Articles Available Online

Lectures & Presentations

The World-System in the 21st Century: 25th Annual Political Economy of the World-System Conference

Digital Archive: Slavery & Emancipation in the Mountain South: Evidence, Sources, & Methods

Digital Archive: Race, Ethnicity, Class & Gender: Women, Work & Family in Antebellum Appalachia

Contact the Author

Wilma A. Dunaway

"Challenging the Dominant Paradigm:

The World Market in Cotton, U.S. Slave Trading, and Destruction of African-American Families."

Paper presented at the Annual Meeting of the American Sociological Association Meeting, August 2005.

This Article Is Copyrighted: When citing this article, refer to this website.

            When Amelia Jones told her story to a WPA interviewer in the 1930s, she described her former owner as a man who routinely traded black laborers. "Master White didn't hesitate to sell any of his slaves. He said, 'You all belong to me and if you don't like it, I'll put you in my pocket.'" Footnote Jim Threat focused on the danger of permanent separation. "We lived in constant fear," he said, "that we would be sold away from our families." In her story, Maggie Pinkard gives us some clue how often black families were disrupted. "When the slaves got a feeling there was going to be an auction, they would pray. The night before the sale they would pray in their cabins. You could hear the hum of voices in all the cabins down the row." Other enslaved women focused more sharply on the mother's perspective. Several of them lamented that they had "no name" to give their children because they must use their masters' surnames, not those of their husbands. "I haven't never had a nine months child," Josephine Bacchus told the WPA interviewer. "I ain' never been safe in de family way." This former slave went on to say that she experienced chronic hunger, sexual exploitation from white males, and quick return to the fields after childbirth. As a result, all her babies, except one, were stillborn (Rawick 1977, vol. 12: 335, 257, Rawick 1972, vol. 2a: 23). Katie Johnson captured the vulnerability of parents when she said: "During slavery, it seemed lak yo' chillun b'long to ev'ybody but you" (Perdue, Barden and Phillips 1976: 159-60). These voices recount experiences that are representative of a majority of slaves of the Mountain South, a region characterized by a low black population density and small plantations. What they have to say is startling because they are reporting a past that contradicts the dominant paradigm. The conventional wisdom is that owners rarely broke up slave families; that slaves were adequately fed, clothed, and sheltered; and that slave health or death risks were no greater than those experienced by white adults.

Methodological Flaws of the Dominant Paradigm

            Why have so many investigations come to these optimistic conclusions? U.S. slavery studies have been handicapped by three fundamental weaknesses. First, scholars who support the dominant paradigm have directed inadequate attention to enslavement in the Upper South. Instead, much of what is accepted as conventional wisdom is grounded in the political economy and the culture of the Lower South. Why is it so important to study the Upper South? In the United States, world demand for cotton triggered the largest domestic slave trade in the history of the world (Patterson 1982: 164-65). Footnote Between 1790 and 1860, the Lower South slave population nearly quadrupled because the Upper South exported nearly one million black laborers (Sutch 1975: 174-75). In a fifty-year period, two-fifths of the African-Americans who were enslaved in the Upper South were forced to migrate to the cotton economy, the vast majority sold through interstate transactions, about 15 percent removed in relocations with owners (Tadman 1989: 19-20).

            Because of that vast interregional forced migration, Upper South slaves experienced family histories that contradict the accepted wisdom in U.S. slave studies. Once the international slave trade closed in 1808, the Upper South operated like a "stock-raising system" where "a proportion of the natural increase of its slaves was regularly sold off." As a result, the chances of an Upper South slave falling into the hands of interstate traders were quite high. Between 1820 and 1860, one-tenth of all Upper South slaves were relocated to the Lower South each decade. Nearly one of every three slave children living in the Upper South in 1820 was gone by 1860 (Tadman 1989: 45, 5, 70, 122). Among Mississippi slaves who had been removed from the Upper South, nearly half the males and two-fifths of the females had been separated from spouses with whom they had lived at least five years (Gutman 1976a: 151). Virginia slave families were disproportionately matrifocal because of the slave trading and labor strategies of Upper South masters (Stevenson 1996: 112-18).

            The second methodological error of the dominant paradigm is its neglect of small slaveholdings. Gutman (1976a: 102) acknowledges this shortcoming of his own work when he comments in passing that "little is yet known about the domestic arrangements and kin networks as well as the communities that developed among slaves living on farms and in towns and cities.” Fogel (1989: 178-82) stresses that “failure to take adequate account of the differences between slave experiences and culture on large and small plantations” has been a fundamental blunder by slavery specialists. Because findings have been derived from analysis of plantations that owned more than fifty slaves, generalizations about family stability have been derived from institutional arrangements that represented the life experiences of a small minority of the enslaved population. In reality, more than 88 percent of U.S. slaves resided at locations where there were fewer than fifty slaves (Genovese 1974: 7, Steckel 1979: 91).

            Revisionist researchers provide ample evidence that slave family stability varied with size of the slaveholding. Analyzing sixty-six slave societies around the world in several historical eras, Patterson (1982: 206) has found that slavery was most brutal and most exploitative in those societies characterized by small holdings because such structural arrangements required “far more contact with (and manipulation of) the owner” and “greater exposure to sexual exploitation." Stephen Crawford (1992: 336-37) has shown that slave women on small plantations had their first child at an earlier age and were pregnant more frequently than black females on large plantations. Compared to large plantations, slave families on small plantations were more often disrupted by masters; and black households on small plantations were much more frequently headed by one parent (Crawford 1980: 149-55). Steckel (1986: 738-39) has argued that hunger and malnutrition were worse on small plantations, causing higher mortality among the infants, children, and pregnant women held there.

