Research Report on the
A CRITICAL ANALYSIS
by Eyler Robert Coates, Sr.
On January 26, 2000, the Thomas Jefferson Memorial Foundation announced that it had reached a conclusion concerning the Jefferson-Hemings controversy. Their basic finding was that there is “a high probability that Thomas Jefferson fathered Eston Hemings, and that he most likely was the father of all six of Sally Hemings’s children.” However, an examination of this report and the methodology used in preparing it, shows it to be an unprofessional, unscientific accumulation of bias and prejudice, and an offense to the memory of the great man that this foundation was chartered to memorialize.
One would expect the Foundation at least to give Thomas Jefferson the benefit of the doubt in the face of the many scurrilous attacks that have been made on his character over the years, for which there is not one shred of direct evidence. But as we shall demonstrate below, the exact opposite is the case. The best evidence was suppressed or ignored, competent persons having opposing views were not consulted, and many alternative but reasonable explanations for the circumstantial evidence were disregarded. As the reader of this analysis will clearly see, it is obvious that the entire controversy was approached, not as advertised (and as Jefferson had written), “to follow truth wherever it may lead.” Rather, there was a deliberate attempt to select and mold the evidence to fit a pre-selected theory and to avoid anything that might resemble genuine balance. The results and conclusions became precise illustrations of something that Jefferson had written on a different occasion:
The moment a person forms a theory, his imagination sees, in every object, only the traits which favor that theory.” –Thomas Jefferson to Charles Thompson, 1787.
This travesty of a report sees in every point only those aspects that favor the preconceived theory. It leaves unconsidered much evidence that would tend to exonerate Jefferson, and it avoids connecting different pieces of evidence that would point away from Thomas Jefferson to some other member of his family or household.
The basic failure of the report derives from there being no rational and reasonable basis for evaluating evidence. Certainly, the report contains no statement of the committee’s policy in that regard. But by examining the decisions that were made when interpreting evidence, it is possible to determine what appears to be the subjective grounds on which the committee made such decisions. Rather than giving emphasis to evidence based on the apparent ability of the testifier to know of his or her own knowledge and experience whether the evidence was true or not, the committee judged evidence on the same basis established by Annette Gordon-Reed in her book, Thomas Jefferson and Sally Hemings; An American Controversy (GR p. 103), who wrote that “the standard for judging the evidence… should be what the declarants say, how they say it, and the amount of extrinsic evidence that exists to support their statements.” And since the “extrinsic evidence” is all based on conjecture and supposition, following that standard means the entire operation becomes a subjective one over which the investigator’s bias, either consciously or unconsciously, has complete sway. The result is a fabricated tissue of hearsay and assumptions, in which gossip and self-serving, handed-down family stories are granted the equivalence of facts and direct testimony. Even the report’s so-called “uncontested historical facts” are so lacking in close and insightful analysis, they become mere tools for the investigators’ bias. The only exception to this incompetent, unprofessional approach to evidence evaluation is the uninterpreted DNA test results, which by their very nature cannot be manipulated.
The chief error committed by the committee was in equating oral tradition — another name for handed-down hearsay — to solid evidence. As David Rafner of Richmond, Va. wrote in USA Today Online, Feb. 1, 2000, “Hearsay is hearsay, no matter how many generations it has been repeated. It can spur investigations but can never be raised to the status of evidence.” Or perhaps “should never be raised” to that status, because that is precisely what the committee has done. Vague standards of evidence, such as those the committee employed, permit the introduction of data from unknown and unverifiable sources, which is what family tradition is. And as we shall see, with the Woodson family tradition, we now have SCIENTIFIC PROOF that even the strongest oral tradition can be absolutely false, especially on the question of paternity.
The evidence of paternity is one kind of evidence that surfaces in several different contexts and that deserves to be singled out for special attention. In the absence of DNA testing, IT IS ALL BUT IMPOSSIBLE FOR THIRD PARTIES TO KNOW WITH CERTAINTY IF A GIVEN MAN IS THE FATHER OF A WOMAN’S CHILD. Often, not even the man himself can know this, and sometimes even the woman cannot be sure, if she has had several different sexual partners in a short period of time. In the case of an unmarried woman, living on a plantation that has dozens of potential partners also living there, and that is visited by hundreds of visitors in any given year, no person can be certain who might be the father of any child that the woman has. Third parties can only guess, based on whatever information might be available to them or observations they may have made. In considering whether there was a relationship between Thomas Jefferson and Sally Hemings, we shall be compelled to consider evidence from a variety of sources, most of whom could not possibly have an inkling (based on their own knowledge) as to whether there was such a relationship or not. But this apparently does not keep them from offering testimony. Some make assertions about the paternity of children that were conceived before the testifier was even born! Others not living on the plantation make assertions but supply no reason or event that would substantiate their assertion, thus making their story pure unsupported opinion and gossip. For others, though their observations may be true, they may mistakenly read into them more than is there. Therefore, in considering all the testimony that has been given, whenever anyone states the exact identity of a person said to be the father of a child, we must keep in mind that it is virtually impossible for them to assert this with absolute certainty, and we must evaluate what they say on the basis of how probable it is that they themselves observed what they are reporting and how likely it was that their interpretation of what they saw was valid.
Historical and Scientific Facts
The Research Findings of the Foundation’s report may be divided into two parts: (1) a series of nine “historical and scientific facts” which the committee considered “uncontested,” and (2) the interpretation of those facts in the form of commentary under each of the nine headings. Three of the so-called “uncontested facts” are easily acknowledged to be true, since they are based on rigorous scientific fact. A couple of the “historical facts” are not very significant even if true, a couple are based on plain gossip and conjecture, and a couple are not entirely accurate, and therefore are actually contestable on that account. There are several historical facts that would tend to contradict the committee’s preformed theory and, therefore, undermine the resulting conclusions of the Research Findings. But these facts were omitted from the list of nine findings, though most were mentioned and briefly discussed and dismissed somewhere in the report. But where the gross transgressions made by this report occur is in the interpretations which the committee included under each “uncontested” fact. We will briefly note first the so-called “uncontested facts,” followed by the facts that were NOT given full consideration, and that will be followed by an examination of the interpretations given by the report to the “uncontested facts.”
Scientific facts. The fact that there was a DNA match between the descendant of Eston Hemings and the descendants of Field Jefferson, and the fact that there was NO match between the descendants of John Carr and the descendant of Eston Hemings, together with the fact that there was NO match between the descendants of Tom Woodson and Field Jefferson, are all scientific facts that are without question indisputable. They are straight-forward and not subject to interpretive manipulation of the scientific findings themselves. In fact, they stand as solid bench-marks, requiring that all other circumstantial evidence not transgress the facts they establish. It is only when elements of gossip, unreliable historical data, and imaginative manipulations of the reliable data are added to these scientific facts that distorted conclusions are reached.
Insignificant facts. The fact that the descendants of Madison Hemings passed down the story that they were descended from Thomas Jefferson and Sally Hemings, and the fact that Eston and possibly other of Sally’s children were said to resemble Jefferson are both facts of no great significance. The Woodson descendants also passed down a similar story, and the DNA evidence proved they were mistaken. Monticello was swarming with Jefferson’s relatives from time to time, and any of them could have been the father of Sally’s children, thus providing the “resemblance” to Thomas Jefferson.
Gossip and hearsay. The fact that several persons not in Jefferson’s actual household believed that Thomas Jefferson was the father of Sally’s children, together with the fact that Madison Hemings told a newspaper reporter the same things, is not EVIDENCE but hearsay and gossip.
Uncontested facts? More uncertain as “uncontested facts” are the assertions that Sally’s birth patterns match Thomas Jefferson’s presence at Monticello, and that Sally’s children were granted a unique access to freedom. Both of these “facts” are not as clearly established as has been assumed. The birth patterns raise several questions. There were many times when Jefferson was present and Sally could have become pregnant but did not. And the access to freedom has a simple and obvious explanation, which will be explained in detail below.
The overriding point that must be recognized with each one of these nine “scientific and historical facts” is, not a single one of them is solid, direct evidence that there was a relationship between Sally Hemings and Thomas Jefferson. All taken together, they are not cumulative, and do not suggest that one piece added to the another produces a convincing combination or a series of steps moving in the same direction. For example, if Sally’s child conceived in France had the Jefferson Y chromosome, and Thomas Jefferson was the only Jefferson there, then that combination of evidence would indicate that Thomas was the father. But if a child of Sally’s had the Jefferson Y chromosome and there were any number of Jeffersons around to impregnate her, then those two facts would not combine to point to Thomas; they are just isolated facts.
Rather than resulting in such combinations, all the above “scientific and historical facts” are just unrelated observations subject to multiple explanations, or they are gossip and hearsay that have no explanation to back them up. In fact, the best kinds of evidence tend to exonerate Thomas Jefferson, but that evidence was overlooked or dismissed for insubstantial reasons and omitted as part of the nine major points!
