13
Decline and Legacy

Finally, when did ancient slavery end? At the end of Roman Antiquity and at the time of the invasions, reply, with one voice, the Marxists, all remarkably faithful on this point to the letter of the writings of Marx. In the eleventh century, according to Georges Duby. At some indeterminate date between the fifth and the eleventh centuries, say, lastly, with some embarrassment (if they say anything at all), most others.

Pierre Bonnassie, “The survival and extinction of the slave system in the early medieval West (fourth to eleventh centuries)” (1991), 14

The competitive system [capitalism] is a system of antagonism and war; ours of peace and fraternity. The first is the system of free society; the other that of slave society. The Greek, the Roman, Judaistic, Egyptian, and all ancient polities were founded on our theory.

George Fitzhugh, Sociology for the South, or The Failure of Free Society (1854), 26

Introduction

Last chapter’s explorations of philosophical and legal justifications, criticisms, and ameliorations of slavery concluded our treatment of slavery in the classical world. This postscript answers two sets of questions about what came afterwards. How, when, and why did classical slavery end? What were and what continue to be its legacies in the modern world? Looking for answers to these questions will highlight continuities as well as differences between ancient and New World slavery.

It is not possible to put a date on the end of classical slavery. None of the successor states to the Roman Empire was a slave society; the importance of agricultural slavery in particular diminished drastically in the centuries after the Empire’s dissolution. Yet all these societies contained some slaves centuries after the fall of the western Roman Empire. Even when, much later in the Middle Ages, slavery died out entirely in some central areas of Europe, it survived in others. For example, wars and hostility between Islamic and Christian states provided the context for slavery to thrive around the medieval Mediterranean.

These conflicts added the element of religious difference to the conception and the institution of slavery, something rarely seen in the classical world. The later racist justification of the enslavement of black Africans was also something new. Greek ethnocentrism – as in Aristotle’s insistence that non-Greeks are natural slaves – provides an analogy, but modern racism was more comprehensive, systematic, and even pseudo- scientific (Isaac 2004, 1–6). And the importance that racist ideology placed on skin color made it much harder for ex-slaves to assimilate or climb socially. Despite these differences, that a continuous history of slavery stretched from antiquity to the New World is one factor that explains why the institution is so easily recognizable from archaic Greece through the Roman Empire to the New World.

This thread of historical continuity and the prestige of classical civilization were preconditions for the powerful legacy of ancient slavery in the New World. Spanish intellectuals and Dominican monks debated the enslavement of Native Americans in Aristotelian terms. Apologists for slavery in the antebellum South sought antecedents in classical culture for their racist and paternalistic justifications of slavery. The legal systems of many European states were based on Roman law. Some of these states enacted law codes governing slavery just as if New World slavery were like Roman slavery and were not based on race.

The Decline of Classical Slavery

Starting in the late third century CE, the eastern and western halves of the Roman Empire became increasingly independent of each other. At times, they even had different emperors. In the late fourth century CE, the history of the two sections diverged dramatically. In 378 CE, the Goths defeated a Roman army at the battle of Hadrianopolis in Thrace. The Empire did not bounce back as it had after similar disasters and invasions in the past. Instead the Goths turned west, ravaging and plundering, and eventually sacked Rome itself in 410 CE. By this time, other Germanic tribes had also invaded the Empire. Although some historians emphasize “transformation” rather than Edward Gibbon’s “Decline and Fall,” the fifth century CE was disastrous (Gibbon 1914, orig. 1776–1789). By its end, kingdoms of Ostrogoths, Visigoths, Franks, Vandals, Burgundians, Anglo-Saxons, and others had replaced Roman government in the West. The Greek-speaking eastern half of the Empire – which historians call the Byzantine Empire even though its people still referred to themselves as Romans – fared better. In the sixth century CE, the Byzantine Empire even succeeded in retaking parts of Italy, Spain, and North Africa. The Byzantines eventually had to abandon those territories, and they lost Syria and Egypt to Muslim invaders in the mid-seventh century CE. For the rest of its long history, until 1452 CE, the Byzantine Empire comprised territories mainly in the Balkan Peninsula and modern Turkey.