            The third methodological error is that U.S. slavery studies have been dominated by the view that it was not economically rational for masters to break up black families. According to Fogel and Engerman (1974: 49-51) households were the units through which work was organized and through which the rations of basic survival needs were distributed. By discouraging runaways, families also rooted slaves to owners. Gutman's (1976a: 11-28, 164, 304-305, 602-603) work established the view that slave families were organized as stable, nuclear, single-residence households grounded in long-term marriages. After 30 years of research, Fogel (1989) is still convinced that two-thirds of all U.S. slaves lived in two-parent households. Many recent studies are grounded in and celebrate these optimistic generalizations about the African-American slave family. Footnote None of these writers believes that U.S. slave owners interfered in the construction or continuation of black families. Fogel (1989: 55) argues that such intervention would have worked against the economic interests of the owners while Gutman (1976b) focuses on the abilities of slaves to engage in day-to-day resistance to keep their households intact. Fogel (1989: 125, 147-49) and most scholars argue that sexual exploitation of slave women did not happen very often. Moreover, the conventional wisdom has been that slaveholders discouraged high fertility because female laborers were used in the fields to a greater extent than male workers. Consequently, the predominant view is that most slave women did not have their first child until about age 21 and that teenage pregnancies were rare. To permit women to return to work as quickly as possible, owners protected children by providing collectivized child care (Steckel 1986: 738-39).

The Target Area for This Study

            In sharp contrast to previous studies, I will test the dominant paradigm of the slave family against findings about a slaveholding region that was typical of the circumstances in which a majority of U.S. slaves were held. That is, I will examine enslavement in a region that was not characterized by large plantations and that did not specialize in cotton production. Even though more than half of all U.S. slaves lived where there were fewer than four slave families (Genovese 1974: 7), there is very little research about family life in areas with low black population densities. Despite Crawford’s (1980: 36-59) ground-breaking finding that plantation size was the most significant determinant of quality of slave life, this is the first study of a multi-state region of the United States that was characterized by small plantations.

            This study breaks new ground by investigating the slave family in a slaveholding region that has been ignored by scholars. I will explore the complexities of the Mountain South where slavery flourished amidst a nonslaveholding majority and a large surplus of poor white landless laborers. In geographic and geological terms, the antebellum Mountain South consisted of 215 hilly and mountainous counties in nine states stretching from western Maryland to northern Alabama and northern Georgia. Footnote In a previous book (Dunaway 1996: 11-12), I have documented the historical integration of this region into the capitalist world-system. The incorporation of Southern Appalachia entailed nearly 150 years of ecological, politico-economic, and cultural change. Beginning in the early 1700s, Southern Appalachia was incorporated as a peripheral fringe of the European colonies located along the southeastern coasts of North America. During the early eighteenth century, the peripheries of the world economy included eastern and southern Europe, Hispanic America, and “the extended Caribbean,” which stretched from the Atlantic colonies of North America to northeast Brazil (Wallerstein 1980: 167, 233-34). As the geographical space for several wars, the Mountain South became one of the major frontier arenas in which England, France, and Spain played out their imperialistic rivalry. Within two decades, the region’s indigenous people were integrated into the commodity chains of the world economy to supply slaves to New World plantations and to produce deerskins to fuel western Europe’s emergent leather manufacturing. After the American Revolution, Southern Appalachia formed the first western frontier of the new nation, so it was quickly resettled by Euroamericans.

            On a world scale, Southern Appalachia's role was not that different from many other peripheral fringes at the time, including inland mountain sections of several Caribbean islands, Brazil, the West Indies, and central Europe. Incorporation into the capitalist world economy triggered within Southern Appalachia agricultural, livestock, and extractive ventures that were adapted to the region's terrain and ecological peculiarities. Yet those new production regimes paralleled activities that were occurring in other sectors of the New World that had been colonized by western Europe. Fundamentally, the Mountain South was a provisioning zone which supplied raw materials to other agricultural or industrial regions of the world economy (Dunaway 1996: 196-97). Footnote

            On the one hand, this inland region exported foodstuffs to other peripheries and semiperipheries of the Western Hemisphere, those areas that specialized in cash-crops for export. The demand for flour, meal, and grain liquors was high in plantation economies, like the North American South and most of Latin America, where labor was budgeted toward the production of staple crops. So it was not accidental that the region's surplus producers concentrated their land and labor resources into the generation of wheat and corn, often in terrain where such production was ecologically unsound. Nor was it a chance occurrence that Southern Appalachians specialized in the production of livestock, as did inland mountainous sections of other zones of the New World. There was high demand for work animals, meat, animal byproducts, and leather in those peripheries and semiperipheries that did not allocate land to less-profitable livestock production.

            On the other hand, the Mountain South supplied raw materials to emergent industrial centers in the American Northeast and western Europe. The appetite for Appalachian minerals, timber, cotton, and wool was great in those industrial arenas. In addition, regional exports of manufactured tobacco, grain liquors, and foodstuffs provisioned those sectors of the world economy where industry and towns had displaced farms. By the 1840s, the Northeastern United States was specializing in manufacturing and international shipping; and that region's growing trade/production centers were experiencing food deficits. Consequently, much of the Appalachian surplus received in Southern ports was re-exported to the urban-industrial centers of the American Northeast and to foreign plantation zones of the world economy. In return for raw ores and agricultural products, Southern markets-- including the mountain counties-- consumed nearly one-quarter of the transportable manufacturing output of the North and received a sizeable segment of the redistributed international imports (e.g., coffee; tea) handled by Northeastern capitalists.

