Harvard professor Annette Gordon-Reed—who has won numerous awards, including a Pulitzer Prize, for her book The Hemingses of Monticello (2009)—is widely acknowledged to be the world’s foremost authority on the nature of the relationship between Thomas Jefferson and his toothsome slave, Sally Hemings. In chapters 13 through 16 of that book and in several subsequent interviews and talks, she goes into considerable speculative detail—e.g., Jefferson seducing or perhaps even raping Hemings in Paris and Hemings, while pregnant, returning to Monticello with Jefferson—concerning how events likely unfolded between the two.

The skeleton she fleshes out for her story is the account of Sally’s son Madison Hemings in the Ohio paper, Pike County Republican (13 Mar. 1873), published by Samuel F. Wetmore. The memoir, titled “Life among the Lowly, No. 1,” was the first in a series of articles designed perhaps to revitalize a slumping newspaper. What bigger bomb could have been dropped than a story about Jefferson fathering a child by one of his slaves? States the 68-years-old Madison Hemings, “During that time [while Jefferson was minister to France] my mother became Mr. Jefferson’s concubine, and when he was called back home she was enciente [sic; Fr. enceinte] by him.” He continues: “Soon after their arrival, she gave birth to a child, of whom Thomas Jefferson was the father. It lived but a short time. She gave birth to four others, and Jefferson was the father of all of them. Their names were Beverly, Harriet, Madison (myself), and Eston—three sons and one daughter.”

Those and other claims of Madison Hemings are today customarily taken at face value by historians. Gordon-Reed, in “Why Jefferson Scholars Were the Last to Know,” asserts boldly, “The most important historical witness in this story is undoubtedly Madison Hemings.” If so, then much rides on the credibility of the story. If Madison is a trustworthy witness, then Jefferson is implicated in the paternity of Madison and his siblings.

There is one unsettling issue. Gordon-Reed has turned doubting the memoir into a racial issue. In Thomas Jefferson and Sally Hemings: An American Tragedy (1997), she notes that prior to the DNA results which showed that Jefferson could be the father of Eston Hemings (Madison’s brother), scholars have not taken seriously the memoir because Madison was black. Blacks have traditionally been marginalized because of their color.

Because of Gordon-Reed’s protestations concerning the marginalization of black figures by historians—and there is much to be said on behalf of those protestations—Madison Hemings’s memoir concerning Jefferson being his and his siblings’ father is nearly universally accepted today by Jeffersonian historians. So too are many other claims. Ought they to be? Is the issue overall one of color? What about evidence?

Much rides on the truthfulness or untruthfulness of key claims in Madison’s memoir. In his Literary Commonplace Book, Jefferson, commonplacing Lord Bolingbroke (§57), says:

A story circumstantially related, ought not to be received on the faith of tradition; since the last reflection on human nature is sufficient to shew [sic] how unsafely a system of facts and circumstances can be trusted for it’s preservation to memory alone, and for it’s conveiance [sic] to oral report alone; how liable it must be to all those alterations, which the weakness of the human mind must cause necessarily, and which the corruption of the human heart will be sure to suggest.

For history to be authentic, Jefferson, continuing to copy Bolingbroke, adds that “these are some of the conditions necessary.” He comes up with 4 conditions, which I’ll refer to as CN1–CN4:

1. it must be writ by a cotemporary author, or by one who had cotemporary materials in his hands. 2. it must have been published among men who are able to judge of the capacity of th[e] author, and of the authenticity of the memorials on whic[h] he writ. 3. nothing repugnant to the universal experience of mankind must be contained in it. 4. the principal facts at least, which it contains, must be confirmed by collateral testimony, that is, by the testimony of thos[e] who had no common interest of country, of religion, or of profession, to disguise or falsify the truth.

What do we get if we apply Bolingbroke’s sensible “conditions necessary” to Madison’s testimony?

CN1 is not met. Madison writes of events—e.g., his mother staying at Jefferson’s hotel in France and not with daughter Maria, Sally Hemings’s son Tom Jefferson’s birth back at Monticello and death shortly thereafter, Dolly Madison insisting on naming Madison at a visit to Monticello at the time of Madison’s birth—at which he was not present. Many of his claims are events that occurred prior to him being born. He was not an eyewitness, and thus could only have heard circumstantially of such events.

CN2 is not met. The publisher, Wetmore, was a Republican Party activist, and therefore, very likely anti-Jeffersonian. His paper would have had much to gain by publishing scandalous material about the deceased president. Moreover, the phrasing throughout strongly suggests either that Wetmore wrote the memoir, given orally to him by Madison, or that he greatly edited what Madison wrote. Gordon-Reed concedes the former. Madison admits in the testimony: “I learned to read by inducing the white children to teach me the letters and something more; what else I know of books I have picked up here and there till now I can read and write.” It seems clear that he did not write the memoir.

CN3 is met. What Bolingbroke aims at are claims that belie common human experiences—for instance, testimony of a human head growing on a tree stump or of Jesus vinifying water. There are questionable claims in the memoir—e.g., that Sally Hemings was pregnant upon returning to Monticello and that she bartered with Jefferson before returning with him (she threatened to stay in France, where she would be free, unless “her children should be freed at the age of twenty-one”)—and claims that are very likely false—e.g., that Dolley Madison “begged the privilege of naming me [Madison Hemings]” (she was likely in Washington at the time of Madison’s birth)—and still other claims that are manifestly false—e.g., that Jefferson “had but little taste or care for agricultural pursuits”—but there are no claims that belie common human experience.

Finally, CN4 is not met. The principal claims of Madison’s memoir are not vetted. They have not been confirmed independently by credible and dispassionate persons. They have instead been contradicted by eye-witnesses: e.g., Monticello overseer Edmund Bacon, who stated concerning Jefferson’s avowed paternity of Harriet Hemings [Sally’s daughter]: “She was not his daughter; she was *****’s daughter. I know that. I have seen him come out of her mother’s room many a morning when I went up to Monticello very early.”

Given Gordon-Reed’s concession that Madison Hemings is “the most important historical witness” of the presumed liaison, we find that there are good reasons for doubting key claims of Madison’s testimony. The testimony fails to live up to three of Bolingbroke’s (and most likely Jefferson’s) four conditions necessary for authentic history.

Gordon-Reed and numerous other pro-paternity Jeffersonian scholars have stated that the DNA evidence by itself does not show Jefferson to be Eston’s father. It cannot. It only picks out Thomas Jefferson as a possible father, and there are other likely candidates such as brother Randolph Jefferson or his sons. Consequently, historical evidence must decide the issue, and as Gordon-Reed concedes, Madison Hemings’s testimony is the key piece of historical evidence. Thus, if there are good reasons for doubting the “testimony,” then historians are in no position to assert categorically or with likelihood that Jefferson was the father of Madison, or Eston, or any of Sally Hemings’s children. There are, I have shown, good Bolingbrokean reasons for doubting the testimony.