One of the interesting aspects of this Research Report, presented to the public as a fair and even-handed consideration of the question of Thomas Jefferson’s paternity of Sally Hemings’s children, is that there are no facts presented and considered in the Findings as items that would tend to exonerate Jefferson. That alone suggests that this was a put-up job, intended not to provide a careful analysis and consideration of ALL the facts, but a one-sided argument in pursuit of political ends, a piece of propaganda intended to persuade and promote a particular idea. That this kind of polemic would issue forth from a foundation established to memorialize Thomas Jefferson is itself one of the incredible ironies of this whole story. The Foundation itself engaged in a sorry attempt to defame one of our great founding fathers.
Nevertheless, we will attempt to remedy this oversight, and present a few salient facts that the Foundation might have included in their list, were they inclined to present a balanced report. Perhaps the Foundation thought that these facts are “contested,” and therefore not proper for consideration. Not so! These are actual historical facts. The fact that they occurred cannot be contested. The contestation comes, not as to whether they actually occurred or not, but on the INTERPRETATION given to these facts. And since the Foundation presented its own interpretations of the “uncontested historical or scientific facts,” one would think that they would at least present these historical facts, and then try to tear them down through interpretation, if that was thought necessary.
Moreover, as we have seen, some of the “uncontested” facts which the Foundation DID present can indeed be contested. But it appears the Foundation did not feel compelled to provide equal consideration to facts that might tend to exonerate Jefferson as they did to those which indicted him. Instead, each of these facts was mentioned off-handedly in relation to other discussions, and dismissed without any serious consideration of the pros and cons. What is fascinating here is the trivial reasons the committee gives for dismissing some of the most important pieces of evidence available relating to the supposed relationship between Sally Hemings and Thomas Jefferson. Also interesting is the arbitrary manner they adopted in evaluating evidence, narrowing or broadening their criteria, depending on whether the evidence supported their theory of the relationship or not.
Jefferson’s denial in his letter to Secretary Smith. The assertion that Jefferson never denied these charges is not entirely true. He thought that to make public denials -only draws attention to the attacker:
I should have fancied myself half guilty, had I condescended to put pen to paper in refutation of their falsehoods, or drawn them respect by any notice from myself.” –Thomas Jefferson to George Logan, June 20, 1816.
That passage includes an implied denial, since Jefferson is suggesting that only those who are guilty become concerned with obviously unscrupulous attacks. And while it is true that Jefferson never denied such accusations publicly, he did deny them in private correspondence. The historical record provides a reference to a denial letter, but the denial letter itself has been lost or destroyed. In a private cover letter to his Secretary of the Navy, Robert Smith (covering a copy of a letter of denial to his Attorney General, Levi Lincoln), he denied all the accusations made against him, except for the singular one related to a Mrs. Walker. This was well after the accusations by Callender, and he wrote:
“You will perceive that I plead guilty to one of their charges, that when young and single I offered love to a handsome lady. I acknowledge its incorrectness. It is the only one founded in truth among all their allegations against me.” –Thomas Jefferson to Robert Smith, July 1, 1805. (DM 1:448)
That response should leave no doubt whatever that he was denying the rumors and charges that he had children by any of his slaves. Nevertheless, some proponents of an affair have suggested that Jefferson was being coy, and that “all their allegations” referred only to the allegations made by Mr. Walker. Walker’s allegations “might not have” included allegations that Jefferson fathered children by a slave, which was one of the allegations made by his other enemies. But the fact is, as the report’s own findings indicate, this was one of several letters that Jefferson wrote “to some close political associates” in response to the “assaults on his character” made by Federalist newspapers, and that is the context in which Jefferson made his denial. It is this kind of twisting of the evidence with imputations of connivance on the part of Jefferson which is typical of the manipulations made by affair proponents. Nevertheless, lest there be any remaining doubt, it should be noted that not too long after writing the letter to Smith, Jefferson also wrote:
There is not a truth on earth which I fear or would disguise. But secret slanders cannot be disarmed, because they are secret.” –Thomas Jefferson to William Duane, 1806. ME 11:94
and a few months before he died, Jefferson wrote the following:
There is not a truth existing which I fear or would wish unknown to the whole world.” –Thomas Jefferson to Henry Lee, 1826. ME 16:179
Of course, never-say-die proponents will probably come back and say that this last quote must be taken in context, and refers only to political subjects, in spite of the fact that Jefferson plainly wrote, “not a truth existing.” But as Jefferson said, when a person clings to a theory, he sees only the things that favor that theory. And, we might add, he twists everything else he sees so that it supports that theory.
Martha Jefferson Randolph’s denial on her deathbed. Shortly before her death, Martha called her two sons, Thomas Jefferson Randolph and George Wythe Randolph, to her bedside and told them of Mr. Jefferson’s innocence of the charges of fathering children by a female slave, citing her reasons, and asking them always to defend the character of their illustrious grandfather. One of the important things to recognize here is that Martha THOUGHT that Thomas Jefferson was not the father of Sally’s children, and since she had lived at Monticello much of the time when Jefferson was there, and was on the closest terms with him, she was in a position to know from her own knowledge and observation whether there was likely a relationship going on or not. Certainly, she was in a better position to do this than someone like Israel Jefferson (see below), who was a child, ran errands, and had only very limited contact with Jefferson. She may have been mistaken as to WHO the father of Sally’s children actually was (and then she may not have been with respect to some of the children). But she either sincerely believed Jefferson was not the father, or she was lying through her teeth on her death bed.
There is one story included in the Foundation’s report that provides good behavioristic evidence that also contradicts the report’s findings. The report states, “There is only one known account of the subject [i.e., a connection to Sally Hemings] being raised in Jefferson’s presence. As Jefferson’s Randolph granddaughters told biographer Henry S. Randall, Jefferson’s daughter Martha Randolph, roused to indignation by Irish poet Thomas Moore’s couplet linking her father with a slave, thrust the offending poem in front of him one day at Monticello. Jefferson’s only response was a ‘hearty, clear laugh.'”
Proponents of an affair say that Martha knew what was going on, but she was “in denial.” But if she were really in denial, she would hardly have angrily presented the poem to Jefferson. Instead, reality would have clashed with her hidden refusal to accept it, and she would more likely have angrily thrown the poem in the trash and blotted it from her mind. But then, having presented the poem to Jefferson, if he were really guilty, he would surely not have laughed (unless he had the character of Al Capone, which he obviously did not), but would have blanched at having his private affairs publicly ridiculed and especially being faced with it directly by his own beloved daughter. But instead he laughed, and his laughing indicates to us that he thought the whole thing utterly absurd. Rather than indicating that there was an affair, and that Jefferson and his daughter both refused to face up to it, this incident clearly indicates to anyone with a sensitivity to human nature that neither Jefferson nor his daughter really believed with any part of their minds that this story was true.
Edmund Bacon’s eyewitness account and denial. Edmund Bacon reported seeing someone other than Thomas Jefferson coming from Sally’s room early in the morning, and offered that as evidence that Jefferson was not the father of Harriet. Bacon was an overseer for Thomas Jefferson, and related the following story about his employer:
He freed one girl some years before he died, and there was a great deal of talk about it. She was nearly as white as anybody, and very beautiful. People said he freed her because she was his own 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.” (JB p. 102)
The report dismisses this account because Bacon was not working for Jefferson at the time Harriet was conceived, and therefore it is assumed that he could not have known who Harriet’s father was. But it doesn’t take much insight to understand what Bacon was saying. Just as the committee assumes that Sally’s children did not have multiple fathers, Bacon was no doubt assuming that whoever he saw coming out of Sally’s room was Sally’s longtime lover, and therefore the father of Harriet. In any case, what Bacon was saying was that Thomas Jefferson was NOT the father of Sally’s children, because he — Bacon — was an eyewitness to the fact that someone else — NOT Thomas Jefferson — was sleeping with Sally. The committee assumes Sally to be “monogamous” with respect to Thomas Jefferson, but apparently cannot consider that she might have been assumed to be “monogamous” when considering Edmund Bacon’s evidence. Here is another example of how the committee deals with evidence that they don’t want to accept: they become very strict, and narrow the focus so that if Bacon’s account is not a perfect fit, it gets thrown out. When they consider something like the Cocke diary (see below), they can broaden the focus and consider his diary as evidence, even though Cocke had no stated reason whatsoever to back up the accusations he made.
Nevertheless, this is one of the most important pieces of evidence in existence bearing on a relationship between Sally Hemings and Thomas Jefferson. BACON WAS THE ONLY OBSERVER TO EXPRESS AN OPINION ON JEFFERSON’S PATERNITY AND THEN BACK IT UP WITH EYEWITNESS EVIDENCE! Every other piece of so-called evidence that identifies Thomas Jefferson as the father of Sally’s children is gossip, hearsay, conjecture, or supposition, and provides no reasons or first-hand evidence to back up the statements that are made. In fact, most of it doesn’t even cite the source or give any indication how the source might have known what it asserts! It is easy to understand why a biased committee would want to dismiss this evidence with a simple “has problems of chronology.”
The fact that the name of the person Bacon identified was later scratched out would tend to suggest that someone was trying to hide the truth or protect the identity of the true perpetrator. We could only speculate on who that perpetrator might be. The essential point for our needs here, however, is that it was NOT Thomas Jefferson.
