Byzantine Culture

History · Article

Slavery in Byzantium

Byzantium inherited Roman slavery and kept it for a thousand years, in reduced and changing forms. Sources of slaves, legal status, the church's ambivalence, manumission, and the slave trade.

Byzantium was a slave-owning society for the whole of its existence. It inherited the institution from Rome, modified it under Christian and fiscal pressure, and never abolished it; slaves are still being bought and sold in Constantinople in the fifteenth century.

The subject has been comparatively neglected, partly because the Byzantine evidence is scattered and partly because the scale was so much smaller than in the classical Roman world that historians have treated it as marginal. It was not marginal to the people involved.

Scale and Function

Roman slavery at its height was a system of mass agricultural production. Byzantine slavery was not. The gang-worked slave estate of the late Republic had largely disappeared by late antiquity, replaced by the tenant farming described in the Byzantine village.

What remained was concentrated in:

  • Domestic service. The main use. Wealthy households held slaves as cooks, servants, nurses, tutors, secretaries and attendants. The number in a great house might run to dozens.
  • Skilled and workshop labour, including in textile production.
  • Imperial and ecclesiastical households, where slaves and freedmen staffed large establishments.
  • Rural labour on some estates, though usually alongside and eventually merging with dependent tenants.

Estimates of the enslaved proportion of the population are guesses. Something in the range of a few per cent is the usual suggestion, higher in Constantinople and the wealthy cities, negligible in the villages.

Where Slaves Came From

War. Prisoners were the traditional source, and every frontier campaign produced them. Byzantine armies enslaved captives; so did their opponents, and the recovery of enslaved Byzantines was a standing subject of diplomacy and of the prisoner exchanges conducted with the caliphate.

Trade. A long-distance slave trade ran through Byzantine territory for centuries. Slavs from the Balkans and central Europe — the origin of the word slave in most European languages is precisely this traffic — were moved through Venice and the Adriatic; Rus traders brought captives down the Dnieper; and the Black Sea ports supplied the Mediterranean. Byzantine legislation regulated and taxed the trade rather than restricting it, except in one respect: the sale of Christians to Muslims and Jews was repeatedly forbidden, and the prohibitions were repeatedly repeated, which indicates how well they worked.

Birth. The child of an enslaved mother was enslaved, regardless of the father.

Debt and self-sale. Legally restricted and practically persistent, particularly in famine years. Parents selling children was a recurring emergency phenomenon, and legislation dealt with the conditions under which such sales might later be voided.

Byzantine law on slavery is Roman law, modified.

A slave was property, could be bought, sold, bequeathed and pledged, could not marry legally under the classical rules, and could not own property except a peculium held at the master’s sufferance. Testimony from slaves was taken under torture, a Roman procedural rule that Byzantine courts retained.

Christian influence produced a series of real but limited changes:

  • Manumission in church (manumissio in ecclesia), recognised from Constantine’s reign, made freeing a slave a straightforward religious act performed before the congregation.
  • Slave marriage was gradually recognised. Justinian’s legislation and later canon law moved toward treating slave unions as marriages; by the time of the Ecloga in 741, a slave marriage blessed in church was valid, and Alexios I in the eleventh century required that slaves be allowed church marriage.
  • Killing a slave was homicide, and cruelty could in principle lead to forced sale.
  • Sanctuary. A slave who fled to a church could not simply be seized; the clergy negotiated, generally for a promise of no reprisal.

The direction of travel is clear and the pace was glacial. None of it questioned the institution.

The Church’s Position

Christian writers condemned the mistreatment of slaves, urged manumission as a pious act, insisted on the spiritual equality of enslaved and free, and did not call for abolition.

John Chrysostom is representative and instructive. He argues that slavery is a consequence of sin rather than a natural condition, tells masters to limit themselves to one or two slaves, to teach them a trade and to free them — and simultaneously tells slaves to obey. The moral energy goes into the conduct of the relationship, not its existence.

The notable exception is Gregory of Nyssa, whose fourth homily on Ecclesiastes attacks slaveholding directly and on principle: the slaveholder claims to own a being made in the image of God, and sets a price on something priceless. It is the most radical statement on the subject to survive from antiquity, Christian or pagan, and it had no discernible influence.

Monasteries owned slaves, though several monastic typika — including major ones — forbade the practice within the community, on the grounds that monks who had renounced property should not hold persons. The record is mixed.

Manumission and Freedmen

Freeing slaves was common and was regarded as meritorious; wills frequently manumit. The freed person retained obligations to the former owner’s family and a degree of social stigma, but full legal freedom was attainable in a way that it was not in later Atlantic slavery.

