Figure 13. The Carr Family Line. Two of the men have an identical haplotype, while C29 has a change at two positions. All three men would be considered to have the Carr haplotype.
Of the five direct descendants expected to carry the original Jefferson haplotype, four had identical patterns, and the fifth had evidence of a single change at one position out of the nineteen specific positions in the DNA being examined (as shown in Figure 12). These results give some assurance about the constancy of the DNA being examined, because it spans more than 200 years and up to eight generations.
DNA from the three descendants of John Carr had a very different haplotype from that of the Jeffersons (Figure 13), varying at almost half the sites in comparison to the Jeffersons (see Figure 16). Because the Carrs were related to Thomas Jefferson by his sister, and they had no known common ancestor, they would of course have a different Y chromosome. This result is therefore quite normal. Again, the three Carr descendants have a haplotype pattern that is very similar to each other, consistent with the general constancy of the Y chromosome.
Five descendants of Thomas Woodson had a haplotype pattern that was significantly different from the Jefferson haplotype (Figure 14). Of these five, four had a haplotype pattern that was almost the same, while W70 had a completely different haplotype. This latter finding strongly supports that for this individual the original Woodson Y chromosome was replaced at some time in the past seven generations by an outside Y chromosome. Thus, the current Woodson descendants are completely unrelated to the Jefferson male line. By extension, their ancestor Thomas Woodson cannot be descended from Thomas Jefferson.
Figure 14. The Woodson/Hemings Family Line. The dashed line from Sally Hemings to Thomas Woodson reflects a lack of consensus as to whether he was her son. Three of the Woodson men (W55, W56, W69) have an identical haplotype, while W61 has a change at two positions. These four men would be considered to have the Woodson haplotype. W70 has a completely different haplotype. H21, descended from Eston Hemings, has the Jefferson haplotype which is shown by J41.
Figure 15. Comparison of the three family haplotypes, using three individuals from Figure 16. Positions where the Jefferson haplotype differs from the other two are marked by a solid circle. Where the Woodson and Carr haplotypes differ is marked by an open circle.
Because there are reports that Peter and Samuel Carr had made claims about fathering Sally Hemings’ children, we can compare the Woodson and Carr haplotypes. If Thomas Woodson was indeed conceived while Sally Hemings was in Paris, then no relationship is expected between the Carrs and Thomas Woodson. The comparison in Figure 15 was done by taking one member from the Carr family (C27) plus a descendant of Thomas Woodson (W55), plus one member from the Jefferson family (J41) for comparison. This directly emphasizes how different the Jefferson haplotype is from the other two. And, while there are only four differences between C27 and W55, the probability of that occurring in a few generations from a common ancestor is highly unlikely, as will be described later. Therefore, it is not credible that a Carr brother was the father of Thomas Woodson.
The two more significant results of the DNA study are the DNA correspondence between the Jefferson haplotype and the Eston Hemings haplotype, and the complete lack of such a correspondence with the Woodson haplotype. That Thomas Woodson, who is presumed to have been conceived while Sally Hemings was with Thomas Jefferson in Paris, did not have the Jefferson haplotype is very significant, because this indicates that Sally Hemings conceived by another male at a time (in Paris) when Thomas Jefferson was the only member of his clan in that city.
The finding that one current descendant of Eston Hemings has a haplotype that is exactly the same as the Jefferson haplotype (Figure 16) strongly implies that some ancestor of this person was sired by a man carrying the Jefferson Y chromosome. Most probably this occurred with the conception of Eston Hemings at Monticello, because Eston Hemings himself moved as an adult out of Virginia, thereby physically separating his descendants from the Jefferson family.
What Is the Likelihood for Changes Within a Haplotype?
Clearly, the benefit of using the Y chromosome for this analysis was the expectation, based on current knowledge, that overall changes in this chromosome are not common. However, a few specific sites have been identified in recent years where variability is sufficiently frequent that changes at these sites may now help to identify individuals. We see clear support for this, because after 6–8 generations, for the three family haplotypes shown in Figures 12–14, only one person in each set has shown changes from the pattern of the others in that family. (W70, in the Woodson family, is excluded from this analysis, because he clearly was not a member of the Woodson haplotype.)
Within the three family haplotypes (all summarized in Figure 16) two current males have a change at one position (J50 and W61) and one has changes at two positions (C29). The frequency for a single change at the MSY1 site has been determined at 2–11%,17 so that having two out of fourteen men (in Figure 16) show a change at this site is quite normal. The frequency for change in the microsatellite STRs is about 2.1 × 10–3 (about one in 500).18 Therefore, seeing two of fourteeen men in Figure 16 (J50 and C29) with one STR change is unexpected for such a small sample. However, these probabilities almost completely exclude a relationship between the Carrs and Thomas Woodson. To change the Carr haplotype to the Woodson haplotype requires two changes in the STRs for a single person, with a combined probability of 4.4 × 10–6 (about one in 227,000). While the Carr brothers may have fathered children by Sally Hemings, the current DNA samples do not support this in regard to Thomas Woodson.
Figure 16. Family lines of descent for the Jeffersons, the Carrs, and two possible children of Sally Hemings. The figure is adapted from Foster et al., 1998. Also included are Randolph Jefferson and four of his sons, for whom the best estimate for their birth year is shown. For some other individuals the year of their birth is shown in parentheses. The dashed line between Sally Hemings and Thomas Woodson indicates the uncertainty as to her being his mother.
Interpretation of the DNA Analysis by Foster et al.
