David braund approaching the Greek slave supply
How did Greeks obtain their slaves, and how did men, women and children come to be slaves in classical Greece? While statistics are elusive, no one would seriously challenge the fact that there were large numbers of slaves in the Greek world of various kinds and origins. This chapter seeks to explore the processes that brought people into slavery in Greece, and also to give some sense of the individual slave’s experience of these processes. The sheer scale of enslavement meant that the supply of slaves was a central feature of the ancient economy, quite apart from its fundamental social significance within and beyond Greek culture. The slave trade was everywhere. At the periphery of Greek culture, slaves were traded all around the Black Sea, in the Adriatic and in the Eastern Mediterranean and North Africa. So too at its traditional centres — in Athens, Aegina, Corinth, Chios and elsewhere. The scattered instances about which our sources tell us are no more than drops in the great ocean of the ancient Greek slave trade, with its ripples reaching far and wide into ‘barbarian’ hinterlands.
The slave trade was also everywhere in ideological terms, explored by philosophers or expressed in art, from Homer through Athenian tragedy and historical and geographical writing of all kinds. Moreover, the slave trade was treated as a fit topic even for comedy, whether in satirical poetry or on the stage. For all the sufferings of individuals and communities which flowed from the Greek demand for slaves, the society and economy of classical Greece was overwhelmingly comfortable with the buying and selling of men, women and children, at least in so far as these ‘goods’ seemed sufficiently alien.
It is in Aristophanes’ last surviving comedy (Wealth) that we find some of the most penetrating comment on the slave trade in Greek literature. At the centre of the play is Wealth itself (acquisition, distribution and lack). The very presence of slave-trading within the play illustrates well enough its importance in terms of property.1 But there is scant sympathy for the
1 Cf. von Reden 1995: 95,198.
II2
Enslaved: in fact, the Athenian protagonist hopes for the proliferation of slaves to do more work. Enslavement and the trade in slaves are presented as facts of life. The slave’s voice is a master’s voice, with no challenge to the slave-owning audience. Aristophanes’ slave bewails his fate but does nothing to suggest that his master is wrong in acquiring and using him. Instead, his enslavement is made an example of the power of money. In the opening speech the slave Carion opens the play by bemoaning his lack of power: slaves must share in the consequences of their masters’ bad decisions. God does not allow the slave to be master of his own body but sets his purchaser in control (Plut. 5—7). The institution even has divine sanction. The difference between Carion’s freedom and slavery is a little silver: he has been sold into slavery for a small sum of money (Plut. 147—8).
There is sympathy for the enslaved barbarian’s plight only in the very limited sense that all may consider themselves at the mercy of Wealth. This is comedy for a society at ease with the purchase of a ‘barbarian’, a taste of prevailing ideology in classical Greece.180
To understand the Greek slave supply, we must come to terms with the normality of enslavement, slave-trading and the commodification of human beings. And also with its fragmented and ramifying geography: while from time to time (especially through war) large numbers of people were enslaved together, there is no sign that the slave trade operated on the grand scale or with the regular routes familiar from the Atlantic trade. Those enslaved in large numbers were immediately bought up by small-scale dealers, who thereby distributed them as they saw fit. This is not to say that the trade was disorganised; only that it was organised around the minimal needs of a fragmented and much-ramified market.181
Almost anyone might buy or sell a slave, anywhere. There were traders who took a particular interest in slave-taking and slave-dealing (andropodis-tai), and there were places particularly favoured for slave-trading at certain periods, but by and large the slave trade was an omnipresent and routine series of small-scale exchanges, made everywhere, by all manner of individuals, and with no sign of any serious challenge to its normality.
In economic terms, slaves were goods like any other. Accordingly, Polybius (4.38), when listing the best and most numerous goods exported from the Black Sea into the Mediterranean world, could list slaves with items such as preserved fish and hides. Moreover, in distinguishing between necessities and luxuries, he located slaves in the former category. Quality and quantity varied from place to place, whether for hides or for slaves. The long and notoriously dangerous voyage to the coasts of the Black Sea will only have been attempted if large profits were to be made. This in turn meant that slaves and other ‘goods’ from the region must have been available in volume and at a unit price advantageous to the trader. Similarly, cabotage required that the price of a slave at source was low enough to allow for a series of profits to be taken.
This was overwhelmingly a private matter. States might decide to purchase slaves for public functions, as with the ‘Scythian archers’ bought by Athens in the mid-fifth century bc to enforce order under the command of the democracy’s officials;182 but the numbers of public slaves were statistically insignificant and were in any case probably obtained through private markets and personal arrangements. Although some states had a particular concern for maritime control and taxation, none of this traffic was in the hands of states, or even of large corporations. Instead, the trade lay with individual entrepreneurs, who duly crop up in the law-court speeches that arose from their activities. Such men might have had personal penchants for particular goods, but we hear rather of mixed cargoes, attested both in texts and through archaeology from the archaic period onwards.183 The archaeology of Greek slavery is of course a difficult affair and can seldom throw light on the slave supply.184 But slaves were a high-value, low-bulk (even self-transporting) commodity, as we shall see, and could well be part of the same exchange as with other commodities (wine, for example, or salt or even fine-wares).185
Exchange was at its starkest on the periphery, in the marginal zones where Greeks encountered significantly different cultures. Strabo (11.2.3) describes the exchanges between pastoralists and the ‘civilised’ Greek world which took place at Tanais near the mouth of the Don, at the north-east corner of the Sea of Azov:
Tanais was the shared trading-centre of the Asiatic and European nomads and those who sailed there from the Crimean Bosporus. The nomads brought slaves and hides and whatever else nomads have, while the others brought for exchange cloth and wine and the other goods which go with a civilised lifestyle.