Map shows Europe and Mediterranean in 530 BCE of Sasanian Empire where Ostrogoths is found in central region, Visigoths in eastern region, Vandals in southeast region, et cetera.

The centuries from 400 to 700 CE also saw a fundamental decline in the importance of slavery. The dozen or more smaller states that ruled what had once been the western Roman Empire were societies with slaves but not slave societies (Harper 2011, 509). The Byzantine Empire too became less dependent on slaves; in particular, their role in agriculture faded (Harper 2010a; contra Rotman 2009, 107–116). In the eighth century, slaves could be found throughout the area of the former Roman Empire, but they were mainly domestic slaves serving the elite, a pattern common in many societies with slaves. It is difficult to be any more precise than this, both because of the probable complexity of the process and because of the scarcity of good evidence.

There may not even be a single answer for when slavery declined. In some contexts, one can speak about slavery in the Roman Empire as a whole. But, especially when the topic is something like the importance of slavery, it is important to be more specific. In Egypt? In Britain? In the city of Rome itself or in the countryside of southern Italy? After the fall of the western Empire, geographic variation became even more pronounced. The practice and extent of slavery in rural North Africa under the Vandals and in Syria under the Byzantine Empire need not correlate at all. There is also no reason to assume that the proportion of slaves politely and steadily declined from 400 to 700 CE: there may have been generations when the number of slaves remained the same, some when it declined precipitously, and some when it increased – for example, in the early fifth century as we’ll see.

Unfortunately, historians do not have the kind of information they would need to trace this complex geography and chronology. This should not be too surprising. Recall the difficulty of quantifying slavery even during the height of the Empire (see Chapter 3), a relatively well-known period. For several centuries after the fall of the western Empire, such estimates are impossible. Historians today tend to avoid the negative connotations of the term “Dark Age,” but it is undeniable that our evidence is thin for many times and places. On balance, I favor the necessarily imprecise position that the sixth century CE was decisive for the decline of slavery in many places. But that argument makes the most sense in terms of another issue, which we need to tackle first: the relationship between the dissolution of the Roman state and the decay of slavery.

These two huge and roughly contemporary changes must have been intertwined in many ways. But how historians conceive of the relationship and what aspects they emphasize vary greatly. Here I’ll sketch out just two contrasting views. One school of thought holds that a crisis in the system of slavery left the Empire vulnerable. The other side emphasizes that, after the destruction of the Empire’s political and economic structures in the West, large-scale slavery was no longer viable there. This dichotomy is a simplification of many more complex and nuanced views, but it gets to the heart of an important historical controversy.

The most prominent adherents of the first school are Marxist historians committed to the primacy of economic factors. They discern a major shift from slavery to serfdom: a “feudal mode of production” superseded the classical “slave mode of production” (e.g., Anderson 1974, 18–19). They argue that this social and economic transition had begun by the second century CE. On this view, slavery became less and less profitable as Roman conquests slowed and stopped: the supply of new slaves thinned and their price increased (compare Chapter 3). Even the encouragement of slave reproduction – one response to a decreased supply – involved costs and concessions for masters. These problems in the economic base of Roman society eventually led to the “crisis of the third century,” a period of civil wars, invasions, and economic decline (235–284 CE). The Roman Empire recovered, but it was no longer a slave society and was fragile to boot.

In this model, once widely accepted, an agrarian system based on a bound peasantry, essentially serfs, had largely replaced agricultural slavery by the Empire’s end. Marxist historians find support in fourth-century references to coloni.¹ The Roman state assigned some peasants and farm workers, coloni, to the owner of the land they worked and deprived them of their right to move. This policy was probably originally intended just to assure the effective collection of taxes, but by 332 CE a law refers to coloni who were “under the legal power” of one landowner or another; it threatens runaway coloni with chains and enslavement (Theodosian Code 5.17.1). All of this suggests that some peasants were losing their independence and becoming more like medieval serfs than Roman citizens. And, from the opposite direction, some slaves and ex-slaves were “hutted up”: instead of closely supervising their slaves, masters settled them on their land and exacted a rent, often in kind and labor. Thus, slaves too were becoming serf-like. Marxists see all these developments as part of an overall transition to serfdom well before the Empire fell.

