More bluntly put, the cities in which individual freedom reached its highest expression – most obviously Athens – were cities in which chattel slavery flourished. . . . One aspect of Greek history, in short is the advance, hand in hand, of freedom and slavery.
M. I. Finley, “Was Greek Civilisation Based on Slave Labour?” (1982a), 114–115
At the death of Messalina, the imperial household was in an uproar. [The Emperor] Claudius didn’t like being single and was submissive to his wives. His ex-slaves fought over who should choose the next one for him.
Tacitus (Roman historian, ca. 56–117 CE), Annals 12.1
Slaves were a relatively powerless group, often at the bottom of the social hierarchy. It would seem that they should have nothing to do with politics, which comprises the exercise of an important type of power, typically by those at the top of the totem pole. In fact, slaves played a part, arguably a crucial one, in classical politics in two very different ways. This chapter will focus first on the indirect role of slavery in the growth of Greek democracy and second on the direct participation of slaves and ex-slaves as bureaucrats in the Roman Empire.
A state’s economic system often shapes its political structure to a greater or lesser extent. In slave societies like Greece where slaves played a large role in the economy, the political system often reflected the importance of this role – not, of course, in the simple sense that slaves dominated politics since their work dominated the economy! Rather, historians have connected slavery with the egalitarian tendency among Greek free males and even the development of democracy. On the level of ideology, slaves played the role of the “other” against whom all citizens could define themselves as a single class. On the level of economics, the use of imported slaves allowed subordinating economic relations between citizens to be minimized. In simple terms, the rich could allow political equality for the poor because they did not need to exploit them; the maintenance of their lifestyle depended mainly on foreign slaves. Although this dynamic may have been true in many Greek cities, our sources of information are overwhelmingly Athenian, and so, in this chapter, I will mainly be talking about Athens.
Slaves have often served the political ends of others, but rarely their own. Their alienation from the rest of society has sometimes made slaves – like mercenaries – the most dependable tools for various types of despots fearful of giving too much power to their own nobles or citizens. This strategy contributed to the crucial role slaves and ex-slaves played in the administration of the vast Roman Empire during two periods. In the early Empire, the emperor needed to delegate power in order to run the Empire, but could find no other group so loyal, competent, and trustworthy as his own slaves and ex-slaves. In the late Empire, eunuchs – male slaves castrated as children – held crucial top offices. Even though they experienced great resentment and sometimes served as scapegoats, under a succession of emperors it was often a eunuch who held the most power in the empire besides the emperor himself; the court eunuchs as a group were also powerful and dangerous to offend.
The bulk of this chapter focuses on these two different aspects of the entanglement of slavery and politics. A shorter section will deal with state-owned slaves in Athens and the participation of slaves in Greek warfare, a role that their low status and exclusion from political power ought to have made unacceptable, but did not.
Karl Marx is most closely associated with the thesis that the economy plays a crucial role in the development of different political systems, but his theories involved over-simplifications about the ancient world and predictions about the modern world that did not come true. When it comes to theory, few historians believe that economics completely determine politics – probably a more dogmatic position than Marx himself held. But just as few would deny important connections between politics and economics. When it comes to ancient Greece, some historians argue that the economic role of slavery had important social and cultural consequences that favored the development of democracy. Paradoxically, slavery was a precondition for a political system in which all male citizens regardless of wealth had equal political rights, an almost unique development in the ancient world and perhaps the greatest of the many contributions of ancient Greece to world culture. More specifically, M. I. Finley, one of the most influential writers on ancient slavery of the past fifty years, proposed a provocative theory: the freedom enjoyed by a citizen in democratic Athens depended in part on slavery. His argument was that the political outcome, democracy, and the rights it assured to free Athenian men was closely related to the socio-economic structure of society including the widespread use of chattel slaves, especially in agriculture.
While today we think of wage labor as the standard arrangement when somebody works for somebody else, in ancient societies wage laborers were rarely central to the agricultural labor supply. Rather, the main alternatives to slaves were peasants of one sort or another. This statement requires a bit of explanation. In general use, peasant just means a relatively poor farmer – and this is the way we used the term in the last chapter. Among social historians, however, a peasant is a farmer who occupies the land but does not fully own it. Peasants are thus subject to a rent: they may need to work for the owner, give him a share of their crop, pay him some money each year, or simply pay high taxes to a government consisting mainly of the upper classes. When peasants are bound (not allowed to leave), they have lost an important part of their freedom and much of their bargaining power. Bound peasants are called serfs.1
When peasants, especially serfs, populate the countryside, conditions are unfavorable to democracy: the rich cannot allow equal political rights and freedom to people whom they need as subordinate workers. Peasants have trouble standing up politically to those who are in control of them economically – though they may revolt. Chattel slavery can make these subordinating ties between rich and poor unnecessary. In classical Athens, the rich did not need to impose labor rents or serfdom on other Athenians, since they could simply buy their labor supply: foreign slaves. The rich could allow democratic rights, and the poor could assert them. The livelihood of the rich did not require that the poor be kept closely in check; the livelihood of the poor did not require that they kowtow to the rich.
This helps explain one of the greatest puzzles about Greek democracy: why the wealthy in many Greek cities, but almost nowhere else, were willing to give up their traditional domination of politics and accept legal equality among the male citizens. The rich were a minority of the population in all Greek states, but their power was great – as is usual in the pre-industrial world. They traveled more and had a wider range of contacts. They were more organized. They had long monopolized state power in a world where tradition counted for a great deal. They had the money to hire foreign mercenaries, if push came to shove in a civil war. For all these reasons, the elite retained all sorts of advantages, and the triumph of the mass of the citizens and the establishment of democracy was anything but a sure thing. Democracy prevailed when much of the elite went along with it or, at least, did not fight it nail and fist.