            Beginning in the 1820s, Great Britain lowered tariffs and eliminated trade barriers to foreign grains. Subsequently, European and colonial markets were opened to North American commodities. Little wonder, then, that flour and processed meats comprised the country's major nineteenth-century exports, or that more than two-thirds of those exports went to England and France. Outside the country, then, Appalachian commodities flowed to the manufacturing centers of Europe, to the West Indies, to the Caribbean and to South America. Through far-reaching commodity flows, Appalachian raw materials-- in the form of agricultural, livestock, or extractive resources-- were exchanged for core manufactures and tropical imports (Dunaway 1996).

            To use the words of Phillips (1987: 802), "the process of incorporation. . . involved the subordination of the labor force to the dictates of export-oriented commodity production, and thus occasioned increased coercion of the labor force as commodity production became generalized." Peripheral capitalism unfolded in Southern Appalachia as a mode of production that combined several forms of land tenure and labor. Because control over land-- the primary factor of production-- was denied to them, the unpropertied majority of the free population was transformed into an impoverished semiproletariat. Footnote However, articulation with the world economy did not trigger only the appearance of free wage labor or white tenancy. Capitalist dynamics in the Mountain South also generated a variety of unfree labor mechanisms. About three of every ten adults in the region's labor force were enslaved. This large land area was characterized in the antebellum period by nonslaveholding farms and enterprises, a large landless white labor force, small plantations, mixed farming, and extractive industry (Dunaway 1996: 87-122).

            Five indicators distinguish the Mountain South from the Lower South (Dunaway 2003b: 241-62). ■Mountain slaves were employed outside agriculture much more frequently than Lower South slaves. At least one-quarter of all mountain slaves were employed full time in nonagricultural occupations. Thus, slaves were disproportionately represented in the region’s town commerce, travel capitalism, transportation networks, manufactories, and extractive industries.

■          In comparison to areas of high black population density, mountain plantations were much more likely to employ ethnically-mixed labor forces and to combine tenancy with slavery.

■          Compared to the Lower South, mountain plantations relied much more heavily on women and children for field labor.

■          Appalachian slave women worked in the fields, engaged in resistance, and were whipped just about as often as men.

■          Mountain masters meted out the most severe forms of punishment to slaves much more frequently than their counterparts in other Southern regions. Appalachian ex-slaves reported frequent or obsessive physical punishment nearly twice as often as other WPA interviewees. There was greater brutality and repression on small plantations than on large plantations, and most of that punishment was meted out for social infractions, not to motivate higher work productivity.

■          Mountain South plantations primarily managed laborers by assigning daily or weekly tasks and by rotating workers to a variety of occupations. Moreover, small plantations relied on community pooling strategies, like corn huskings, when they needed a larger labor force. Since a majority of U.S. slaves resided on holdings smaller than fifty, like those of the Mountain South, it is likely that gang labor did not characterize Southern plantations to the extent that Fogel (1989: 193) has claimed.

■          Mountain slaves almost always combined field work with non-field skills, and they were much more likely to be artisans than other U.S. slaves.

Methods, Sources, and Definitions

            To research this complex topic, I have triangulated quantitative, archival, primary and secondary documents. To be as inclusive as possible in history production, I have grounded this study in analysis of narratives of nearly 300 slaves and more than 400 white Civil War veterans. I derived my statistical analysis from a data base of nearly 26,000 households drawn from nineteenth-century county tax lists and census manuscripts. In addition to those samples, I relied on archived records from farms, plantations, commercial sites, and industries. A majority of the slaveholder collections utilized for this research derived from small and middling plantations. Large plantations account for less than one percent of all the citations and details provided in this study. I employ the definitions that are typically applied by U.S. slavery specialists. A planter or large plantation held 50 or more slaves while a middling plantation or slaveholder owned 20 to 49 slaves. Thus, a small plantation was one on which there 19 or fewer slaves. Footnote

Challenging the Dominant Paradigm

            Since 1980, four important investigations have called into question the dominant paradigm. Crawford (1980: 71-87) and Patterson (1982: 218-30) contend that size of plantation is a major determinant of the quality of slave life, but not in the manner that popular mythology would lead us to predict. Patterson argues that world wide, over time, small slaveholdings have always been more brutal in living conditions. Using the WPA slave narratives, Crawford found that there were many more female-headed households, more severe punishment, greater malnutrition, and higher fertility on small plantations. By studying slave trader records, Tadman (1989: 211-12) has established that one of every three Upper South slave marriages was broken by a master’s intervention, one in every five disruptions being caused by a sale. Moreover, one-half of the sales involved the separation of children from their parents. Most recently, Stevenson (1996: 222) has made a strong case that most Upper South slave households consisted of women and children and that black children had “fewer options for daily, familial-like contact with adult male slaves than those who lived among larger numbers of men generally found in more populous quarters.”

