other emperors that established rules for legal cases in which slaves were alleged to have killed their owners. In the event of a slave-owner’s death, all the slaves in a given household were with few exceptions liable to torture until the offenders were discovered, and in the meantime the owner’s will remained unopened so that no slave could or would benefit from it. If the will provided for the manumission of certain slaves, they could be subjected to torture anyway and be punished as necessary. Even those slaves owned separately by a victim’s wife were liable to examination. The extreme severity of the legal provisions is notable and perhaps points to a certain deterrent value in the senatorial decree. But developed and extended as its harsh provisions were, the law points above all to what can only be taken as genuine and reasonably frequent assaults by slaves on slave-owners, and to the practical steps that had to be taken once assaults proved murderous. To slave-owners, it appears, slaves were always a threat to life and limb, and owners were always alert to the possibility of revolt. Seneca (Clem. 1.24.1) tells of a proposal that was once made in the senate to make slaves easily recognisable by having them wear distinctive clothing — at Rome slavery was not associated with skin colour of course — but the idea was abandoned once the senators realised that, if implemented, the proposal would make slaves conscious of their numbers and induce them to make common cause against their owners. The anecdote hints at the permanent state of hostility that some thought existed between slave-owner and slave (cf. Sen. Moral Epistles 4.8.), at the equation that was often made between slaves and enemies (cf. Dionysius of Halicarnassus 10.59.6; Sen.

Ep. 18.14, 47-5; August. Conf. 9.8.), and at a constant awareness of the potential in slaves for violent resistance to slavery.

The anxiety detectable in Seneca’s story must sometimes have become almost palpable. Livy (4.44.13—4.45.2) and Dionysius of Halicarnassus (12.6.(6); cf. 5.51.3, 5.53.3—4) wrote, more or less simultaneously, of a plot slaves were thought to have once formed in Rome’s early history to secure the Capitol and other key points of the city, first setting fires at night to divert attention, then appealing to slaves throughout the city to seize their freedom, kill their owners, and assume possession of their owners’ wives and property. The historicity of the event is dubious. But the report might well be taken as evidence of a chronic suspicion of revolt — and of how revolt was thought likely to unfold — that was deeply embedded in Roman society of the central historical age. From an establishment point of view, the report had a reassuring and even comforting end: the plot was betrayed from within and the chief culprits were crucified. But this was hardly enough to dispel the fear that trouble might erupt at any moment. Livy elsewhere (21.41.10) makes clear that it was insulting for a free man to be attacked by a slave, and that he should respond indignantly and angrily. Assault was predictable. The slave-owner as much as the slave could pose the question: ‘Should slaves be obedient to their masters or refuse the wishes of those that possess their bodies?’ (Philostr. VA 7.42).

For present purposes the most significant point about the record of revolt is that it provides unambiguous evidence that slavery was resisted. Some slaves at Rome under some circumstances found the means to contest the authority of slave-owners directly, no matter what the threats to personal safety. But whereas no one could forecast when a distressed slave might attack a slave-owner in the heat of the moment, slaves could calculate as well as anyone else that to plan revolt was to enter dangerous territory: the risks of betrayal and the prospects of adverse effects, in the event of failure, on family relations for instance were high. In the slave regimes of the New World, except for the single case of the rebellion led by Toussaint L’Ouverture in St Domingue at the end of the eighteenth century, insurrection of a revolutionary kind was unknown and revolt at large was sporadic rather than constant, a desperate strategy likely to result in punishment or death. So too at Rome. As the New World record also makes clear, however, there were less dangerous methods of expressing discontent, by running away, by sabotaging property and by finding ways to deceive slave-owners. These methods are equally noticeable in the Roman record.

One illustration of the high profile in everyday Roman life of running away appears in a title of the Digest (11.4) headed ‘De fugitivis’ — fugi-tivi being the usual, and pejorative, Latin word for runaway slaves. The title is only a fragment of the full body of Roman law that dealt with fugitive slaves, but as a sample it is enough to point to the sorts of circumstance that running away involved, and to show the limits of what can be understood.

The law considered a thief, and thus criminally liable, any person who harboured a runaway slave. Under penalty of a fine, farm bailiffs and private estate agents (procuratores) were forbidden to protect runaways and to allow them to enter wooded areas; the fugitives were to be surrendered to owners or civic officials within a period of twenty days. A person who refused to allow his property to be searched or who found fugitives on his land and did not expose them was also subject to a fine; and a person who apprehended a runaway had to hand the slave over in public to civic authorities. Searches for runaways could be conducted by military or civilian personnel, and civic magistrates were obliged to co-operate with those investigating. A provincial governor was likewise required to assist a slave-owner pursuing runaways on a third party’s property by giving him a letter authorising entry and use of an assistant if necessary. The governor could penalise any landowner who refused to give permission to enter, and the property of all social classes was subject to search, including that of the emperor and members of the senate. A slave-owner wishing to return apprehended runaways and to punish those who had harboured them was entitled to assistance from provincial governors, civic officials and troops, who together with harbour-masters had powers of arrest and were obliged to guard recaptured runaways carefully in order to prevent further escape, with chains if needed. Runaways who had been recaptured were to be brought before thepraefectus vigilum or a provincial governor, returned to their owners and, if they had pretended to be free while at large, punished severely. To help magistrates apprehend them, the law said that physical descriptions and owners’ names were to be posted in public places and temples, with special attention to identifying features such as scars.