This was partly because the rich continued to provide the most influential politicians and speakers even when a state gave equal political rights to all its citizens. It is also because less was at stake for them economically: the rich held onto their land and wealth, since the Athenian democracy never attempted the radical policies – abolition of debts and redistribution of land – sometimes instituted in other cities. And slaves could provide the labor to work the land, workshops, and mines of the rich. This acquiescence in democracy does not mean that the rich didn’t look down on the poor or resent democracy: one elite writer described the poor as ignorant and shameless and another called democracy, “obviously ridiculous.”2 Some rich men plotted to overthrow the democracy and institute an oligarchy. What is striking is how many went along with the democratic compromise, in which the wealthy could maintain their incomes and lifestyles but ceded political rights to the masses. Slavery made this compromise possible.
Finley also argued that two basic types of social structure existed in the ancient world. First, the people in some societies were spread out among many different status levels: from slaves to serfs to peasants of different sorts to more affluent tenant farmers to big landowners to members of the governing elite. Very few people were really free, since each class was bound in various ways to those above. Second and much more rarely, we find societies where people were concentrated at two points in the social spectrum, either they were completely free, possessing equal political rights and subordinate to nobody, or they were utterly un-free, chattel slaves, whose only recognized tie was one of subordination to their masters. According to Finley, Athens in the classical period was among the rare states that fit into this second category. Athens was marked by social dichotomy between free and slave and also by a lack of subordinating economic relationships among the citizens. This does not mean that there were no differences in wealth; rather, poor and middle-class citizens were not economically and thus socially bound to the rich, since they didn’t usually work for them.
One phenomenon in Athenian society highlights this lack of subordination among the citizens: the relative infrequency and unimportance of patron/client relations compared to other ancient societies.3 Patronage exists when wealthy men (the patrons) and poor men (the clients) form bonds based on reciprocal services. Although both patron and client do things for each other, the relationship is not equal. The client sacrifices much of his independence: he is expected to show respect, deference, and even obedience to his patron in return for the benefits he receives. Patronage is an essential ingredient in the way many societies operate. At Rome, for example, it was largely via patronage that some of the poor obtained help when they needed it and the rich exercised informal power over some proportion of the poor. In Athens, however, the poor sought assistance more often from the government and from equals than from patrons. In contrast to Rome, Athenian friendships were supposed to be between equals, not between a patron and his clients (Konstan 1998, 279–280). The avoidance of patronage was an aspect of the lack of subordinating economic ties between rich and poor made possible by the institution of slavery. This phenomenon was part of a self-reinforcing cycle working in both directions: the economic independence and political rights of the poor pushed the rich towards an increasing use of slave labor and slave labor made the independence and political rights of the poor palatable to the rich.
Finley’s explanation for democracy makes sense in that it proposes two types of social structure and shows that democracy is more compatible with one of them. But why did Athens go down one path instead of the other? The process was long and complicated, but one crucial step both towards democracy and towards the dependence on foreign slaves rather than peasants within may have come with the reforms of Solon in the early sixth century. Unfortunately, although we possess some of Solon’s own assertions about his programs – in poetry! – most of our sources about him are late, unreliable, and hard to interpret. Consequently, historians disagree both about his actual reforms and about their social context. According to one reconstruction, Solon instituted two agrarian reforms that gave more independence to many Athenians in the countryside: he abolished debt bondage and freed some sharecroppers from their obligations. These reforms played an important role in what was no doubt a complicated long-term process in which the countryside became more egalitarian and subordinating ties between rich and other farmers became less common.
Debt bondage remains today a common way that poor farmers fall into a subjection that is close to slavery (Bales 2012, 9). It is one of the reasons that agrarian societies sometimes drift towards greater and greater inequality and subordination until some violent and radical change becomes inevitable. In archaic Athens before Solon, not only was there debt bondage, but people in that condition began actually to be sold abroad – a practice sometimes distinguished as debt slavery. Perhaps this was a new abuse and led to the social conflict that Solon was called upon to mediate.
At the time of Solon, Athens did not even have its own coinage, so the debt for which people had been losing their freedom could have been a debt in kind: for example, in a bad year peasants sometimes end up eating all their grain and have to borrow the seed to sow for the next year’s harvest. The debt might also have come as a result of peasants failing to pay the rent on land that they farmed but did not own outright. This possibility brings us to the hektemoroi, a class of peasants in archaic Attica, whose name means “sixth-parters.” All we know about the hektemoroi is their name, one common interpretation of which is that it implies a sharecropping arrangement. It is a sign of our ignorance that historians can’t decide whether the sharecroppers needed to pay over one sixth of their harvest (a surprisingly good deal if the land is somebody else’s), only got to keep one sixth of their harvest (a bad deal and probably not enough to survive on), or some third alternative (e.g., one-sixth of a fixed rent is paid in advance).
In any case, Solon’s reforms somehow liberated the hektemoroi from their obligations. Presumably the hektemoroi were given full, unencumbered property rights to the land they lived on and had been farming. Solon also eliminated debt bondage; he even claimed to have brought back to Attica and restored to freedom some people who had been sold abroad. Lending on the security “of the body,” that is debt bondage, became illegal, at least for citizens. By the classical period, we find that many independent middle-class farmers – typically even owning a few slaves of their own – and a few rich farmers populated the Attic countryside. It is almost impossible to find any peasants, in the social-historical sense, in Attica. Rather, the real poor seem to have lived mainly in the city of Athens. The agrarian social structure of Attica was one in which democracy could thrive.