Marxist historians connect this shift in the mode of production with the vulnerability of the Roman Empire that led to its fall. The treatment of G. E. M. de Ste. Croix provides an example.² Rather than depending on slaves, the elite turned more and more intensely to the exploitation of the peasantry, who constituted a large majority of the population and were the traditional source of Roman military manpower. Of course, the peasants resented this. Signs of their discontent, resistance, and even revolts dot the history of the late Empire. The peasants had little stake in resisting the barbarian invasions and even joined them on occasion. In sum, de Ste. Croix argues that a problem with the economic base, slavery, led to political consequences, the most important of which was the conquest of much of the Empire by foreign tribes.

This whole picture – which I have admittedly simplified – is liable to several lines of criticism (Wickham 2005, 259–265). First, the Marxist model has just one trend in ancient slavery, downward, and one trend in serfdom, upward. Most historians today prefer a more complicated timeline. Evidence attests to large-scale slavery in the fourth and fifth centuries CE: for example, a fourth-century tax record lists over 150 slaves on a single farm on the Greek island of Thera; other sources refer to rich men and women like Melania the Younger possessing thousands of slaves (Harper 2011, 167–168, 192–195). Some historians infer that slavery never declined and that reproduction entirely filled the gap when the pace of conquest slowed. But the evidence does not allow us to rule out more complicated scenarios: slavery may have declined in the first and second centuries CE when Rome stopped expanding. The late Empire, however, witnessed a growing concentration of wealth as well as increased warfare and lawlessness. These two circumstances could provide both a cheap supply of and continued demand for slaves. For example, from a letter of Augustine of Hippo, historians learn that slave raids had become endemic in Roman Africa in the early fifth century and that the people enslaved were being sold in other parts of the empire (Augustine, Letter 10 [Divjak] in Eno 1989; Harper 2011, 92–94).

However common bound coloni may have been in the late Empire, medieval serfdom did not directly replace slavery. For example, Chris Wickham presents a radically different picture of early medieval social structure. He argues for a period marked in many places by a “peasant mode of production”: relatively unburdened peasants dominated the countryside in much of what had been the western Empire and the aristocracy endured a period of reduced power and wealth.³ This is a controversial thesis, but medieval historians agree that the large-scale reduction of peasants to serfdom was a much later development. The popular notion that serfs replaced slaves is false (e.g., Bonnassie 1991, 58; cf. Davies 1996).

Finally, the organization of labor in the countryside was almost always mixed, including slaves, tenant farmers, sharecroppers, and bound peasants of various sorts. Even at its peak, slave labor only dominated agricultural production in some areas of the Roman Empire (e.g., Lenski 2017). And few medieval historians today believe that serfs ever constituted the majority of the rural population in western Europe. Even in the high Middle Ages (1000–1300 CE), most peasants were not subject to the full obligations of serfdom (Freedman 1991, 1–17, 214). A more complicated rural social structure also undermines the idea that slavery varies inversely with serf-like status. In the late fourth century CE, for example, there may have been many slaves and many serf-like coloni; two centuries after the sack of Rome, there may have been few of either.

In addition to these criticisms, a second paradigm reverses Marxist economic determinism. On this view, the political and military catastrophes of the fifth century led to the decline of slavery. The economic complexity that had made the large-scale use of slave labor profitable did not survive the fall of the western Roman state and the associated wars and chaos (Harper 2011, 497–506). The dissolution of the Roman Empire greatly reduced trade. Cities too had shrunk: the population of Rome in the seventh century CE was only about one-tenth its former size. Trade and cities had provided the markets for produce that made agricultural slavery profitable. The overall movement away from market agriculture made slaves, who require closer supervision, a less attractive labor force than peasants or serfs, who can provide labor and rents without requiring much attention.