Evidence such as this does not allow the incidence of flight by Roman slaves to be measured statistically; but the mere existence of law establishing procedures for dealing with runaways is enough to indicate the seriousness of flight as a social issue. The various provisions suggest that runaway slaves were expected to appear in any and all regions of the Roman world, and that sympathetic third parties could be expected to render aid and assistance at any time — which is interesting as a reflection perhaps of how widely attitudes towards slavery may have differed in society as a whole. Equally, however, the law created the firm expectation, if not obligation, that all respectable members of society were to co-operate in capturing fugitives and restoring them to their owners, asserting thereby the primacy of conventional property rights no matter what a nuisance it was, as Columella noted (On Agriculture 1.5.7), to have someone travelling across a private estate.

Viewed from the perspective of the slave, the law also communicates a sense of how fraught the enterprise of running away was. Advantageous collusion with harbourers was possible, it seems, and difficulties in implementing the law might sometimes have worked in slaves’ favour: institutional resources for ensuring that the law was carried out effectively were always relatively limited at Rome. Also, the law recognised that runaways might claim to be free, a claim presumably based on real experience, and evidently enough skin colour was not an automatic impediment to successful escape for Roman runaways, as it was for slaves in the New World. Overall, however, the chance of betrayal must have been high in view of the penalties stipulated for concealing them, while the prospects of help from third parties, and of success in general, cannot often have seemed great in view of the constraints on the free to collaborate in finding and capturing fugitives. Also, because of the powers of search the law allowed, runaway slaves cannot have failed to know that pursuit was inevitable and that discovery would mean re-enslavement if not more extreme consequences. Flight may have been less dangerous than revolt, but it was dangerous nonetheless.

Whether Roman law accurately reflects historical norms or simply refers to bizarre oddities or possibilities is always a question to be considered. One provision on flight (Dig. 11.4.1.5) stipulates that a child born to a fugitive slave woman was not to be considered a runaway slave, while a second (Dig. 11.4.5) states that runaways remained under the authority (potestas) of their owners even if they volunteered to fight in the arena, and that they had to be returned to their owners — presumably on detection — because they might have embezzled money from or committed some other crime against them that needed redress. Both provisions exemplify the sort of close attention to detail that Roman jurists relished. One real case of a runaway slave woman giving birth to a child might have been enough to produce the first provision, and one real case of a runaway volunteering to fight in the amphitheatre might have been enough to produce the second. But both items could equally well have presented themselves to jurists as matters of speculation alone. It is impossible to tell. On balance, the law’s assumption that flight was a common problem that from time to time could involve very complicated circumstances seems more important than any particular circumstances described, but there is surely also an inherent plausibility to the two sets of circumstances underlying the provisions even if specific cases were rare.

What cannot be recovered at all, from the law or any other evidence, is the inner debate about the decision to flee that many Roman slaves who ran away must have had with themselves, or the debates that might be imagined they had with one another or with sympathisers. Hints emerge in the sources of the immediate causes of a decision to run away — perhaps a flogging that was the proverbial last straw — but it is a different matter with the calculations and preparations for flight that must often have been necessary. Did slaves for instance think of the slave-catchers who would be sent to pursue them? How would they make contact with potential harbourers, or plan to evade detection by city magistrates, troops and provincial officials? How would they decide on a destination to which to flee? Did they consider what it would be like to wear one of the iron collars that were placed around the necks of recaptured runaways if they were caught, advertising the fact that they had once tried to escape? Questions like this are raised in Apuleius’ novel, the Metamorphoses (6.26), when an opportunity for running away from his owners presents itself to Apuleius’ comically deliberating slave-like ass (the transformed hero, Lucius): ‘But where in the world will your flight be directed? And who will provide sanctuary for you?’ They have enormous point. In the absence of direct testimony from slaves, Lucius’ fictional dilemma is helpful for understanding the dilemma of the slave who was on the verge of running away, no matter what its comic character.