Finley’s picture gains some support from the reverse set-up in some other Greek states, most famously Sparta, which controlled two large areas of the Peloponnese: Laconia and Messenia. Under Sparta, the agrarian economy focused on the production of subsistence crops and grain. Rather than depending on foreign slaves, closely supervised, the Spartans depended on a serf-like class called the Helots to work their farms, mainly on a sharecropping basis.4 Helots were not slaves but were rather natives who had lost their freedom to the extent that they were bound and subject to labor services as well as owing a share of their crops. How this came about in the poorly known archaic period is obscure and controversial, although the conquest of Messenia by Sparta played a role in the case of the Helots there. The results are clearer: in Laconia and Messenia, the Spartans imposed a harsh subjugation on many of the inhabitants, reducing them to serfs with no political rights; they had no need for – and possessed few, if any – imported chattel slaves. In contrast, would-be peasants in Attica ended up with political rights and economically independent, either as farmers or as working men of various sorts in Athens and other towns in Attica. Chattel slavery played a crucial role in this process, but it would be a mistake to claim that it was the only factor. The institution of slavery and the cheap availability of foreign slaves constituted a sine qua non for Athenian democracy, but many other factors – cultural, individual, economic – were also necessary.
Indeed, an apparent problem with Finley’s whole theory is that there existed a third class of Greek city-states, for example Chios, that employed chattel slaves on a large scale, but remained oligarchies for most or all of the classical period (cf. Jameson 1992). Two approaches can help us deal with what seems at first like a devastating set of counter-examples.
First, many Greek oligarchies are classified as broad rather than narrow: many people possessed political rights rather than just a few. In a narrow oligarchy, an elected council of ten men from the richest class might wield almost all political power. In contrast, a broad oligarchy might make decisions in an assembly open to the richer half or third of the free male population, numbering in the thousands. In the latter case, the oligarchic citizens were mainly the free, middle-class farmers whose independence was made possible by slavery. In these states, the original narrow aristocracy of birth and wealth had shared its political power, but just not as widely. They had not taken the step of granting active political rights with the poor, who were – at Athens at least – mainly an urban group.
The second and more important response to the objection is to be clear about Finley’s argument: slavery was a necessary condition for Greek democracy; it was not a sufficient one. Slavery made possible the democratic compromise by which the elite’s wealth and lifestyle survived the sharing of political power with poor Athenians. This is far from a full explanation of democracy. That would require a detailed consideration of the individual statesmen, their backers and opponents at each critical step. It would also require a consideration of Greek intellectual culture, the political impact of colonization, the fear of tyranny, and the type of wars the Greeks fought. This is not the place to treat these topics, but historians have invoked them all as partial explanations for democracy, a phenomenon not likely to have just one cause.
Our focus on Solon also tends to compress a longer and more complicated process. Slavery could not have been a new phenomenon in 591 BCE: after all, Homer and Hesiod depict or mention slaves at least a century earlier. Instead, we should imagine that, as Solon’s ideas for reform were passing around and various rich landowners were deciding how strongly to oppose them, the option of using foreign slaves to work their large holdings made concessions more palatable. We have already discussed in the last chapter the reasons that slaves were so cheaply available in Athens in the classical period; some of these factors were already in place at the time of Solon. The rich probably calculated that as long as they prevented a general redistribution of land, they could do well enough using slave labor on their own lands – as they did.
In addition, more and more farms may have been turning to mixed crops and market farming instead of just harvesting grain for subsistence. As we have seen in the case of Roman Italy, such farms are able more efficiently to use slave workers. Market rather than subsistence agriculture allowed farmers to make money by selling their produce: mainly fruit, olives, and grapes. These shifts in agricultural practices were long-term trends that made the use of slaves become more attractive over generations rather than suddenly and as the result of something Solon did.
Slavery not only affected the political economy of Athens, but also the way Athenian men thought about their social world, their ideology. In particular, slavery was a double-edged sword in its effect on the way Athenians conceived of hierarchy and equality among the citizens.
On the one hand, Orlando Patterson, whose definition of slavery we have met already, argues that slavery commonly leads to a cultural emphasis on honor and a keen sense of rank. Slave owners are used to ordering people around; they do not necessarily lose this habit when they are interacting with people other than their slaves. As a reaction, in a society with slavery, people are in general sensitive to anything that smacks of being treated like a slave and are quick to take offense. Slavery also opens another prospect for inequality in that not all free people are slaveholders. This can become an important distinction between free people. For example, in the American South, it became a mark of respectability to own at least one slave – a mark of respectability that most of the population could not attain.
On the other hand, slaves can also become a group against whom people can define themselves as a unity: “we are all alike – and friends and allies – since we are all superior to them.” Historians find this interpretation of the effect of slavery persuasive in part because it is easy to find other examples of this social and psychological phenomenon. The scapegoating of Jews by the Nazis allowed them to emphasize the essential unity of the Aryan people; one view of racism against blacks in the United States – not a fully accurate one – insists that racism is attractive to poorer whites because it gives them somebody to look down on and allows them to imagine themselves as part of the ruling (white) class. This way of thinking has been associated with the concept of the other, an external group whose existence and difference allows another group to think of itself as united and superior.
