When the Past Speaks to the Present: Thomas Jefferson and Sally Hemings

Annette Gordon-Reed is the Charles Warren Professor of American Legal History and a professor of history at Harvard University. Her books include The Hemingses of Monticello: An American Family (2008), which received the Pulitzer Prize, the George Washington Book Prize, and the Frederick Douglass Book Prize, and “Most Blessed of the Patriarchs”: Thomas Jefferson and the Empire of the Imagination (2016), co-authored with Peter S. Onuf. Her honors include a Guggenheim Fellowship in the humanities, a MacArthur Fellowship, the National Humanities Medal, and the National Book Award.

Recent years have witnessed an explosion of interest in, and historical scholarship about, American slavery. Both in the academy and outside of it, Americans have come to realize that part of our national consciousness was shaped between 1619 and 1865, when racially based slavery flourished in North America. Historians have always written about slavery, of course, even when the institution was in place. But with the exception of the works of black scholars like W. E. B. Du Bois and Carter G. Woodson, most historical writing about slavery in the nineteenth and twentieth centuries was produced by white southerners, who focused on slavery from the perspective of the slave-owning class rather than the enslaved.

This focus on slavery from the slave owners’ perspective no longer dominates the field. Beginning with Kenneth Stampp’s The Peculiar Institution, scholars have produced works that strive to examine slavery from the perspective of those who suffered under it and to reconstruct their lives, their values, their customs—and, above all, the role they played in shaping the world in which they lived. Yet it is important to realize that in reconstructing that past, scholars face multiple challenges, among them evaluating two quite different forms of evidence: document-based evidence and the evidence based on oral traditions. Equally important and challenging is the need to subject claims made by masters and those made by slaves or the children of slaves to the same rigorous standards of proof. For, as the history of the debate over the relationship between Thomas Jefferson and Sally Hemings demonstrates, scholars may still be inclined to give more weight to the records and accounts of the powerful, the famous, and the educated, and to their descendants, rather than to the stories told by those such as slaves, women, and the poor who stand outside the circles of power and authority.

No better example of this pattern can be found than the long and highly contested debate over the relationship between Thomas Jefferson and Sally Hemings, who was one of Jefferson’s slaves. Although eighteenth-century gossip and newspaper stories suggested that Jefferson was the father of Madison, Eston, and the other Hemings offspring, twentieth-century historians proved reluctant to credit these stories. Two written documents—a letter and a memoir created by grandchildren of Thomas Jefferson—seemed to provide adequate proof that the claims of Jefferson’s paternity were false. One of these grandchildren stated that Jefferson’s nephew, Samuel Carr, fathered all of Hemings’s children; the other insisted that Samuel’s brother, Peter Carr, was their father.

But a third account also existed. This came from Madison Hemings, Sally Hemings’s son. Madison Hemings provided his version of the relationship between Jefferson and his mother in an interview with an Ohio newspaperman in 1873. It was his claim that his mother, Sally, was Jefferson’s longstanding mistress, and Jefferson the father of all of her children.

Clearly, the written "evidence" on the relationship between Thomas Jefferson and Sally Hemings was contradictory. The reliability of all three accounts required investigation. For, as scholars know, the written word is not necessarily inherently truthful; people can make things up and put them in family letters, declare them in written speeches, record them in diaries, or provide them in newspaper interviews. Putting a story on paper does not turn falsehoods into truth. Regardless of the author’s social status, his education level, or the passion of his conviction, the historian must examine the author’s motives, consider how the author acquired the information contained in the document—and then must search for corroborating evidence before what is written is accepted as what is true.

Yet a bias led historians to give more weight to the grandchildren’s accounts than to the account of Madison Hemings, despite the fact that the grandchildren’s accounts were themselves contradictory. No doubt the high esteem in which Thomas Jefferson was held by the public and by the historians contributed to the acceptance of the grandchildren’s claim that Jefferson did not father the Hemings children. But no doubt racism—sometimes conscious and sometimes unconscious—played a role in privileging the grandchildren’s claims over the claims of Madison Hemings. A further complication comes from the fact that the type of corroborating evidence historians are accustomed to look for—other written documents—was rarely produced by enslaved African Americans. The vast majority of enslaved people were illiterate, and even those who were literate rarely contributed to the creation of official records that often give historians information about past lives. Slave marriages were not recognized in law, and slaves did not write wills or deeds—although they appeared in them as items of property. Most of the personal experiences of enslavement live on in a very different form: in the stories passed from one generation of African Americans to the next through an oral history tradition.

This tradition differs from modern-day forms of oral history interviews in which scholars pose questions and record the answers of eyewitnesses to historical events or collect autobiographical materials from those being formally interviewed. It is through this oral history tradition that stories are passed privately and informally from one generation to another. Scholars have been slow to accept oral tradition as historical evidence, for they are aware that stories can be changed or embellished as they pass from one generation to another. This concern that a story told to many people across decades may produce error cannot be lightly dismissed.

Whatever the reasons, this much is clear: Most scholars did not pursue the evidence about the paternity of the Hemings children thoroughly enough. Instead, the grandchildren’s claim that one of the Carr brothers fathered Sally Hemings’s children was accepted. But in 1998, a new, remarkably modern form of evidence corroborated Madison Hemings’s account. In that year, DNA testing on Jefferson, Hemings, and Carr descendants disproved any connection between the Carrs and the Hemingses, and confirmed a connection between the Jeffersons and the Hemingses. The results of the DNA tests, combined with other historical analysis of evidence, demonstrated that the Jefferson family documents were insupportable.

The unraveling of the Jefferson family’s written explanations about who fathered the Hemings children provides a cautionary tale for those who accept at face value written accounts of events. Yet evidence drawn from oral tradition must be scrutinized just as carefully. For example, the descendants of a man named Thomas Woodson also claimed descent from Jefferson and Hemings. No corroborating statement comparable to Madison Hemings’s narrative has been found, however. What we know of the Woodson link to Jefferson and Hemings comes exclusively from generations of Woodsons, who passionately asserted the truth of their oral family tradition. Their claims could not be dismissed out of hand, however, for different branches of the family, who had no contact with one another, had preserved the same account of Woodson’s paternity. Once again, DNA testing was determinate: The DNA results that bolstered the Hemings family tradition totally discounted the Woodsons’ claim.

There are three lessons to be learned from this story. First, we owe it to ourselves and to history not to privilege the masters’ documents over the documents produced by slaves or the descendants of slaves, not to let a bias toward prestige or social standing prevent the thorough investigation of claims. Second, we must not privilege one form of evidence over another. Especially in reconstructing the story of slavery we must remember that slave owners controlled the production of documents and prevented those whom they enslaved from producing their own records of their experiences during slavery. Certainly there was much about their relationships with slaves that masters simply did not want outsiders to know. Finally, we must view both orally transmitted history and document-based history with a trained, critical eye, remembering that both are created by human beings and both are therefore fallible.