Here slaves are among the few ‘products’ of the non-Greek world which were desirable to Greeks. The fact that Strabo calls attention to the provision of slaves and hides at Tanais seems to echo Polybius’ notice on the export of goods from the Black Sea region as a whole. Wine is a principal item of exchange for both authors, and although we can hardly presume to generate statistics, a strong pattern emerges. The peoples of the region (both Greeks and non-Greeks) had a thirst for wine that could only be met by substantial imports. Archaeology shows how substantial those imports were and how embedded Greek wine had become in the ideology and practice of local non-Greek elites by the sixth century bc — for Greek wine soon became a regular feature of the grave goods of major local burials. Indeed, we probably underestimate its penetration and distribution: much was evidently carried in skins, not cumbersome pottery.186 But unless simply seized it had to be paid for, whether by services or exchange. In large part, that meant skin, whether hides or slaves. In other words, the wine trade in the region reflects the trade in slaves. As Gavriliuk (2003; cf. 1999) has recently observed, it was the export of slaves (and hides) that drove exchange on the north coast of the Black Sea far more than the much-vaunted export of grain.
Furthermore, it would be rash to assume that the non-Greeks gained a good price for their slaves and hides. Greeks who had made the dangerous voyage from the Aegean sought exorbitant profits, and the history of exchange tends to suggest that the pastoralists would have been overimpressed by the wares and trinkets proffered by traders from the ‘civilised world’.9 Polybius’ observation that the Black Sea takes ‘every kind of wine’ encourages further scepticism, as do complaints among the Greeks of Olbia about the ‘cheap rags and foul wine’ that traders bring there from the Mediterranean (Dio Chrysostom 36.25). At the same time, we must consider the effect of a ready market for slaves at the coast on the societies of the hinterland. Chattel slaves were not of enormous use to pastoral societies, but the development ofGreek communities and trading posts on the coast from the seventh century bc had made them extremely useful as a means of acquiring luxury goods, however tawdry. Accordingly, there seems every reason to suppose that the slave trade at the coast served to generate instability and conflict in the interior.
Greeks and barbarians: ransom, booty, colonialism
War had a power to change everything quickly and totally in the Greek world. Accordingly, when populations went to war, they did so in the knowledge that the outcome could very well entail enslavement, whether of the individual, part of the community or even the whole population. The main issue was who would win, which at its most extreme meant who would have the power to enslave whom.
However, the horror of potential enslavement was not without some hope of evasion. For within Greek culture the treatment of the defeated was a major ethical issue in its own right. While there is scant sign of
Deep concern about the fate of defeated ‘barbarians’, the victor had more difficult decisions to make about defeated Greeks. Enslavement was a real option and was carried out, but the victor paid a price for that in the shaping of his reputation, which might well be taken to outweigh any financial gain, vengeful satisfaction or warning to others. We should take seriously the ethics implied in the tradition that Agesilaus’ troops were driven to unstoppable fury by their discovery that the people ofLampsacus had gone so far as to consign to their mines the Greeks whom they had taken as slaves. Accordingly, we know of no Greeks among the slaves in and around Laurium.187
The famous Athenian debate over the treatment of defeated Mytilene in 427 illustrates the dilemma well enough, though it is wholesale execution that is primarily at issue (cf. Thucydides 3.47). Enslavement had been an obvious option from the first (Thuc. 3.28). And it is most unclear that the women and children were ever to be executed, for the focus is very much on the men. In the same year, the Spartan forces took the city of Plataea: while Thucydides (3.68) gives some detail on the execution of many of the men, the enslavement of the women and children is mentioned only in passing. Similarly, the set-piece debate between Athenians and Melians, which prefigures the defeat of Melos, is centred upon the issue of enslavement. From the Melians’ standpoint, the issue is the de facto collective enslavement that they will suffer if they kowtow to Athenian imperialism as against the uncertain (and at once horribly predictable) outcome of resistance. Their defeat means the execution of the men whom the Athenians catch and the actual enslavement of the women and children. There is a pattern here. When cities fell, there was a recurrent tendency for the victor (even when dealing with Greeks) to kill the men and enslave the women and children. At the same time, the discourse of a more metaphorical enslavement bolstered resistance to imperial control and made the decision to enslave all the more charged. The Melian enslavements confirmed the negative image of Athenian imperialism as one of the enslavement of Greeks, not of their liberation. Accordingly, the discourse of enslavement and liberation remained a powerful factor in the ideological context ofimperialism in the Greek world well into the Roman period.