It is here, also, if only for imaginative purposes, that comparative material from New World slave societies proves useful, particularly accounts of flight produced by men and women who had once themselves been slaves. The autobiographical Narrative of the Life of Henry Box Brown is a case in point. Henry Brown was born of slave parents in Virginia in the early nineteenth century. As a child and youth, he lived with his family in comfortable circumstances in the ownership of a relatively caring and generous master. But on the owner’s death, the estate of which he was a part was divided among the owner’s several sons, and Brown found himself separated from his parents and siblings, experiencing for the first time, as he wrote in his work, the intense emotional pain and suffering that slavery was always likely to cause. A second incident some years later when he had grown up and had a family of his own proved a turning point in his life. Brown returned to his house in Richmond one day from the tobacco factory where he worked to find that his wife and their children had been sold and were about to be taken away to North Carolina — despite his owner’s promise that he and his wife would never be parted. Taken away his wife and children were, and Brown witnessed their agonising departure. Shortly afterwards he determined to avenge their loss by running away. He knew that he could secure freedom if he escaped to the North or to Canada. The question was how to do it without being caught. He hit on the ingenious idea of mailing himself in a crate to Philadelphia, and with the help of sympathisers carried out his plan on 29 March, 1849, travelling the 350 miles from Richmond in a wooden box and miraculously surviving the journey of twenty-seven hours with just a little water to sustain him. Thereafter Brown made a living in the abolitionist cause by capitalising on his adventure, giving lectures and composing the memoir of his life in slavery and his daring flight to freedom. A first edition was published in 1849 and a second in 1851. But Brown apparently never saw his wife and children again.

Like all slave narratives, the Narrative of the Life of Henry Box Brown is a highly tendentious work, designed to dramatise the horrors of slavery and to promote the cause of emancipation. Yet it allows resistance, and especially the human factors of familial disruption and forced separation that provoked resistance, to be understood in a way that sources produced by slave-owners cannot. Flight was a common activity in all New World slave regimes, well known, for example, from advertisements offering rewards for the return of runaways in newspapers from eighteenth-century Virginia and nineteenth-century Brazil, and consistently regarded by slave-owners as evidence of miscreant and irresponsible character on the part of those who fled. A slave narrative such as Brown’s, however, reverses the negative images constructed by slave-owners and exposes a positive view of what flight involved from the slave’s vantage point: the arrival of a moment when submission and its destructive consequences could no longer be tolerated; the power of the desire to live in freedom and security of person; the courage required to defy authority; the audacity needed to act. Despite its tendentiousness, Brown’s writing humanises and justifies what might otherwise at best be regarded, superficially, as a slave’s mere disobedience.

The Roman evidence on running away is copious, extending beyond references in law codes to a multiplicity of allusions in literature and notices of individual runaways preserved on papyrus that resemble the newspaper advertisements of more modern times. But there is nothing resembling the modern slave narrative, which means that a dimension of knowledge about slave behaviour is automatically lost to Roman historians. The motives prompting flight, however, can scarcely have been different from those of runaways in modern slave regimes, and the acts of flight themselves hardly less courageous. Roman slaves were physically and psychologically coerced like their modern counterparts, and as in New World societies their families lacked security and were broken by sale or the division of estates. Fronto (AdM. Caes. 2.1) knew of the saying attributed to a fugitive messenger that, although he might have to run sixty miles for his master, he would run a hundred in order to free himself from slavery and the punishing conditions it imposed: the Roman slave’s will to escape slavery that the saying implies can readily be granted. A richly detailed modern narrative such as that of Henry Brown can therefore serve a valuable purpose for understanding antiquity. The actions of Roman slaves that the sources portray no more than obliquely become more comprehensible as human actions when read against the narratives of later slaves who underwent comparable experiences and left a record ofthe emotional states and practical decisions surrounding

Them. Brown’s method of escape was unorthodox. But the moment of conscious resolution that instigated it, which in Brown’s case was, as he tells it, almost like an act of religious conversion, was a moment that must surely have been commonly known in the Roman slave experience.

To draw on cross-cultural analogies in this way admittedly runs the risk of eliding important differences between ancient and modern societies — differences of time, place, scale and social values, not to mention the difference between racially and non-racially based slave-owning. There is also a danger of foisting onto the classical past an anachronistic and colonialist image of slavery. A distinction should be drawn, however, between superficial phenomena such as varying rates of manumission from one slave regime to another, and structural constants in the history of slavery at large such as the commodification of chattel slaves, their denial of kin ties, per-sonhood and legal status. It is also important to recognise that comparison is valuable not only for highlighting similarities between historical societies, but also for detecting contrasts that can reveal the specific variations on a general historical theme. The historically unique and culturally specific features of individual slave societies obviously have to be kept in mind, but the advantage of a wider outlook can scarcely be gainsaid. A reductionist essentialism can be avoided even as the universal features of chattel slavery are recognised.

