Assessing T. J. Randolph as a Source

I’ve been quoting and analyzing two statements about Thomas Jefferson, Sally Hemings, and the relationship between them. Both documents—letters from biographer Henry S. Randall and from Jefferson granddaughter Ellen Randolph Coolidge—relied on oral statements from the President’s oldest grandson, Thomas Jefferson Randolph (shown at left). How well does that information stand up to scrutiny? Not well. In fact, I don’t think we can avoid the following conclusions.

1. Thomas Jefferson Randolph was not a reliable source on the question of Sally Hemings’s children.

Randolph told Randall and Coolidge several things about why his grandfather couldn’t have been the father of any of the Hemings children, and who secretly was the father of them all. But those statements are contradicted by:

• documentary evidence: Sally Hemings and Thomas Jefferson were not separated for fifteen months before the birth of any child.
• architectural evidence: Monticello had no bedroom close enough for Randolph to hear Jefferson’s breathing at night.
• biological evidence: Patrilineal descendants of the youngest Hemings child, Eston, don’t share the same Y chromosome as patrilineal relatives of the Carr brothers, Peter and Samuel.
• chronological evidence: Randall understood that “At the periods when these Carr [i.e., Hemings] children were born, he, Col. Randolph, had charge of Monticello.” Randolph, born in 1792, was three years old when Hemings had her first immediately documented child, and still only sixteen when she had her last.
  
• each other: Different people recorded hearing Randolph identify two different Carrs as having had a long-term, exclusive sexual relationship with Sally Hemings and fathering all of her children.

It’s possible to come up with explanations for some of those contradictions. For example, Randolph might have consistently named Samuel Carr as the father, but years later Randall wrongly remembered hearing the name of Peter Carr instead. It’s also possible that Randolph gave out some accurate information, but stretched the truth to seem more authoritative or to make his explanations cover a longer time. But those suppositions all come back the point above: we can’t rely on the statements that ultimately come from T. J. Randolph.

Most of those statements’ contradictions were apparent when the Randall and Coolidge letters were first published in full. Nevertheless, until Annette Gordon-Reed published Thomas Jefferson and Sally Hemings: An American Controversy, most Jefferson biographers took Randolph’s statements about the Hemings children as the most authoritative on the subject.

2. There’s no reliable evidence pointing to Peter or Samuel Carr as father of Sally Hemings’s children.

Only two people appear to have ever mentioned the Carrs as having any contact with Hemings at all: Henry S. Randall and Ellen Randolph Coolidge. And they got their information from Randolph—contradictory and unreliable information. Knowing that, we must conclude that either Peter or Samuel Carr is no more likely to have fathered the Hemings children than any other white man living within seven miles of Monticello. They just happened to be the householders related to Jefferson who lived closest.

In fact, the Carrs are less likely to have fathered the Hemings children than other candidates because:

• Eugene Foster’s genetic study showed that the Carr Y chromosome doesn’t match the Hemings Y chromosome. (The Jefferson Y chromosome does match.)
• No one in the Hemings family and no one in the vicinity of Monticello except for Randolph told people that either Carr was the father. (Madison Hemings and some neighbors said that Thomas Jefferson was the father; others denied that, but didn’t record alternative names.)

Of course, given the paucity of evidence two hundred years later, and the difficulty of proving any paternity without D.N.A. from parents and child, it’s still possible to imagine ways that one of the Carrs could have had some children with Sally Hemings. For example, she could have had children with multiple men, but so secretly that no one at Monticello remarked on that possibility, and the only genetic test available happens to apply only to a child fathered by a Jefferson instead of a Carr. That’s possible. But it’s a less likely explanation than what’s already on the table: Madison Hemings’s recollection, supported by other statements from the time.

3. We have to ask what Thomas Jefferson Randolph might have been hiding.

Normally historians treat people’s statements and recollections as reflecting how they honestly see their world—unless there’s evidence to the contrary. Then we have to consider whether they were mistaken, or shading the truth, or deliberately lying. For example, Henry S. Randall was probably sincere when he wrote:

It so happened when I was afterwards examining an old account book of the Jeffersons I came pop on the original entry of this slaves birth: and I was then able from well known circumstances to prove the fifteen months separation—but those circumstances have faded from my memory.

Still, Randall was mistaken. Possibly influenced by Randolph’s assurance that there was a fifteen-month gap, he overlooked one of Jefferson’s trips to Monticello nine months before Sally Hemings gave birth.

Similarly, we can consider that Randolph was sincerely mistaken about that same fifteen-month gap. But it’s harder to find a simple explanation of why he thought he could hear his grandfather breathing at night, or why he’d tell Randall about his Carr cousins wailing over what they’d done to his grandfather’s reputation and not tell his sister. And then there’s the fact that he wanted to keep his statements from being subjected to public scrutiny.

So without concluding that Randolph definitely was stretching the truth or lying, we should consider that possibility. What would that imply about the whole situation?

What would have motivated Randolph to lie? He said that shortly before his mother died he’d promised her that he would “defend the character” of Thomas Jefferson. Randolph was clearly trying to shape how Randall and subsequent historians wrote about his grandfather. He offered an explanation of the Hemings children alternative to the one discussed most in the press until then: that they were Jefferson’s children.

Perhaps Randolph believed in that explanation, but thought it needed more evidence behind it and tried to fill in some holes. And perhaps he was trying to conceal facts that he feared would reflect badly on his grandfather, his family, and/or himself. In the latter case, what might he have been trying to hide?

Though Randolph defended his grandfather’s character, he wasn’t so careful about his Carr cousins’ reputations, or the Hemings sisters’. He said that there was a lot of sex between white men (“Irish workmen,” “dissipated young men in the neighborhood”) and black women at Monticello. Randolph said his grandfather “was extremely indulgent” toward some young male relatives, and “the idea of watching them for faults or vices probably never occurred to him.” That’s a long way from the eighteenth-century ideal of a patriarch in control of his household (not that many patriarchs truly were). So if Randolph was trying to deny or conceal something, his statements would be evidence of something even more embarrassing than that.