These three papers were among those presented at a conference entitled “The Atlantic Slave Trade: Who Gained and Who Lost?” held under the auspices of the Frederick Douglass Institute of the University of Rochester in October 1988. Other papers from this conference will be published in Social Science History over the next several issues, and the full set of papers will be published by Duke University Press.

From the point of view of the preindustrial world, the development of the English cotton textile industry in the eighteenth century was truly revolutionary. The industry was established early in the century as a peasant craft (section 2; note 2), and by 1850 it had been almost completely transformed in terms of the organization and technology of production. Of the total work force of 374,000 employed in the industry in 1850, only 43,000 (approximately 11.5 percent of the total) were employed outside the factory system of organization. In terms of technology, the industry was virtually mechanized by this time: there were 20,977,000 spindles and 250,000 power looms in the industry in 1850. What is more, steam had become the dominant form of power used in the industry—71,000 horsepower supplied by steam as opposed to 11,000 supplied by water (Mitchell, 1962: 185, 187). Value added in the industry by this time exceeded by about 50 percent that in the woolen textile industry, the dominant industry in England for over four centuries. This rate of development was something that had never been experienced in any industry in the preindustrial world. Indeed, the Industrial Revolution in England, in the strict sense of the phrase, is little more than a revolution in eighteenth-century cotton textile production.

The “numbers game” (Curtin, 1969: ch. 1; Darity, 1985) remains a favorite event in academic jousting over the Atlantic slave trade, not only because unexpectedly detailed quantitative records continue to turn up in archival repositories but also, more recently, because of the suppleness with which scholars have applied data discovered by the first generation of researchers to new, and increasingly more sophisticated, historical problems. Old, relatively formal, analytical categories—decades; large, internally diverse stretches of the African coast; colonial/national aggregates on the American side of the Atlantic—although comparable among themselves, now seem more revealing of the data than of the history of the trade and are very salutarily giving way to questions and issues arising more directly from the experience itself: the causes of slave mortality, the economic strategies of slavers, age and sex distinctions among the slaves, the meaning of slaving for specific regions in Africa, and the trade’s contributions to events in Europe and the Americas. Use of quantitative data now presupposes the discovery of historically relevant categories of analysis and at the same time informs the meaning of the categories employed.

Accounts abound of slaves suffering from lack of water during the middle passage, many of them collected during testimony before the British Parliament and by British abolitionists. A Captain Hayes spoke of a cargo “labouring under the most famishing thirst. . . being in very few instances allowed more than a pint of water a day” (Buxton, 1844: 154–155). Thomas Clarkson (1969 [1789]: 573) claimed that he had seen slaves “almost dying from want of water,” and Thomas Buxton (1844: 151-152) alleged that “there is nothing which slaves during the middle passage suffer from so much as want of water.”

Recently, I attended a seminar at which a historian of women presented a dazzling interpretation of the polemical writing of Olympe de Gouges and its (not to mention her) reception during the French Revolution. A crusty old historian of the Revolution rose during the question period and inquired, in his own eastern twang, “Now that I know that women were participants in the Revolution, what difference does it make!” This encounter suggested to me what I will argue are two increasingly urgent tasks for women’s history: producing analytical problem-solving studies as well as descriptive and interpretive ones, and connecting their findings to general questions already on the historical agenda. This is not a call for integrating women’s history into other history, since that process may mean simply adding material on women and gender without analyzing its implications, but for writing analytical women’s history and connecting its problems to those of other histories. Only through such an endeavor is women’s history likely to change the agenda of history as a whole.

Louise Tilly’s thoughtful and provocative essay “Gender, Women’s History, and Social History” makes a genuine contribution to our thinking about the research that has been done in women’s history by proposing new categories for analysis and by addressing the current debate over the usefulness of deconstruction as a methodological tool for historians. She demonstrates that we have been engaged in “the study of women in time,” with some success. We have studied women’s occupations, sexuality, personal relationships, marriage choices, family roles, and education. We have examined changes in women’s legal and political rights, their engagement in social and political causes, their successes and failures in the political arena. We have analyzed the ways in which states have tried to regulate women’s behavior, the social and economic constraints under which women have functioned, and the intersection of patriarchy and capitalism. And we know infinitely more than we did two decades ago about these subjects.

In her complex and subtle paper, Louise Tilly raises a host of intriguing, debatable issues. She tells us, pace Joan Scott, that women’s history has “arrived” in terms of both institotionaliza-tion within the academy and development into a separate historical specialty. She tells us that we must strive towards a women’s history that is not only descriptive (seeing the task “of retrieving women’s lives and achievements ... as sufficient unto themselves”) but also analytical (“connecting its problems to those of other histories”). She tells us, again pace Joan Scott, that a literary approach to gender downplays human agency and offers no constructive means of historical explanation. And she tells us that social history, as it has been informed and altered by women’s history in the last twenty years, is the best place for connection between women’s history and “the agenda of history as a whole.”

Judith Bennett has persuaded me that in the oral version of this paper I conflated five points: (1) All women’s history has a feminist motivation and message. (2) Descriptive women’s history has discovered valuable evidence about women in the past; this has now been accepted as historical “fact.” (3) Sociological use of gender as a concept adds an analytical edge to descriptive accounts. (4) Social history that makes gender and women’s experience problematic and analyzes it systematically can add to the achievements of descriptive/interpretive women’s history. (5) Both types of women’s history (descriptive/interpretive and analytical) can only benefit from explicitly demonstrating the ways in which their findings contribute to answering questions already on the historical agenda. I have revised my paper somewhat to clarify these points.