Victors might well choose to do otherwise than enslave. Indeed, the lack of Greek slaves in the historical record for classical Greece suggests that Greek victors usually preferred another course.11 A major example is offered by Xenophon, who shows the Spartans resisting the calls of their allies to enslave the city of Athens, finally defeated after a long and gruelling war in 404. The decision illustrates very well that the victor had choices to make, which would reflect also upon the victory and have consequences for the future.188 For while it might be generally accepted in the abstract that the persons and property of captured cities belonged to the victor, the more important question was how the victor chose to use or abuse his dominance. It was the particular decision that really mattered, both to the parties concerned and to external observers. It is not so much law as ideology that matters in the relationships between master and slave, whether in the dramatic context of a captured city or in the everyday experience of slaves in general.189
As Xenophon (Hellenica 2.2.20) tells it, in 404 ‘the Spartans refused to enslave a Greek city which had performed a great benefit amid the greatest perils that had fallen upon Greece’. What Xenophon’s Spartans really mean here is the Athenian role in defeating the Persians earlier. On this view, enslavement was for barbarians, but not for a Greek city which deserved so much from other Greeks. Clemency of this kind was a powerful strategy: it was much easier and more convincing now for the victor to erect a triumphal monument at the Pan-Hellenic religious centre of Delphi. And Xenophon himself most certainly approved: his favoured Spartans had shown magnanimity towards his own city with a rhetoric of implicit enmity towards the Persians against whom he personally waged war. More generally, the fact that the inhabitants of a defeated city could hope to escape the most severe consequences, including enslavement, must have done something to mitigate the communal terror of outright war. If defeat was even envisaged, then it might be endurable. However, warfare between Greeks could still generate slaves. An Athenian man taken as a child during the last decade of the Peloponnesian War was sold into slavery far away on the island of Leucas. We know of his fate only because he was ransomed and because his son became embroiled in a court case at Athens (Demosthenes 57.18—19). Other specific instances happen to be known through the survival of inscriptions, in particular honouring those who arranged and paid the ransom: what better beneficence could there be?190
Ransom was a major brake on the slave supply, but also a very profitable strategy for the enslaver. We may be sure enough that around the periphery of the Greek world there were substantial prospects of ransom, so that both Greeks and barbarians captured there might be bought back by relatives or others interested in them. And even when barbarians had left the periphery and been taken deeper into the Greek world, they might still be able to arrange a ransom, as did the Thracians who were travelling back on the boat which saw the death of Herodes (Antiphon, On the Murder of Herodes, 20). Meanwhile, especially in view of the ideological difficulty in the holding of Greeks as slaves, we may also be sure that many Greeks were ransomed: in their case communication was all the easier in every regard. Of course, almost everything depended on the wealth of the enslaved individual, or at least his access to wealth. The case of Nicostratus illustrates the process and economics well. Taken by pirates in the fourth century, this Athenian was sold into slavery close to home, on Aegina. He had written to his brother, who could not or would not raise the funds needed to bring off the deal and collect him. The ransom was reportedly 26 minae, that is roughly ten times the average price of a slave. Captors and subsequent purchasers had a powerful economic incentive to take on the trouble of ransom.191
The profits of ransom apart, Greek victors had much less reason to decide against the enslavement of defeated barbarians. The argument that had saved Athens in 404 augured badly for any Persians subsequently taken. Accordingly, when the Spartans set about the invasion of Asia Minor, they did so with the rhetoric of the liberation of its Greeks, but with the strong prospect ofthe enslavement oftheir barbarian enemies. And these were not simply Persians. A striking feature of the campaigns of the early 390s is the repeated Spartan decision to plunder other ‘barbarians’, who might better have been conciliated against the Persians: Carians to the south-west ofAsia Minor and Bithynians to the north-west. These were not so much acts of war against the Persians as attempts to gather booty from barbarians, with whom local Greek communities tended to have awkward relationships. Chief among such booty were slaves.192 And there was nothing new in the Spartan mix of state-supported war and plunder for gain, perhaps on a more private basis. In 409, for example, as Xenophon (Hell. 1.2.4—5) himself tells us, an Athenian force had got into difficulties while raiding the Lydians for slaves and other goods. As for the Bithynians, Xenophon himself (Anabasis 6.6.38) led his own mercenary army on a detour through their lands, specifically to garner booty consisting in part of slaves.
For Greeks the enslavement of barbarians was not problematic. The well-known discussion of Aristotle in the first book of his Politics describes the distinction explicitly (cf. Peter Hunt’s chapter in this volume). Aristotle insists that the acquisition of slaves is a key task of the master, for whom they will serve their purpose as living tools, active pieces of property (Pol. 1253b, 1255b). The just acquisition of slaves is like a ‘science’ of war or animal-hunting (1255b). However, he is much less clear about the justice of enslavement through war, reflecting the dilemma of Greek enslaving Greek seen in the Spartan stance in 404 and elsewhere. Justice depends on the causes of the process of enslavement (notably the just cause of a war), while it is barbarians not Greeks who are slaves by nature (Pol. 1255a).193 In the Politics the institution of slavery is fundamental to the household and thus to human social organisation, so that the acquisition of slaves is no small matter. This concern takes Aristotle from slavery into a broad study of trade — a theoretical discussion substantially of the slave trade, couched in the discourse of property and exchange. That in turn raises a question to which we shall return: if the household is to be as self-sufficient as possible, what of the supply of slaves?
At a more empirical level, war made slaves and slaves in turn featured prominently among the spoils to be gained through war. Yet students of antiquity have been reluctant to explore the probable corollary, that is whether campaigns were launched with the specific purpose of acquiring slaves and the profit that came with them. If so, slaves and the slave trade were not simply a by-product of war, but even its objective. Garlan (1987: 10) simply asserts that ‘Greeks are not seen going to war in order to increase their stock of slaves.’ And yet, while it is true that accounts of the causes of substantial wars in the Greek world do not focus on the acquisition of slaves, so much as on higher and juster causes, it is also true that the booty to be gained by a war was a substantial consideration, for the state as for the individual (cf. Thuc. 6.15.2).
Xenophon’s detour among the Bithynians exemplifies the tendency for commanders to enrich their men (and themselves) by slave-raiding. Certainly, it might be argued that this was more an act of banditry or piracy than an act of war,194 but the distinction would be a fine one. The supply of slaves was a benefit of imperial power, to be included with the other goods available by trade or seizure. Accordingly, in the 420s when the comic poet Hermippus (Fr. 63) lists goods flowing into imperial Athens in his Basket-bearers, slaves are included among them.19 Throughout it must be remembered that slaves were one kind of commodity circulating with and exchanged for other goods such as wine that were of a completely different order. At the same time it is Athenian imperial power which makes all these goods available, so that the connection between war and the slave trade is implicit, Later, Aristotle (Pol. 1257b) builds war into his discussion of slave acquisition to argue that the art of war should be directed not only against wild animals but also against natural slaves who refuse to accept their natural subordination.195 In essence, it seems to be Aristotle’s position that war is properly waged by Greeks against barbarians for the purpose of acquiring slaves. There is no indication that such a position is controversial, and Xenophon’s account of his own activities in Bithynia confirms well enough that it was not. Theory and practice coincide.
