There is no evidence of any other slave at Monticello named “Tom” who might have been the son of Sally Hemings,54 and indeed no record either of Thomas Woodson. Jefferson did not maintain his Farm Book records between 1783 and 1794,55 so the absence of a record of birth may not be dispositive. (However, most, if not all, births during that period are presumably reflected in later lists in the Farm Book.) Of greater significance is the absence of any listing of a slave child named “Tom” between 1795 and well after Callender’s 1802 allegations. Some have suggested that Jefferson might well have decided to conceal the birth of his own illegitimate child,56 but he routinely recorded the births of all of Sally’s other children.
We may never know whether there was ever a “Tom Hemings” at Monticello. Other than James Callender’s allegations, the only meaningful evidence for such a child has been the strong oral history of the Woodson family. Whether one concludes that there was no “Tom,” or that the Woodson family tradition is accurate and their ancestor was the young man repeatedly mentioned by Callender and Federalist critics, one thing is now clear beyond reasonable doubt: Based upon DNA testing of six descendants of three of Thomas Woodson’s sons, Thomas Jefferson could not have been his father.57 That may be the most significant finding of the DNA tests. Professor Gordon-Reed argued: “If he [Callender] was lying about that [Jefferson’s paternity of a ten- to twelve-year-old Hemings child named “Tom”], he was lying about the Jefferson-Hemings liaison. … ”58
Finally, of course, it is clear that the DNA tests addressed only the issue of the paternity of Eston Hemings and Thomas Woodson. No continuous male-line descendants of any of Sally Hemings’ other children had been identified when the tests were done, and thus the tests provide no scientific evidence concerning the paternity of Harriet I, Beverly, Harriet II, or Madison Hemings. Specifically, the Foster study does not rule out the possibility that one or more of these older children were fathered by Peter and/or Samuel Carr, who reportedly admitted having fathered children by Sally Hemings. This issue will be addressed in Chapter Ten.
Other Possible DNA Testing—William Hemings and
Thomas Jefferson
It is worth noting that the grave of Madison Hemings’ son, William Beverly Hemings, was recently discovered in the U.S. Military Cemetery at Ft. Leavenworth, Kansas. According to Herbert Barger, there are no known descendants of William; and other descendants of Madison Hemings reportedly have refused to cooperate in having the body exhumed to search for usable DNA that might shed further light on this issue. That is presumably their option, although in the absence of known direct descendants of the deceased, it may be that authority to make this decision rests within the government. In any event, our understanding about this issue might well be enhanced by such an effort. That said, before any steps are taken to exhume anyone, some serious thought needs to be given to the ethical implications of disturbing gravesites and human remains to satisfy historical curiosity.
I am told that some members of the Hemings family have suggested that they will consider the exhumation of William’s remains only if the Jefferson descendants agree to dig up Thomas Jefferson. At first this may seem only equitable, but the proposal that Thomas Jefferson’s remains also be exhumed misunderstands the science involved in Y-chromosome DNA testing. It is theoretically possible, of course, that Thomas Jefferson was of illegitimate birth and thus might not share the Y chromosome found in the descendants of his cousins. But no serious scholar has suggested such a possibility (which, if true, would exclude him as a candidate for Eston’s paternity), and if usable DNA could still be extracted from Thomas Jefferson’s remains, it would almost certainly possess the same Y chromosome as the other male Jeffersons. Those who argue that Thomas Jefferson was probably not the father of Sally Hemings’ children do not deny that he was the legitimate son of Peter Jefferson and carried the same Y chromosome extracted from the blood of modern-day descendants of his cousins.
In contrast, I am told by experts that DNA testing of William Hemings’ remains probably would provide enough information to establish that he (and thus Madison) was a descendant of: (1) a male Jefferson, (2) a male member of the Carr family, or (3) someone other than a Jefferson or a Carr. To be sure, the presence of a Jefferson Y chromosome in William’s remains would not conclusively establish that Thomas Jefferson fathered Madison Hemings. It would presumably be interpreted by some as strengthening the case for that conclusion, but there were two dozen other potential fathers carrying the same Y chromosome. On the other hand, if William were shown to carry the Y chromosome of the Carr family or to be unrelated to both the Carrs and the Jeffersons, that would strongly suggest that Sally Hemings was not monogamous and thus would undermine much of the circumstantial case against Thomas Jefferson.
