Flawed Evidence

The Resemblance Claim 

According to an anecdote in a Chillicothe, Ohio newspaper in 1902, a group of men from Chillicothe thought a statue of Jefferson on the lawn of the White House bore a striking resemblance to Eston Hemings. They were observing a statue by French sculptor Pierre-Jean David (d’Angers), placed in the 1840s by President James Polk. While these men may have thought Eston looked like the statue, there is some conflict whether the statue looked like Jefferson. The best-known “face” of Jefferson is the Houdon bust which Jefferson sat for while he was in Paris. 

The d’Angers statue was commissioned by Uriah Phillips Levy (who would later buy Monticello) and sculpted in 1833. He apparently used the 1821 Thomas Sully portrait, painted when Jefferson was 78 as his model. It was provided by the Marquis de Lafayette.

To peruse the portraits and statues made of Jefferson during his lifetime is to see a thin thread of resemblance among a widely divergent set of faces. Of course, we know that all are supposed to be Jefferson.

The “resemblance argument” was first raised by James Callender in his September 1,1802 claim in the Richmond Recorder that Jefferson had a slave son. He asserts that “Tom,” who was about 12 years old, bore a “striking although sable resemblance” to Jefferson, who would have been 59. This claim evaporated along with Tom when the DNA tests proved that the descendants of the family who claim descent from Tom were not the descendants of Thomas Jefferson. 

Ten years after he published his biography of Jefferson, Henry Randall wrote a letter in 1868 to James Parton, who was preparing a new biography of Jefferson. Randall related a conversation that he had some 15 years earlier with Thomas Jefferson Randolph, Jefferson’s grandson. Randall has Randolph say “she had children which resembled Mr. Jefferson,” and that a dinner guest was “startled” at the resemblance of a servant.  Actually, Randolph is not trying to imply that these “children” had Thomas Jefferson’s blood, but that of “a” Jefferson.

However, the circumstance of time make it highly unlikely Randall had correctly conveyed the context of Randolph’s remarks. The “servant” is not identified, nor is the “startled” guest. Randall does not state when this incident supposedly occurred, but it would have to have been some 35 years old. If it were Hemings’s oldest son Beverley even at age 12, Jefferson would have been 67. There would have been little apparent resemblance between the two – certainly not a startling one. By the time the youngest son Eston was 12, Jefferson would have been 77. The disparity in appearance would become more pronounced as Jefferson aged. Not a single record survives contemporary with the years the Hemingses lived at Monticello that asserted any of them had a likeness to Jefferson. 

The “Single Father” Theory

The DNA tests had implicated a Jefferson male as the father of Eston Hemings, which could have been any anywhere from eight to two dozen men. Prior to the DNA tests, Monticello found the paternity claim against Jefferson “unpersuasive” and “entirely circumstantial.” Although none of this evidence had changed, Monticello’s position after the DNA tests was that “the best evidence available suggests a strong likelihood that Thomas Jefferson and Sally Hemings had a relationship over time that led to the birth of one, and perhaps all, of the known children of Sally Hemings.” 

However, the descendants of Beverly and Harriet were not known, and there were no known living descendants of Madison, so they were not included in the DNA tests. One device to fill this gap in the evidence was the “single father theory.” Monticello declared that if Jefferson were the father of Eston, and since all of Sally’s children had the same father, then Jefferson must be the father of Beverly, Harriet and Madison.

The rationale for this claim was based on the “closeness of the family” as established by Sally’s children naming their children after their siblings. Eston did name his second son Thomas Beverly.  Madison had sons Thomas Eston and William Beverly and a daughter Harriet. It is worth noting that these names were prominent in the Randolph family (Jefferson’s mother was Jane Randolph) rather than in Jefferson’s.

Monticello assumes this “naming” practice was a sociological paradigm among slave families but it cites no empirical studies or references for such an important assumption. 

It is also asserted that this “single father “theory is valid because there is no record that Sally Hemings had more than one lover. Of course, Callender claimed otherwise, but on this point, he is ignored, as is Ellen Coolidge. So, Sally about whom nothing is known is assumed to be monogamous precisely because nothing is known about her having multiple partners. 

The proximity equation  

Pearl Graham first advanced the equation of proximity and the “conception period.” It was more carefully detailed in Jefferson Vindicated of the exact dates of Jefferson’s presence at Monticello. This study reflects that Jefferson was absent during half of the conception windows for Beverly, and there were days he was away during a conception window from Madison and Eston, and a week for an unnamed third child. Since Jefferson was not present during the whole of the conception period, Sally Hemings could have gotten pregnant by someone other than Thomas Jefferson. 

During a thirteen-year span, Hemings gave birth to four children who reached adulthood. The DNA test was on the descendents of one son, who was found to have a Jefferson ancestor. There is no record that Sally Hemings was at Monticello continuously during her conception periods. No observation survives, oral or written, that give any indication Jefferson and Hemings had any physical contact or show of intimacy during all of her childbearing years.

The Monticello Report and William and Mary Quarterly even published an article on a Monte Carlo simulation that concludes Thomas Jefferson was the father of all the children because he was at Monticello each time Hemings conceived. The expertise of the author who conceived this exercise is not revealed, nor his success with similar mathematical simulations. Why his assumptions and conclusions provide no probative value to the paternity issue is analyzed in “Sally Hemings, Thomas Jefferson, and the Authority of Science.”  

The Family Relationship Proposition

 There is also the unproved assumption that Sally Hemings was Martha Jefferson’s half-sister. It has been routinely repeated by historians that Sally Hemings was the daughter of John Wayles, father of Thomas Jefferson's wife, Martha. This would make Sally the half sister of Martha Jefferson. No one who repeats this claim has ever provided any supporting evidence that it is true, other than rumor and innuendo. This issue was examined in Anatomy of a Scandal.  The authors reveal the source of this rumor is no more substantial than an 1805 letter to the editor by an unidentified writer, which caused a new flurry of the Callender slanders. 

This leads some writers to fantasize that Thomas Jefferson saw in Sally Hemings a resemblance to his dead wife. A new genre of Jefferson books have emerged in which the authors use their imagination to construct an intimacy between Jefferson and Hemings, which is not supported by historical fact.