Moreover, the small scale of Xenophon’s raiding demands attention. Small-scale conflicts and raids were far more usual between Greeks and barbarians than great wars. Around the periphery of Greek settlement, there was a tendency to conflict between Greek communities and local barbarians which, taken together, will have generated a significant number of barbarian slaves new to the Greek world. While we cannot pretend to have viable statistics, there cannot be much doubt that this kind of low-intensity warfare brought barbarians into slavery in the Greek world. Similarly, ever-present banditry and piracy also contributed substantially to the pool of slaves.
Slaves were a routine part of the booty of warfare.196 They are often listed with livestock and other goods seized from the enemy. And it is perhaps worth adding that also in Greeks’ wars with Greeks, barbarian slaves were again a key part of the booty: these were not additional to the slaves available in the Greek world, but redistributed out of the looted community. Once acquired, booty was usually sold, though particular items might be kept. So too with slaves: sale (or ransom) was the norm, not least for practical purposes. Not only did slaves need to be watched, fed and marched about, but they also presented other problems. One of Xenophon’s men, for example, had only limited use for an adult and hostile male Bithynian or Thracian. Meanwhile, it is worth stressing that traders (in slaves and other goods) would also turn their hands to raiding when opportunities arose: the familiar categories of soldier, bandit-pirate and trader might be applied to the same individual or group, depending on time and place.197 Xenophon’s men are a well-documented example of this merged identity. When Xenophon’s force demanded to be able to trade for supplies with the cities of the north coast of Asia Minor, from Trapezus westwards, by what means do we suppose that they expected to trade? After their long and embattled march from Cunaxa, they had little else than the booty they had acquired. That included slaves, who at least, for all the difficulties of maintenance, had the virtue of carrying themselves along and no doubt portering other goods as well (cf. Menander, The Shield 155).
And what of ‘colonisation’ in all this? Xenophon and his men were potential colonists as well as soldiers, slavers and slave-traders. For the archaic period it is no longer usual to worry whether colonies grew up on the basis of agriculture or trade, for we now understand that any settlers will have maximised the economic advantages of any circumstance and will therefore have engaged in both. Yet it is worth taking the argument a little further. The trade surely entailed trade in slaves, for, as we have seen, local peoples might well have little else to offer in exchange for the imported goods they desired from the Greek world. That can only have encouraged the production of more slaves in the region, whether through raiding or some other exploitation of the weak.198 Meanwhile, the military advantages which settlers seem to have had, at least when in concert with some part of the local population, also made slave acquisition a likely prospect. Finally, it is not unreasonable to ask whether the availability of slaves, not only for purchase but also for capture, was a significant attraction for Greeks who settled around the periphery of the Greek world. On Thasos, for example, Archilochus harps on warfare with Thracians, probably in satirical vein. But was that warfare not simply a problem in the colonising process, but even part of the attraction for settlers (alias slavers and traders)? Thracians might indeed pose a serious physical threat, but they also represented a profit that could be taken. Their bodies were an important part of the booty available.
Finley (1980: 85) objected that Greek slave-raiding of this kind does not square with the fact that Chios and Corinth, though cities picked out by the sources as centres of slave-dealing, were not conquering or imperial states. However, Corinth may indeed be seen as such a state by virtue of its colonialist and indeed imperialist activities, both in Magna Graecia and up the Adriatic coast. Many of the Illyrians known to have been sold as slaves into the Greek world presumably came via Corinthian interests in that region. But the main weakness of the objection is precisely that cities with large slave markets did not need to carry forward colonialism in their own right in order to derive benefit from the process. Such is the case with Chios, which had the dubious honour of being thought the first Greek state to have invented the institution of chattel slavery. The tradition is intriguing in that among the earliest known imports into the Black Sea region was wine from Chios, found there from the late seventh century bc. Once again, wine for slaves and slaves for wine. Here the process is all the more suggestive because it seems to fit very well the hypothesis that the slave trade was important to Greek colonialism. It does not matter that the resultant colonies did not see themselves as Chian foundations: we need not even suppose that the wine and slaves were traded by Chians, but merely that Chios was a principal counterpoint (together with other places, notably Lesbos) for this trade. In short, there is no necessary direct linkage between a city’s reputation for slave-dealing and its activities at the periphery, though such linkage can certainly occur. Accordingly, the objection has no real bearing on the fact that Greeks did engage in recurrent conflict with local barbarian neighbours at the periphery (as also with neighbouring Greeks), nor upon the strong likelihood that such conflicts produced slaves for the Greek world. More broadly still, the involvement of any state in the process of foundation was usually (and perhaps always) preceded by individual contacts and experiences, which tended to be overwhelmed by the traditions that developed around colonial settlement, not least in their concern for an appropriately grand origin-tale.199
Meanwhile, although we may speak of‘slave-markets’ to mean the slave trade in general, it is all too easy to overstate the prominence of particular places as markets for slaves in the Greek world. There is in fact very little evidence that particular Greek locations had especially busy markets in slaves before the Roman period. Diversity seems to have been more the rule: slaves were marketed everywhere in Greece. So too within each community. There is no real indication at all in our sources that in Attica, the region we know best, there were specific slave markets or special slave-sales. In Athens, where the main agora has been studied most intensively, it seems that slaves were traded along with other goods, and that only particular areas within markets might be used for slave-selling, rather as other goods tended to be gathered in other areas. At Athens we are told of ‘the circles’ where slaves were exhibited for sale, possibly with other domestic ‘items’, raised up on tables.200
It may be concluded that the slave supply to Greece flowed in large part from a reservoir of slave labour which could be obtained from the periphery of Greek culture. Warfare on a large scale was no doubt significant from time to time. Greek victors had reasons to be slow to enslave defeated Greeks, but they had ample scope for enslaving barbarians in occasional major conflicts — in the Persian Wars and at Eurymedon, for example, in the fifth century (cf. Diodorus Siculus 11.62). However, the slave supply came more regularly and, as it seems, in larger numbers from relationships between Greeks and non-Greeks at the periphery. Certainly, there was recurrent conflict at this periphery, which fed the supply. But there were also more peaceful relationships of collaboration and exchange. While Greeks settled at the periphery might seize slaves, they were also placed very well to trade for them. For the barbarians, the Greek demand for slaves, together with their own elite’s demand for Greek goods, constituted a powerful drive for exchange in slaves — especially because slaves were among the few commodities available there that were wanted by Greeks. Local elites had a special interest in raiding for slaves, or simply buying them in from further inland. There can be little doubt that it was not so much slaveholding as slave-trading which drove the process of wealth concentration and social stratification observable at the periphery. Moreover, the development of strong hierarchies gave local elites another source of slaves to trade with the Greeks: if all else failed, they could sell some of their own subjects.201 Meanwhile, the availability of cheap slaves and other goods can only have encouraged Greeks to come and to settle at the margins. However unpalatable the thought may be, there is not much room for doubt that the slave supply was a significant factor in driving the extended process of Greek colonial settlement.