            In line with the work of those four writers, this investigation also found that most Appalachian slaves did not reside in nuclear households with both spouses present. In the Mountain South, masters threatened the persistence of enslaved households in three significant ways. First, masters routinely interfered in marriage formations and child rearing, weakening the bonds between family members and causing high child and maternal mortality. Second, owners repeatedly disrupted households, as they removed members and structured the absence of adult males. On the one hand, "abroad" spouses, children, and other kin kept dropping into and out of households, as they moved to and from hire-outs or neighboring owners. On the other hand, frequent sales and high mortality rates meant that household composition altered every few years. High death rates and forced labor migrations necessitated the absorption of unrelated persons or extended kin whose former households had disintegrated. Mountain slaveholders placed black families at risk in a third way. Owners expropriated most of the production, allocating to their slaves survival rations below the actual subsistence requirements of household members. In order to maximize profits, mountain masters externalized to slave households:

■          most of the costs of their food, clothing and health care,

■          as well as the costs associated with reproducing surplus laborers for export (Dunaway 2003a).

            The myth of the stable nuclear slave family was derived from analysis of conditions that are representative of only a small minority of U.S. slaves. Since I do not want to replicate Gutman’s (1976a: 11-28, 164, 304-305, 602-603) methodological error, I am not suggesting that we make the theoretical quantum leap of generalizing family patterns for all U.S. slaves from another restricted sample. What I am arguing is that the findings about the Mountain South vary so dramatically from the accepted wisdom that we have strong indicators that the dominant paradigm is just flat wrong about a majority of U.S. slave families! On the one hand, this study documents broad variation from the dominant paradigm for 177 counties in six Upper South states and 38 counties in three Lower South states. Since findings about this large diverse region contradict most of the accepted family theses, we have strong justification for questioning the universal applicability of the dominant paradigm. On the other hand, the life situation of slaves in the Mountain South was more similar to that of a majority of U.S. slaves than were those Lower South large plantations on which supporters of the dominant paradigm have grounded their generalizations. Southern Appalachia is not just a regional peculiarity, for my findings are in line with other studies about small plantations and about the Upper South. It is clear from my research and from the work of Crawford (1980), Tadman (1989), and Stevenson (1996) that the three best predictors of slave family instability and disruption are:

■          residence of the slave in a slave-selling region,

■          frequent slave hiring for distant nonagricultural occupations,

■          and ownership by a small slaveholder.

All three of those risk factors were at play in the Mountain South, a slave-selling region characterized by small holdings, with about one-quarter of its black laborers assigned to nonagricultural occupations. What I have learned about the Mountain South provides more reliable clues about threats to the persistence of a majority of U.S. slave families than does the dominant paradigm. Like black Appalachians, a majority of U.S. slaves were threatened by one or more of these disruptive factors by 1820 (Tadman 1989). Since the Mountain South was typical of enslavement circumstances for a majority of U.S. slaves, several findings of this study call into question the most important theses of the dominant paradigm.

Slave Trading and Family Breakups

            In the seventeenth and eighteenth centuries, tobacco production absorbed two-fifths of the U.S. slaves. By the end of the American Revolution, however, prices for the three major Southern exports-- tobacco, indigo, and rice– had declined in the world economy (Wallerstein 1989: 167). Simultaneously, cotton was required to fuel expansion of the English textiles industry, so prices for that commodity escalated and underwent fewer cyclical drops. Between 1810 and 1840, U.S. cotton production increased nearly tenfold, as plantations pushed westward to become concentrated in a long belt stretching from South Carolina through Texas. In almost every decade from 1810 to 1860, Lower South cotton production expanded three times faster than the agricultural output of the Upper South. As a result, the Lower South demand for slaves expanded 1800 percent, more than twice the rate of increase in the U.S. slave population (Fogel 1989: 63-65, 70). Between 1790 and 1860, U.S. enslavement underwent a major transformation such that two-thirds of the country’s slaves were reconcentrated in the region producing the most profitable staple for the world economy. Over this seventy year period, the Lower South slave population nearly quadrupled because the Upper South exported two-fifths of its African-Americans (Sutch 1975: 174-75), the vast majority sold through interstate transactions, about 15 percent removed in relocations with owners (Tadman 1989: 9-20).

            The world demand for cotton drove the demand for black Appalachian laborers. In addition to surplus foodstuffs, salt, whiskey, timber, and iron (Dunaway 1996: 225-48), Southern Appalachians exported surplus laborers to the cotton South. In 1860, the Upper South counties of the Mountain South were populated by only 71 percent of the slaves who would have been there if no forced outward migrations had occurred. More than 40 percent of the 1840 cohort aged 10 to 59 had disappeared from Southern Appalachia by 1860. Males were exported from the region at a slightly higher rate than were females. In the 10 to 59 age group, 1.1 males were exported to every female, most probably leaving the region before the age of 29. Footnote

            Interstate sales caused the vast majority of the interstate relocations of Appalachian slaves. Consequently, Southern Appalachians contributed significantly to the interstate traffic in slaves. Between 1840 and 1860, more than 63,000 Appalachian slaves were exported from the region through interstate sales. Thus, black Appalachians were the victims of nearly one-fifth of all the interstate sales that occurred between 1840 and 1860. To put it another way, Appalachia was home for one of every eight of the slaves involved in the interregional black diaspora between 1840 and 1860. The chances of a slave from the Upper South falling into the hands of interstate traders were quite high. According to Tadman, "teenagers from this area would have faced rather more than a 10 percent chance of being traded, and those in their twenties perhaps an 8-10 percent chance. By their thirties, the danger of falling into the grasp of interregional speculators would have eased to about 5 percent." Nearly one of every three slave children living in the Upper South in 1820 was "sold South" by 1860. Each decade between 1820 and 1860, traders transported one-tenth of all Upper South slaves to the Lower South by. After 1840, Appalachian slaves were being exported at a rate slightly higher than the Upper South average. Footnote

