Peter Hunt
Thesis: The slaves listed in the Attic Stelai are unrepresentative of the Athenian slave population: the proportion of female slaves is too low, because the confiscations from the Hermokopidai did not usually include property attached to their wives. This property contained a high proportion of female slaves, but would often have been retained by the wives or re-possessed by their families before the auctions recorded in the Attic Stelai.
The Attic Stelai provide social historians with valuable evidence about the slave population at Athens (Pritchett 1953, 1956, 1961): prices, probable origins, some occupations, and, seemingly, the gender ratio. The Stelai would have a unique value if, as usually assumed, all property and hence all slaves were originally listed and the surviving fragments provide a random, if small, sample from them. Admittedly, the lists do not appear to include mine slaves, brothel slaves, or large concentrations or craftsmen—like the 120 slaves of Lysias and Polemarchus (Lysias 12.19); conceivably the types of slaves owned by the elite Athenians and metics on the Attic Stelai are atypical. These caveats notwithstanding, most historians assume that the Stelai include a small but random sample of the slaves most individual Athenians owned. That 80% of the slaves listed are male seems to confirm the impression from more obviously unrepresentative sources, such as comedy, that most slaves in Athens were male (e.g., Andreau and Descat 57, Schmitz 77-78, Tordoff 11 n. 66;)—an important result for understanding the function of slavery at Athens, the probability of family life, and the slave supply. But one aspect of Athenian law, in conjunction with one common social practice, has distorted this gender ratio.
Legally, a wife’s personal property, in all cases, and her dowry, often, would have been sequestered from a condemned man’s property before it was confiscated and sold. Athenian women possessed some property of their own, which would naturally remain in their possession regardless of the condemnation of a husband. This included the parapherna, with which they entered the marriage, as well as whatever they acquired as gifts or otherwise during the marriage. The treatment of dowries presents a more complicated and difficult issue, but the general tendency of the law is clear: creditors were allowed their claims against the property of a condemned man before it was auctioned off (e.g., Rhodes and Osborne 36.14-35); a wife was treated like a creditor and could demand her dowry (Harrison 54; Harris 233; contra Cox 119 on, e.g., [Dem.] 42.27, [Dem.] 47.57). Even if a wife had a son—to whose inheritance the dowry attached—her natal family might well have insisted on a (probably collusive) divorce to regain control over the dowry rather than having it confiscated.
In Athenian society, dowries and women’s personal property naturally included slaves. The close relationships between affluent women and their female slave attendants are well attested in drama, law-court speeches, vase painting, and grave reliefs. Other evidence suggests that these close relationships often included ownership: for example, when Neaira settled her dispute with Phrynion, she retained her three slave attendants as personal property, “since they had been purchased for her. . .” ([Dem.] 59.46); Pasion’s will included female slaves among the property he allocated to the dowry of his widow, Archippe ([Dem.] 45.28; cf. Theophrastus 22.10). In neither of these cases, did the woman’s property or dowry include any male slaves, perhaps a common situation.
The attested dowries of daughters constituted a much smaller proportion of a family’s wealth than did the property to be inherited by the sons, so the property attached in any sense to women was a fraction of the whole (Wolff 140 ff.). Nevertheless, even a couple of missing female slaves per wife could distort the gender ratio on the Attic Stelai. Other evidence for a slave population composed more of men than of women remains, but the Attic Stelai no longer provide support for a large differential.