Returning to ancient Greece, the Athenian democracy has been described as a free male citizen’s club in that democratic rights excluded slaves, women, and foreigners, and the self-conception of the democracy depended on oppositions between the free male citizens and these others (Vidal-Naquet 1986, 206). One striking piece of evidence of the position of slaves as the other was the tendency, when Athenians categorized the men in the city, to refer to two groups only, “slaves and citizens.” In addition to forgetting about the resident foreigners and women, this dichotomy is not a natural one. One would expect slaves versus masters. But all citizens, masters or not, were thought of as a single group in opposition to slaves. Similarly, speakers addressed juries in Athens as if all the jurymen owned slaves. This has puzzled social historians, since they know that many of the jurymen were poor and didn’t own slaves. All jurymen were citizens, however, and seem to have felt that they were at least potential slave owners and, as one speaker insists, all had a stake in the good behavior of slaves ([Demosthenes] 45.86). In some ways, this was just an ideology: it served to disguise actual conflicts of interest between rich and poor among the citizens – which sometimes did break out into civil war. In contrast, the rich often viewed the mass of poor Athenian citizens as distinct and inferior to themselves, hardly different from slaves. Nevertheless, the view of the males as divided between citizens and slaves was no more than a simplification of an actual state of affairs: in terms of legal and political rights, classical Athens was a society dominated by one big dichotomy rather than a more complex hierarchy.
A recent book by a French ancient historian, Paulin Ismard’s Democracy’s Slaves: A Political History of Ancient Greece, argues that slavery was crucial to the development of Greek democracy in another way as well (2017 with Hunt 2016a). Public slaves are slaves purchased and owned by the state. Best attested in Athens, they helped build and fix roads, worked in the mint, served as prison guards and executioners, and checked the authenticity of coins in the marketplace. Some even assisted the magistrates and supervised the public archives. These slaves contributed to the smooth running of government: the magistrates themselves were citizen amateurs, usually chosen by lot and serving for a single year, while the slave “assistants” had probably been working at the same post for many years and may have had a much better idea of how things needed to be done.
Ismard goes further and maintains that the longstanding Greek ideal of transparent and direct rule by the community, an important democratic tenet, was only possible because, out of sight, slaves provided the expertise the government required. The citizens did not have to delegate political power to an administrative class from within their ranks. Rather they bought foreign slaves to perform government functions. And, in contrast to a governing elite, public slaves gained little power from their administrative experience and knowledge. Although they enjoyed a higher status than private slaves in several respects, they remained slaves and thus isolated outsiders – not to mention that they could still be whipped.
To modern ways of thinking, the Scythian archers represent the most startling use of public slaves at Athens (see Map 2 for Scythia). These armed slaves served some of the functions of a modern police force, for example, keeping order in the assembly and seizing or removing people at the order of a magistrate. This use of slaves as police was in part a result of the egalitarian ethos among the citizens: for one Athenian to lay hands on another was perhaps too provocative; it symbolized inequality and was likely to lead to violence or at least violent resentment. The Scythian archers, foreign slaves, were somehow out of the game. Of course, individual Athenians resented it when they were manhandled by these foreign slaves (Hall 1989, 47). But such treatment did not threaten democratic equality as similar treatment by a fellow citizen would have.
The oddity of these armed slaves within Athens leads us to the military role of slaves, a problematic one, since military service was a key aspect of being a citizen, slaves were in theory the opposite of citizens, and this dichotomy was central to the way the Athenians categorized people. At first blush, it would seem that slaves could play no role in the many wars that Greek city-states fought against each other. The risks of entrusting such oppressed and often resentful men with weapons are obvious. And many Greeks felt that military service gave a person a claim to political rights, something from which slaves were definitely excluded. The stereotype of slaves as cowardly and childish would hardly recommend using them as warriors. Finally, although wars are the main topic of contemporary historians, references to slave involvement are few and far between. Nevertheless, in my first book, Slaves, Warfare, and Ideology in the Greek Historians (1998), I argued the opposite of what one would expect: slaves were important to classical Greek warfare, but their participation was an awkward topic and contemporary historians systematically neglected it as a result.
The patriotic citizen soldier may be an ideal, but reality did not – and does not – always live up to it. From the Persians who (reportedly) drove their soldiers into battle with whips in the Persian Wars (Herodotus 7.223.3) to early modern armies that could not camp in wooded areas for fear their men would desert, armies have often made do with resentful and oppressed soldiers. We cannot rule out the possibility of disgruntled non-citizens serving as soldiers on that count: the frequent military service of the serf-like and sometimes rebellious Spartan Helots provides a striking example. And slaves often served in the navy, where they did not have to be given weapons but rather rowed. Finally, slaves could be motivated by the promise of winning their freedom, in addition to all the other incentives or threats that have motivated previously unenthusiastic soldiers throughout history.
There was nothing impossible about using slaves in warfare. It could also be highly advantageous, since slaves were numerous, could bolster a city’s forces considerably, and might escape to the enemy otherwise. For example, the Athenians promised freedom and even citizenship to the slaves they recruited to man a navy in the desperate last years of the Peloponnesian War, a navy that went on to a great victory in the largest sea battle of the war. The grant not merely of freedom but also of citizenship, usually a closely guarded right, may have been a tactic to encourage these freed slaves to remain with the Athenian navy even after the battle rather than deserting to the Peloponnesians, who were paying higher wages at that time (Hunt 2001, 366–370). The practical advantages of enlisting more men – often a lot more – encouraged Greek states to overlook the contradictions between the low status of slaves and the high prestige of military service. When push came to shove, when a navy had to be manned, military advantage trumped ideology.