The confession of Peter Carr. One of the more startling omission in the committee’s report is the absence of any serious consideration given to the confessions by Peter Carr. In a criminal trial, if there exists a confession by a perpetrator, that alone can determine the outcome of the trial. Yet in the case of who was sleeping with Sally, evidence of a confession is completely brushed aside. There are several reasons why this crucial information was disregarded, of course, but none of them are valid.
There are two stories regarding the Carr brothers and the alleged fathering of Sally’s children by Peter Carr. In one of the two stories, historian Henry S. Randall relates that Thomas Jefferson Randolph told him of the Carr brothers crying when Randolph confronted them with a newspaper article accusing Thomas Jefferson of fathering mulatto children. They wept because their uncle was being disgraced for something that was their own doing. In another story, Ellen Randolph Coolidge wrote that T. J. Randolph had told her that he had overheard Peter Carr laughing, and telling a friend that “the old gentleman had to bear the blame of his and Sam’s misdeed.” Proponents of an affair note a big difference between the two stories, and suggest for that reason that the stories are made-up and not true. But that is foolish and undiscerning. Nothing about the difference in the two stories suggests that they can’t both be true. The circumstances were entirely different for each, and that would easily account for the different reactions. Aren’t we all familiar with the man who will laugh and tell dirty jokes to his friends, but is embarrassed to do so before the minister of his church? There is nothing uncommon about such duplicitous behavior, and the fact that biased persons want to disregard the confessions on that basis only means that they will grab at any grounds, however insubstantial they may be, to eliminate evidence that undermines the case they are trying to make.
The other reason for dismissing the Carr confession is that it was demonstrated wrong in the case of the father of Eston, since the DNA evidence demonstrates decisively that it was a Jefferson, not a Carr, that was Eston’s father. Here again, biased investigators are trying first to suggest that Sally had all her children by one male — something for which there is no evidence whatsoever — and then having made that assumption, they try to say that the DNA test proves that neither of the Carr brothers was that male. But this is fallacious thinking. Even though the DNA tests demonstrate conclusively that SOME Jefferson was the father of Eston Hemings, that in no way excludes the possibility that one of the Carr brothers was father of some of Sally’s other children. Until DNA tests can be conducted on the descendants of Sally’s other children, the Carr brothers are still prime suspects as father for her children other than Eston, since there is evidence for this which on its face is highly compelling. This is not mere gossip or other kinds of guess-work, but evidence of actual confessions overheard and related by Thomas Jefferson’s grandson. And a confession, like eyewitness testimony, is one of the strongest kinds of evidence that is available to us. Only the most biased and incompetent investigator would dismiss such evidence on such flimsy grounds, and accept in its place gossip and hearsay.
The interesting thing regarding the possibility that Sally’s children had multiple fathers is the way affair proponents are avoiding any further testing of Hemings descendants. After seventeen months of search, Herbert Barger, a Jefferson family historian and genealogist, was able to locate the body of William Hemings, son of Madison Hemings, whose gravesite was long-forgotten by other Hemings family members. William died in 1910, and it is highly possible that usable DNA can be taken from his burial site. The information gained from this testing would be infinitely more valuable than Dr. Eugene Foster’s testing of yet another of the Woodson descendants (see below). It should be borne in mind that there was only ONE SINGLE LINE OF DESCENT from Eston that was tested. That single line is the ONLY scientific link between the Hemings and the Jeffersons. Nevertheless Dr. Foster, Dan Jordan, president of the Foundation, and now the living relatives of William Hemings, all express no particular interest in having these tests made on William Hemings. Dan Jordan says he doesn’t think the family should be “pressured.” The family at first gave their oral consent, but then have refused to sign a written document to permit the exhumation. But as someone remarked recently, if we have no qualms about digging up a former president to see if he died from cherry jubilee, why should this be a problem?
So why this profound lack of interest? Why wouldn’t these people, who are so interested in following truth “wherever it may lead” not want to pursue this possibility for adding more scientific knowledge to help settle this controversy? The reason is obvious, because if those tests showed someone other than a Jefferson to be the father — perhaps even a Carr! — then that would blow this whole Jefferson-Hemings Relationship theory to smithereens. Of course, if the tests showed a match to the Jefferson Y chromosome, that would not seal the case on Thomas Jefferson, but it would surely make it a little stronger. But none of these assassins of Thomas Jefferson’s reputation want to take that risk. They would rather hold on to the gains they think they have now than take the risk of having the whole side-show sent down the tubes. It is interesting to note that the research report mentions that “the committee is aware that further DNA testing, coordinated by Dr. Foster, is in progress.” They bother to note that utterly futile attempt to make Jefferson guilty, but they don’t even mention this potentially explosive DNA testing that Herb Barger is trying to promote. So much for the Foundation’s commitment to “follow truth wherever it may lead.”
It is interesting at this point to note that in order for proponents of an affair to dismiss all the above first-hand evidence, and to substitute in its place the hearsay and gossip that constitutes most of the “historical evidence” which proponents treat as authentic, it is necessary to consider all the above solid, first-hand narratives as coming from people who, for one reason or another, all lied. Any error about paternity, which, as we have stipulated, is an extremely difficult fact to ascertain, is cited as sufficient cause for dismissing everything the person had to relate. But we should realize that such a tactic is silly. A person can be wrong about precisely who the father is, but yet be right about other related elements. Paternity is something that only DNA testing can tell us with certainty.
The Woodson family tradition. The Woodson family story is the strongest oral tradition associated with this controversy. If Tom (Hemings) Woodson left Monticello at age twelve, he was old enough to know OF HIS OWN KNOWLEDGE who his mother was and where he was living before he went to stay with the Woodsons. As with all oral tradition, we cannot accept all of it indiscriminately. It must be carefully evaluated, and one of the bases of evaluation is what Tom Woodson himself could have known, and what he could not have known firsthand. What he could not have known firsthand was the name of the person who was his father, and that is the precise bit of information the DNA tests demonstrate the Woodsons were mistaken about.
The Woodson family tradition poses problems for the theory which the research report adopts, and the committee’s way of dealing with this is to throw it in the bin labeled ‘things that will probably never be completely understood.’ But when one realizes that Thomas Jefferson was NOT the father of any of Sally’s children, then the Woodson family tradition fits neatly into place. There is no problem dealing with it. Thus we have this situation: the researchers could not make the Woodson oral tradition fit into their theory, especially since the DNA indicates no Jefferson was the father of Tom Woodson. So, What to do? The answer is simple, if Sally is the mother and Jefferson is NOT the father. But that wouldn’t fit the theory. So, the committee just pushes it aside, in hopes, no doubt, that the Woodsons would just go away. This does pose some problems for a “38-year affair” and all the accusations by Callender. But, the only solution the committee has is to call it a mystery.
In any case, the Woodson family tradition tells us two things loud and clear: (1) Whether Tom Woodson was Sally’s first child or not, there is no question whatsoever that Thomas Jefferson was NOT his father. (2) And whether Tom Woodson was Sally’s first child or not, we have a perfect, scientifically-proven example of how utterly worthless even the strongest family oral tradition can be on matters of paternity.
We must remember that the fact that Tom Woodson was NOT a son of any Jefferson male was much more firmly established by the DNA evidence than the supposed paternal relationship between Thomas Jefferson and Eston Hemings. Eston’s father was proven to be any one of 25 Jeffersons, which still admits of a large number of possible men as the father, with about eight as probable. But Tom Woodson’s father was proven to be from NONE of the Jefferson males living at that time, which positively excludes each and every individual Jefferson, including Thomas.
It is interesting to note that the Woodson family tradition is so strong, the family itself refuses to accept the results of the DNA tests. Those tests, performed on FIVE lines of descent, demonstrate beyond any reasonable doubt that Thomas Jefferson was NOT the father of Tom Woodson. The living persons tested were all descended from two sons of Tom Woodson, who were born twelve years apart. In order for the DNA tests to be invalid, those two sons would need to have been fathered illegitimately by the same man, twelve years apart! An absurdly unlikely possibility.
In what seems like a pointless and hopelessly futile effort to somehow secure a match with the Field Jefferson descendants, Dr. Foster is in the process of testing yet another line of descent from Tom Woodson. Such indefatigable efforts in the face of such absurd odds can only be taken as a measure of the insistence of affair proponents to prove that Thomas Jefferson was the father of Sally’s children. He is doing this because IF Tom Woodson is Sally’s Paris-conceived son and could be shown to have the Jefferson Y chromosome, it would then be almost certain that Thomas Jefferson was his father, since Thomas was the only Jefferson in Paris at the time who could have impregnated Sally. One thing the DNA results does demonstrate is, whoever Tom Woodson’s father was, his Y chromosome was one common amongst European whites, not sub-Saharan blacks, and if Tom was Sally’s first child, his father was almost surely someone she met in France. Nevertheless, if somehow Dr. Foster were able to get the results he is seeking with this sixth Woodson descendant, those results would call into question Dr. Foster’s whole study, because if there was anything that was proved almost irrefutably and unquestionably by the previous results he obtained, it is that the Woodson DNA, as previously certified, can be traced all the way back to Tom Woodson himself. This fact makes us wonder whether Dr. Foster himself really understands the science of the DNA testing and the results obtained thus far.