In fairness to the Hemings descendants, the suggestion that William’s remains should be exhumed and examined may seem like a “no-win” proposition. It could undermine their claim to such illustrious ancestry if William did not carry the Jefferson Y chromosome; but, if he did, the doubters would simply note that Randolph Jefferson and two dozen other men remain possible fathers. But this is not a game of chance or a sporting contest; it is a search for truth involving the reputation of one of America’s most beloved Founding Fathers. And ultimately, as a pragmatic matter, if the Hemings family is seen as blocking the search for truth by denying access to potentially important evidence, they may find it difficult to persuade others of the certainty of their conviction that they are, in fact, descendants of Thomas Jefferson.59
Conclusions
In summary, contrary to conventional wisdom, the DNA study conducted by Dr. Eugene Foster and colleagues and reported in Nature did not “prove” that Thomas Jefferson fathered any of Sally Hemings’ children. Rather, it excluded the reasonable possibility that Thomas Woodson was the child of Thomas Jefferson or any other male member of the Jefferson family, and it established the very strong probability that Sally’s youngest son, Eston, was fathered by one of the more than two dozen adult Jefferson men who were in Virginia at the time he was conceived. As will be discussed in Chapter Ten, there is documentary evidence to support the conclusion that at least seven Jefferson men may well have been present at Monticello when Eston was conceived; and apparently no evidence one way or the other concerning the whereabouts of the remaining theoretical suspects.60 Because of his advanced age (sixty four), health, and character, Thomas Jefferson may arguably have been the least likely of the group to have fathered a child by Sally Hemings61—although some of the relatives lived sufficiently far away from Monticello to be less likely candidates on that basis alone. But the DNA tests clearly show that it is possible that Thomas Jefferson was Eston’s father. Since the tests do not address the issue of the paternity of Sally’s other children, they tell us nothing at all about that paternity.
* * *
Footnotes
1. E. A. Foster et al., Scientific Correspondence: Jefferson fathered slave’s last child, NATURE, Nov. 5, 1998, at 27. This article is reprinted in Thomas Jefferson Memorial Foundation Research Committee, Report on Thomas Jefferson and Sally Hemings, Jan. 2000, Appendix A (hereinafter “Monticello Report”).
2. When our report was initially released I saw no problems with the DNA testing. However, to the extent that the purpose of the testing was to determine the paternity of Eston Hemings, the tests used obviously could not do that. A team of eleven scientists, lawyers, and historians writing about biohistory ethics in a 2004 issue of Science magazine observed: “Often, investigators fail to pose an investigative question capable of resolution by genetic testing. For example, Eugene Foster’s 1998 comparative Y-chromosome study of the descendants of Thomas Jefferson and his slave Sally Hemings was intended to establish whether the President had fathered Hemings’ children. Yet the study protocol was inappropriate for determining the paternity of Hemings’ children—the only possible conclusion was that some Jefferson and Hemings male-line descendants had common relatives.” Lori B. And
rews, et al., Ethics: Constructing Ethical Guidelines for Biohistory, SCIENCE, 9 Apr. 2004 vol. 304, at 215, 216. Even accepting that the protocol could not have scientifically established the paternity of Eston Hemings, it is not my personal view that the tests should not have been undertaken. They did contribute to our understanding of this issue by establishing that Thomas Woodson could not have been the son of Thomas Jefferson, and it could have ruled out Thomas Jefferson as the father of Eston Hemings as well had Eston’s descendants not carried the Y-chromosome found in the Jefferson family. Thus, while I don’t dispute the observation by Andrews, et al., that Foster’s protocol could not have determined Eston’s paternity, I believe the tests might have contributed to our understanding of these issues had they not been inaccurately sensationalized by Nature and the popular press. Dr. Foster was very candid in acknowledging this inherent shotcoming in any test in the absence of a sample of Thomas Jefferson’s DNA, so I cannot fault him for doing a useful if imperfect study.
3. One might try to be charitable and assume they were just asserting that some “Jefferson” was Eston’s father, but the title is so likely to be perceived as referring to the former President as to be irresponsible even without the less ambiguous references discussed below.
