The World Wide Web is old enough that some pages now preserve the historical consensus of the past, the way libraries have long done. For example, from 1990 to 1997, acclaimed filmmaker Ken Burns worked on a public-television documentary about Thomas Jefferson. He interviewed several leading historians and other voices about the man.
The website created to accompany that broadcast has become a time capsule of that moment in intellectual history, and not just because it still uses frames. That film was made before the 1998 Nature study which matched the surviving Y-chromosome from Sally Hemings’s children and the surviving Y-chromosomes from Thomas Jefferson’s patrilineal line.
Among Burns’s interviewees was Merrill D. Peterson, author of Jefferson in the American Mind. In that book and his interview for the documentary, Peterson said that having children with Sally Hemings would be out of character for Jefferson:
Was Sally Hemings his mistress?By the same arguments, William Jefferson Clinton would never have a sexual affair with an intern in the Oval Office after adultery had been a problem for him in the 1992 election, and Strom Thurmond would never have impregnated a sixteen-year-old maid working for his family. In 1996, those arguments seemed more solid than they do now.
Well, I can’t just say no. Because I can’t... No, I do not believe Sally Hemings was Jefferson’s mistress.
Well, the reasons for that go partly to the fact that I think it would have been a moral and psychological impossibility for him to have engaged in that kind of relationship with one of the Monticello slaves, especially a person of that age, of that vulnerability, in the way that it’s generally described. It is said that the relationship began when Jefferson was in Paris and Sally had come to Paris bringing Jefferson’s younger daughter. And she would have been 15 or 16 at the time the seduction took place. And it is said that she came home pregnant by Jefferson and that the first child, one who was said to be named Tom, was born. There is no record of the Tom. But that is another question. There is no historical documentary evidence to support any of this. There is oral evidence. But the oral evidence probably came after the written evidence—after the story was written down by [James] Callender and then by others—so that it could have entered into oral memory after it was written down.
But you don’t think he did.
No, I do not think that she was his mistress, no. That would have required—just to pursue this a little further—that he continue this relationship for a period of 25 years and that two of the children would have been born after the Callender story came out and while he was president of the United States. And that would have been a tremendous presumption on the American public, I think, and public sensitivities in this area as well as his own sensitivities. And so I think it's quite impossible.
In the same series of interviews, historians dismissing the idea of a sexual relationship between Jefferson and Hemings included Andrew Burstein, Joseph Ellis, Lewis Simpson, and Garry Wills, as well as Dan Jordan of Monticello, children’s author Natalie Bober, and translator Stephen Mitchell.
Scholars who were undecided, some saying that the unanswerable question of sex distracted from the undeniable history that Jefferson owned some of his relatives, included James Cox, Paul Finkelman, John Hope Franklin, James O. Horton, and Jan Lewis. Lucia Stanton of Monticello spoke of Hemings and her children’s reminiscences without actually stating what they’d said about their father, at least as transcribed.
Notably, the voices saying they believe the Hemings family account came from outside the ranks of historians: activist Julian Bond, Hemings descendant Robert H. Cooley III, impersonator Clay Jenkinson, and novelist Gore Vidal. (In the rest of the interview transcripts, the name “Hemings” never comes up.)
After the D.N.A. findings, there was a sea change in the historical consensus. Annette Gordon-Reed’s analysis Thomas Jefferson and Sally Hemings: An American Controversy had already shown the relative robustness of Madison Hemings’s memoir and the weakness of the Jefferson family’s explanation. Many historians changed their minds about the likelihood of a relationship. Some might even have done a little scholarly penance, with Wills publishing “Negro President”: Jefferson and the Slave Power in 2003, and Burstein Jefferson’s Secrets: Death and Desire at Monticello in 2005.
Peterson, who died in 2009 at the age of eighty-eight, never wrote or spoke publicly about the Jefferson-Hemings question after the Nature study. His Washington Post obituary reported that he still thought as he had before:
“He did not believe in any sexual connection between Jefferson and Sally Hemings,” Paul M. Gaston, a longtime U-Va. colleague, said yesterday in an interview.Diehard deniers continue to list Peterson as a scholar in their camp, but he never engaged openly with the new evidence.
When the evidence for such a relationship became more persuasive in recent years, Gaston said, Dr. Peterson “didn’t argue with it. He just distanced himself from that discussion.”
The shift in consensus about the Jefferson-Hemings relationship might be a rare example of Thomas Kuhn’s paradigm shift in history rather than science. The P.B.S. website and Burns’s film (now available on D.V.D.) capture the state of the field just before that happened.