The Clinton Impeachment Factor
The story also may have been influenced by the fact that President William Jefferson Clinton was facing impeachment in the Congress on grounds related to sexual misconduct when the Nature story was rushed into print7 only days before the 1998 congressional elections.8 At least one of the contributing authors, Professor Ellis, was an outspoken critic of those proceedings,9 and the Nature article he coauthored emphasized the “striking” parallels of the two cases.10 He wrote that the “dominant effect of this news” would be “to make Clinton’s sins less aberrant and more palatable.”11
Four months after the Nature stories were published, an editorial in Natural Science magazine cautioned:
Both the media and the public at large should be skeptical about all scientific claims until they have been evaluated, not only by peer-reviewed journals, but also in the open forum of scientific and public discussion. In particular, the public should be skeptical about scientific claims that support political interests. When such claims lack intrinsic scientific significance (as in the case of those made in the Foster paper), their publication in a scientific journal should be recognized for what it is: an abuse of the scientific press.12
The DNA Tests
A more important contributing factor than the Nature articles to the misunderstanding may have been the widespread confusion about the nature of the Jefferson-Hemings DNA tests. Most Americans learned about DNA testing during the period leading up to and during the 1995 murder trial of O.J. Simpson, and they read in USA Today and other major papers that DNA “genetic fingerprints” are “99.9% accurate,”13 or even “99.99 percent accurate.”14 When the Jefferson-Hemings DNA story broke four years later, it was not surprising that many people assumed scientists had matched Thomas Jefferson’s DNA with that of one of Sally Hemings’ children, and conclusively established Jefferson’s paternity by this remarkable new technology. But that is clearly not the case.
First, there exists no known DNA from Thomas Jefferson to analyze. And since the only Jefferson children to have children of their own were daughters, there exists no unbroken male line of Thomas Jefferson’s known descendants to test. So Dr. Foster used DNA taken from the blood of descendants of two of Thomas Jefferson’s cousins.15 For the testing that was done, this should not have been a problem. The tests were not designed to place Thomas Jefferson at a crime scene beyond reasonable doubt, but rather simply to ascertain whether one of Sally Hemings’ children was likely fathered by any “male Jefferson.” Since, in the absence of illegitimate birth at some stage, Thomas Jefferson should have had the same male Y chromosome as his paternal male ancestors and their direct-line male descendants—and, very important, his brother and his brother’s sons as well—Dr. Foster’s approach in using descendants of Thomas Jefferson’s cousins was scientifically sound and unobjectionable.16 The problem has been the widespread confusion by scholars, press, and public alike of this more general test—designed merely to ascertain whether some male Jefferson likely fathered a Hemings child—with the far more precise “genetic fingerprint” technique made famous by the Simpson trial.
Considered by itself, the results of the DNA tests in the Jefferson-Hemings controversy suggest that the statistical probability Thomas Jefferson was the father of Sally Hemings’ youngest child range from as high as seventeen to as low as four percent17—or perhaps a bit lower if the possibility of a slave father carrying the Jefferson DNA is considered.18 The DNA tests could not discriminate among the more than two dozen adult male Jeffersons in Virginia at the time Eston Hemings was conceived, and there is reasonable evidence to suggest that at least seven19 of those men (including Thomas Jefferson) may well have been at Monticello when Sally became pregnant with Eston.20 It should be obvious that these DNA tests say nothing about the paternity of any of Sally Hemings’ children whose descendants were not tested.
When one considers such factors as Thomas Jefferson’s advanced age (sixty-four at the time of Eston’s conception), his reputation, his health, and the fact that—unlike Thomas Jefferson—at least one of the other key suspects had a propensity, documented in Memoirs of a Monticello Slave, “to come out among black people, play the fiddle and dance half the night”21 when he visited Monticello, the odds that Thomas Jefferson was Eston’s father would seem to decrease even further. But if that is true, why have so many people across the country and around the world been so misinformed?
Overstating the Results
The problem lies not only with a news media prone to over simplifying and sensationalizing complex stories.22 Numerous prominent scholars have contributed to the misunderstanding by characterizing the DNA study as “confirming”23 or “clinching”24 the case for Thomas Jefferson’s paternity.
In the article accompanying the Foster DNA study in Nature, the respected Professor Ellis announced that, when considered with the circumstantial evidence, the DNA evidence “seems to seal the case that Jefferson was Eston Hemings’ father.”25 Writing later in the William & Mary Quarterly, Ellis concluded that “Jefferson’s paternity of several [sic] Hemings children is proven ‘beyond a reasonable doubt’”26 by the DNA study. This is an absurd statement.
In an introduction to this same issue of the Quarterly, Rutgers University Professor Jan Lewis added that “virtually all professional historians” now accept “that Jefferson was the father of at least one of Sally Hemings’s children. … ”27 I am aware of no reliable survey of opinion on this issue, but to the extent that it is true that professional historians accept the allegations, I suspect it has as much to do with misunderstanding the significance of the DNA tests and widespread respect for Professor Ellis as it does with the actual merits of the case for Jefferson’s paternity. This is being written before most of the professional historians on the Scholars Commission have voiced their final conclusions; but I will be greatly surprised if—after spending nearly a year looking carefully at all of the evidence—anything like “virtually all” of them find it probable that Thomas Jefferson fathered Eston Hemings.