            In the Upper South and in Southern Appalachia, slave trading occurred much more frequently than has been assumed and was more economically significant to slaveholders than Fogel (1989) has contended. In addition to the interstate slave trade, Appalachian masters routinely sold slaves locally, marketing about 20,000 in their local communities in 1860. In addition to selling slaves, Appalachian owners hired out annually one–fifth or more of their surplus slaves, usually to employers at distant sites. All together, Mountain South plantations disrupted slave families through six types of forced labor migration.

■          sales of slaves (59 percent of family disruptions)

■          long-term hire-outs (16 percent of family disruptions)

■          abroad marriages (15 percent of family disruptions)

■          owner migrations (4 percent of family disruptions)

■          estate settlements and presents to kin (3.5 percent of family disruptions)

■          labor allocation to an owner’s distant work site (less than 3 percent of family disruptions)

Consequently, breakup of slave families by masters and the structured absence of black fathers occurred in the Mountain South more often than in the Lower South or on large plantations. For the South as a whole, slave trader records have been used to estimate that one of every three or four sales triggered the separation of spouses. However, Appalachian slaves reported marriage breakups more often than the national average. In reality, one of every three Appalachian slave marriages was destroyed by the masters' structural interference to maximize profits from forced labor migrations. A majority of family disruptions were caused by sales that permanently severed kinship ties, and nearly two-thirds of all Appalachian slave sales separated children younger than fifteen from parents and siblings. Only one-fifth of the Appalachian slave households were complete families in which parents and children resided together, and female-headed households occurred four times more often than Gutman (1976a) claims. Footnote

            Small plantations structured the absence of fathers in several ways. Even when Appalachian masters thought of themselves as preserving family units, they rarely sold fathers with their spouses and children. Almost one-fifth of the family disruptions resulted from forced labor migration of members, disproportionately males, to distant sites that kept them away most of the time. When they were hired out on a yearly basis or when they were assigned at their owners’ distant enterprises, family members were reunited only a few times each year. More than 15 percent of all Appalachian slave family separations occurred when women had "abroad" husbands who were owned by neighboring slaveholders. We must be careful not to romanticize abroad kin arrangements as day-to-day resistance in which slaves forced upon their masters family linkages that worked against the owners' economic interests. Indeed, abroad marriages were necessary if small plantations were to maintain high fertility among their slave women, so mountain masters encouraged or required such arrangements. Mothers and children were expected to retain their masters' names, further obscuring the identity of fathers. Very few Appalachian ex-slaves were able to maintain permanent surnames that reflected the identities of their biological fathers.

Structured Malnutrition and Mortality as Threats to Family Survival

            Hunger and malnutrition were also a serious threat to the survival of black families on small Appalachian plantations. In order to export large quantities of foods to distant markets, Mountain masters provided grains and meat at levels that would have permitted adult slaves only two-thirds of their needed daily calories and even less of the nutrients essential to good health, still they exported large quantities of these foods. Thus, Appalachian ex-slaves reported that their masters supplied them inadequate food more than twice as often as did other U.S. ex-slaves, and they reported food stealing three times more frequently. In addition, small mountain plantations were twice as likely as other U.S. owners to require slaves to produce much of their own food supply, thereby adding to the workload of family members. Three quarters of Appalachian slave children were fed a cornbread-buttermilk mush; the rest received cornbread blended with "pot liquor" from boiled meats. Less than one-fifth received fruits, molasses, vegetables, or any regular ration of meat. As a result, slave children faced serious deficiencies of protein, iron, amino acids, and several vitamins and minerals essential to growth and good health. Footnote

            As a result of shortfalls in basic survival needs and masters’ exploitative labor strategies, a mountain slave was 1.4 times more likely to die than other U.S. slaves, enslaved adult women experienced a higher mortality rate than males, and more than one-half of all slave children died before age ten. In the United States, slaves died at a frequency only slightly higher than that experienced by whites. In sharp contrast to national trends, mountain blacks had a much lower life chance. In 1850, a mountain slave was 1.4 times more likely to die than other U.S. slaves, and the mortality risk of mountain slave children exceeded the national average. Footnote

Revisiting the Slave Breeding Debate

            In addition to risks associated with slave trading, malnutrition, and ecological hazards, mountain slave families were threatened by four patterns of systematic reproductive exploitation.

The primary structural mechanism by which mountain slaveholders directly controlled the reproductive process was through their interference into slave marriages. In nearly three-fifths of the cases described in the narratives, slaves were required to obtain their masters' permission before selecting their spouses. Mountain masters strongly shaped about one-third of the spouse selection decisions and unilaterally matched spouses in about one of every thirteen marriages.