This model implies a later date for slavery’s decline: after the political catastrophes rather than before them. Kyle Harper has recently argued that the sixth century CE was probably the pivotal era in many regions. One of his arguments rests on the price differential between male and female slaves (Harper 2010b; cf. Saller 2003). In the fourth century CE, male slaves typically commanded higher prices than female slaves. This is consistent with Harper’s view that slavery was still thriving in the late Empire, for male slaves generally outnumber females in slave societies and command higher prices – despite the reproductive capacity of female slaves. In contrast, in the vast majority of societies with slaves, females outnumber males and are priced higher. They are valued in terms of their contribution to their master’s lifestyle and prestige not only in terms of the products of their labor. By the eighth century, female slaves cost more, the pattern for the rest of the medieval era and up to the reinstitution of plantation slavery. Harper concludes that the transition from the one big slave society of the Roman Empire to the many societies with slaves of the medieval world was essentially complete by 700 CE. And in the medieval period, slaves were more expensive, suggesting a luxury item rather than a productive investment.

The atrophy of the slave trade provides additional evidence for a decline of slavery after the fall of the western Empire. In the late sixth century CE, Gregory the Great, then a bishop in Rome, still possessed slaves. But on two occasions he had to request somebody to buy slaves elsewhere and to bring them to him (Harper 2011, 498). Such a scanty slave trade is almost inconceivable in a slave society such as Rome in the late Republic or early Empire. The direction of the early medieval slave trade was also significant: slaves went east to the Byzantine Empire – and later to the Muslim world (Harper 2010b, 237). The East was richer and western kingdoms had little other than slaves to trade anymore, another dramatic reversal from the centuries when Rome was the center of a vast slave trade.

The picture in the Byzantine Empire was different and closer to the Marxist model in some ways. The eastern Mediterranean remained prosperous, with large urban markets still thriving in late antiquity. Nevertheless, slavery and especially agricultural slavery withered (Harper 2010a, 949; 2011, 505). Slavery may have declined because other sources of labor became easier to obtain. Wage labor had always been more common in the East, and, backed by a strong state, rich Byzantine landowners subjected the peasantry more and more tightly (Banaji 2009, 78–86). The terms of peasant oppression varied, but many peasants ended up in a position little different from serfdom. Here peasant vulnerability to exploitation may have contributed to the decline of slavery – a reversal of the dynamic that linked the growth of citizen rights and of slavery in classical Greece and the Roman Republic (see Chapter 5).

In some places, slavery merely declined, but in large areas of Europe it almost disappeared by the high Middle Ages. Marc Bloch makes a pointed contrast:

Under Charlemagne [reigned 768–814 CE] and under Philip Augustus [reigned 1180–1223 CE], there lived and labored on the soil of France men who in both reigns were designated in Latin as servi, men who in both reigns were considered to be deprived of that juridical characteristic known as freedom, but what a contrast between their actual conditions. (Bloch 1975, 34)

In the time of Charlemagne, France was a society with slaves. There may have been far fewer of them than in the fourth century, but the Latin word servus (plural: servi) still denoted a chattel slave. By the end of the twelfth century, slavery was so marginal an institution and a certain type of un-free peasants was so common that servus was applied to the latter. Historians translate this use of servus as “serf,” to distinguish these bound peasants from the chattel slaves for whom servus was earlier used. An entirely new word denoted chattel slaves: many medieval slaves came from the Slavic peoples on the eastern frontier of Europe and this led to the use of their ethnic name to designate slaves, a usage that began in the ninth century (Oxford English Dictionary, “Slave.”). Hence the English word slave – and variants of slave in other modern European languages.

Slavery may have almost died out in the heartland of Europe, but it continued on its periphery. Most obviously, the wars between Christians and Muslims in Spain and across the Mediterranean gave medieval slavery a religious aspect. In theory, Christians only enslaved Muslims and vice versa, but the reality was more complicated. Medieval Italian city-states, such as Genoa and Venice, played a large role in this trade, which in turn contributed to their wealth and power. For the Mediterranean, now divided along religious lines, continued to provide a conduit for an active slave trade as it had in classical times. And, as we mentioned, the eastern land frontier of Europe was a source of Slavic slaves for centuries; these often ended up in Muslim territory. I can do no more here than mention these places where slavery persisted, but one point is important. Although slaves became rare in the core of Europe, slavery persisted. Areas such as Spain and Portugal had continuous experience of slavery from the fall of Rome to the establishment of Indian and then African slavery in their colonies. The religious conflict that provided the basis of much medieval slavery and the racist basis of New World slavery were essentially post-classical developments, but in other ways New World slave systems were direct, if distant, descendants of classical slavery.