Randolph Jefferson’s presence when Eston was conceived. Randolph was invited by Thomas to come to Monticello to visit him and Randolph’s twin sister, who had just arrived one day earlier. In a letter dated August 12, 1807, slightly more than nine months prior to Eston’s birth on May 21, 1808, Thomas Jefferson wrote his brother Randolph, advising “Our sister Marks [Randolph’s twin sister, Mrs. Hastings Marks] arrived here last night and we shall be happy to see you also.” (BM p. 21) This twin sister was at Monticello at the time of the writing, and this suggests that Randolph, probably with some of his sons, would not likely delay the journey to see his sister in order to arrive before she departed from her visit. Since the trip to Monticello from Randolph’s home could easily be done in less than a day, that would almost certainly put Randolph at Monticello at the correct time to be the father of Eston. And since Randolph had the same Y chromosome as all the Jefferson males, he would be completely capable of supplying that chromosome to Sally’s son, Eston. Any child of Randolph would also have the same physical attributes as Thomas. Nevertheless, the committee report rejects the possibility that Randolph was at Monticello at the time Eston was conceived, which we will examine in greater detail below.
The committee’s “Research Findings and Implications” are a study in bias. Instead of presenting the various contradictory interpretations in their best possible light, weighing them against one another, and then providing reasons for choosing the one that is the most likely, the committee presented its findings as conclusions, and dismissed opposing possibilities, evidence, and interpretations out of hand. We examine below the interpretations given under each “uncontested historical or scientific fact.”
1. DNA Match between descendants of Eston Hemings and Field Jefferson. The report says that this match “provides scientific support for the statements of Madison Hemings and Israel Jefferson.” But this is a gross overstatement. Both Madison and Israel made numerous and extensive statements completely unrelated to the DNA match. At most, it can only be said that the statements by Madison and Israel that Thomas Jefferson was the father of Sally’s children were not contradicted by the DNA match. But there are many other possible explanations that are also not contradicted by this match. Their exaggeration, however, illustrates the committee’s bias. The committee also acknowledges the “scientific possibility” that Randolph Jefferson or one of his sons could have been the father. Notice they do not say “provides scientific support,” although science supports that possibility just as much as it does the Madison version. But the committee adds, “the preponderance of known historical evidence indicates that Thomas Jefferson was his father.” That “preponderance” is presumably presented somewhere else, because it is only stated here as a conclusion. But when we examine that “preponderance” of evidence elsewhere, we find that it is the familiar mix of gossip, hearsay, and speculation.
The report then states that “Randolph Jefferson and his sons are not known to have been at Monticello at the time of Eston Hemings’s conception.” But that statement is demonstrably false! As we mentioned above, Randolph was invited to come to Monticello to visit with his twin sister at the very time that Sally conceived Eston. It is true, of course, that this letter is the only record of Randolph’s probable presence at the time of conception. Nevertheless, the Foundation’s report dismisses the possibility of Randolph’s presence because “A search of visitors’ accounts, memorandum books, and Jefferson’s published and unpublished correspondence provided no indication that Randolph did, in fact, come at this time.” But there is not likely to be any other record of Randolph’s presence except as incidental notes in correspondence just like the note inviting him. The real question is, were there notations in visitors’ accounts, and memorandum books indicating all family visits? Were all these visits in fact formally recorded? Would we, for example, know that Randolph’s sister had come for a visit to Monticello if it had not been mentioned in a letter to Randolph, or possibly in previous correspondence with the sister herself? Was there a formal register that listed casual visits by near relatives? Apparently not. In other words, it is highly likely that such notations were not made except incidental to some important activity or incidentally in correspondence.
To dismiss the possibility of a visit because there was no particular reason for writing about it becomes merely a device to eliminate evidence that is not wanted because it muddies the water. The research report states in Appendix J that existing correspondence between Thomas and Randolph from the year 1807 suggests that similar invitations may have been extended previously, but Randolph “may not always have acted on these invitations. In his post-1807 letters, ill health, the poor state of the roads, and other circumstances were often cited as reasons to postpone his Monticello visits.” But the question naturally arises, Was there a letter in which Randolph cited reasons for not coming to visit when his twin sister was at Monticello? No there was not. And the report states that Randolph’s only recorded Monticello visit in this time period was “on his own business.” Well, of course it was! Randolph had Thomas make out his last will and testament in 1808, a short time after Eston was born by the way, and so naturally we have a formal recording of the date! Here again, the committee raises or lowers its requirements for “documented” evidence as is necessary to admit or dismiss evidence it wishes.
Moreover, it should be noted that Randolph would be more likely to have a sexual encounter with Sally than would Thomas. Randolph was known to socialize with the black slaves at Monticello when he visited there. Isaac Jefferson, in his “Memoirs of a Monticello Slave,” as dictated to Charles Campbell, made the following statement:
Old Master’s brother, Mass Randall, was a mighty simple man: used to come out among black people, play the fiddle and dance half the night; hadn’t much more sense than Isaac.” (JB p 22)
And we know that social occasions can easily lead to sexual encounters, and that they are more likely to lead in that direction than a cold contact. And since Randolph was a widower at the time, it is easy to understand how he could become involved with one of the beautiful house servants. The report in Appendix J dismisses this evidence of Randoph’s socializing, saying Isaac left Monticello in 1797, and the story “most likely refers to the 1780s” — a time when Randolph had a wife — rather than later, after the evidence indicates Randolph was widowered, though the report offers no evidence for its assumption.
Then there follows a most revealing statement. The report says, “nor has anyone, until 1998, ever before publicly suggested them [Randolph and sons] as possible fathers.” One could hardly expect to see such a naive declaration of incompetent thinking from a supposedly professional source. Notice that the committee requires for this evidence a PUBLIC announcement. That means the statement by Rebecca Lee McMurry cannot be considered as evidence. (McMurry testified in a notarized statement that her family’s oral tradition affirms Randolph Jefferson as the father of Sally’s children.) Their approach means the committee disqualifies any evidence that was not generally known at an earlier time. It is as if, when one is trying to solve a problem or get to the bottom of an enigma, one cannot look for a solution except amongst the theories and suppositions that have already been proposed and publicly aired 200 years ago! It would be just as intelligent to suggest that DNA evidence should not be offered because it was unknown in the 19th century! This is arbitrary nonsense! But it is a good example of the incompetent, unprofessional approach which the committee took in this research study.
The report lists in Appendix J the names of several Jeffersons beside Randolph who were present at Monticello during the time when Sally conceived Eston and some of her other children. Thomas Jefferson Jr., Randolph’s son, was born in 1783, and was a “resident at Monticello for extended periods of schooling in 1799, 1800 and possibly 1801.” That means he apparently was present when Sally’s unnamed child (who may have been named Thenia) was conceived in March 1799, and when Harriet II was conceived in August or September 1800.
Robert Lewis Jefferson, another son of Randolph, was born in 1787 and was known to have carried a letter to Monticello intended for Thomas Jefferson “in July or August 1807,” according to the report. Jefferson arrived back at Monticello on August 4. The letter was dated July 9, but Jefferson did not receive it until August 8. And since Sally conceived Eston in this same time period, Robert Lewis Jefferson apparently was present at Monticello during conception time.
We must bear in mind that both or these young men had the Jefferson Y chromosome and physical characteristics. Thomas Jr. would have been 17 when Harriet II was conceived, and Robert would have been 20 when Eston was conceived. It is odd that the committee considered Thomas Jr. and Robert “unlikely fathers because of their youth and very intermittent presence,” at a time when Sally was 26 or 27, i.e., ten years older than the former, and 34, or fourteen years older than the latter, yet they did not consider Sally too young for Thomas Jefferson, who was thirty years older than her, and a genuinely old man of 64 when Eston was conceived. When we consider that Thomas Jr. and Robert were of an age that placed them at the height of their sexuality, one or both of them would seem to be much more likely a father for Sally’s children than Thomas Jefferson himself. And as for the “intermittent presence” of Randolph’s sons, we can be sure that even occasional visits would be enough to establish a sexual relationship between two persons, and even a one-night stand would be sufficient to result in the conception of a child. The committee, however, seems fixed on the idea that there had to be an on-going relationship, and for that reason refused to consider the perfectly reasonable possibility that Sally’s children were the result of various encounters.
George Jefferson, Jr., a descendant of Field Jefferson, served as Thomas Jefferson’s commission agent in Richmond. It is reasonable to assume that he would have made at least occasional visits to Monticello to consult with Thomas Jefferson on business. Moreover, these visits would have occurred during the time period when Sally was bearing her children, “although no reference to such visits has yet been found,” according to the Research Report. Is it possible to imagine that George Jefferson, Jr., served as Thomas Jefferson’s business agent in Richmond, roughly 70 miles away, and NEVER came to Monticello to discuss business? But if we demand formal records for a visit, that is the conclusion that must be drawn.
John Garland Jefferson, studied under the direction of Thomas Jefferson in the early 1790’s, and did not marry until 1800. His visits to Monticello would surely have coincided with Thomas Jefferson’s presence there.