In the composition of written histories, however, ideology remained dominant. Contemporary historians did not dwell on slave participation, an awkward topic to Greek ways of thinking. Oblique statements occasionally reveal a whole pattern of slave use that contemporary historians otherwise neglect. Thucydides, for example, explains why the crews of the Syracusan and Thurian ships were particularly vocal in demanding their back-pay from their admirals: “since free men made up the majority of their crews” (8.84.2). First of all, this passing reference tells us that some of the members of the Thurian and Syracusan crews were slaves. More startling, it also implies that among the other contingents in this Peloponnesian fleet, slaves made up a majority of the crew, something we would not have suspected without this passage. My arguments probably shifted scholarly opinion in the direction of accepting a greater role for slaves in Greek warfare – and for Spartan Helots as well – but particular cases remain controversial.
Some historians believe that slavery played an indirect role in Roman politics, specifically in the fall of the Roman Republic and the establishment of a monarchy in its place, the Roman Empire. If it were true that large estates worked by slaves displaced small farmers, traditionally the backbone of the Roman army, then it would follow that more and more soldiers did not own land or only very little – which seems to have been the case. Such landless soldiers, likely enough, had less of a stake in the Roman Republic and less loyalty to the state. They would be more closely bound to their general, especially when he promised to arrange for retirement bonuses, often land, for his veterans – something the state did not regularly provide. As competition among the elite grew more intense, ambitious generals eventually wanted and were able to lead their men even against Rome and its lawful government, to seize power for themselves, and to obtain the rewards they had promised their troops. This led to a period of civil wars, and eventually one of the competing generals, Augustus, succeeded in establishing a monarchy, in part by regularizing retirement bonuses for the soldiers (Brunt 1988).
According to this reconstruction, slavery’s economic impact may have been a partial and indirect contributory cause in these momentous changes, just as it played a role in the Greek development of democracy. The growth of slavery cannot come close to providing a complete explanation for the fall of the Republic, but that is not even the main problem. Rather every step in the theory I outlined above is disputed, either fundamentally (Were landless soldiers any less patriotic?) or with respect to crucial details (When exactly did the soldiery cease to consist mainly of independent middle-class farmers?).5 To examine all the links in this chain would take us far from Roman slavery and would require a book of its own. This section, therefore, considers slaves who played a direct role in the Roman political system, similar but more conspicuous than Athenian democracy’s employment of slave administrators: the slaves and freedmen of the Roman emperor. In the early Empire, these slaves and ex-slaves constituted a powerful and large section of the imperial bureaucracy. In the late Empire, eunuchs, either slaves or ex-slaves, gained immense power from their control of access to the emperor.
The slaves and ex-slaves of the emperor’s bureaucracy were known as the familia Caesaris: the familia of Caesar, that is, of the emperor. Familia originally denoted a person’s entire household, but it was eventually mainly used of slaves and ex-slaves. Slaves had long acted as accountants or household managers for their masters. The households of rich Romans could include extended families, various clients and guests, and slaves in the hundreds; expenditure and income was on a grand scale, so household managers and accountants held skilled and responsible jobs. By the time a smart, ambitious, and loyal slave had advanced to such a position – in his thirties or forties perhaps – he had often already gained his freedom. As we shall see in Chapter 8, a Roman ex-slave was still attached to his master by bonds of patronage and would often continue to work for him, so in this section we will be talking as much about the emperor’s freedmen as about his slaves. At the top levels, we will mainly be dealing with freedmen.
The emperor was not merely any rich Roman. From the beginning of the Empire, the emperor’s personal wealth was incomparably larger than that of any other individual and included property in far-flung provinces. It became larger and larger as time went on. In addition, the emperor personally controlled much of the state’s revenue and expenditure, including, for example, the salaries of the entire Roman army, more than 300,000 men. Fiscal matters were only part of the story: the emperor was responsible for managing an empire, including more than 25 provinces, with borders thousands of miles away from Rome, and a population of perhaps sixty million. Even a conscientious and hard-working emperor couldn’t read and keep track of all of his correspondence on local political matters with the hundred-odd provincial governors, army commanders, and the managers of his estates throughout the empire. Not to mention that some emperors were not, in fact, remotely conscientious and hard working. Ruling this empire was an immense task and early emperors naturally turned to their slaves for help. So the use of slaves and ex-slaves was both a natural extension of their role in managing a private household and essential for the administration of the empire.
Some branches of government were delegated to the Roman elite, the senatorial and the equestrian class. These classes provided military commanders and provincial governors, for example. The emperor wanted to supervise other aspects of the administration more closely. Central government functions, such as managing the imperial budget, needed to be taken care of at Rome. It was natural for the emperor to supervise these tasks himself, but his subordinates could not be senators. For, in the early Empire, Augustus and his successors tried to maintain the façade that important institutions of the Republic continued to exist and to matter. Crucial to these attempts was the emperor’s respect for the dignity of the senatorial class. It would have been awkward, to say the least, for an emperor, ostensibly just the first-man, the princeps, of the traditional Roman aristocracy, closely to supervise senators and equestrians, to have them read his mail and take dictation, to keep track of his finances, and to draft responses to all the requests that came to him. The sensibilities of the aristocracy would be offended by such direct, daily, and long-term subordination to the emperor.
Finally, relations between the emperor and Senate were sometimes tense – something of an understatement in the case of emperors who had prominent senators executed because they suspected or had discovered plots against themselves. Such emperors might hesitate to give the power that went with these administrative positions to a senator out of fear about how that power might be used for the senator’s own gain, for his allies, or even to undermine the emperor.
Thus, for both positive and negative reasons, the emperor turned to his slaves and ex-slaves to constitute his bureaucracy. The slaves who took advantage of this opportunity could expect social advancement. This was the case for a large proportion of the familia Caesaris down to minor secretaries and accountants and certainly for the top administrators who enjoyed immense power and wealth, but were also exposed to the bitter resentment of the Roman aristocracy.