The route through the imperial household was the most spectacular. Freedmen and, especially, enslaved boys who were castrated and entered palace service could rise to the highest offices in the state, as Byzantine eunuchs describes. The parakoimomenos Basil Lekapenos, who effectively governed the empire for decades in the tenth century, was the son of an emperor by a Bulgarian slave.

Decline

Slavery contracted over the Byzantine millennium without ever disappearing. The reasons are economic more than moral: dependent tenancy was cheaper than slave-holding for agricultural labour, since the tenant fed himself; and the supply from warfare diminished as the empire stopped conquering.

In the late period, the Italian merchant colonies of the Black Sea revived the trade on a large scale, buying in the Crimea and selling in Italy, Egypt and elsewhere — a traffic conducted through Byzantine waters and largely outside Byzantine control, connected to the wider commercial shifts described in Byzantium and the Islamic world.

The institution passed to the Ottomans, who ran a slave system of considerable scale in the same city, including the devşirme levy of Christian boys for the janissary corps and the administration — a practice with clear structural echoes of what the Byzantine palace had done with enslaved and castrated foreigners.

How to Judge It

Two errors are worth avoiding.

The first is to treat Byzantine slavery as a mild, almost benign institution because the church softened it and manumission was common. Enslaved people were bought and sold, tortured for testimony, sexually available to their owners and separated from their children.

The second is to read it through the lens of the Atlantic system. Byzantine slavery was not racialised, was not hereditary in the absolute sense, was not concentrated in plantation production, and permitted a route out. It was a different institution with the same central fact at its heart.

The Prices

Slave prices appear in enough documents to give a rough series, and the figures put the institution in economic perspective.

A healthy adult slave in the middle Byzantine period cost in the region of twenty to thirty nomismata. Skilled slaves — a trained scribe, a physician, a musician — cost considerably more; children and the elderly less. Eunuchs commanded a premium, sometimes several times the price of an intact male, which is one reason the trade in castrated boys persisted despite repeated prohibition.

Set against the figures in Byzantine coinage, a slave cost roughly two to three years of an unskilled labourer’s wages, or about the price of a good horse and rather more than a modest provincial house.

That is expensive. It is the fundamental economic fact about Byzantine slavery: at those prices, slaves were a capital investment suited to domestic service in wealthy households and to skilled work, and not to agriculture, where a paroikos who fed himself was far cheaper. The shift from slave to tenant labour on the land, described in the Byzantine village, follows from the arithmetic rather than from any moral development.

The gradual recognition of slave marriage is the clearest case of Christian influence actually changing the law, and the sequence can be followed.

Roman law: a slave union was contubernium, cohabitation, not matrimonium. It created no legal kinship, could be dissolved by the owner at will, and produced no legitimate children.

Justinian: prohibited the separation of slave families in certain circumstances and allowed slaves to contract marriage with their owner’s consent, though the union remained legally weaker than a free marriage.

The Ecloga (741): a slave marriage blessed in church is a marriage. This is the decisive step — it brings slave unions inside the sacramental system, and once a union is a sacrament the owner cannot dissolve it.

Alexios I (late eleventh century): legislated that slaves must be permitted church marriage, explicitly criticising owners who refused on the grounds that slaves were property. The novel is unusually direct: it says that slaves share the same nature and the same faith, and that denying them the sacrament is not permissible.

The trajectory is real and took six centuries to travel. It did not touch the institution itself, and Alexios, who legislated that a slave might marry, did not suggest that a slave might not be owned.

Comparison and Its Limits

Placing Byzantine slavery against other systems clarifies it, provided the comparison is made carefully.

Against classical Roman slavery: much smaller in scale, not concentrated in agricultural production, and with a legal position slightly improved by Christian legislation. The essential structure — property, saleability, judicial torture, sexual availability — is unchanged.

Against Atlantic slavery: not racialised, not permanent across generations in practice, not tied to plantation production, with manumission common and freedmen able to rise, in the imperial household, to the highest offices of state. These are large differences and they matter.

Against contemporary western European serfdom: the Byzantine paroikos and the western serf are closer to each other than either is to a slave, and by the late Byzantine period the practical distinction between a paroikos and a slave engaged in agriculture had narrowed considerably.

Against Islamic slavery: structurally similar in several respects — domestic and military rather than agricultural, with manumission encouraged and routes to power through the household — and the two systems traded with each other continuously across the frontier.

The comparison that should not be made is the reassuring one: that because Byzantine slavery was smaller, milder in law and escapable, it was not really slavery. People were bought, sold, tortured for evidence, used sexually and separated from their children, in a Christian empire, for eleven hundred years.

Continue exploring