Everyday exploitation: being bought and sold
Slaves were not all the same. Ethnicity, age and gender, as well as physical and intellectual capacities, varied enormously. We may be entirely sure that in the buying and selling of slaves these differences were of fundamental importance. It is enough to look at the description of slaves inscribed c. 414 in the list of property confiscated from Athenians convicted of mutilating statues of Hermes and sold by the Athenian state. There we find mention of forty-five slaves. They appear as property, alongside land and animals: the list is concerned particularly with their market value. Prices are given for each slave. Further, slaves are listed with their ethnicity, gender and, in some cases, special skills (notably a Carian goldsmith). Two children are picked out. The ethnic spread is worth observing: of the 35 slaves whose ethnicities have survived, 12 are Thracian, 7 Carian, 3 Scythian, 3 homebred, 2 Syrian (if not ‘white Syrians’ from Pontic Cappadocia; cf. Strabo 12.3.9), 2 Illyrian, i Colchian, i Lydian, i Phrygian, i Cappadocian (or another Illyrian), as well as i Macedonian and i Messenian (?). The general absence of Greeks is notable. The dominance of the north in our small sample may be significant, for it is borne out by the other evidence we have, particularly in contemporary comedy and later in the clear statement of Polybius, on the situation in 220, that the most and best slaves came from the Black Sea region. At the same time, however, there is also a significant presence from Asia Minor, especially Caria.202 Not that we can assume that the ethnicities (whether explicit or implied in the name) are completely reliable: it is salutary to recall Daos from Phrygia (Men. Asp. 242, allegedly), whose name would usually be taken to indicate an origin towards the Danube. Further, quite apart from distortions in the process of sale, the application of ethnics to slaves may result from a range of factors, including place of sale into the Greek world.203 Nonetheless, Syrians are expensive: they were from further afield, perhaps more rare and considered more talented than the average. The Carian goldsmith is most expensive, surely by virtue of his skills. Meanwhile, there is only a limited sign of the tendency, often suggested at a theoretical level, to have slaves from a variety of origins: the best-attested group, belonging to a metic in Piraeus, consists of seven or eight small ethnic units, though it is unclear whether he or anyone else conceived of them in such a fashion (see T. E. Rihll’s chapter in this volume). Certainly they are not listed according to ethnic grouping in our inscribed record. If we knew the metic’s own origin, we might have some deeper insight into the particular composition of his slaveholdings.
As for prices, the main point must be that these slaves were costly commodities. It is not hard to see how slavers could turn a fine profit, even if the slave had initially to be bought and passed through many hands on long journeys from Thrace, Colchis or Syria. For the average of the prices listed, taken in whatever form, falls in the range around 155—175 drachmas. Importantly, that figure is not much out of line with known prices of the fourth century, though it may be a little low. State-sale may not have been the best way to achieve the highest possible price, especially in the peculiar conditions of the confiscations which may have carried some sense of religious taint for some potential buyers. But our evidence strongly indicates that the major impact on pricing was not ethnicity but the skills which a slave could offer.204 The point is clearly illustrated from Thracian males. The skilled Thracian overseer Sosias is said to have been bought by Nicias for a huge sum (a talent = 6,000 drachmas), probably by virtue of skills he had acquired in the mines of Thrace. By contrast, Thracians bartered into slavery in exchange for salt were proverbially cheap and therefore suitable for a short and brutal life of hard labour, for example in the mines under the likes of Sosias, whose price was equal to some forty of these lesser beings.205
Estimating the regular payment of a free skilled labourer at this time as around one drachma per day, we may conclude that a slave with no particular skills would cost something approaching half a year’s pay. This was no trivial amount, especially when taken together with the need to feed and otherwise maintain the slave. Although the initial investment and maintenance costs might be offset by any income generated by the slave’s work, there was still a significant sum to be found. Small wonder that ancient texts (e. g. Arist. Pol. 1252b)206 allow the possibility that a man might not be wealthy enough to own a slave.
Moreover, cost provides a context for the presence in our list of homebred slaves. It made sense for owners to have slaves bear children, even allowing for the cost of rearing and no doubt grim rates of infant mortality and death in childbirth. The younger child in our list has a price of only 72 drachmas, while the older child bears an adult price of 174 drachmas: the lower price presumably allows for the extra costs and risks of ownership set against the capacity of the child to work.
There is every likelihood that we underestimate the extent of slavebreeding in Greek society. Apart from the economic advantages, there was also the brute fact that the master had sexual access to his slaves. It is enough to consider Xenophon’s account (Oeconomicus 10.12) of an ideal household where the master will prefer sex with his wife to sex with a slave, in part because the slave has no choice but to submit to his demands, being a woman ‘compelled to give service’. Beyond his own whim, the only control on the master in this regard was the attitude of his wife.207 Children surely resulted, as also from the deliberate breeding of slaves. Xenophon presents slave-breeding as an entirely unremarkable practice. The only point he seeks to stress is the desirability of control in the matter: the sexes should be kept apart in their respective quarters, so that well-behaved slaves should be bred with other good slaves, resulting in still better-behaved offspring. The general rule is stated also in the negative: bad slaves who breed with bad make still worse children. On that logic, there is little to be gained by breeding slaves who are not ‘good’. Accordingly, the home-bred slave is commonly imagined in antiquity as especially loyal and trustworthy. Meanwhile, there is also an implied reward for good behaviour: sex and reproduction. At the same time, we may reasonably wonder how many masters kept the kind of close control of their slaves’ sexual activities that Xenophon’s account recommends. In any event, we may presume that slaves found opportunities for exercising their own initiative in the matter, resulting in unplanned offspring who would enhance the stock of slaves.