Moreover, they routinely separated males from wives who were barren or who bore few children. The second pattern of systematic reproductive exploitation was sexual abuse. Mountain masters, their male offspring, and their managers resorted to sexual exploitation of slave women three times more often than did their counterparts in the rest of the South. Nearly 15 percent of the regional narratives describe acts of white sexual exploitation of enslaved women, and most of those instances involved force or violence. Moreover, one in ten mountain slave families was headed by a woman whose children were the outcome of sexual exploitation by white males. Even though supporters of the dominant paradigm (Fogel and Engerman 1974: 132) claim that only one of every ten U.S. slaves was a mulatto with Caucasian heritage, the manuscript slave schedules do not support this low incidence of sexual exploitation. In fact, more than one-third of the Appalachian slaves enumerated in the 1850 census were identified as mulattoes. Footnote

            The third pattern of reproductive exploitation lay in pressures toward early childbearing.

Slave women on small plantations had their first child at an earlier age and were pregnant more frequently than women on large plantations. Using the 1870 census manuscripts, we can glean significant information about women's reproductive decisions. Using the enumerated information, I calculated backwards the ages at which women delivered their children and how far apart these women spaced their pregnancies. Nearly half the enslaved mothers delivered their first child before they were seventeen. Three-quarters of them endured their first pregnancy before the age of nineteen. Moreover, Appalachian slave women were averaging fewer than two years between their pregnancies. These reproduction strategies are startling when compared with fertility averages for other populations. For the U.S. South as a whole, the average age of enslaved women at the birth of their first child was about 21, and these mothers were spacing more than two years between pregnancies. Emancipation signaled a dramatic change in the reproduction patterns of the region's African-American women. Among those women who delivered their first baby between 1865 and 1870, fewer than 2 percent were younger than 17-- a pattern that was directly opposite to child bearing trends during slavery. Footnote

            The fourth pattern of systematic reproductive exploitation lay in the child nursing practices structured by mountain masters. Appalachian slave women were trapped by their owners in a vicious cycle of early weaning, high child mortality, and high fertility. To maximize women's productive labor in the fields or at hired locations, masters required mothers to return to work within a few weeks of childbirth and to begin to wean their babies by the sixth month. Because breastfeeding acts as a natural deterrent to pregnancy, owners could also insure higher fertility through early weaning. Modern medical research documents the powerful biological linkage between breastfeeding and child survival. Infants who are breast-fed longer than one year have the highest survival rates and higher natural immunization (Maher 1992: 24). Since people of African heritage are disproportionately lactose intolerant to animal milk (Savitt 1994: 119), mountain babies died in great numbers from the diet substitutes that were mixed with contaminated water. Paradoxically, child mortality acted as a spur to high fertility. When their infants died, mountain slave women were pregnant again within less than a year. Footnote

            While masters required early weaning of slave children, their own infants were typically breast-fed for at least two years. To accomplish that, owners employed black mothers to serve as wet nurses and care-givers for white offspring. At the same time that mountain slave women were weaning their own infants early, one-fifth of them worked as wet nurses for white babies. At some point during their enslavement, two-thirds of the females were employed as care-givers to white children, often requiring them to leave their own offspring without adequate nutrition and nurture. The masters' strategies of malnutrition, high fertility, and inadequate work release during pregnancy led to high mortality. During their childbearing years, enslaved females suffered higher mortality rates than their male peers. In fact, black Appalachian women labored under the cloud of death rates that were 1.5 times higher than national slave averages and 1.8 times higher than the mortality rates of the white Appalachian males who so often raped them. Footnote

            While not orchestrating barnyard breeding methods, mountain plantations did indeed structure an exploitative reproductive system aimed at generating surplus slaves whose labor could be sold or hired. On the one hand, many Appalachian slaveholders profited as much, if not more, from enslaved women's reproduction of surplus laborers as they did from female production of market crops. On the other hand, mountain masters treated enslaved women and children as "resources to be optimally manipulated for productive ends” (Patterson 1982: 5-9), so they restructured the most fundamental aspects of childbirth and parenting. Owners' control of breastfeeding and child rearing weakened the woman's role as mother and established her children as bonded property. Such reproductive interference was "designed to teach both of them their place in the patriarchal system" (Gutman 1976a: 75-76).

            After their children reached age ten, Appalachian slave mothers lost two of every five of them through sales or other forced labor migrations. However, these mothers lost an even higher percentage of their offspring to premature deaths. Attenuated breastfeeding, mothers' early ages at first births, and ecological risks caused high infant mortality rates. Malnutrition, environmental hazards, and occupational accidents also took a high toll of black children between ages one and ten (Steckel 1979: 100-101). Paradoxically, that high child mortality spurred increased fertility among slave women, a pattern that parallels contemporary population trends in several poor countries (Maher 1992: 164). When the first child died before age one, four-fifths of the mountain slave women were pregnant again within less than two years. After 1800, infant and child mortality rose steadily on Appalachian plantations, in exactly the same historical era when the number of slave exports to the Lower South was peaking. As the market demand for slaves escalated, the reproductive rates of Appalachian slave women increased, as did the mortality rates among slave children. It is not just coincidental that Appalachian slave women began to reproduce at younger ages or that they shortened the waiting periods between pregnancies. Far from being natural or the result of women's unilateral decisions, such high fertility was imposed on mountain slave women, to no less degree than frequent pregnancies were inflicted upon the wives of slaveholding males. Rising mortality spurred significant declines in the mothers' average age at first birth and in the amount of time women spaced between live births. Consequently, the number of children born to slave women increased over time, as a result of earlier ages at first birth and shorter average spacing between pregnancies (Stevenson 1996: 245-48.