4. Eric S. Lander & Joseph J. Ellis, Founding Father, NATURE, Nov. 5, 1998, vol. 396, issue no. 6706 at 13.
5. JOSEPH J. ELLIS, AMERICAN SPHINX 366 (1996).
6. JOSEPH J. ELLIS, FOUNDING BROTHERS: THE REVOLUTIONARY GENERATION (2000).
7. See, e.g., Eliot Marshall, Genetics: Which Jefferson Was the Father? SCIENCE, Jan. 8, 1999 at 153–54 (“Foster agrees that the headlines were ‘misleading’ because they suggested that the data were conclusive. He attributes this ‘unfortunate’ slipup to the haste with which his article and the Lander-Ellis essay went to press. …Nature staffer Rosalind Cotter agrees that ‘the whole thing really was rushed through.’”) See also, Andrew Cain, Journal backs off on Jefferson Report: Says there is no way to prove he had child with slave, WASH. TIMES, Jan. 7, 1999 at A1. In his discussions with me, Dr. Foster asserted that the rush to publish the articles was unrelated to the upcoming congressional election but rather was necessary because Professor Ellis had given the story to U.S. News & World Report, which was about to make it public.
8. The issue was scheduled for release on Tuesday, November 3, which was Election Day in the United States, but was released instead on Friday, October 30.
9. For example, Professor Ellis was one of “four hundred professional historians” who signed an advertisement in the New York Times on Oct. 30, 1998, denouncing the impeachment effort.
10. “Politically, the Thomas Jefferson verdict is likely to figure in upcoming impeachment hearings on William Jefferson Clinton’s sexual indiscretions, in which DNA testing has also played a role. The parallels are hardly perfect, but some are striking.” Lander & Ellis 13.
11. Quoted in Sean Wilentz, What Tom and Sally Teach Us, NEW REPUBLIC, Nov. 30, 1998, at 14.
12. “Jefferson fathered slave’s last child”—journal article raises a question of credibility, NATURAL SCIENCE, Mar. 19, 1999, available on line at: http://naturalscience.com/ns/articles/edit/ns_ed05.html.
13. Kevin Maney, DNA Test Basks in Simpson Spotlight, USA TODAY, Dec. 2, 1994, at 2B.
14. Andy Soltis, Evidence Bolsters Cops’ Case, BOSTON HERALD, June 18, 1994, at 2.
15. It is of course equally correct to describe the sources of the DNA as descendants of Thomas Jefferson’s paternal uncle (Field Jefferson, elder brother of Thomas Jefferson’s father, Peter Jefferson), as Dr. Foster does, or as descendants of Thomas Jefferson’s paternal grandfather, as Professors Lander and Ellis do in Nature. The key point, conceded by all involved when pressed, is that the data from the DNA study do not point towards Thomas Jefferson’s paternity any more strongly than they do towards any of his male relatives. To be sure, there are other variables that should influence a reasonable judgment in this matter—including proximity and opportunity when Eston was conceived, age and health of the possible fathers, character, and the like—that will be discussed below.
16. Some questions have been raised about the failure to follow all of the sample control procedures required for DNA testing to be admissible in a court of law. I personally am not especially troubled by this. Realistically, while there may have been somewhat greater risk of samples being inadvertently mixed up, such an error would presumably be far more likely to produce a false negative than a false positive result. My strong sense is that Dr. Foster and his colleagues in this enterprise are highly professional individuals whose work warrants a strong presumption of validity.
17. The difference is based upon whether one considers all Jefferson males of sufficient age to father a child and believed to have been in Virginia around the time Eston was conceived, or narrows the field of suspects to only those likely to have been in the vicinity of Monticello at the time—such as brother Randolph Jefferson and his sons, who lived near Monticello and were invited to visit shortly before Eston’s estimated conception.
18. Thomas Jefferson inherited at least two slaves from his father who were thought to have been fathered by a white man. If their father (or grandfather) was their owner (Thomas Jefferson’s father or grandfather), or another visiting member of the Jefferson family, they would have carried the same Y chromosome as Thomas Jefferson. We do not give this possibility any weight in our analysis, but we also cannot absolutely rule out the possibility that older Monticello slaves carried the Jefferson chromosome.
19. When this chapter was written in 2000, I was working from the data provided in the Monticello Report about Randolph Jefferson’s children. They estimated James Lilburne Jefferson’s year of birth as circa 1789. (Monticello Report, Appendix J at 3.) I am now less confident that James was seventeen or eighteen years old; so elsewhere you will find me counting only the four of Randolph’s children (all sons) we know were well past their eighteenth birthday—an age which certainly would make them eligible suspects for the paternity of Eston Hemings.