One other son of Randolph Jefferson — Isham Jefferson — was born in 1781, and was reported to have been “reared” by Thomas Jefferson. He would have been age 26 when Eston, the last of Sally’s children, was conceived. Little more than that is known about his presence at Monticello. The Research Report states that “no reference to him [Isham Randolph Jefferson], however, has yet been found in Thomas Jefferson’s papers.” In fact, there is indeed a reference to Isham coming to Monticello. It is in Jefferson’s deposition of 1815, given with reference to the probate of the last will and testament of Randolph (senior). Jefferson states in the deposition that Isham Randolph Jefferson came to speak to him about some financial problems related to his father and his father’s second wife.
The point of all this is, however, that there were a number of Jeffersons who were around Monticello much of the time, and the fact that no notice of their presence appears in a formal register means nothing. Visits from nearby relatives were not likely to be noted unless there was something special going on in connection with the visit, especially if those relatives were frequent visitors. We would never have know that Isham was at Monticello in 1815 except for the fact that his visit had a material relationship to the subject of Jefferson’s deposition.
Thus, we see from all the above that Monticello was virtually crawling with Jeffersons who could have been the father of one or more of Sally’s children, all of whom had the tell-tale Y chromosome. And only a rigid, naive perspective would assume that the owner of a plantation was the only person who could father children by black slaves. The plantation functioned like a social microcosm, not a rigidly controlled personal harem, and the young white men who lived or visited there were just as likely as the owner to have a sexual relationship with the unmarried slave women. In the case of an owner who was up in age, such as Thomas Jefferson, the younger men were undoubtedly MORE likely to have such illicit affairs. In fact, a visiting white workman at Monticello was thought to have fathered a child by one of Sally’s sisters, indicating casual relationships developed easily on the plantation.
One interesting point made by one of the DNA consultants, and duly ignored by the Foundation, was a concern that the match between the Eston Hemings descendant and the Field Jefferson descendants was predicated on the results gathered from a single Eston Hemings descendant. According to one consultant, there is a danger that there could have been a “bookkeeping” error — a mix-up of test tubes, a slip in mislabeling — that might have resulted in the single match that was reported. Since an error of this magnitude would have astounding results on the conclusions drawn from the testing, one would think that a dedicated researcher would be interested in following the advice given and have the single Eston descendant retested. Or perhaps extraordinary efforts could be made to locate another Eston Hemings male-line descendant so that this enormously important finding might not rest on the blood from a single living person. The testing of a descendant of Madison Hemings would lend some collateral support to the original findings. But no. The researchers have gotten the results they were looking for, and they would rather waste time testing yet another Tom Woodson descendant (thus making a total of six!) rather than do something to put their investigation on more solid scientific grounds.
2. No DNA match with descendants of Eston and Carr. The report states that Jefferson’s grandchildren claim that the Carrs were the fathers of Sally’s and one of her sister’s children, and “The DNA study contradicts these statements in the case of Sally Hemings’s last child, Eston.” Which, of course, is true. But that does not necessarily mean that the Carr brothers were not father to some of the other children of Sally. Anyone who has known unmarried women who had multiple children knows that as a rule, they have those children by multiple fathers. And in fact, Sally’s own mother and two of her sisters each had multiple children by multiple fathers. So this would be nothing unusual in the Hemings family. But the report then states, in effect, that because Sally’s patterns of conception match Jefferson’s presence at Monticello, that should be taken as evidence that ALL her children were by the same father, i.e., Thomas Jefferson. In this way, the committee uses one assumption to support another.
But this is still a dubious claim. It should be plain to anyone who gives it a second thought, that other Jeffersons were many times more likely to come to Monticello for a visit when Thomas Jefferson was there, rather than when he was away. In the January 2000 issue of William and Mary Quarterly, Fraser D. Neiman dismisses the possibility of some other father being at Monticello only when Jefferson is there with a footnote, rigidly adhering to the dictates of his statistical model:
Because the model outcomes are tabulated against Jefferson’s arrival and departure dates, the probabilities that result apply to Jefferson or any other individual with identical arrival and departure dates. The chances that such a Jefferson doppelganger existed are, to say the least, remote.”
A less pedantic approach to the problem would recognize that “any other individual” could be the father of Sally’s children without necessarily having “identical arrival and departure dates” with Thomas Jefferson. Such an individual only need be there when Sally conceived. Thus we see that the author naively reveals his bias and relates the alternative possibility, not to Sally’s conception dates, but to Thomas Jefferson’s presence! Moreover, the statistical model does not even take into consideration the fact that Thomas Jefferson was present at Monticello on as many occasions when Sally could have conceived but did NOT, as he was when she did conceive. It begins to appear obvious that the model was set-up to produce the desired results, and was not realistically designed to take into consideration other possible explanations. It was designed to indict Jefferson, and it should come as no surprise that the conclusion is such a high probability as was “discovered” by such a ruse.
Thus in saying all the children were fathered by Thomas Jefferson, the report seeks to support one dubious supposition with another equally dubious. And by a similar slight-of-hand, it tries to say that since all Sally’s children had one father, and since there is no evidence of anyone other than Thomas Jefferson being at Monticello on such a constant basis, therefore “evidence of the sort of sustained presence necessary to have resulted in the creation of a family of six children is entirely lacking.” That could be called begging three questions at the same time. And incidentally, we note how the committee uses the assumption of a “sustained presence” to eliminate the possibility of multiple fathers, but forgets about that kind of assumption when evaluating Edmund Bacon’s eyewitness evidence.
3. No DNA match with descendants of Tom Woodson and Field Jefferson. This was perhaps the most salient result to emerge from the DNA tests. Had there been a match, then it would have been as conclusive a proof as possible that Thomas Jefferson was the father of a child of Sally, since he was the ONLY Jefferson in Paris when Sally’s first child was conceived. BUT THERE WAS NO MATCH! This forced proponents of an affair to backtrack and say that, well… maybe Tom Woodson was never at Monticello, even though he was specifically named by James Callender in 1802! It is important to remember that just because someone lies or makes an error, that does not mean that every word they wrote is necessarily a lie or an error. And it must also be remembered that paternity is frequently an indiscernible fact, whereas the existence or non-existence of a person is something that IS definitely knowable. Paternity is perfect for gossip, because who can prove it one way or another? But saying there is a 12-year old boy living at the plantation is not that kind of a statement.
Madison Hemings, whose Memoirs never mentions James Callender or his accusations, claims that Sally’s first child died, even though there is no record of the death. Of course, all of this happened years before Madison was even born, at a time when he could not possibly know the facts he was relating of his own knowledge. Therefore, his evidence was little more than gossip. And what about the strong oral tradition of the Woodson family? Their oral tradition should be given more credence than anyone else’s, because much of it was based on Tom Woodson’s own knowledge and observation, not on what he had been told by someone else.
Nevertheless, the report states that “No documents have yet been found to support the belief that Woodson was Sally Hemings’s first child, born soon after her return from France.” And here we see how the committee arbitrarily selects or rejects evidence based on how well it fits their preconceived theory. Never mind that Callender specifically identified Tom. Never mind that the Woodson oral tradition is more firmly established and from a more reliable source (Tom Woodson) who was there and who knew first hand about most of what he spoke, as compared to Madison’s account, the essentials of which occurred many years before he was even born. This is all considered as nothing worthy of consideration by the committee. But we are compelled to ask, in the wake of the above supportive evidences, “Were there any documents found to support the committee’s belief that Tom Woodson was some other woman’s child?” No, there were not.
4. Sally’s birth patterns and Jefferson’s presence. The report states that “the conception windows for Hemings’s known children is far more likely if Jefferson or someone with an identical pattern of presence at and absence from Monticello was the father.” But as we saw above, any other possible father is much more likely to be at Monticello when Jefferson was there, rather than when he was away.
One of the great mysteries related to the reputed birth patterns is, If Thomas Jefferson had an affair with Sally that started in France, and if that union produced a child that was born after their return to Monticello, the birth occurring sometime in early 1790, then why were there no more children born for almost six years? The remaining group of Sally’s children were all closer together. Moreover, this being supposedly the start of an affair, one would not expect to find such a great amount of delay between births at the very beginning. Of course, if we accept as fact that Thomas Jefferson was NOT the father of Sally’s first child, then this makes much more sense: the child’s father was some other person in France, and he did not accompany Sally on her return to America, so there were no more births immediately afterward. Moreover, that possibility finds support in the DNA evidence, and in the Woodson family history.
The committee solved the problem of Sally’s long hiatus at the beginning of the affair by declaring Sally’s first child, reportedly conceived in France and born in this country in 1790, and everything connected therewith to be a mystery. They begin calculating Sally’s pregnancies with the first child conceived IN THIS COUNTRY, beginning in December of 1794. Whereas such amateurish manipulation of the evidence may avoid certain problems in dealing with a difficult, confusing, and contradictory situation, we think that handling fundamental aspects of the overall problem in that manner disqualifies the committee from proposing any kind of solution or position on the issues. We feel that the non-match between Jefferson descendants and Tom Woodson descendants demonstrates that Thomas Jefferson was NOT the father of at least one of Sally’s children, and for the committee to simply avoid grappling with the issue demonstrates their incompetence and their inability to deal with any evidence except that which coincides with their biases and their preselected theory.