That slave-breeding was a commonplace is further illustrated by a rather neglected passage of the Hippocratic Airs, Waters, Places from c. 400. The author first stresses the low reproduction rates of Scythians in their native land and explains it, on the female side, by the fatness derived from their inactivity. He triumphantly observes:
And the (Scythian) slave-women offer substantial testimony to this. For they cannot come into contact with a man without falling pregnant, on account of the hard work they do and the leanness of their flesh.
His point is that by being put to hard work, Scythian slave women cease to suffer from the restrictions on their fertility which result from their lifestyle and location. For the medical writers, the slave trade offered a ready insight into a range of problems, which include the impact of location on health and wider questions about the bearing of children (cf. Diseases 2.4.5). Indeed, their writings offer a series of intriguing insights into all aspects of slavery, showing, for example, that even a sickly slave woman with multiple symptoms might be bought and soon taken in search ofa cure and restored menstruation. Presumably she had been sold cheap as damaged stock but was now restored and able to reproduce. Or perhaps she had been passed off as healthy, for the suspicion of fraud which hung over all exchange in the Greek world was a particular concern in the purchase of slaves, whose health and history might well be concealed. That is why Plato took care specifically to design legislative controls on the sale of slaves in his Laws (9ioa-c).208
Childbearing, and wet-nursing, were evidently commonplace in the experience of female slaves. What percentage of the slave stock consisted of home-bred slaves is beyond our knowledge on the meagre statistics available: around i0 per cent has been suggested, but if childbearing was so common, the figure may well have been higher.209 And at its highest, we may presume, for domestic slaves, for it was in the household that masters could make most use of the loyalty, perfect Greek and other talents of such individuals. It was surely only the most unfortunate of home-bred slaves who would find themselves sent to the mines. As for how many were marketed under normal circumstances, we have no figures at all, but their perceived virtues suggest that they were less likely to be sold than other slaves.
More usually, however, slaves purchased in Greece had come from alien cultures around the Greek world, especially from the north and from western Asia Minor. A few others came from further afield. Black slaves seem to have been comparatively few and were presumably prized all the more for their rarity. The fact that the makers of Greek terracottas had the fancy to depict black Africans affects the issue only in that it may mislead.35
It is hard to imagine the experience of these individuals. Some slaves had been seized violently, having seen their communities ripped apart and their families slaughtered, whether by Greeks or by barbarians in the hinterland. Those with wealthy friends to call upon might be ransomed and return home: ransom was certainly frequent enough. The poorer and less well-connected had no recourse, except the risks of running away. For that reason, they might well be penned up, like sheep, or shackled and marched long distances. Those who could not keep up would be abandoned or killed as a warning to the others. They might well be loaded with other goods, like beasts of burden. Throughout, they were entirely at the mercy of their captors. Meanwhile, others had been sold into slavery by their own families, who may even have hoped that a better life for their children would follow among the prosperous Greeks. Herodotus (4.95) relates the story of Zalmoxis, which was told by the Greeks of the Hellespont and Black Sea coast: he was a former slave of Pythagoras who had been sold out of Thrace in his youth and subsequently returned to his native land as a man of wealth, knowledge and consequent power. The historicity of the story matters much less than the potential of its myth: here was a model, related by the Greeks of the region, which might encourage the idea that children sold out of Thrace were off to a better life and might even return one day to show their success. Where such misplaced optimism was not the driver, raw economic and political pressures, whether drought, food shortage or simply the exploitative desires of the local elite, can only have encouraged and expanded this kind of trade. This may be seen as a collective counterpart to the abandonment of children by the poor and disadvantaged within Greek communities, on which we are told very little by our sources.
Physically, the newly enslaved were wholly at the mercy and disposal of their masters, to whom their bodies belonged. Rape was likely, as were all kinds of other abuse, for men, women and children alike.210 Some slaves were marked, largely by tattooing but perhaps also by being branded like livestock: our texts are sometimes unclear about the precise method used, but both kinds of marking were painful. They are also unclear about when marks were applied: in some cases on enslavement, though there seems always to have been an element of punishment entailed.211 Slaves so treated were often marked on their foreheads. However, that practice is to be distinguished from evidence for the tattooing of Thracians (male and female) over much of their bodies, which seems to be part of local social practice. Even so, it remains unclear whether particular marks were also applied to Thracian slaves, perhaps by other Thracians. Were they marked in a special manner as they were sold from the Danube to the Black Sea coast or reached the sea at Pagasae in the north-west Aegean in the hands of Thessalian slavers, from whom traders would take them on for sale in Athens? Perhaps.212 Yet Thracians were not alone: others are credited with tattooing as a social practice, including Illyrians, who were certainly traded into slavery in Athens (Strabo 7.5.4). Meanwhile, we happen to hear of marked Persian slaves too. That barbarians might be marked follows from the fact that sometimes, under circumstances of particular bitterness, Greeks might even mark fellow Greeks as a punitive measure (Plutarch, Nicias 29; Pericles 26). However, there is no reason to suppose that marking was routinely done to slaves. It would hardly enhance their market value.213
Castration might also be attempted, though it required a certain skill: the apparent enhancement of value presumably offset the cost of the procedure and the likely deaths, while the satisfaction of a score settled may have been more important still in isolated cases.214 Meanwhile, other barbarian slaves kept within Scythian society were blinded: they could not run away but could perform their repetitive labour well enough without vision (Hdt. 4.2).215
Mentally, all was confusion and desperation, especially once the newly enslaved had come into the hands of Greeks and others whose language they did not understand. It is small wonder that barbarian slaves are often portrayed in Greek texts as stupid. How were they to understand the commands that were given them? The climate of terror is stupefying, rendering initiative a hazard. It was wise not to be too clever, so that Greeks might well consider that ‘Zeus removes half the mind of men when they are taken into slavery’ (Pl. Leg. 777a). We may recall the Greek anxiety about clever slaves expressed, together with satisfaction over slave stupidity and slave punishment, in Greek comedy of all periods. Meanwhile, some slaves had talents which might give them hope for a better life. Special skills enhanced the value of the slave and therefore the price. That in turn offered some slight protection against abuse: the slave was a commodity, but the skilled slave was a relatively valuable commodity. In addition to all kinds of craft skills, we must consider also entertainment of every kind, some more wholesome than others. Xenophon’s mercenaries proudly displayed to a Paphlagonian ruler a dancing-girl who could do a party-piece. We are not told of her subsequent fate (Anab. 6.1.13).