20. We know that Thomas Jefferson was at Monticello at least part of the period in which Sally Hemings probably conceived Eston, and we know that he had invited his younger brother, Randolph (a widower who had at least four sons old enough to father a child and lived only about twenty miles from Monticello) to visit during this same period. This issue will be addressed in greater detail in Chapters Five and Ten.
21. MEMOIRS OF A MONTICELLO SLAVE, reprinted in JEFFERSON AT MONTICELLO 22 (James A. Bear, Jr. ed., 1967.) The description by former slave Isaac Jefferson is of Thomas Jefferson’s younger brother, Randolph, who will be discussed in Chapter Ten.
22. For an excellent summary of the numerous articles and columns that misreported the DNA evidence in the Washington Post, see E. R. Shipp, Reporting on Jefferson, WASH. POST, May 30, 1999, at B6. An even earlier critique of media coverage of the story in general was provided in the same newspaper by David Murray, Paternity Hype Visits Monticello, WASH. POST, Nov. 15, 1998 at C1.
23. Jan Ellen Lewis & Peter Onuf, Introduction to SALLY HEMINGS AND THOMAS JEFFERSON 11.
24. Rhys Isaac, Monticello Stories Old and New, in SALLY HEMINGS AND THOMAS JEFFERSON 119.
25. Eric S. Lander & Joseph J. Ellis, Founding Father, NATURE, Nov. 5, 1998, at 13, reprinted in Monticello Report, Appendix A.
26. Joseph J. Ellis, Jefferson: Post-DNA, 57 WILLIAM & MARY QUARTERLY 126 (2000).
27. Jan Lewis, Thomas Jefferson and Sally Hemings Redux: Introduction, in SALLY HEMINGS AND THOMAS JEFFERSON at 121. See also the assertion that Jefferson “almost certainly” fathered at least one of Sally Hemings’ children by Yale University Professor David Brion Davis, in his Preface to LUCIA STANTON, FREE SOME DAY: THE AFRICAN-AMERICAN FAMILIES OF MONTICELLO 12 (2000).
28. Wilentz, What Tom and Sally Teach Us at 14.
29. In 2001 this organization was renamed “Thomas Jefferson Foundation,” or TJF.
30. The full text of this insigh
tful interview can be found on the PBS web page at http://www.pbs.org/ jefferson/archives/interviews/Jordan.htm.
31. Statement of Daniel P. Jordan, Ph.D., president, Thomas Jefferson Memorial Foundation, DNA Press Conference at the International Center for Jefferson Scholars, November 1, 1998, reprinted in Monticello Report, Appendix D.
32. Eugene A. Foster, The Thomas Jefferson Paternity Case, NATURE, Jan. 7, 1999, vol. 397 at 32.
33. Eugene A. Foster, In Jefferson-Hemings Tie, a Family’s Pride; Tenable Conclusions, letter to the editor, N.Y. Times, Nov. 9, 1998, at A24.
34. Indeed, all of the letters on this issue in the January 7, 1999, issue of Nature challenged the interpretation that the DNA study pointed to Thomas Jefferson as Eston’s father. Professor David M. Abbey, of the University of Colorado Health Science Center, noted “the authors did not consider all the data at hand in interpreting their results. No mention was made of Thomas Jefferson’s brother Randolph …or of his five sons.” Dr. Gary Davis, of Evanston Hospital, in Evanston, Illinois, added that “If the data of Foster et al. are accurate, then any male ancestor in Thomas Jefferson’s line, white or black, could have fathered Eston Hemings.” The Thomas Jefferson paternity case, NATURE, vol. 397, Jan. 7, 1999 at 32.
35. I am not personally critical of Dr. Foster for his role in this controversy, but it may be worth noting that he did not enter into the project as a “neutral.” In his discussions with me, Dr. Foster acknowledged that he originally undertook the project hoping to prove that Thomas Jefferson fathered Sally Hemings’ children.
36. Foster et al., Jefferson fathered slave’s last child at 27 (emphasis added).
37. Id. (emphasis added).
38. Excerpted in Monticello Report, Appendix B (emphasis added).
The Jefferson-Hemings Controversy Page 12