If we are to consider “someone with an identical pattern” of attendance and departure at Monticello as Jefferson had, we should remember that common sense dictates that Monticello would be flooded by persons whose presence was almost surely to coincide with Jefferson’s. Moreover, there were as many times when Jefferson was present and Sally could have gotten pregnant, but did not, as there are when he was present and she did. And the calculation of those times omits the first year after childbirth, when lactating women are usually not fertile. Putting all those factors together, we begin to see that Sally’s conceptions had a certain amount of randomness to them, within the confines of the times that she would be likely to conceive at all, that is, when Jefferson and his entourage were there. And any statistical study of Sally’s conceptions and Jefferson’s presence are on faulty premises if those studies do not take into account the fact that other possible fathers are much more likely to be present when Jefferson was also present then when he was away.
An interesting pattern occurred the year Eston was conceived. Thomas Jefferson had access to Sally at two different times in the year in which she conceived Eston: he was at Monticello between April 10 and May 16, and between August 4 and October 3. But Sally did not conceive Eston during the first time period, when there is no definite evidence that any Jefferson other than Thomas had access. Rather, Sally conceived in August of 1807, when there is very good evidence that his brother Randolph was there.
More precisely, if we take 267 days as the normal time of gestation (from impregnation to birth) with a normal range from 250 to 285 days, then the normal date of impregnation for a birth occurring on May 21, 1808, would have been August 28, 1807, with a normal range between August 11 and September 15. Randolph had been invited to come to Monticello to visit with his twin sister on August 12, therefore his likely time of visiting would have been exactly within the range of time in which Sally would normally have conceived. This combination of evidence would tend to suggest that Thomas Jefferson was NOT the father of Eston, since Eston was NOT conceived at a time when evidence suggests only Thomas was at Monticello. Rather Eston was conceived only after Randolph was added to the list of those present.
The report states in Appendix F that “There is no record that Sally Hemings was anywhere but at Monticello from 1790 to 1826.” This is a curious way of stating it, because we suspect that documentary evidence for her presence OR absence was incidental at best, as it would have been for most of Randolph and his sons’s comings and goings. Yet the committee seems willing to assume Sally’s presence without documentary evidence that she was there, and unwilling to accept Randolph’s presence, even when incidentally inferred. But this is just another example of how the committee adjusted the standard of evidence in order to support their predetermined theory.
In addition, we know, as stated in Appendix F of the report, that shortly before Martha Jefferson Randolph’s death, she “reminded them [her sons] that the Hemings who most resembled Jefferson could not have been his child, since he and Sally Hemings were ‘far distant from each other’ for fifteen months before the birth.” That has always been assumed to be an error, since we know where Thomas Jefferson was every day of his adult life, and he was at Monticello when Sally conceived all of her children. But perhaps the error was in the assumption that relates this to Thomas Jefferson’s presence. The research report assumes that Sally was at Monticello constantly from 1790 to 1826. But Martha’s statement suggests that she may have had information indicating that Sally was NOT present at Monticello when one of the sons was conceived, and therefore Thomas Jefferson’s presence at Monticello would NOT have meant that he could father Sally’s child. This is a possibility that apparently has not been explored, since proponents of an affair assume that Martha was either mistaken or lying.
5. Several neighbors believed that Jefferson fathered Sally’s children. The word for this, of course, is GOSSIP. None of these people had any way of knowing from their own knowledge that Thomas Jefferson and Sally Hemings had a sexual relationship. And if they did, that explanation was never a part of their story. At best, they passed on a story that they had gotten from someone else, or that was based on their own or others imagination and supposition. There are many fathers who are not sure that their legal children are actually their natural children. Persons outside the bedroom can only be even less sure of paternity. So, without some kind of accompanying explanation for why an outside person happens to “know” that some man is the father of some woman’s child, the accusation cannot be taken as anything other than their own suppositions or the repeating of hearsay and gossip.
All of this is the weakest kind of evidence, and should not be considered evidence at all by a serious investigative committee. It was the product of overactive imaginations, just as it has continued to be in our time with the books by Fawn Brodie and Annette Gordon-Reed. Both of these authors filled their pages with scores of “might-have-been” suppositions and assumptions. And that, no doubt, is what was done 200 years ago. Indeed, the source of all this gossip was inadvertently explained in the report, under the section on Family Resemblances: “It was evidently their very light skin and pronounced resemblance to Jefferson that led to local talk of Jefferson’s paternity.” And there in a nutshell is the whole story. This was the source of the gossip and of Callender’s accusations, and this was what led to the present-day controversy. When Callender wrote, “There is not an individual in the neighbourhood of Charlottesville who does not believe the story, and not a few who know it,” he was talking about gossip and hearsay, not evidence. Whatever the individuals in the neighborhood believed about Sally being a concubine of Thomas Jefferson, they could not possibly know it of their own knowledge. It was conjecture created by wagging tongues, based on a prurient imagination.
In appearance, Sally’s children were near white, and were thought to resemble Jefferson. That alone was enough to set the tongues wagging. The next step was to assume that they must be his children. And once the gossip gained the authority of appearing in print and being voiced abroad, it was a simple matter for it to be incorporated into “Family Oral Tradition.” Since slaves of the time had a great tendency to attach themselves to their masters by adopting their names, it was a simple thing for an unmarried mother to tell her children that the master of the plantation was their father. If this was already talked about and published in the newspapers, this made the claim almost unavoidable. Under such circumstances, if a mother were to tell her children that the master was NOT their father, the children would probably reply, “But everybody says he is!”
So it is easy to see how this sort of gossip could get started. All of that, plus the example of the Woodson family tradition and their belief that Jefferson was the father of Tom Woodson — something conclusively proved to be false by the DNA tests — should make us very wary of any kind of neighborly gossip and family oral tradition.
John Cocke claimed that Jefferson fathered Sally’s children, but he provided no explanation for WHY he thought this. Unlike Edmund Bacon, who had a reason to back up his statement, Cocke could only be repeating gossip that he picked up from neighbors. There is absolutely nothing intrinsic to his “evidence” to make it worthy of any consideration at all. And Israel Jefferson could hardly have had more to substantiate his backup of what Madison had said. Israel was not a house servant, and most of whatever knowledge he had of what was going on there was no doubt gained from others. He was a friend of Madison Hemings, and was probably prompted by the interviewer when giving his statement. Israel was about ten years old at the time Eston was conceived, so we can well understand what he could have known of his own knowledge about an affair. In addition, it would be inconceivable that he would contradict the statement that Madison had previously made. This stuff is as close to non-evidence as we can get.
6. Madison claimed that Thomas Jefferson was the father of all Sally’s children. The statement made by Madison Hemings in 1873, as already noted, contained information that he could not have known firsthand, and that he had to have gotten from some other person. It contains nothing to indicate how he knew the things he was relating, and even the Research Report in Appendix F includes the statement that Madison “did not specifically mention when or how he learned the identity of his father.” Apparently, Thomas Jefferson acted towards him in no way that would indicate that he was Madison’s father. Moreover, his statement was intended for a publication with political motivations that were anti-Jefferson. The statement itself was not Madison’s original writing, neither did it attempt to reproduce his exact words (as was done with the story of Isaac in Bear’s Jefferson at Monticello). Rather, it was rewritten by the editor, and for that reason the exact words cannot be taken as authentic.
This story by Madison is treated as something unique, as an unusual piece of evidence. But the truth is, there are other descendants of Monticello slaves who also claim that Thomas Jefferson was their forefather. And recently, another black family has claimed to be descended from George Washington, whereas the best evidence indicates he was sterile, since his wife had children by a previous marriage, although George and Martha were not able to have any. These attempts to claim a great man of previous times as one’s grandparent are common, and should be scrutinized carefully before being accepted as evidence. At best, they can only be used to lead one to substantial evidence, and if that substantial evidence is not present, then this kind of hearsay should be disregarded.
But always, as previously noted, it must be remembered that it is not possible for third parties to make an absolutely certain assumption about who is the father of an unmarried woman’s child, unless the mother and father are living together openly, or the father acknowledges the child, or some certain evidence supports the claim. As a result, statements like Madison’s concerning the parentage of himself and his siblings — facts that he could not possibly know of his own knowledge — should be viewed at most as evidence of what was believed by the speaker, and not as real evidence establishing the facts spoken.
The Report repeats the same fallacious point seen before: “Even the statements of those who accounted for the paternity of Sally Hemings’s children differently (Thomas Jefferson Randolph, Ellen Randolph Coolidge, and Edmund Bacon) never implied that Hemings’s children had different fathers.” SO WHAT? The investigators do not seem able to comprehend that, when seeking the solution to a problem, the fact that people in the past did not anticipate a given solution is irrelevant and should not hinder people living today from reaching a reasonable conclusion just because it was not previously anticipated. We must also bear in mind that investigators today often have available to them evidence that was not available to individuals living at the time events occurred. For example, how many of the people at that time knew that Randolph had been invited to be at Monticello in the time period when Eston was conceived? Later investigators are often able to make connections that would never have occurred to persons at the time.