The Legacy of Ancient Slavery

Slavery was intrinsic to daily life in the classical world; it was also apparent in many aspects of the enduring legacy of Greek and Roman civilization. So, wherever admiration for classical culture thrived, slavery was a familiar institution, often from the books people read, if not in their daily lives. In some eras and places, modern slavery even gained a specious respectability through its association with classical Greece and Rome. This connection was often vague and implicit, but before the American Civil War, advocates for the South and for slavery regularly pointed out that classical civilization was based on slavery. In their opinion, the same institution of slavery lay behind the accomplishments of Greece and Rome and behind their own culture (DuBois 2009, 72–74). In some cases, it was not classical culture in general, but a particular legacy of Greek and Roman slavery that had the greatest impact on New World slavery. Two influential aspects of ancient slavery were the Aristotelian theory of natural slavery and the Roman law of slavery.

Aristotle in the New World

As we saw, Aristotle’s theory of natural slavery drew upon prejudices against slaves and against foreigners. Aristotelian philosophy in general came to dominate medieval thought in western Europe after his works were translated into Latin in the late eleventh century and twelfth century. For example, Thomas Aquinas, the most important philosopher and theologian of the medieval period, reconciled Aristotelian philosophy with Christian dogma in the thirteenth century. His Aristotelian views have remained central to Catholic theology through to the present day; they were particularly dominant among the Dominican order. Aristotle’s high status among Catholics provided the background for a fascinating and consequential debate in sixteenth-century Spain about the treatment of those New World peoples whom the Spanish had conquered or planned to conquer.

This debate featured a Dominican Friar, Bartholome de Las Casas, on one side, and Juan Gines de Sepúlveda, a famous intellectual and translator of Aristotle’s Politics on the other. There were several issues at play including the American Indians’ supposed sins against nature and the need to convert them to Christianity and to protect the weak among them from oppression by the strong.4 The main debate, however, revolved around whether they were natural slaves according to Aristotle’s ancient definition. Not only would this determine their treatment after the Spanish conquest, but Aristotle had also identified the reduction of natural slaves to slavery as legitimate grounds of war (Politics 7 1333b39–1334a3). If the Indians were natural slaves, Spain need not worry about the justice of any wars they started against them – at least as far as Aristotelian theory was concerned.

This debate, turning on the application of a philosophical theory almost two millennia old, had profound and concrete results. The early Spanish conquests in Mexico and South America were relentless and brutal, but the horribly oppressed natives had found sympathetic allies, especially among the Dominicans. For example, when he was bishop in Chiapa in southern Mexico, Las Casas had refused confession to Spaniards holding Indians in servitude. Some of these powerful and wealthy Spaniards had endured for years the state of being unconfessed and thus liable to purgatory or perhaps damnation. They were enraged and retaliated in whatever ways they could. In Peru, the local conquistadors actually killed a viceroy who tried to enforce laws protecting the Indians – and they carried his head around on a string as a display of their defiance (Hanke 1959, 34).

In 1550, Charles V, the Holy Roman Emperor and King of Spain, ordered all Spanish conquests to cease (!) until it was determined whether the enslavement of the Indians was just. The king appointed judges to make this decision. Before them, Sepúlveda argued that the Indians ought to be enslaved, “On account of the rudeness of their natures, which obliged them to serve persons having a more refined nature, such as the Spaniards” (Hanke 1959, 41). Although earlier in his career Las Casas had pictured Aristotle burning in hell, by 1550/1551 he accepted that the debate had to be conducted in terms of Aristotle’s theory. Las Casas had served the church for almost fifty years in the New World and provided copious evidence that the American Indians were not natural slaves. He claimed that they were in many respects superior even to the ancient Greeks and Romans. Although the debate was inconclusive and the judges did not return a decision, the tide was turning Las Casas’ direction. Eventually, the Spanish “basic law of 1573” governing the treatment of Indians included many of the reforms Las Casas had fought for (Hanke 1959, 86–87). This was far from the end of the oppression of Indians, but the “basic law” was superior to earlier rules and encouraged treatment less harsh than before.

