It is, of course, difficult to assess the role of women in ancient and medieval societies because of the overwhelmingly masculine tone of the literary sources which have been preserved and the absence of extant records about women's attitudes towards their own environment and activities. Women were not entirely silent in the ancient world, but when they did speak their voices have often been silenced during textual transmission. Most notable here are the poetesses of ancient Greece Telesilla and Praxilla (Sappho is a notable exception), whose works have been lost, and even in the case of Kassia, the renowned Byzantine hymnographer of the ninth century, much of her output has been reattributed to other (male) writers. We should not allow such accidents to colour our view of the ancient world: it is easy to remember the anecdote of Egnatius Metellus who supposedly took a cudgel and beat his wife to death because she had drunk some wine (Valerius Maximus 6.3.9), and ignore the overwhelming epigraphic evidence for affectionate marital relationships in Rome, such as the first-century BC 'Eulogy of Turia'. In addition, the historian's task is made more challenging by the fact that sources in general concentrate on the aristocratic, royal and wealthy classes and are frequently preoccupied with adultery and divorce, a circumstance which radically affects women's legal status. We should not allow the fact that in the ancient world women were legally considered for the most part in terms of their relationships with or services to males to affect our views of women's activities in general. The contribution of the lower classes, prostitutes, and even slaves and freedwomen to their societies has also to be considered - after all they played a vital role, and many slaves, as in Sumer and ancient Greece, had the right to engage in business, borrow money and buy their freedom. In Byzantium the 'third sex', eunuchs, also played an important part in the court, bureaucracy and aristocratic households.

Women in the ancient world were not necessarily dependent on men for a livelihood. The Old Babylonian Code of Hammurabi (109-10) deals in detail with women wine-sellers, and clearly such a woman was master of her own establishment, for she is required to arrest any outlaws who congregate in her establishment and take them to the palace. In Sumer a woman could hold property, engage in business, and qualify as a witness: Innashagga, wife of Dudu, bought a house with her own money and proved the fact on oath (Kramer, Sumerians, 1964, 335). The fact that wives could be divorced on relatively light grounds should not be taken to outweigh these factors. Historians have to be prepared to argue ex silentio and extrapolate from the existing evidence: information on Egyptian women even of royal rank is slight; nevertheless the case in the eleventh century BC of Princess Nebt, heiress of the Elephantine island at Aswan as well as a patroness of the arts, demonstrates that women in Egypt could maintain their own estates and hold rank in their own right or as regents for young sons.

Even in classical Athens, where women are often thought to have been 'secluded', some women were employed in the market, as sellers of bread or garlands, and could be found as innkeepers; the Caputi Hydria may even show a woman vase painter. We know that in Byzantium aristocratic women could be shop-owners, and women were deeply involved in retail trade and could be bakers, cooks, innkeepers, bathkeepers, washerwomen, gynaecologists, midwives, dancers, prostitutes, matchmakers and sorcerers, as well as silk-workers (Oxford Dictionary of Byzantium, 3.2202). Danielis, a widow from Patras, in the ninth century, controlled 'not a small part of the Peloponnese', and reportedly owned innumerable slaves, 3,000 of whom Leo VI freed and settled in southern Italy (Theophanes Cont. 317-21).