By the time the newly enslaved reached the sea, the main highway of this world, they may very well have been sold several times already. At the coast, they could be sold on in any direction, including to another part of the periphery: we happen to know of a slave with the non-Greek name Phaylles (SEG xlviii. 1024)216 bought at Olbia in the north-west Black Sea and sent on to Phanagoria, in the Asiatic Bosporus to the north-east, despite the substantial numbers of slaves bought and sold more locally there. For we happen to be told of a substantial slave trade to the east of the Crimea. There slave-raiding was an organic part of an otherwise impoverished economy: the neighbouring Bosporan kings might choose to provide a market for the slaves thus taken, presumably in order to derive taxes and other benefits from the trade (cf. Strabo 11.2.12). We should not suppose that the slave trade was only centripetal, running on Greece and the Aegean. There were markets everywhere and many routes to market.
Contrast, for example, the story related by Aelian, in which the trader Dionysius bought a Colchian girl from a neighbouring people of the eastern Black Sea. He was an experienced and greedy trader, but not specifically a trader in slaves. Dionysius cheated on a deal to ransom her to her family and instead put her on sale in the market on Chios. The story’s pattern is realistic enough: we may be sure that Colchian girls did find their way into the Greek world in such ways.
Greek owners sometimes gave their slaves Greek names.217 Yet slaves might also be known by an ethnic formed into a name: ‘Thratta’ (‘Thracian female’) is often found, whether or not the bearer was strictly Thracian in origin. Or a name associated with their supposed place of origin might be chosen: a Thracian might be called Daos, or a Phrygian Manes, for example. Appearance was another prompt to naming, easy for the master to remember: for example Xanthias (‘Blondy’). Such names show little concern for the individuality of the slave, while the slave’s response to them can only be guessed at. All the more so, when names were changed, according to the master’s fancy, upon purchase or resale perhaps.218
An adult male taken into slavery with no special skills might well find himself allotted to hard and repetitive labour, whose mixed blessing was no doubt a short expectancy of survival. If not the mines, then there were other gruelling processes: for example, it would be good to know more about the men who slaved in quarries. It is unedifying, but perhaps telling, to realise that a whole comic play of Aristophanes was set in a flour mill worked by slave drudges, whether or not they walked the treadmill itself.219 Slaves, especially cheaper ones, were ripe for mindless hard labour: a comic Thracian is so deluded as to take pride in the fact that the mills are full of his people, the ‘salt-bought’ no doubt (Men. Asp. 245). Other adult males might be worked in the fields, perhaps in chain-gangs or, if they were lucky, in some rather more liberated capacity, as herdsmen for example. Only men with skills might have the dubious pleasure of working in the heat and danger of an arms-factory or supervising at the mines.
By contrast, the adult female had a wider range of occupations, including childbearing and wet-nursing but also with a range of domestic functions which varied in their demands. A female slave might well be bought specifically to perform work which could be sold at a profit, with other female slaves or she herself taking it to market. We are told (Aeschines 1.97) of such a woman inherited by Timarchus, who had the special skill of making the finest textiles and who herself sold the goods she made.220
The violence and confusion of enslavement must have been all the more disturbing for those enslaved as children. They might have no idea of their origins at all, like the adult Macronian with Xenophon who only discovered his identity when passing through the land of the Macrones, an unlikely return to his homeland (Anab. 4.8.4—7). Subsequently, as we have seen, children might be trained for any work, but the master’s investment in their upbringing would tend probably to give them a less painful future by virtue of the skills they were made to acquire and perhaps a bond that had formed over the years. We may speculate that the individuals who turn up from the late sixth century as potters and painters of Athenian fine-wares and have names redolent of slavery (Skythes, Kolkhos, Lydos, (Egyptian?) Amasis and others) had been enslaved young and taught their trades from an early age. But what of the anonymous labourers who prepared their materials, or those who made the coarse-wares and tiles that dominate the archaeological record for ancient Greece?
When the master of the house went to market for a slave, there is little sign that he gave serious thought to the slave’s experiences except in so far as they affected the price and the particular kind of role which he had in mind for his acquisition. However, a slave was a big-ticket purchase, so that real care must have been taken in weighing slaves’ prices against their various potentialities for work and even a return on the investment in terms of profit from their labours. Presumably, especial care was taken in the choice of slaves who were to work as servants in the house, as many
Did.221
But that was scant comfort to the slave, whose individual personality meant nothing beyond the tenor of his or her character, as in the purchase of an ox or mule. Practice in the markets seems to have varied. The adult male slave might be displayed naked to the shopper’s gaze, perhaps on a raised dais, as apparently at Athens. And the slave was made to exercise, presumably to show physical condition (cf. Men. Fr. 150). In Menander (Sicyonians 7—16) we find a girl of four kidnapped from the coast of Attica with her male slave. They are taken for sale to Mylasa in Caria, where many slaves were marketed, but the market practice described for Mylasa presumably reflects the Athenian usage familiar to Menander’s immediate audience. No doubt shoppers’ attention was caught by the particular individuals on sale and their appearance, but overwhelmingly exchange of this kind was everyday routine. In any case, the market was full of slaves going about their business, whether buying, selling or simply passing through. In all likelihood, as with other commodities, a proportion of those engaged in the buying and selling of slaves were themselves slaves.