We know a fair amount about the social prospects of the familia Caesaris – or at least those resident at Rome – from what was engraved on their gravestones, their epitaphs. Roman custom was to engrave not only the name of the person who had died on a gravestone but also the name of the person who had put up the gravestone and his or her relationship to the deceased. In addition, Roman names often reveal legal status: whether the person was a Roman citizen, a foreigner, a slave, freeborn, or an ex-slave. The epitaphs of ex-slaves also reveal the name of the person’s former master; for the Roman name of an ex-slave included his or her single slave name as a first name and then the family names of their former masters. When the family names are those of an emperor, we can be fairly certain that we are dealing with a member of the familia Caesaris. Several hundred such gravestones have survived for study.
From this evidence, P. R. C. Weaver identified one startling pattern: slave men from the familia Caesaris often married freeborn women.6 Indeed, at least two-thirds of the women married to members of the familia Caesaris were freeborn. True love may perhaps strike anywhere, but ancient marriages, largely arranged by male relatives, tended mainly to respect the social hierarchy. Normally, the last thing a free family wanted was for one of their women to have sex with a slave, let alone marry one. In addition to the disgrace, the children from such a pairing – since it would not be considered a marriage – would be illegitimate.
The main exception to marriage taking place between status equals is hypergamy, where a woman marries up the social hierarchy, a common pattern in many societies and in part explicable in terms of evolutionary biology. Paradoxically, it turns out that the pattern Weaver discerned is just another case of hypergamy. Imperial slaves possessed a high enough status that freeborn women were willing to marry slaves, albeit slaves with prospects. Furthermore, many of the wives also carried imperial names, names shared by emperors. This suggests that they were descended from imperial freedmen of earlier generations, a group less likely to consider free birth the decisive indicator of social status.
This surprising marriage pattern was most pronounced among the familia Caesaris living in Rome, many of whom must have been accountants and secretaries in the emperor’s service. Such marriages had one consequence detrimental to the imperial bureaucracy. By Roman law – and what they claimed was the law of all nations – the status of any child not born from a legitimate, official marriage followed the status of the mother. Since marriages of slaves were not legitimate, the children of male slaves of the familia Caesaris and free women were free. This deprived the emperor of the next generation of slave bureaucrats, and so a law was passed to discourage such pairings or, at least, to ensure that the offspring would become the emperor’s slaves. The details are controversial and obscure, but the important point was that free women who had children with slaves might be liable to enslavement; thus their children would be slaves and potential secretaries and accountants for the emperor. The author of this new law was himself a freedman, Pallas, one of the great administrators and advisors of the emperor Claudius and confidante (and supposedly lover) of Claudius’ last wife, Agrippina. We turn now to men like Pallas, the powerful and infamous freedmen at the top of the imperial bureaucracy.
At the head of each of the different departments of the imperial bureaucracy at Rome stood a high-ranking administrator, like Pallas, certain to do much better than merely to marry a freeborn woman. Until the reign of Hadrian (117–138 CE) these officials were typically ex-slaves. The three most important were the a rationibus, the ab epistulis, and the a libellis. The Latin a/ab means “from” so the a rationibus means something like “[the man put in charge] from accounting.” This particular position was something like the Secretary of the Treasury in the United States today, and was like a cabinet-level position in that the a rationibus answered only to the emperor, as was the case with all of the major bureau chiefs. The ab epistulis handled the emperor’s official correspondence (epistulae) – with governors throughout the Roman Empire, for example – and seemed to have played an advisory role in making appointments and promotions. In fact, the future emperor Vespasian owed his first major appointment to Narcissus, the ab epistulis under the emperor Claudius. The ab epistulis has been likened to the Secretary of State. Finally, the a libellis administered a large and important part of the emperor’s responsibilities, his responses to letters that petitioned for a favor or requested the emperor to redress some grievance. These requests were answered with a reply on the bottom of the original petition. The a libellis or one of his subordinates wrote the responses and had them signed by the emperor. Given the responsibility and discretion of all three of these posts and that they answered only to the emperor, they arguably held more practical power, if lower prestige, than all but the most important governorships or magistracies.
We gain an idea about the scope of their powers from the poet Statius in an elegy addressed to the wife of the deceased freedman Abascantus, ab epistulis to the emperor Domitian:
Your husband’s job was to send out, far and wide into the great earth, the orders of the emperor, the successor of Romulus, and to control with his hand our imperial armies and the methods of our empire . . . he must disclose to the emperor who deserves to be a centurion and to ride on horse among the companies of infantry, who is to command a cohort, who merits the higher status of noble tribune. (Statius, Silvae 5.1.86–97)
Similar praise of Polybius, a libellis, comes from the hand of Seneca, an exceptionally rich and powerful senator, an intellectual and philosopher, and top advisor to Nero – by whom he was eventually executed:
You must hear so many thousands of men and decide so many petitions. You have to examine such a mass of problems streaming in from the four corners of the world, for the purpose of submitting them in the due order to the judgment of our supreme ruler. (Seneca, Consolation to Polybius 6.5, trans. in Duff 1928, 155)
Seneca thus provides us with a description of the work of the a libellis, but this passage’s interest to historians does not end there. Seneca had been exiled for (purportedly) committing adultery with a member of the imperial family – any sex with an unmarried woman was considered adultery. The Consolation to Polybius, written in response to the death of Polybius’ brother, was part of Seneca’s campaign, eventually successful, to have himself reinstated at Rome. That he composed a lengthy work for an ex-slave in hopes that he could intercede with the emperor shows clearly the power that the top freedmen were thought to possess.