Writing in The New Republic, Princeton historian Professor Sean Wilentz began: “The DNA test proving that Thomas Jefferson fathered at least one child with his slave Sally Hemings was good news. … ”28 Consider also the case of Dr. Daniel Jordan, the respected President (1985–2008) of the Thomas Jefferson Memorial Foundation (TJMF)29 that runs Jefferson’s home at Monticello and promotes scholarship about the famous President. Few if any organizations have done more over the decades to promote the legacy of Thomas Jefferson. When interviewed for the 1997 Ken Burns PBS video, Thomas Jefferson, Dr. Jordan asserted there was “no historical evidence that there was a relationship between Thomas Jefferson and Sally,” noted that such a relationship would be “totally out of character,” and concluded that it was “morally impossible for that relationship to have occurred.”30 And yet, in a press statement on November 1, 1998, Dr. Jordan asserted: “Dr. Foster’s DNA evidence indicates a sexual relationship between Thomas Jefferson and Sally Hemings, an African-American woman who was one of his slaves.”31
Dr. Foster deserves credit for his efforts to correct the misunderstanding. He published letters in both Nature32 and The New York Times33 noting that the headline had overstated the actual conclusions of his work.34 But he also may bear some of the responsibility for the misunderstanding.35 Rather than saying that his findings “provide evidence that he [Thomas Jefferson] was the biological father of Eston Hemings Jefferson”36—which is not inaccurate, but may have been subject to misinterpretation—a more precise phraseology might have been to report that Thomas Jefferson “may have been” Eston’s biological father.
Distinguishing History from Science
Another problem with the scientists’ historical interpretation of their DNA findings was the exclusion of any alternative candidates for Eston’s paternity other than Thomas Jefferson or his nephews by his sister Martha, Peter and Samuel Carr. Thus, the Nature article concludes: “The simplest and most probable ex
planations for our molecular findings are that Thomas Jefferson, rather than one of the Carr brothers, was the father of Eston Hemings Jefferson. … ”37 No consideration was given to any of the two-dozen other Jefferson males in Virginia at the time, each of whom carried the same Y chromosome as Thomas Jefferson and Eston Hemings. But this distinction was not picked up in the press.
More admirable were some of the observations of prominent scientists contacted by the Thomas Jefferson Memorial Foundation and asked to comment on the Nature articles. Particularly impressive were the comments of Dr. Kenneth K. Kidd, Professor of Genetics at Yale University, who recognized some “controversy” over the “interpretation” of the data and wrote:
I think Eric Lander and Joseph Ellis in their News and Views commentary over-interpreted the results as proving that Jefferson was the father of Eston. …How many other male-line relatives of Thomas Jefferson were alive at that time? …[A]s with modern day paternity testing, we can prove a man is/was not the father, but we cannot absolutely prove a man is/was the father.
So the proof ultimately rests on demonstrating that Thomas Jefferson was present at the time Eston was conceived and that no other male relative with the same Y chromosome was hiding in the bushes. That is something I have no knowledge of. …38
Similarly, Dr. David Page, of the MIT Center for Genome Research, “felt that more thought and attention could be paid to the ‘competing hypotheses’ in interpreting the results.”39
In an e-mail to this writer on October 1, 2000, Dr. Foster elaborated on his position at some length.40 He wrote that “The DNA tests do tell us that Eston Hemings was very likely fathered by a member of the Jefferson family,”41 and added: “The scientific evidence alone would have told us that it was possible that Thomas Jefferson was Eston Hemings’ father. …[Y]ou are right that our DNA findings say nothing directly about the paternity of any of Sally Hemings’ children other than Eston Hemings.”42
Dr. Foster has repeatedly voiced the opinion that “it is very likely that Thomas Jefferson was Eston Hemings’ father.”43 Such statements may have contributed to public misunderstandings of the results of his study—in spite of his clear efforts to clarify the record. It is important to understand that his “very likely” conclusion is largely an historical rather than a scientific judgment based upon Dr. Foster’s understanding of a wide range of historical data that are addressed in the chapters that follow. Responding to inquiries posed by the current writer, Dr. Foster on October 3, 2000, emphasized that he was relying heavily on historical evidence and clarified:
You are perfectly correct that my conclusion “ …it is ‘very likely’ Thomas Jefferson was the father is based not solely upon …[our] scientific DNA inquiry but involves interpreting those data in the light of historical evidence. … ” It is also true that “ …there is nothing in …[our] DNA study that itself would lead …[us] to suspect Thomas Jefferson as the father versus Randolph [Jefferson] or his sons.”44
In fairness to Dr. Foster, he has followed this issue closely and it is certainly legitimate for him to voice conclusions based upon more than his scientific expertise.45 One need not intend criticism of him to observe that some of his public comments may nevertheless have contributed to a broad misunderstanding of his scientific research.