Some advocates for slavery in the United States South adopted a version of Aristotle’s theory in the nineteenth century. For example, George Fitzhugh, an outspoken apologist for slavery before the Civil War, explained that when he read Aristotle, he had realized with great pleasure that their views were identical (1857, xxi). He went on to appeal to Aristotle’s authority at least a dozen times in his two books. In one place, Fitzhugh attempted to rebut the Declaration of Independence with a crude paraphrase of the Aristotelian view of slaves:

[Men are not] born entitled to equal rights! It would be far nearer the truth to say, that some were born with saddles on their backs, and others booted and spurred to ride them, – and the riding does them good. They need the reins, the bit and the spur. (Fitzhugh 1854, 179)

In other words, slavery was in the interest both of natural slaves and of masters. In the ancient world, Aristotle was not influential with intellectuals, who tended to prefer some variant of the Stoic attitude. Antebellum racist ideology found his most ambitious claims plausible. Most Southern racists did not, of course, treat non-Greeks as natural slaves, as Aristotle did; rather they justified the enslavement of black Africans on the grounds of their supposed inferiority.

Fitzhugh often appealed to the prestigious, slave-based civilizations of Greece and Rome to highlight what he saw as the failures of the capitalist, free-market Northern states; he sometimes even found common cause with socialism. His main argument was an extreme and eccentric version of paternalism, a way of thinking too common to attribute to classical antecedents. One of his favorite tactics was to contrast the supposed familial affection and mutual care of slaves and masters in the South to the capitalist North, where unabashed competition meant that nobody cared for the workers – as in the passage at the start of the chapter denouncing “free society.” His views occasionally went even beyond racism: he believed that poor white laborers would also benefit from slavery.

In defending slavery, Fitzhugh had to justify the ownership of people, which does not seem the most organic or familial of institutions. He was undismayed and claimed that ownership is the key to affection and benevolent treatment. He repeated a description of an English farm:

The cattle, the horses and the sheep are fat, plentifully fed and warmly housed . . . two freezing, shivering, half-clad boys, who have to work on the Sabbath, are the slaves [metaphorically] to these animals . . . if the boys had belonged to the owner of the farm, they too would have been well-treated, happy and contented. (Fitzhugh 1854, 46–47)

Fitzhugh’s criticisms of the early period of unregulated capitalism were not always wrong. But historians in recent decades have emphasized the extent to which Southern plantations too resembled capitalist enterprises much more than extended families (e.g., Fogel and Engerman 1974). Ownership was obviously no guarantee of good treatment. Fitzhugh’s paternalistic justification of slavery was obviously tendentious, and vulnerable to all sorts of objections and rebuttals, which abolitionists naturally made. His arguments were thus more transparently self-serving than those of ancient paternalists, like Seneca and Pliny, who did not have to anticipate an audience likely to question their professions of fatherly care for their slaves.

The Roman Law of Slavery and Modern Slave Codes

In the sixteenth century, Roman law constituted the basis of most continental legal systems, a circumstance that would persist until the early nineteenth century. Consequently, the Roman law of slavery played a large role in the law codes of Spanish, French, and Portuguese slave societies in Latin America and the Caribbean. In contrast to most of continental Europe, Great Britain retained its common law even as Roman law gained influence after the twelfth century. The influence of Roman law on British law and thus on the United States was much less, though not negligible.

France instituted the first version of the Code Noir, “Black Code,” for Caribbean islands like Saint-Domingue (Haiti) in 1685 and later for Louisiana, still a French possession. Alan Watson argues that the Code Noir incorporated the Roman law of slavery wherever applicable (Watson 1997; contra Palmer 1995/1996). For example, it preserved the Roman legal procedure of noxal surrender: if Marc’s slave ruined Jean’s property, Marc could either pay damages or surrender his slave to Jean. Despite its name and its topic, slavery, the Code Noir was rarely racist. In particular, manumitted slaves became colonial subjects of France with similar rights to what whites possessed. This policy was out of line with prevalent racism, but French jurists followed Roman laws, originally designed to regulate a non-racist and more open system of slavery.