Seneca’s flattery of Polybius is thus easily understood; his admiration was probably feigned. When he had risen to become one of Nero’s top advisors, Seneca represents the deceased Emperor Claudius as a dupe of his own ex-slaves, as Tacitus does in the quotation with which we started this chapter. In a satire on Claudius’ supposed divinization after his death, Seneca describes Claudius trying to give an order in heaven: “They paid him no more attention than his freedmen had” (Seneca, Apocolocyntosis 6).
Pliny, writing under the emperor Trajan (98–117 CE), gives us another and more blatant example of senatorial resentment of powerful, imperial freedman, in this case of Pallas, the a rationibus under Claudius and Nero, more than a generation earlier. Pliny had noticed a monument to Pallas with an inscription describing some honors that the senate decreed for him: the insignia of a praetor and fifteen million sesterces. Pliny, himself a wealthy and prominent senator, was moved to go through the Senate’s official records and located the actual decree, which was filled with flattery of Pallas. He was indignant and ashamed of the Roman Senate:
But who is so crazy as to desire advancement won through his own and his country’s dishonor, in a State where the chief privilege of its highest office is that of being the first to pay compliments to Pallas in the Senate? I say nothing of this offer of the praetorian insignia to a slave, for they were slaves themselves who made the offer. (Letters 8.6, trans. Radice 1969)
The historian Tacitus was also aghast at these proceedings: “And the senatorial decree was made public, inscribed in bronze: an ex-slave, now worth three hundred million sesterces, was loaded with praise for his old-fashioned thriftiness” (Annals 12.53). Members of the senatorial class were bitterly competitive and quick to resent any of their peers who surpassed them in rank and prestige; it’s easy to imagine the smoke coming from their ears when ex-slaves eclipsed them in the emperor’s eyes and in public honors. It probably never occurred to them to consider how competent, useful, and loyal to the emperor such ex-slaves were.
Powerful freedmen were particularly prominent under Claudius, whose relations with the Senate were tense and whose claim to the throne was at first tenuous. Even reputedly good and strong emperors used them high in their administration until the time of Hadrian, whose reign started in 117 CE. At that point, the Empire had been established for many generations and service to the emperor had become an honorable profession. Although many of the minor bureaucrats and secretaries were still slaves and freedmen, the top positions were reserved for members of the equestrian class, the wealthy class next to the senators in rank. Many of these equestrian officials in the second century CE carried the same family names as earlier emperors. Since they were definitely not actual relatives, they were almost certainly the descendants of the familia Caesaris of earlier emperors. Indeed, the consul of 167 CE could claim the great freedman a rationibus Pallas among his ancestors. Even though the honor of the consulship was one of the pinnacles of the most successful senatorial careers under the emperors, this consul’s actual power was likely only a shadow of that of his infamous ex-slave forebear.
Beginning in the fourth century and lasting through the Byzantine Empire, eunuchs played a large role in the imperial government, arguably even greater and certainly longer lasting than the role played by early imperial freedmen. For example, the Grand Chamberlain of the Emperor, typically a eunuch, possessed the fourth highest rank in the whole imperial administration. The bribes that even lesser eunuchs regularly demanded and received reflected the fact that they could get things done or prevent them from happening. The resentment that they provoked among the traditional aristocracy was bitter and probably proportional to the influence they had on the emperor. Although some emperors were criticized for allowing the eunuchs too much power, all sorts of emperors – the good, the bad, and the mediocre – used eunuchs in their administration.
Eunuchs are men or boys who have been castrated. This surgery was generally only performed on slaves, although naturally most high-ranking eunuchs had received their freedom earlier in their careers and were thus freedmen when they became, for example, Grand Chamberlain. The exact physical and biological results of castration depend on the age at which the operation is performed and whether just the testicles or the penis too is excised. One certainty is the danger of the operation, performed without sterile instruments, antibiotics, or modern painkillers. According to one story, the emperor Justinian (ca. 482–565) was outraged at hearing that only three children lived out of ninety undergoing the operation. This would seem like sensationalism, except that better statistics from nineteenth-century Africa also reveal a 90 percent mortality rate for slaves undergoing castration (Lovejoy 2012, 35). The mortality rate associated with castration had an economic consequence too: eunuchs must have been extremely expensive slaves – since it might require the waste (painful deaths) of about ten slave children to produce one eunuch. Paradoxically, this may have contributed to their impressiveness as expensive and exclusive status symbols.
Ancient eunuchs are described as effeminate, smooth skinned, and fat. Their voices might have been high or unusual, depending on the age of castration. Their inability to procreate may have been their original attraction: the Greek word eunuch means “those who guard the bed” and one early function was as trustworthy supervisors of the harems of eastern monarchs. In such monarchies, eunuchs also possessed great power; Roman emperors may in part simply have been imitating a model they encountered in states such as Sassanid Persia.
In a patriarchal society, the sterility of eunuchs marked them as unnatural and outside of society, except for their bond with the king or emperor. Paradoxically, some eunuchs purportedly retained a sexual appetite; indeed, some were described as insatiable. Without the possibility of paternity, this only made them seem unnatural and monstrous. Most eunuchs probably had their origins outside the Empire, so they were barbarians too. As sterile men, as biologically altered, as slaves, and as barbarians, eunuchs would seem far from ideal holders of power in the Roman world. But it was eunuchs’ difference and alienation that provided the rationale behind their power. Before we can understand this claim, we need to explore the nature of that power.