By contrast, most of the slaves on sale had not seen anything like it. Athens was a famously dazzling city for the outsider new in town: how much more so for newcomers who had been brought on journeys like theirs? They could only hope for the best. One of the few things familiar to them was the institution of slavery itself, but its very normality, encompassing slave, seller, purchaser and owner, tended to obviate any critical engagement.
Meanwhile, those slaves who emerged successfully could find a voice, albeit small and mediated, in their epitaphs. In particular, a scatter of epitaphs and other inscriptions have survived in south-east Attica around the mine-workings there.222 It seems likely that the few who had memorials were like Sosias: the ‘salt-bought’ mine-fodder were far removed from their exalted status. One unusually extended epitaph (IG ii2.ioo5i), from c. 350—300, takes a Paphlagonian called Atotas back to his roots, omitting only (and tellingly) the manner in which he had lost his freedom. This evidence of a slave’s retained connection with his homeland is all the more significant when we recall the massive dislocation and cultural rupture of enslavement and de facto exile far from home.223 And the epitaph is an ambitious composition, which uses grandiloquent and punning couplets to place the deceased in the glorious and mythical context ofthe heroic age. There is something pathetic and perhaps even humorous in the attempt, but also a sense of optimism.224 Atotas had retained his association with his distant homeland and (if we can assume that he shared the sentiment of his epitaph) found more than a measure of personal pride in that origin:
ATOTAS, MINER
From Pontus Euxinus, Paphlagonian great-hearted Atotas, In a land from which, my body rested from toils.
In skill no-one supplanted me, from the family-tree of Pylaemenes am I, who, mastered by the hand of Achilles, met his death.
Atotas had probably come into slavery with skills of a miner, which would account for his apparent prominence at Laurium. He may even have cost as much as Sosias the Thracian. The importance ofthe skills he had gained in Paphlagonia helps to account for the pride that his epitaph shows in his work: there was potential satisfaction in that. It accounts also for the grandeur of his claims, namely that he was no less than the scion of the royal family, linked to the most famous Paphlagonian of them all, Homeric Pylaemenes. And yet there is no mention of his father: perhaps the commissioner of the epitaph did not know his name, perhaps Atotas found more respect in leaping from banal realities to heroic myth. While we are given no hint as to the real master of Atotas, we are invited to think of the pair after the manner of Pylaemenes ‘mastered’ by Achilles here. That was good for Atotas, no doubt, in death as perhaps in life, but it was also good enough for Atotas’ real master who may have honed the epitaph. In reality, as the epitaph also acknowledges, Atotas had died after a life of toil far from his Paphlagonian home.
Bibliographic essay
There is no book-length treatment of the Greek slave supply. Traditionally, this issue has been considered as a preface to general studies of slavery in Greece, among which Westermann (1955: 5—12) is still useful, while Garlan (1988) is fresh and full of ideas. Thompson’s lively study (2003: 1—46) gives a new twist to the tradition, by bringing material culture to the fore. Cartledge (2001c) offers much in short compass.
Significant attention has been given to violent enslavement (Ducrey 1999; cf. Pritchett 1991 and, on piratical slave-raiding, De Souza 1999). Much attention has also been given to the impact ofthe Greek demand for slaves on warfare and much else within barbarian societies (e. g. Arafat and
Morgan 1994; Taylor 2001). Probably most important is Garlan’s exploration (1987) of the disjunction between evidence of the apparent enslavement of defeated Greeks and the absence of such persons from our data on the Greek slave population. Ransom and extensive sale overseas would help to explain the phenomenon (cf. Rosivach 1999).
The slave supply continues also to be included in studies of the Greek economy (e. g. Foxhall 1998; cf. Osborne 1995), but the economics ofslave-trading require much fuller treatment. Among regional studies the Black Sea and Danubian areas stand out. The pioneering study of Finley (1962; inspiring e. g. Braund and Tsetskhladze 1989) is best understood as part of a sustained engagement with the slave trade of these regions, usually by the scholars of the area itself. The latest study is available in English (Gavriliuk 2003), but Gavriliuk (1999), on Scythia’s economics as a whole, has to be read in Russian.
Scholars concerned with texts, inscriptions and papyri consider the slave supply from a range of angles. On slaves in comedy see, e. g. Krieter-Spiro (1997). Philological concerns sometimes extend valuably into social history (notably Babler 1998) and commentary (e. g. Pomeroy 1994). Female slaves have benefited from more recent concern with the representation and reality of ancient women (notably Faraguna 1999). From the study of nomenclature, the largely negative conclusion emerges that names need not tell much about slave origins (e. g. Fraser 2000). For the ideology of which these texts, names and practices are expressions, see especially Garnsey (1996).
Very little attention has been given to the personal experience of enslavement and sale. Beyond the battlefield, it only features in isolated studies (notably Jones 1987 on slave-marking). However, slaves were ultimately bought for a purpose, and progress is possible by starting with the end of their journey and their various work roles (e. g. Jones 2004). Here Lauffer (1979) is particularly valuable (cf. Conophagos 1980). For all that, the nature of the evidence is such that empirical studies alone will never give much insight in this area: informed hypothesis and controlled inference (embracing also comparative material) are especially necessary if we are to gain some understanding of what it felt like to be bought and sold; cf. Morris (1998c), sympathetic to the dislocation of slave experience.
World History