As noted already, slaves like Henry Brown who decided to run away knew that there was somewhere to run to where freedom could be gained and enjoyed. In Brown’s case refuge in the North was, as it happened, jeopardised by the Fugitive Slave Act of 1850, and Brown found himself forced to move elsewhere. Canada could have been an option, but like the Jamaican fugitive Catherine Brown in Cyrus Francis Perkins’ novel, Busha’s Mistress or Catherine the Fugitive (evidently based on first-hand knowledge of slavery in the nineteenth-century Caribbean), Henry Brown judged England a more attractive destination and eventually took up residence there. In contrast, Roman slaves who ran away could never have known the assurance that there was a specific destination that would guarantee them freedom if they reached it. There was no region of the Mediterranean in which slave-owning was outlawed, and nothing is heard of a Roman underground railroad. Roman slaves’ chances of passing as free may have been better than those of New World slaves, and runaways might well have formed communities like those of New World maroons: the career of the Greek slave Drimacus, known from Athenaeus (6.265c—266e), who founded a community of runaways on the island of Chios, suggests so, as does the elder Pliny’s intriguing notice (Natural History 6.172—3) on the Ethiopian city of Adulis, on the western shore of the Red Sea, which, he says, was formed and populated by runaway slaves from Egypt. But in the absence of any extensive intellectual or moral debate about the propriety of slavery, and in the absence of any physical area in which slavery was legally banned, in a very real sense there could be no escape for runaway slaves, no hope of permanent release. The freedom that flight conferred on the slave was always precarious, not legitimately acquired (cf. August. Conf. 3.3). To the extent, therefore, that running away was one of the most prevalent but hazardous forms of slave resistance in the Roman world, the actions of those who undertook it stand as an eloquent statement on the harshness of the institution they wished to flee, and on the motivating force of the will to reject slavery.

Next, theft. Some lines sung by slaves in Brazil capture very well the notion of how in the sharply asymmetrical relationship between master and slave the act of stealing was a matter of relative moral significance and depended on who was stealing from whom and why: ‘The white man says: the black man steals. / The black man steals with good reason. / Mister white man also steals / When he makes us slave.’ To the slave, as an act of revenge for the loss of freedom, theft could have no morally negative connotations and was perfectly justifiable as the result of‘good reason’.

Stealing from the master was one of the many activities that make up what historians of New World slavery commonly refer to as everyday resistance to slavery. They include wilful sabotage of property, deliberate dilatoriness at work, truancy, pretending to be ill, and other forms of deceit and dissimulation. They could be carried out with far less danger and risk to the wellbeing of slaves than revolt or escape, and they allowed slaves effectively to assert their independence. Success depended on slaveowners learning that their interests, especially their material interests, had been damaged in some way, but not necessarily knowing who among their slaves had caused the damage concerned; success depended, that is to say, on slaves outwitting masters in contests of ingenuity. Slaves were well aware of what was at issue. In a section of his autobiography, Frederick Douglass wrote that it was the slaves’ habit when he was a field hand in Maryland to see who could take in the largest crop each day, the winner enjoying a certain claim to masculine distinction. But ultimately the realisation set in that male competitiveness, ‘racing’ as Douglass termed it, was counterproductive to the slaves’ interests, because if through personal rivalry they increased the amounts of the crop collected, the amounts assigned them would be increased and they would all have to work harder; it was obviously best therefore to abandon racing altogether, and to work at a much slower pace if the workload were to be kept down. Such tactics of evasion that countless slaves adopted in the New World slave regimes were met with incessant complaints from their owners of slaves’ congenital laziness, general unreliability and moral reprehensibility.

In the Roman slave system, there was a similar undercurrent of petty, everyday resistance to slavery. The evidence comes once more from sources that represent the attitudes of the slave-owning establishment, and it is highly prejudicial. It takes the form of constant complaints that slaves, of every type and description, were lazy and troublesome, that they were thieves, pilferers, arsonists, embezzlers, dissemblers and truants. Recall, for instance, the reference to slave gladiators as suspect embezzlers in the law on fugitives described earlier (Dig. 11.4.5); or notice Columella’s dyspeptic remarks (Rust. 1.1.7, 1.1.20, 3.10.7, 7.4.2) on the examples of pillage, sabotage, unreliability and evasiveness to be found among farm slaves. The negligence of the vilicus on the farm was indeed proverbial (Val. Max. 4.1 ext. 1). There is Seneca’s complaint (On Anger 3.34.1) that idle slaves were a stock cause of anger in slave-owners, the elder Pliny’s characterisation (HN 33.26) of slaves’ pilfering of food and wine as rapine, Plutarch’s story (Mor. 75pf—760a) of the slave who tried to steal his master’s wine, or Martial’s condemnation (11.54) of his poetic slave Zoilus as fur and fugitivus. Galen even complained (41K) that domestics stole the books he had written. Every slave was a thief, it seemed, the perpetrator offurta parva atque servilia in Apuleius’ pregnant phrase (Met. 4.8).