Keith Hopkins has argued persuasively that the eunuchs of the late Empire derived their importance primarily from their own intimacy with the emperor and from their control of access to him (1978b). The titles of the great eunuchs reflected this role. The position translated as “Grand Chamberlain” in Latin is praepositus sacri cubiculi, which literally means, “the man put in charge of the sacred bedchamber.” Such intermediaries became more important as the emperor represented himself more and more as divine and an increasingly elaborate court ceremonial tended to separate him from mere humans. Eunuchs thus provided, and sometimes controlled, communication between the emperor and his subjects. They often provided the emperor’s most trustworthy intelligence about the empire he ruled. They would communicate requests from his subjects or set up audiences – often only upon receipt of a hefty bribe. But, given that the emperor needed – as all rulers do – to communicate with his subjects and needed courtiers to moderate, channel, and ensure this communication, why did the late Roman and Byzantine emperors choose eunuchs for this function?
Recent interpretations of eunuchs in the Byzantine Empire view them from the perspective of gender studies: Were they a third sex or gender? How did their sterility affect how they were perceived as men? Was their supposed androgyny part of their power? (Stevenson 1995, 495–511; Ringrose 2003). But our focus is instead on eunuchs as slaves and thus Hopkins’s political interpretation of their power is most apt and, in my opinion, most convincing overall. Hopkins makes the argument that the eunuchs were particularly trustworthy and thus useful to the emperor because their allegiance was undivided (1978b). They were slaves, natally alienated: their prior families, alliances, and other social bonds had been severed. They had no loyalties other than to the emperor, upon whom their entire position depended. Their biological difference from other men, that they seem to have been regarded with disgust and suspicion, severely limited their ability to ally with (or potentially to plot with) the imperial aristocracy. Most crucially, their sterility meant that they could not contract marriage alliances. One main, if not the main, purpose of elite marriage was alliance between families, and children or the expectation of children cemented the alliances. Unable to contract marriage alliances, eunuchs remained, individually and as a class, in isolation from the rest of society and in dependence on the emperor. They were his perennial tools.
This position could bring great power, but it also had its risks. In particular, eunuchs often provided scapegoats for unpopular imperial policies. If an emperor’s policies evoked too much resistance – widespread riots, for example – the emperor could blame his eunuch chamberlain and execute him. Again the isolation of eunuchs was key to their usefulness as scapegoats: there were no families and allies who would have to be placated, destroyed, or distrusted if an emperor dismissed or executed a eunuch. It might worry other eunuchs, but the matter was essentially between the emperor and his victim.
This scapegoating of eunuchs means that it was sometimes in the emperor’s interest to attribute power to a eunuch when he was actually just carrying out an order that the emperor later decided to wash his hands of. In the fourth century, a eunuch of Constantius II reportedly prevented the future emperor Julian from obtaining an audience with the emperor, but perhaps it was Constantius himself who didn’t want to meet with Julian and later found it convenient to blame his eunuch (Hopkins 1978b, 173). Such shifting of blame may leave us with an exaggerated view of eunuchs’ actual initiative and importance. Nevertheless, because of their incapacity to be assimilated into the aristocracy, eunuchs provided a crucial tool for semi-divine and otherwise isolated emperors and accrued much practical power as a result.
That classical Athens, like many of the cities of ancient Greece, was a slave society meant that its politics were profoundly affected by slavery. Just as some historians trace the segregation and growing use of African slaves in the United States South with the growth of the notion of equal citizenship for whites, the equality of citizens in the Athenian democracy may have roots in slavery. Economically, slavery reduced subordination among citizens; on the level of ideology, slaves provided the other against which all citizens, be they rich or poor, could define themselves as a unity, however diverse their real interests.
Although slaves by definition lack power with respect to their owners, they can obtain political power through their bond with a powerful master, especially when that master is the emperor of Rome. In the early Empire, slaves and ex-slaves provided a bureaucracy for emperors who needed one. In the late Empire, eunuch slaves watched over the divine emperor’s person and facilitated his relations with his subjects. In theory both groups of imperial slaves were merely tools, extensions of their master, whose power they merely transmitted. In practice, quite a bit of power tended to stick to these men, so regularly used as the tools of such a powerful monarch.
Finley 1982a and 1982b present the main arguments about democracy and chattel slavery that I summarize in the section on Athenian democracy and slavery. I strongly believe that Finley had the big picture right and his basic conception of the relationship between Athenian democracy and slavery remains influential (e.g., Scheidel 2008). But, for example, Vlassopoulos 2016a provides a critical overall assessment and Harris tries to attack important aspects of this model (2002 and 2012). Divergent interpretations of Solon’s reforms and their socio-economic context and impact are as numerous as what we know for sure is little. The topic is, nevertheless, a fascinating one and essays collected in Blok and Lardinois 2006 and the commentary of Noussia-Fantuzzi 2010 provide an entrée. Hunt 2006 provides a short and accessible summary of some main arguments from Hunt 1998, my book on slaves in Greek warfare. On the slaves of the emperor, Weaver 1972 is detailed and scholarly. His breakthroughs resulted from his statistical analysis of the epitaphs of members of the familia Caesaris rather than on trying to reinterpret the well-known literary evidence. Duff expresses a contempt for powerful freedmen that resembles the attitude of the Roman aristocracy: “imperial freedmen gained an ascendancy in the Empire the like of which has never in another nation fallen to a series of low-born upstarts” (1928, 174). Yet his descriptions of the growth and decline of the power of ex-slaves, the various offices they held, elite resentment, and the lives of famous or notorious freedmen are clear and detailed. My treatment above closely follows Hopkins 1978b on the political context for the rise of powerful eunuchs. Tougher 2008 provides a more general treatment of eunuchs in the Byzantine Empire.