            To offset the dampening economic effect of high child mortality on their slave exports, mountain masters externalized to enslaved women the responsibility for producing adequate numbers of surviving children. Appalachian slave women were deprived of autonomy in motherhood roles through their masters' structural interference in pregnancy, breastfeeding, and child-rearing. While those affluent white parents celebrated and mourned deaths of their own children, they rarely noted with any empathy the mortality toll that took one-half or more of the black children they owned. Footnote While reeling under the emotional strains of childhood deaths in their own families, Appalachian masters burdened their enslaved women with repeated incidents of emotional and psychological distress associated with the removal of their offspring. Even though they were economically valuable to their masters, mountain slave women were treated as objects of disdain. Thus, an above-average incidence of white male sexual exploitation was also foisted upon mountain slave women. By blaming mothers, Appalachian masters ideologically camouflaged the degree to which their malnutrition, reproductive, and overwork strategies caused the ill health or deaths of children. Because black women experienced a higher incidence of miscarriages than whites, owners often accused expectant mothers of destroying the fetuses. If babies died prematurely of SIDS, owners indicted the mothers for infanticide. By bestializing the mountain slave mother in these ways, the Appalachian master received community legitimation for his structural interference in slave reproduction. By violating the only areas of possible reproductive autonomy enjoyed by the enslaved mother, the mountain master cemented his structural dominance over the slave family. Footnote

            While mountain masters may not have engaged in direct breeding practices, they did commodify the sexuality of mountain slave women. In an attempt to maximize profits from all levels of slave women's labor, Appalachian masters extended their power over the black woman's body, so that they controlled her both as an asexual labor unit and as a reproductive laborer. Consequently, Appalachian masters externalized to enslaved women unmanageable structural tensions among their different work roles. In their daily lives, mountain slave women juggled the conflicting labor demands associated with market commodity production, subsistence production for their own families, manufacture of goods consumed by their masters' households, and reproduction of the labor force. On the one hand, slave girls reached puberty one to three years earlier than boys because of their high-fat diets (Frisch and McArthur 1974: 949-51). At the same time that girls were undergoing the added physical stress of heavier chores, they were reaching menarche. However, females were probably fed less well due to their lower sales value (Steckel 1996: 44-56). Simultaneously, mountain masters shifted teenage girls to harder physical labor and began to pressure them to reproduce. Because of the health risks associated with their conflicting productive and reproductive roles, mountain slave girls were more likely than boys to be underweight, and they experienced a rising mortality rate after fifteen. On the other hand, the duties associated with reproduction and child care did not relieve women from other work. Pregnant women and nursing mothers continued their long hours of evening household production. In addition, there was a complex relation between annual cycles of crop production, hiring contracts, pregnancies and deliveries. The agricultural cycle of production and the regimen of annual hires were deeply intertwined with the biological cycle of reproduction. One-third of all pregnancies began in the months after the harvest, so many women faced the final trimester of pregnancy at a time when their labor would be of greatest value in the fields or at hired locations. The woman's advanced stage of pregnancy did not usually release her from much of her normal work load. In short, women gave birth in a strong seasonal pattern during the months of highest labor demand and greatest health risk from lowered food supplies and epidemics. To complicate matters, those newborns needed to be weaned to a controlled feeding schedule in time for the mothers to return to fall harvests. Consequently, conception, pregnancy, birth, and breastfeeding often conflicted with the seasonal cycles of productive labor (Cody 1996: 72-73). That structural tension between reproductive and productive labor increased the incidence of maternal and infant mortality. That increased child mortality, in turn, kept mountain slave women trapped by their owners in a vicious cycle of early weaning, high child mortality, and high fertility. Perhaps one mistress explained the paradox most aptly. In order to secure his economic "station," she wrote, the slaveholder had to be "master in all things on his plantation, everything, nothing excepted" (Lumpkin 1971: 22).

Conclusion: Toward a New Paradigm of the U.S. Slave Family

            In his slavery capstone study, Fogel (1989) argues that enslavement was morally reprehensible because owners denied to African-Americans freedom from domination, economic opportunity, citizenship, and cultural self-determination. Fogel is narrow-minded and sexist in his exclusion of the forcible removal of kin and masters’ disruptions of black households from his moral indictment. While I strongly endorse his call for an “effort to construct a new paradigm on the slave family,” I would hope to see writers assign greater priority to the human pain of family separations than has occurred over the last three decades. Celebration of resistance and cultural persistence to the exclusion of investigations of those forces that broke families will not advance a new school of thought in directions that are any more accurate and reliable than previous generalizations.

            The world market in cotton drove the interstate slave trade in the United States and stimulated the export of slaves from small plantations. The dominant paradigm fails to take into account the real impact of that export of laborers on slave families. Indeed, the best rationale for a new paradigm can be heard in the painful voices of African-Americans. Elderly ex-slaves mourned the loss of parents, spouses, children, siblings, and grandparents. Even when they had been separated from kin at very early ages, they sensed that a significant element of their souls had been wrenched from them. A mountain slave, says it best: “We never met again. . . . That parting I can never forget” (Hughes 1897: 6). In the minds of black Appalachians, poverty, illiteracy, and racial inequality were not the worst legacies of enslavement. Bad as those structural factors were, it was the forced removals of family that broke their hearts and generated a community wound that was not healed by liberation. Moreover, half the Appalachian ex-slaves carried into the twentieth century the structural impacts of past diasporas, exacerbated by new family separations borne of a chaotic war and an inhumane emancipation process. Footnote

Bibliography of Archival and Published Sources

Bailey, John L. Collection, William L. Perkins Library, Duke University Manuscripts, Durham, North Carolina

Berlin, Ira, Marc Favreau and Steven F. Miller, eds. 1998. Remembering Slavery: African Americans Talk about Their Personal Experiences of Slavery and Emancipation. New York: New Press.