Slave law in the southern United States was less subject to Roman influence. It tended to be more racist and to reflect more precisely the experiences and needs of Southern slaveholders, who dominated the region. How strong a dichotomy we should draw between the institution of slavery in Latin America and the United States is a complex and controversial topic. And the large role that Watson attributes to the Roman antecedents of the Code Noir is also contested (Schwarz 1991 on Watson 1989). That it had some influence is clear.

Conclusion

The legacy of ancient slavery paradoxically includes the image of its most bitter enemy, Spartacus, who remains today a symbol of heroic opposition to oppression. Communist admiration for Spartacus is the most conspicuous and well-known case. Already Marx admired Spartacus, and a group of German communists after World War I called themselves the Spartacists. In the United States, the Spartacist League still publishes the Workers Vanguard, a socialist newspaper. The classic movie Spartacus (1960) was directed by Stanley Kubrick and starred Kirk Douglas. It won four- academy awards, clear evidence of Spartacus’ appeal beyond the far left. The historian Art Eckstein sums up:

The screenplay [of Spartacus] was written by the blacklisted Communist Party writer Dalton Trumbo (his name appeared as a credit for the first time in almost ten years); the pamphlet handed out at the film’s gala premier was red. Spartacus, then, served as a signpost to the rebellious 1960s. But Spartacus the person was also a hero to Voltaire, to Garibaldi, to Ze’ev Jabotinsky, one of the founders of Zionism, and even to Ronald Reagan, who, speaking to the British parliament in 1982, employed Spartacus as a symbol of sacrifice in the struggle of freedom against totalitarianism. (Eckstein 2010, 1)

Indeed, President John Kennedy ignored a hostile demonstration by the American Legion to watch the movie Spartacus (New York Times, February 5, 1961). His action symbolized the end of the 1950s era of persecution of communists, suspected communists, and their sympathizers, most notoriously by Senator Joseph McCarthy. In one of the movie’s last and most famous scenes, the defeated slaves are offered their lives in return for turning in Spartacus. None of the slaves points out Spartacus; nobody “names names” as people had before the House Un-American Activities Committee in the 1950s. Rather, one after another the slaves claim to be him. They say “I’m Spartacus” and go bravely to their deaths.

This is, of course, the Hollywood version, not a happy ending but a heroic one. Actual events and people are messier. The aim of this book has been an intellectual one: to understand, as far as possible, classical slavery in all its messiness and complexity. But this does not mean that history has no ethical side or that I have never ventured judgments of persons, practices, policies, and ideas. Nonetheless, I do not pretend that to condemn the slaveholders of antiquity requires any courage today. Nor does it do ancient slaves any good. Our understanding may have the past as its object, but our moral and political aims need to be present ones.

The world would be a better place if, in one way or another, we all said “I’m Spartacus,” more often. One obvious way to do this is to oppose slavery today. Although slavery is illegal everywhere in the world, the anti-slavery organization, Free the Slaves, estimates that something between twenty and forty million people live in slavery today. Despite the high total and the immense human misery involved, the proportion of slaves in the whole population is probably the lowest in history.5 Free the Slaves defines slavery in terms of coerced work for another’s profit. So major categories of slavery include debt bondage (especially among migrant workers), involuntary prostitution, domestic servitude, and forced child labor. Not all these forms of slavery today fit the historian’s definition of slavery in terms of property or natal alienation. Many do and, in any case, all are wretched and oppressive conditions for any person to endure; there do not seem to be modern equivalents to, say, the Athenian slave bankers.

Modern slavery is far from the only injustice in our world worth bravely fighting. Yet opposing injustice is not really the main purpose of a book on ancient slavery. The varied and complicated aspects of ancient slavery we have tried to understand, often from poor evidence, have made this an ambitious enough undertaking in any case: our topics have ranged from economics to politics, from daily resistance to bloody rebellion, from family life and sex to manumission, from slaves in drama to philosophy and law – and throughout the alternation of Greek and Roman cases with parallels and contrasts between them. No single intellectual approach is appropriate for this large and motley range of subjects involving slavery. Nonetheless, whenever possible I have tried to imagine actual slaves and masters and picture their lives and interactions rather than formulating theories in abstract terms. Such an effort of imagination is critical for all students of history, but it is just as important not to mistake what you imagine for what you know.