Similarly, Dr. Eric S. Lander, of MIT, who co-authored the “Founding Fathers” interpretive essay in Nature with Professor Ellis, wrote in a December 27, 1998, e-mail: “The DNA evidence strongly indicates that Eston’s father was either Thomas Jefferson or another male-line relative of Thomas Jefferson. …The DNA evidence obviously does not distinguish among male-line relatives. I leave it to historians to weigh the evidence.”46
Excluding Thomas Woodson
One of the most important, but largely overlooked, findings of the Foster DNA study was that Thomas Woodson could not have been Thomas Jefferson’s child. This finding undermines both the original 1802 James Callender story and the strongest oral history claim that Jefferson fathered children by Sally Hemings. As will be discussed in Chapter Three, Callender’s allegation that President Jefferson was sexually involved with Sally Hemings was largely premised upon the alleged existence of a ten- to twelve-year-old light-skinned slave named “Tom” whose “features are said to bear a striking although sable resemblance to those of the President himself.”47 This child has long been presumed by many to be Thomas Woodson, and no evidence of any other “mulatto”48 child named “Tom” at Monticello has been found.
When Professor Annette Gordon-Reed wrote her 1997 book, Thomas Jefferson and Sally Hemings, more than half of the chapter on James Callender was devoted to the question “Was There a Tom Hemings?” She notes that:
The nonexistence of Tom serves two functions for opponents of the Jefferson-Hemings story. First, it suggests that Callender lied about an essential item of his story. If he was lying about that, he was lying about the Jefferson-Hemings liaison. …
The second reason it is important for Tom not to exist is that it would suggest that Sally Hemings was not pregnant when she came back from France. …The notion that Sally Hemings was not pregnant when she came back from France is crucial to Jefferson’s defenders because it makes it much easier to argue that someone other than Jefferson fathered all of Hemings’s children.49
She then notes the strong oral history of the descendants of Thomas Woodson, who assert that he was the “Tom” of Callender’s stories:
Additional confirmation that Tom may have existed comes from a family that claims descent from this individual. According to the oral history of the Woodson family, which has been accepted by the Thomas Jefferson Memorial Foundation as accurate—except for the part about Jefferson being Tom’s dad—Tom was sent to live with a family called the Woodsons after the scandal broke. He dropped the name Hemings in favor of Woodson. …The family has been interested principally in establishing that Tom Woodson existed rather than proving that he was the son of Thomas Jefferson.
After the publication of her biography of Jefferson, [Professor Fawn] Brodie did additional research based upon the oral history of the Woodson family and found that a number of the details the family had passed down about Tom Woodson could be verified. …[S]he was able to determine that he had been born in 1790, the year that the mysterious Tom Hemings would have been born.50
Professor Gordon-Reed considers arguments on both sides of the “Was there a Tom?” issue, arguing in the process that Callender would have been foolish to manufacture such an allegation:
The point for Callender’s story was that the existence of children who looked like Jefferson would tend to prove that he was having sex with Sally Hemings. A boy old enough to compare closely to Jefferson and one named after him, too, would strengthen Callender’s case; “President Tom” was an effective device. But was it so effective that Callender would have employed it knowing that Jefferson’s supporters could easily point out there was no such boy?
The defenders of Jefferson never offered this rebuttal to Callender’s claim. …[T]he fact that there was no twelve-year-old child at Monticello named Tom would have been an innocuous enough bit of information to pass along to those who were defending him. This alleged child appeared as a prominent and regular feature of Callender’s articles. What if there had been no Sally Hemings at Monticello? Would the supporters of Jefferson have mounted a defense against the allegation of an affair with her without mentioning that she did not exist?
Other newspapers were investigating the story. If Tom did not exist, it seems likely that someone among Jefferson’s supporters or anyone who could be described as neutral would have stumbled upon the fact that Tom did not exist, if he did not.51
In her 1974 book, Professor Fawn Brodie adds that “[a]t least two other editors, after checking, corroborated the story” of Tom Hemings.52 My own Jefferson collection includes an original copy of the October 1, 1802, issue of the Philadelphia Aurora, published by William Duane. A page two article defending President Jefferson against C
allender’s charges, reprinted from the Richmond Examiner, stated: “That this servant woman [Sally Hemings] has a child is very true. But that it is Mr. Jefferson’s, or that the connection exists, which Callender mentions, is false. I call upon him for his evidence, I challenge him to bring it forward.”53
Surely Jefferson’s friends and supporters discussed Callender’s attack, and how best to counter it, among themselves. Is it probable that William Duane—or, for that matter, Meriwether Jones, editor of the Examiner and another Jefferson ally—would have gratuitously confirmed the existence of “Tom” without reason? Possible, yes—but hardly probable. Nor does it seem likely that none of Jefferson’s many friends would have bothered to challenge this key point in Callender’s argument if there was in fact no “Tom.”
The Jefferson-Hemings Controversy Page 11