Evidently enough the representation of servile behaviour in statements like these is stigmatic, and not at all likely, to judge from Frederick Douglass’ example, to match slaves’ own views of their actions. The complaints were grounded on a real base of slave activity: many slaves over time did deceive and steal from their owners and cause all kinds of damage to their property. But it was not because of irresponsibility. The similarities between the actions of Roman slaves that underlie the querulous way their owners wrote about them and the actions taken by slaves in other slave regimes is so striking that there can be no doubt that they added up to small-scale resistance to Roman slavery on a major scale. Everyday resistance was an endemic feature of Roman slavery that exposed slaves to minimal danger while opening up ways to express frustration, anger, revenge and other emotions that could not be expressed openly.

Two items are of special note. First, the elder Pliny’s record (HN 14.78) of the Coan custom of adding seawater to a Greek wine called ‘Bios’ that was used for medicinal purposes, a practice whose origin Pliny attributed to a deceitful slave who had once put seawater into the wine he was making as a ruse to yield the quota of wine he had been ordered to produce. This is a precise parallel to the kinds of deceptive practices abundantly documented in the slave societies of the New World. It does not matter whether Pliny’s explanation is true (and presumably it was a Greek slave who was the culprit here): its assumed plausibility is the significant point. The explanation could stand for Pliny and his Roman audience because Roman slave-owners expected trickery from their slaves and habitually characterised what they saw, moralistically, as poor work performance. The ‘failing’, however, is better understood not in terms of an absolute morality, but of a standard of comportment slaves created as a response to slavery. It was the result not of weakness, but of decisions consciously taken to vex, annoy and defy slave-owners, to lighten workloads, to protest against servitude. It happened all the time.

Secondly, Columella’s diatribe (Rust. 3.10.6—7), when giving instructions on how to take cuttings from a vine, against what he saw as the contemporary practice of giving the task to a useless slave who scarcely had the knowledge or was strong enough to do the job properly. Even if he had a modicum of knowledge, Columella said, the slave pretended otherwise because he was too physically weak to do the work; his sole concern was to finish the task, to meet the allocation assigned by the vilicus, which meant he was neither careful nor conscientious in doing what he had to do. Columella was writing, in the age of Nero, from personal experience of farm management, and his remarks have real credibility. Again, however, the question is not whether what he says was literally true, so much as how he construed the behaviour he had observed. The one-sided, moralistic stereotype of the conniving, irresponsible slave made sense to the slaveowning author and his audience. But the action, or inaction, underlying the stereotype is again better and preferably understood as the result of a deliberate choice by slaves not to perform as instructed, in order to save themselves unnecessary labour and to withstand their owners (or their owners’ surrogates) at the same time. Slaves, after all, must often have had little or no personal stake or interest in the work they did. Seneca said (Ira 3.29.1) that newly enslaved prisoners of war, mindful of the freedom they had recently lost and hardly able to adjust to their sudden reversal of fortune, refused to work at all.

Frustratingly little can be seen of slaves making decisions about their work performance or indeed of their general comportment towards those they served. But there are indications that suggest something of what was involved. Seneca (again) once rhetorically asked (Ira 3.24.2), ‘What right have I to make my slave atone with beatings and manacles for too loud a reply, too rebellious a look, a muttering of something I do not quite hear?’ Implying that slaves should not be punished unjustifiably, from pique or irritation, the question is superficially laudable, the mark of a man devoted to self-improvement through the management of temper. But imagine the quandary of the philosopher’s slave (and Seneca was a slave-owner): how, when performing a service about which he normally had no choice, did the slave calculate whether the pitch of his voice or a facial expression would please the owner or drive him into a rage? In every aspect ofcomportment, the slave had to anticipate the owner’s reaction, and an error of judgement could bring severe and horrible retribution. Seneca himself came close to seeing the dilemma. But exploring the psychology of the relationship between slave and master, as the slave understood it, had no appeal to him, or anyone else. If Roman slavery is to be fully understood, however, the balance has to be redressed and the acknowledgement made that slaves always had to contend with the pressure of anticipating their owners’ reactions to the most ordinary forms of human intercourse. It cannot have helped when they heard the lament, as it appears in a deliciously cynical poem from Martial (9.92), that the obligations which flowed from the possession of wealth and social standing made slave-owners’ lives many times more burdensome than the lives of slaves. Slaves were simply expected to know their place, and that meant among other things imposing selfrestrictions on their speech. Anything that smacked of insulting frankness towards a slave-owner was out of the question. But if, against expectation, a slave were suddenly set free, a dramatic contrast could occur, because language previously impossible to use immediately burst forth as restraint disappeared (Livy 39.26.8). Meantime it was politic certainly for slaves to avoid gossip: in the household of Augustine’s mother, his father had some slaves flogged because they had caused a dispute between mother and mother-in-law through loose talk, and he promised more of the same if the gossip continued (Conf. 9.9). And yet, incongruously, some slaves were encouraged to be impertinent to their owners and their owners’ guests, as a source of amusement (Sen. Constant. 11.3).