But this same point about something not having been mentioned previously is made at several different places in the report. In the Assessment of Possible Paternity of Other Jeffersons it is said that “In almost two hundred years since the issue first became public, no other Jefferson has ever been referred to as the father.” Again, What difference does that make? As we have pointed out, paternity is the kind of thing that persons not a part of the relationship can only guess at. And sometimes, even the parties to the relationship are not sure. Is the Foundation incapable of entertaining a new idea, a new way of looking at things? The committee’s attitude on this issue is a most unscientific one, and borders on being silly. What it clearly reveals is the workings of a biased mind set on throwing up any excuse in a dogged effort to support its theory. This uninformed assumption should not be given credence in a search for truth, whether it is assumed by people today, or was assumed by people in the past.
The fact that the Hemings family seemed to be close-knit, and named their children after one another is another example of an irrelevancy. No doubt Sally, as the matriarch of the family, and a family leader, is responsible for their closeness, and paternal consanguinity has nothing whatsoever to do with it. Closeness is a function of the family culture, not of blood. It is not uncommon for social workers to see a mother with five children having five different fathers, and being just as close and supportive of one another as a similar family with one father. And if Sally’s children did not know that they had different fathers, that would produce the same result. In any case, such flimsy, ambiguous evidence is hardly worth serious attention when the endeavor is to find evidence that supports paternity. That kind of evidence can only leave the impression that the committee has a theory, and are marshaling every kind of inconsequential evidence they can find to support it.
The report then adds, “there is no documentary evidence that Thomas C. Woodson was Sally Hemings’s son.” We wonder at the arbitrary standard the committee adopts for evidence. Apparently, if someone had jotted down in a diary, “Tom Woodson is Sally’s son,” that would be considered evidence by the committee. But if James Callender publishes in a public newspaper the same information, and no one refutes it, even though they might negate other things that Callender wrote, then that is NOT considered “documentary evidence.” There are few things in this report that make less sense.
7. Unique access to freedom. No one has ever denied that Sally Hemings, her children, and many of Sally’s close relatives, held a unique position in the Jefferson household. They were house servants, and they had status and privileges above all other servants on the plantation. The fact that these special considerations were expressed by giving two of Sally’s children their freedom and allowing them to “run away,” and that the remaining two were given their freedom in Jefferson’s will (together with three other Hemings-related servants) should not be viewed as anything extraordinary. It should be noted that Sally herself was NOT given her freedom in Jefferson’s will, so these manumissions show no evidence of special favoritism to Sally. It is easy to conceive of all of this being done without any special promise to Sally, or any extraordinary attempt to reward her remaining two sons. Many reasons exist for Jefferson to do this, and those reasons do not suggest that Jefferson was their father.
In addition to being favored house servants, Sally was supposedly the half-sister of Jefferson’s wife, which would make Sally Jefferson’s sister-in-law, and her children his niece and nephews! There is no evidence, other than the statements by Madison Hemings and Isaac Jefferson, for believing that John Wayles was the father of Sally Hemings, and some historians do not believe this was true. But if it is true, we know that the Carr brothers, who lived for some time at Monticello, were Jefferson’s nephews, and he was careful to look after them. If Sally’s children were the children of his wife’s half-sister, should we be surprised that Jefferson made special allowances for these closely-related servants? Indeed, we might expect them to be virtual family members, and to receive special consideration in training and the assignment of tasks.
Should it then seem unusual that Jefferson, when dying a bankrupt and knowing that all his property (including slaves) was to be auctioned to the highest bidder, would do everything in his power to prevent such close relatives/servants from being put on the auction block as slaves? One would be horrified were he to do otherwise. There’s no telling where they might end up if they were simply auctioned off to the highest bidder. And even before that final event, when two of them reached adulthood, might he not easily be agreeable to allowing them to leave the plantation and make lives for themselves on their own? And none of this need remotely suggest that they were his own flesh-and-blood children.
The fact that “Jefferson gave freedom to no other nuclear slave family” is meaningless, because it is only to be expected with this particular family. They were not just another slave family. Moreover, Jefferson had enormous debts, and freeing all his slaves was beyond his power, since all his property was mortgaged, and slaves were considered property in those days. But it is interesting to note that in freeing these uniquely-related servants, he did so roughly in accordance with a plan he outlined in his “Notes on Virginia” when Sally was less than ten years old. What Jefferson was doing with the Hemingses, therefore, was a modified form of the emancipation he had proposed for ALL slaves before Sally even went to France, years before any of her children were even born. It was the children of slaves who were to be freed, and that was to be done after they received a level of training that enabled them to continue as free persons.
Moreover, Sally’s children were not the only servants freed. Burwell Colbert, Joe Fosset, and John Hemings were also freed in Jefferson’s will. But that doesn’t make Jefferson their father. All were freed for several discernible reasons, but as Elizabeth Langhorne points out, there was one thing that all those who were freed had in common: “By no means did these five represent the whole Hemings connection; they were those whom Jefferson believed could go it alone in the white world.” (EL p. 256) That was important to Jefferson, who no doubt had second thoughts about freeing slaves when he saw what happened to James Hemings. James was freed in 1796, and even though he had training for a trade, he nevertheless had trouble finding and keeping a job, ended up an alcoholic, and committed suicide five years after being freed, in 1801.
8. Madison Hemings family history. The committee constantly put enormous emphasis on oral family tradition. The fact that Madison’s family story was printed in the Pike County (Ohio) Republican in 1873 no doubt tended to stabilize this story as a piece of family tradition to be passed down from generation to generation. But why the report should list this fact as a separate, and presumably important, “uncontested historical fact” apart from Madison’s statement itself seems curious. Perhaps this results from the Foundation’s emphasis on oral family tradition with their “Getting Word” project, and their obsession with oral tradition. This Madison family history, the report is careful to point out, was passed down “despite a climate of hostility and denial,” as though this level of tribulation somehow adds veracity to the story.
In genealogical circles, oral family tradition is given no credence at all, except perhaps to direct an investigator to substantial documented sources that are considered the only reliable genealogical materials. The present experience with the Woodson family tradition well illustrates how unreliable such stories can be. The Woodsons are adamant that they are descended from Thomas Jefferson, and absolutely refuse to accept the DNA evidence which proves conclusively and irrefutably that they are mistaken. But if the Woodsons are wrong about being descended from Thomas Jefferson in spite of their extremely strong tradition and beliefs, is it not equally possible that the other Hemings descendants are equally mistaken? On questions concerning the paternity of an unmarried woman’s children, oral family tradition is probably the most unreliable evidence that exists.
9. Family resemblances. Here again is another weak piece of evidence. The fact that Eston almost certainly had Jefferson blood in his veins, as indicated by the DNA tests, would cause one to expect that he might have a resemblance to the Jeffersons, including Thomas. The 12-year old boy Tom (Hemings) Woodson was also said to resemble the President, and we now know from the DNA tests that there was no connection to the Jefferson family at all. These “facts,” in and of themselves are ambiguous, have no great significance, and do not necessarily establish a paternal connection to Thomas Jefferson. If Thomas were the ONLY Jefferson that could have been the father, and the question was whether it was he or some other non-Jefferson-looking male that was the father, THEN perhaps this “look-alike” evidence might be significant. But when there is a whole handful of Jeffersons that cannot be eliminated as the possible father, look-alike evidence is meaningless. Any one of them could have been the father and have produced the same result. And when there is some evidence that another Jefferson indeed was at Monticello at the time of conception, then this look-alike evidence turns out to be worthless.
The report notes that “Thomas Jefferson Randolph told Henry S. Randall in the 1850s of the close resemblance of Sally Hemings’s children to Thomas Jefferson.” In fact, he told Randall that Sally “had children which resembled Mr. Jefferson so closely that it was plain that they had his blood in their veins.” So what are we to make of this? Later on in the very same interview, T. J. Randolph told Randall “that there was not the shadow of suspicion that Mr. Jefferson in this or any other instance ever had commerce with his female slaves.” This could only mean that what he meant in the first statement was, it was plain TO THOSE WHO SPREAD THE RUMOR that Sally’s children had Jefferson’s blood in their veins, or perhaps he thought the children had JEFFERSON blood in their veins, but not that of Thomas Jefferson himself. He went on to explain in that interview that he actually lived at Monticello, and he saw nothing that suggested to him that there was any familiarity between Mr. Jefferson and Sally Hemings. And, as we now know, there is good evidence that Randolph Jefferson could have been the father of at least one of Sally’s children, therefore what we now know confirms BOTH that at least one of her children had Jefferson blood, and that this blood was likely not from Thomas Jefferson.
Summary. As we can see from the above nine major pieces of “uncontested scientific or historical facts,” there is NOTHING to directly link Thomas Jefferson as father of Sally’s children. None of this is “overwhelming,” and all of it can be explained easily and simply. The “evidence” that identifies Thomas Jefferson as the father is all gossip, imaginative supposition, and oral tradition, and all of those are examples of weak evidence indeed. The very strongest possible kinds of evidence — eyewitness testimony, and the perpetrator’s confessions — are both on the side of Jefferson’s innocence. Therefore, weighing the two sides, and judging according to the quality of the evidence, one feels compelled to conclude that the best evidence says that Jefferson was NOT involved in an affair with Sally Hemings.
