forms of slavery 273
protection “to their own communities.” These communal identities were a continuation
of the religious identities of the medieval communities. In the late medieval and early
modern period, “color,” “origin,” and “ethnicity,” in addition to religion, became
the main identifiers of slaves, and was used as means of segregation.
One of the most interesting characteristics of the status of the slave in the
Mediterranean world was the development of access to the law for slaves.
Detectable already in Late Antiquity, it was more common in the medieval period,
because of the slave’s religious identity, becoming quite frequent in the early
modern period. This gave slaves a legal persona through which they could act
independently in order to attain freedom. Thus, although slaves were normally
acquired in order to improve the socio-economic position of their master’s house-
hold, slaves and manumitted slaves sometimes found the means (both legally and
illegally) to become independent. The same variety and social versatility applied to
the Ottoman Mediterranean.
Like contemporary Iberian societies, Ottoman society imported slaves from
North African ports as well as from east Africa, eastern Europe, the Balkans, and the
Caucasus. Estimates of the trans-Saharan slave trade to North African markets under
Ottoman rule alone are between three to four million (Austen, 1992).^1 War and
political circumstances were also major sources for slaves. The Circassians, for exam-
ple, were mainly used as rural slaves on the large estates of Anatolia following their
expulsion from Russia in the nineteenth century. In the Middle Ages, Turks had
formed one of the main sources of slaves in the eastern Mediterranean, but now
under Ottoman Turkish hegemony, captives from the Caucasus and the Balkans
replaced them. Ottoman society continued to use imported slaves in order to fill the
military and civil-service elites of their empire, perpetuating the institutions of the
Janissaries and Ghulāms (young male slaves trained in the palace to serve in Ottoman
administration). Some of these rose very high and became local governors, while a
few famous manumitted palace slaves became viziers. But the majority of captives
and imported slaves were used by private owners in private households. Simply nam-
ing these slaves “domestics” is to attribute a common socio-economic position to
them defined solely according to their habitat; it ignores their economic use and
social function.
As in medieval Byzantine and Muslim societies, the backbone of Ottoman social
life was the private household. Placing slaves in urban manufacture as artisans, for
example in the silk industry, and manumitted slaves as businessmen, traders, bro-
kers and fiscal agents (Inalcık and Quataert, 1997) enabled the private household to
become a socio-economic enterprise. Unlike that for workers of free status, the
slave-owner relation ensured total dependency and a permanent social hierarchical
structure. The price of a slave, although substantial, was equivalent to a couple of
years’ wages of a regular employee in both Spain and the Ottoman Empire (Stella,
2000; Inalcık and Quataert, 1997). Investing in a household slave thus gave the
owner a worker for life, for whom he paid the equivalent of a two years’ wages
plus expenses, and whom he could later free (in Ottoman society normally after
7–10 years) and use as both an agent and a member of the extended family. This
socio-economic reasoning stood in contrast to the new notion of “wage labor” as
it was conceptualized and practiced in northwestern Europe following the industrial
revolution.