A story from Apuleius’ Metamorphoses (8.31) suggests how the slave in real life might have experienced pressure. A desperate cook prepares to hang himself and mournfully bids farewell to his little son when a dog makes off with a choice piece of meat the cook is supposed to prepare for his master’s dinner. Dejection and terror arise from the prospect of the master’s angry response to his loss of a special meal, but happily the cook is saved from an untimely death by his wife’s quick thinking: a substitute piece of meat can be found if a conveniently available ass (the ass of course who tells Apuleius’ story) is slaughtered. A fiction, quite clearly, and no compelling reason therefore to believe that a slave cook was ever really driven to the panicky verge of self-destruction like this. Except that Apuleius’ story has much in common with a medical case-history recorded by Galen (De praecognitione 6.11—13), in which a wealthy man’s steward has fallen into a state of agonised distress because his owner was due to review with him the financial accounts for which the steward was responsible. Preparing for his owner’s arrival, the slave, an older man with an evidently strong sense of responsibility, had discovered that a sum of money was missing. He fell into a state of debilitating anxiety as a result, and grew worse as he anticipated the coming day of reckoning. Recourse to suicide is not mentioned, but the state of anxiety into which Apuleius’ cook and Galen’s steward fell is essentially the same, and there is every reason to think that the steward’s experience was common among the Roman slave population at large, especially among those whose occupations brought them into direct personal contact with their owners. (Galen was able of course to cure the steward, which is the only reason why the incident is known.) This is once more the world of Plutarch’s anecdote about the slave of Pupius Piso, though a world where humour has given way to stern realism. Galen’s slave had to face what Apuleius calls erilis.. . comminatio (Met. 9.19), the threat of both psychological and physical force that the slave-owner embodied and that was part of every slave’s life. Coping with this threat was a prime cause of slave resistance.

Suicide was the ultimate means of resisting slavery, the one incontestable way to deprive slave-owners of their power and property. It is well documented in the wider, global history of slavery. In the early eighteenth century, an Italian Jesuit advising sugar planters on slave management in Brazil admitted that suicide was a foolproof way for recaptured runaways to avoid punishment, and an American traveller of the mid-nineteenth century observed that in Rio de Janeiro reports of slave suicides were regularly issued by the police, but in disproportionate numbers:

Those who plunge into the Bay and float ashore come under the cognizance of the authorities. Of such as sink and never rise, and all that pass out to sea, or are devoured by sharks before they reach it, no account is or can be kept, nor yet of those who destroy themselves in the secret places of the city or dark recesses of the neighboring forests. Many are advertised as runaways who have reached the spirit land. Suicides, it is said, have greatly increased during the last three years.

Conversely, and against expectation, an Austrian merchant travelling to the Sudan in the mid-nineteenth century on a Turkish slave-raiding expedition commented on the unusual absence of suicide among the captives taken: ‘They saw the impossibility of offering resistance, and became reconciled to their hard lot.’ In this context Seneca’s outburst (Ira 3.4.4), ‘How many slaves a master’s anger has driven to flight, how many to death!’, should not be dismissed as rhetoric but taken as an accurate reflection — and again almost a depressingly sympathetic recognition on Seneca’s part — of the extremes to which Roman slaves were sometimes driven by slavery. (Glimpses of what could physically happen are occasionally allowed by Roman jurists when they speak of situations in which slaves threw themselves down from a height.) Seneca (Ep. 4.4; cf. 70.19—26; Marc. 20.2) imagines suicide as a final means to escape a slave-owner’s distemper or to avoid recapture after running away: a self-inflicted death lightened the captive’s chains and set the slave free whether the master liked it or not.

Prisoners of war in the Roman world seem regularly, and grimly, to have preferred death to enslavement. In Augustine’s view (De civ. D. 18.12), the will to survive was a fundamental feature of human nature, so in defeat submission to the enemy was the choice to recommend over death. But Cassius Dio (72.14.2) reports that under Caracalla a group of captive German women informed the emperor that they would rather be executed than sold as slaves, and when Caracalla did in fact sell them the women committed suicide en masse, some killing their children as well. (Compare the striking remark of a nineteenth-century Sitka captive woman: ‘It is just as well to die as to be enslaved’, and note Plut. Mor. 242d of a Spartan woman.) The report has parallels in other war accounts and indicates at the very least the plausibility of suicide by prisoners of war who were suddenly thrust into slavery. Scenes of Dacian prisoners on the Column of Trajan killing themselves point in the same direction: sculptors, or their patrons, knew what a Roman audience would find credible.