Cody, Cheryll A. 1996. "Seasonality in Women's Lives on Low Country Plantations." Pp. 61-78 in More than Chattel: Black Women and Slavery in the Americas, edited by David B. Gaspar and Darlene C. Hine. Bloomington: Indiana University Press

Coleman, J. Winston Papers on Slavery, University of Kentucky Special Collections, Lexington, Kentucky

Crawford, Stephen. 1980. "Quantified Memory: A Study of the WPA and Fisk University Slave Narrative Collections. Ph. D. diss., University of Chicago.

-----. 1992. "The Slave Family: A View from the Slave Narratives. Pp. 331-50 in Strategic Factors in Nineteenth Century American Economic History, edited by Claudia Goldin and Hugh Rockoff Chicago: University of Chicago Press.

Dunaway, Wilma A. 1996. The First American Frontier: Transition to Capitalism in Southern Appalachia, 1700-1860. Chapel Hill: University of North Carolina Press.

-----. 2003a. The African-American Family in Slavery and Emancipation. Cambridge: Cambridge University Press.

-----. 2003b. Slavery in the American Mountain South. Cambridge: Cambridge University Press.

Fogel, Robert W. 1989. Without Consent or Contract: The Rise and Fall of American Slavery. New York: W.W. Norton.

----- and Stanley L. Engerman. 1974. Time on the Cross: The Economics of American Negro Slavery. Boston: Little, Brown and Co., vol. 1.

Frisch, Rose and J.W. McArthur. 1974. "Menstrual Cycles: Fatness as a Determinant of Minimum Weight and Height Necessary for the Maintenance of Onset." Science 435: 949-51.

Genovese, Eugene D. 1974. Roll, Jordan, Roll: The World the Slaves Made. New York: Random House.

Gutman, Herbert. 1976a. The Black Family in Slavery and Freedom, 1750-1925. New York: Pantheon Books.

-----. 1976b. “Persistent Myths about the Afro-American Family.” Journal of Interdisciplinary History 6: 181-210.

Hubard Family Papers, Special Collections Department, Alderman Library, University of Virginia, Charlottesville, Virginia

Hughes, Louis. 1897. Thirty Years a Slave: From Bondage to Freedom. Reprint. New York: Negro Universities Press, 1969.

Lenoir Family Papers, Southern Historical Collection, University of North Carolina, Chapel Hill, North Carolina

Lumpkin, Katherine D. 1971. The Making of a Southerner. Westport, CT: Greenwood Press.

Maher, Vanessa.1992. The Anthropology of Breastfeeding: Natural Law or Social Construct? Oxford: Berg Publishers.

Patterson, Orlando.1982. Slavery and Social Death: A Comparative Study. Cambridge: Harvard University Press.

Perdue, Charles L. T.E. Barden and R.K. Phillips, eds.1976. Weevils in the Wheat: Interviews with Virginia Ex-slaves. Charlottesville: University Press of Virginia.

Phillips, Peter D. 1987. "Incorporation of the Caribbean, 1650-1700." Review of the Fernand Braudel Center 10: 781-804.

Rawick, George P., ed. 1972. The American Slave: A Composite Autobiography. Westport, CT: Greenwood Press, 19 vols.

-----, ed. 1977. The American Slave: A Composite Autobiography, Supplement I. Westport, CT: Greenwood Publishing, 12 vols.

Ross, David Letterbook, 1812-1813, Virginia Historical Society, Richmond, Virginia

Savitt, Todd L. 1994. Medicine and Slavery: The Diseases and Health Care of Blacks in Antebellum Virginia. Urbana: University of Illinois Press, 1978.

Steckel, Richard H. 1979. "Slave Mortality: Analysis of Evidence from Plantation Records." Social Science History 3 (3): 86-114.

-----. 1986. "A Peculiar Population: The Nutrition, Health, and Mortality of American Slaves from Childhood to Maturity." Journal of Economic History 46: 721-41.

-----. 1996. "Women, Work and Health under Plantation Slavery in the United States." Pp. 43-60. In More than Chattel: Black Women and Slavery in the Americas, edited by David B. Gaspar and Darlene C. Hine. Bloomington: Indiana University Press.

Stevenson, Brenda E. 1996. Life in Black and White: Family and Community in the Slave South. New York: Oxford University Press.

Sutch, Richard. 1975. "The Breeding of Slaves for Sale and the Westward Expansion of Slavery, 1850-1860." Pp. 173-210 in Race and Slavery in the Western Hemisphere: Quantitative Studies, edited by S. L. Engerman and E. D. Genovese. Princeton: Princeton University Press.

Tadman, Michael. 1989. Speculators and Slaves: Masters, Traders, and Slaves in the Old South. Madison: University of Wisconsin Press.

Wallerstein, Immanuel. 1980. The Modern World-System II: Mercantilism and the Consolidation of the European World-Economy, 1600-1750. New York: Academic Press.

-----. 1989. The Modern World-System III: The Second Era of Great Expansion of the Capitalist World-Economy, 1730-1840s. New York: Academic Press.