The report states that “Many aspects of this likely relationship between Sally Hemings and Thomas Jefferson are, and may remain, unclear, such as the nature of the relationship, the existence and longevity of Sally Hemings’s first child, and the identity of Thomas C. Woodson.” The fact is, the whole hypothesis that there was a relationship is built on inference, supposition, and handed-down gossip. There is no solid evidence except that opposed to the existence of a relationship. Therefore, it is to be expected that there would be any number of loose ends and contradictions that can never be resolved, mainly because it is all based on false assumptions. That the committee would issue a report saying the evidence was “overwhelming” in favor of a relationship when such fundamental aspects, such as the issues surrounding Tom Woodson as Sally’s first child, are unresolved only attests to the committee’s wish to promote their findings regardless of the evidence. On that issue alone, a reasonable person would conclude that no position should be taken unless those questions were settled.
Absurdities and Contradictions
It became necessary for the Research Committee to gloss over all the many absurdities and contradictions engendered by the theory they adopted, which assumed that indeed there was an affair between Thomas Jefferson and Sally Hemings. Unlike real scientific solutions, their “solution” to this controversial question does not provide a KEY to understanding. It does not make all or most of the many elements of this story fall into place and make sense. Rather, just the opposite occurs: it is a forced and contorted and distorted explanation, disregarding key pieces of evidence for insubstantial reasons, varying standards of evidence to accommodate their own theory, making certain assumptions for evidence favoring their theory and reversing those assumptions for evidence not favoring their theory, failing to view with common sense the likelihood of persons actually knowing the rumors and gossip they were repeating, and themselves repeating the unscientific fallacies and misconceptions of biased proponents of an affair. True explanations clarify a situation; they don’t make it more confusing, nor do they rely on a belief that a large number of reputable people were liars. Here are some of the absurd and contradictory results of this disgraceful example of scholarship:
•Thomas Jefferson himself, one of the great Founding Fathers of this great nation, is made out to be a liar, a hypocrite, a child molester (if Sally had a son in 1790), a fornicator, a miscegenist, and a de facto law breaker over a period of 38 years.
•Martha Jefferson Randolph is made out to be a liar on her death bed.
•James Callender is acknowledged to be a liar, but then is also viewed as telling the truth (which was based solely on gossip).
•Oral tradition is accepted as equal or superior to personal testimony, even if the source of the tradition is unknown and cannot be verified.
•Tom Woodson, around whom the accusations started, is made never to have lived at Monticello.
•Thomas Jefferson is made to continue with this “affair” even after being exposed in the public press.
•Sally Hemings is assumed to be “monogamous” with respect to Thomas Jefferson, but is not “monogamous” when considering Edmund Bacon’s eyewitness evidence.
•The Carr brothers confession is not considered valid because of the fact that Eston had a Jefferson as his father, and that proves Sally was not monogamous with respect to the Carr brothers.
•The Carr brothers confession is also not considered valid because they made confessions on two different occasions, and they laughed at one and wept at the other.
•The Foundation’s Committee set itself up to determine the facts of this controversy, yet they refused to consult with notable opponents, such as Herbert Barger or Willard Sterne Randall.
•The Foundation and its committee refused to actively pursue additional scientific evidence in the form of DNA from the deceased William Hemings that would add a greater measure of certainty to their investigations.
With a committee of this sort, composed only of staff members of the Foundation, one would expect that they should bend over backwards in an attempt to demonstrate some kind of neutrality, open-mindedness, and even-handedness. We might even forgive them for showing a bias toward Thomas Jefferson, demanding a strict adherence to “guilty until PROVEN innocent,” and not accepting the probable existence of Jefferson’s paternity of Sally’s children without at least some solid, conclusive evidence that there indeed was such a connection. But that this committee, consisting of members of the Thomas Jefferson Memorial Foundation, should adopt a position supporting the enemies and calumniators of Jefferson, and that they should do this on the basis of the shoddiest, most incompetent interpretations of vague and inconclusive evidence, which relies almost entirely on implications and innuendos and handed down gossip — such a betrayal of the person and reputation of the great man their Foundation was established to uphold, is dastardly and unforgivable.
As the complete Research Report so carefully documents, ALL of the existing evidence related to these accusations was available to the committee. But their biased theory blinded them to the significance of the material. And this only proves that all the best evidence in the world is of no help when the minds of the investigators are closed to new leads and possibilities and to the input of ideas from those who disagree with their chosen theory.
It is understandable why the Foundation, after recording the oral history of dozens of descendants of ex-slaves associated with Monticello, might wish to validate their story. But the fact is, those stories each started from one person. And if that one person exaggerated, or switched the identity of a forebear, that became established as “oral tradition.” This switching the identity of a forebear was well illustrated in the case of Eston Hemings’s descendants, who apparently did it to make a break with their black heritage. Naturally, the Foundation feels a great sympathy with the persons who cooperated in their Getting Word project, and would find it repugnant to turn their back on these people and seem to consider such contributions worthless, even if that were true. BUT THAT IS ALL THE MORE REASON WHY THE FOUNDATION SHOULD NOT HAVE CONDUCTED THIS RESEARCH THEMSELVES. Or, if determined to do it in-house, they should have been especially careful to engage and listen to outside consultants and other persons with opposing view. This they evidently did not do, and this is something that would naturally tend to make their report unconsciously biased. The only “outside” consultants appear to be those who advised on the DNA results, the science of which, as one of the consultants stated, was so routine that “the DNA study does not merit a scholarly conference.” Thus we have the picture of a research team, carefully engaging multiple outsiders on questions upon which everyone agrees, and just as carefully avoiding outsiders on those aspects where there is real controversy.
While we can understand the need to validate the black experience and to help all those once-discriminated-against men and women feel they are an authentic part of American life, nevertheless we feel that to do this by destroying the honor and reputation of one of America’s greatest Founding Fathers, and discredit the man who was the intellectual father of American democracy — a political system that has become the rallying point for freedom-loving people around the world — this can do neither blacks, nor this country, nor the cause of liberty any good whatsoever. This is not the time to be denigrating our most eloquent spokesman for liberty, democracy, and freedom. That this should be done on the basis of such shoddy evidence and corrupted methodology only attests to the incompetence of those who organized this study.
Bias is very tricky, and an honest investigator must take extraordinary steps to prevent it from sneaking in and displacing other views that might lead to a solution. As Nobel laureate and scientist Richard Feynman wrote, “The first principle is that you must not fool yourself — and you are the easiest person to fool.” This describes the scientific approach, the approach of true investigators. They deliberately seek opposing views and carefully test them against their own theory, because it is through that clash of differing ideas and opinions that the truth emerges. If investigators do not do this, they run the risk of fooling themselves and ultimately undermining their own best intentions. Input from different viewpoints is needed in every endeavor. Shutting such inputs out is a prescription for disaster.
But this also points up the poverty of the approach taken by the historians on the Foundation staff and elsewhere who deliberately exclude those who disagree with them, who call them “crazy,” and “flat-earthers.” It should be plain to even the simplest of minds, that if we are dealing with a controversy — and certainly the questions surrounding Thomas Jefferson and Sally Hemings are controversial! — then reasonable, intelligent people can take different sides, and one cannot hope to reach an intelligent conclusion by treating opposing views with contempt. Indeed, an intelligent investigator cherishes such views, and studies them more carefully than the views of supporters, because it is through understanding the opposition that one is likely to gain new insights.
The committee report suggests that the preponderance of the historical evidence indicates that Thomas Jefferson was the father of Sally’s children, but this is patently false. There is not a single stitch of direct evidence linking Thomas Jefferson to the paternity of Sally’s children, and every piece of so-called “evidence” can be easily explained in simple, ordinary, uncomplicated ways. No one suggests that Jefferson ever publicly or privately acknowledged Sally’s children to be his — not even Madison in his statement, or Cocke in his diary, or any of the oral family traditions. From the very beginning, the whole story has been the product of gossip based on inference, supposition, and imagination. Nevertheless, there are several pieces of good evidence indicating that Jefferson was NOT involved with Sally, and to counteract these, proponents of an affair are compelled to make all the responsible persons into liars. The only response that an open-minded, intelligent person could have to this tissue of twisted inferences is to say that it is all poppycock.
BM Bernard Mayo and James A. Bear, Jr.
“Thomas Jefferson and His Unknown Brother”
University Press of Virginia, 1981.
DM Dumas Malone
Jefferson and his Time, 6 vols.
Boston: Little, Brown, 1948-1981.
EL Elizabeth Langhorne
Monticello: A Family Story
Chapel Hill: Algonquin Books, 1987
GR Annette Gordon-Reed
Thomas Jefferson and Sally Hemings: An American Controversy
Charlottesville: University Press of Virginia, c. 1997, 1999.
JB James A. Bear, Jr., ed.
Jefferson at Monticello
Charlottesville: University Press of Virginia, 1967.