Roman slavery was a complex institution. The slave population was heterogeneous and its members provided labour and services of many different kinds. The range of slave statuses was enormous. At one extreme there were those, under the Principate, belonging to the emperor who were rich, influential and far better situated materially than much of the free population. Galen (50K) noted the paradoxical circumstance, which he ascribed to the operations of Fortune, that even a slave could be much wealthier than the respectably freeborn. At the opposite extreme there were those who provided heavy labour for agriculture and mining, whose living conditions were by any standards miserable and whose prospects of longevity were minimal. In between, status was controlled by many factors: the slave’s location in an urban or rural environment, the relative standing in the household hierarchy, the education or skill level required for the work done or service provided, the degree of proximity to the owner, sex, age and reputation. The results could be bizarre. Drusillanus Rotundus, a slave who held an imperial administrative position in Spain under Claudius, was remembered long after his death for his magnificent silver plate, as the elder Pliny records (HN33.145; cf. 34.160). In a burial collegium such as that of Diana and Antinous at Lanuvium (ILS 7212), slave members reclined at dinner in the company of the free (and freed), as if there were no distinction between them — and provision was made for a special celebration to mark a slave’s manumission: the slave was required to provide an amphora of good wine. Over time many Roman slaves who were set free prospered, having apparently suffered no ill consequences from their experiences as slaves, and as their epitaphs and funerary monuments above all show (in the thousands), they integrated themselves into Roman society with ease and often in turn became slave-owners themselves. The freedman C. Furius Chresimus, known also from Pliny (HN 18.41-3), stood out for his exceptional success as an independent farmer and the good care he showed the slaves he himself possessed. The phenomenon of the ex-slave who himself became a slave-owner is the ultimate symbol of the structural centrality of slavery in Roman society, which as far as can be told never aroused opposition even among those who were its victims.

To concentrate on slave resistance in isolation, therefore, is to risk minimising or blurring Roman slavery’s complexity and presenting an overly schematic impression of historical reality. Rome’s slaves cannot all be lumped together, regarded as a cohesive, homogeneous class, imagined as always occupied in acts of resistance. Such notions are far too simplistic. At any moment there must have been thousands of slaves who unquestioningly obeyed their owners as they were expected to, or in whom consent, no matter how grudging, was elicited by the generous application of munera and the forceful infliction of poenae. Consent and coercion went hand in hand. The evidence leaves no doubt, however, that there was a deeply textured pattern of resistance in the fabric of relations between Roman slaves and slave-owners, that consent, coercion and resistance were threads woven inextricably together all of a piece, and that the modes of resistance can now be arranged in a typology that runs from grand-scale revolt of the Spartacan kind to the simple time-wasting and shirking of which owners complained so much. Whether there developed as a result a slave subculture of the sort characteristic of New World slave societies it is difficult to tell, but among the members of large domestic households in the city of Rome, or among the members of contiguous rural familiae, it may not have been impossible. Race will not have bound people together as it did in later history. But there remains the tantalising hint in Phaedrus (3 Prol. 33—7), writing on the origin of the fable, of a secretive means of communication among slaves that assumes a special, and timeless, bond and understanding among them, no matter what the differences and distinctions.

Did modes of resistance change over time? After Spartacus there were no major slave revolts at Rome, which implies that the three major wars of the late Republican period were aberrant episodes, and that the idea of premeditated large-scale revolt was recognised as the least hopeful means of improving slaves’ lives. Spontaneous violent actions arising from immediate acts of provocation could never of course be ruled out, and particular incidents captured the fearful imagination of historians and social commentators. Whether verbally or physically, the slave might lash out at a slave-owner at any time (cf. August. De civ. D. 21.11). Slave suicide was also to be anticipated at any moment and was perhaps frequent, but for obvious reasons the rate ofself-destruction among slaves cannot be known. Flight, however, was seemingly common at all times if the distribution of surviving evidence is any guide at all to slave behaviour. It could be the case on a conventional view of Roman slave recruitment that running away was more prevalent under the middle and late Republic, when wars of conquest brought masses of first-generation slaves to Rome, than under the

Principate, when natural reproduction was an important supplier of new slaves, the reason being that first-generation foreign slaves were likely to try to return to the regions of the Mediterranean from which they had been torn, and, it could be argued, the political upheavals of the revolutionary age created special opportunities to do so. It is doubtful, however, that the pattern of slave recruitment over time can be reduced to a simple formula of importation on the one hand and natural reproduction on the other, and unquestionably there is an overwhelming amount of material on running away from the imperial age. Flight should be regarded accordingly as a persistent, largely unvarying form of slave resistance, though perhaps greater in volume at moments of political crisis or uncertainty, as suggested, for example, by Dionysius of Halicarnassus (5.26, 6.50.3, 7.1.2). The same is true of the activities that make up the category of everyday resistance. The guise of truculence under which petty resistance is coded in the sources is evident as early as the comedies of Plautus and never disappears from the literary and legal record: Plautus’ servus callidus is Martial’s fallax ancilla (11.49) and Ulpian’s servus onerosus (Dig. 17.1.8.4). Once more, everyday resistance occurred every day.

Did resistance have an economic impact? In eighteenth-century Africa violent resistance to enslavement at the points where slaves were recruited reduced the numbers of Africans enslaved by as many as half a million and substantially increased the costs of the enslavement process. The impact of Roman slave resistance cannot be measured in a comparable way, but even on a rudimentary estimate there can be little doubt that its effects were felt. Roman slave-owners were sensitive to the costs of maintaining their slaves, who as property always represented an investment. Their views emerge as another set of highly charged complaints, this time complaints about the burdens ofhaving to feed and clothe slaves (Sen., On Tranquillity 8.8; Ep. 17.3, 96.2), about the time that had to be spent in managing them (with punishments, for instance [Sen., On the Shortness of Life 3.2]), and about their inefficiency as a labour force (Plin. HN 18.21). Because slavery was maintained for such an enormously long time at Rome, it is unlikely that slave labour in the prime areas of agriculture and mining really was inefficient, in the sense that slave-owners repeatedly failed to draw enough profits from their operations for their own purposes; and the social capital that accrued to elite owners from maintaining large entourages ofdomestic slaves was in any case incalculable. The complaints nonetheless reflect the real expenses that slave-owning involved, the most serious of which was the loss caused by the death of a slave, which, as even Martial recognised (6.33), was a true catastrophe.

Another medical history from Galen (632—4K) offers a simple illustration of the practical issues concerned. A slave was injured in a wrestling-school as a result of a blow to his chest. There was damage to his sternum.

After four months the slave had not recovered. A surgeon performed an operation but infection set in. The slave’s owner brought in other doctors, one of whom was Galen. He successfully removed the infected part of the sternum, though before operating he made it clear that he could not guarantee success, and at one point the slave’s life seemed to hang in the balance. Eventually, however, the man made a full recovery and the owner’s investment was saved. But during the period of injury and treatment, the outlay for maintenance and medical personnel was not matched by any obvious return. So there was a burden on the owner, and financial loss in a situation like this was inevitable.

Given the structural character of petty resistance within the Roman slavery system, slave-owners must always have had to reckon with a steady trickle of financial loss, no matter whether individual slaves consciously set out to inflict damage on their owners — whether from anger, hatred, desire for revenge, or simple mean-spiritedness — or to make their own material lives easier by supplementing their rations of food and clothing. Slaves themselves must often presumably have been unable to separate one strand from another in an intertwined bundle of motives. Those costs, moreover, would necessarily be raised when slaves ran away: arranging to recover fugitives with slave-catchers took a toll, and the services of runaways were obviously lost, at least temporarily, while they were at large. If not recovered, the slave property was lost altogether, and this was emphatically true when slaves committed suicide. The costs of damage to other forms of property have also to be factored in. Widespread negligence and shirking or malingering must have increased the costs of agricultural production; and when slaves burned down a suburban villa, as they did with a property belonging to M. Aemilius Scaurus, the praetor of 56 bc (Plin. HN 36.15), the impact had to be felt even by a rich senator. The provocative activities that hostile sources attribute to slaves — stealing grain from the threshing-floor, falsifying account books, feigning illness — had the permanent effect of eating away their masters’ wealth.

From a modern perspective, every act that defied the authority of a slave-owner could be construed as an implicit rejection of slavery. But as far as can be told, the challenges to enslavement that Roman slaves mounted were not challenges to an institution so much as protests against the enslavement of individuals or groups. Resistance was directed towards mitigating the hardships of slavery, concerned with gaining respite and release and with damaging the interests of slave-owners, but not with changing the structure ofsociety. Again, the willingness offormer slaves to become slave-owners themselves is both noteworthy and remarkable, and it cannot be emphasised too strongly that the idea of abolishing slavery was a development unique to a much later age. It was not impossible at Rome of course to imagine a world without slavery. At the annual celebration of the Saturnalia in December, everyone was reminded of that long ago and far away mythical Golden Age when all men had been equal and slavery did not exist. When slaves enjoyed the holiday feasts their masters provided, and sometimes even served, they had to be aware, perhaps disturbingly aware — how could they not be? — of the social inversion suddenly there in their midst, and of the possibility of a different life it revealed. Yet the briefly shining moment of the Saturnalia never seems to have inspired a true ideological challenge to slavery, an intellectual imperative on which to build a movement implicating a sizeable proportion of the slave population with militant leaders and coherent organisation; for most slaves most of the time, the demands of work and survival made the luxury of thinking in such terms impossible.