The title of the article published by Foster et al. to report their results was: “Jefferson fathered slave’s last child.” This title is not a statement of fact. It therefore represents an unfortunate lapse, because Foster et al. omitted any presentation of the number of adult male Jeffersons living in 1807 in that region of Virginia (Figure 12). This has been corrected (Figure 16) to include Thomas Jefferson’s younger brother Randolph, and Randolph’s five sons, who all lived nearby. Four of Randolph Jefferson’s sons would have been old enough to beget a child. This would have been possible at their age in 1807, which can be determined from their birth year (Figure 16), obtained from family records.19 A title that would have been more correct, given the available data, would be: “A member of the Jefferson clan fathered slave’s last child.” Even this would not be an absolute statement of fact, but simply a more reasonable interpretation.
The above article in Nature was preceded by an opinion piece which was meant to focus attention on the DNA article, and to emphasize its dramatic conclusion.20 This opinion piece was written by two American professors: Eric S. Lander, a molecular biologist and geneticist, and Joseph J. Ellis, a historian. In their article these two authors state: “DNA analysis by Foster et al. shows that he [Thomas Jefferson] fathered at least one child by his slave, Sally Hemings.” Again, this statement misrepresents the actual results.
What was actually demonstrated by the DNA analysis is that a male with the Jefferson Y chromosome was the father of Eston Hemings. Even the DNA results do not absolutely establish this connection with a Jefferson male because the statistical probability that a non-family member near Monticello would have this haplotype is less than 1%, but it is not zero. Therefore, while the probability that a member of the Jefferson clan fathered Eston Hemings is very high, greater than 99%, it is not 100%. However, for most people these results confirm that a Jefferson male sired Eston
Hemings.
It is of interest that Nature, a very prestigious journal that carefully limits the size (page space) of accepted research articles, gave slightly more space to the opinion piece by Lander and Ellis than was used for the actual DNA paper itself,21 because Foster et al. in a later commentary claimed “Space constraints prevented us from expanding on alternative interpretations.”22
Clearly, the journal was willing to give extra page space to promote an exciting interpretation of the DNA results, and this was helped by the actions of Foster et al. in omitting to mention that Jefferson had a brother and nephews living nearby. As shown in Figure 16, little extra space is required to include Randolph Jefferson and his sons in the family tree. In responding to criticisms of his paper, Foster stated: “we hoped to obtain some objective data that would tilt the weight of evidence in one direction or another.”23 However, one should not “tilt” the evidence by omitting important facts.
Two final possibilities need to be considered:
1. While it is normal to assume that Thomas Jefferson shares the Jefferson DNA haplotype, there is a small chance that he in fact had some other DNA, in the same way that we see in Figure 14 that one of the current Woodsons has a Y chromosome different from the rest of his family. This is an unlikely scenario, and there is no historical support for this.
2. An ancestor of Thomas Jefferson may have produced male children by a slave woman. Such a male slave on the estate would then have had the same Y chromosome DNA haplotype as the Jefferson family. If such a slave produced a son by Sally Hemings, this would also lead to this child’s descendants having the Jefferson Y haplotype.
Conclusion
The DNA test has several important results:
1. Thomas Woodson could not have been related to Thomas Jefferson. Then if Thomas Woodson was conceived in Paris, it had to be by a different father. This in turn argues for Sally Hemings conceiving children by more than one partner, because the Woodson descendants have a different haplotype from the descendant of Eston Hemings.
2. The Carr brothers could not have been related to Thomas Woodson. Also, the Carr brothers have a haplotype significantly different from the Jefferson haplotype found in a descendant of Eston Hemings. These results show that they could not have sired the first or last child of Sally Hemings. The DNA gives no evidence regarding the paternity by either Carr brother of the remaining children born to Sally Hemings.
3. One current descendant of Eston Hemings had the Jefferson DNA haplotype. This strongly supports that Eston Hemings, the last child of Sally Hemings, also had the Jefferson Y DNA haplotype. People who could have been responsible for this child’s paternity include Thomas Jefferson, Randolph Jefferson and four of Randolph’s sons, other Jefferson males in the Field Jefferson family, or even male slaves sired by ancestors of Thomas Jefferson. It will require some careful documentation by historians (if appropriate records exist) to determine which of these males could have been at Monticello during the few weeks in 1807 when Eston Hemings could have been conceived.
Notes
1. E. A. Foster, M. A. Jobling, P. G. Taylor, P. Donnelly, P. de Knijff, R. Mieremet, T. Zerjal and C. Tyler-Smith, “Jefferson fathered slave’s last child,” Nature, 396 (1998), 27–28.
2. S. Jakubiczka, J. Arnemann, H. Cooke, M. Krawczak and J. Schmidtke, “A search for restriction fragment length polymorphism on the human Y chromosome,” Human Genetics, 84 (1989), 86–88.
3. P. Malaspina, F. Perischetti, A. Novelletto, C. Iodice, L. Terrenato, L. Wolfe, M. Ferraro and G. Prantera, “The human Y chromosome shows a low level of DNA polymorphism,” Annals of Human Genetics, 54 (1990), 297–305.
4. A. Spurdle and T. Jenkins, “The search for Y chromosome polymorphism is extended to negroids,” Human Molecular Genetics, 1 (1992), 169–170.
5. Because nucleotides contain, and are defined by, one of the four bases, the term base is as frequently used in describing a DNA sequence.
6. In 1998 this represented an enormous amount of DNA sequence, and at that time not a single human Y chromosome had been completely sequenced.
7. N. Mathias, M. Bayés and C. Tyler-Smith, “Highly informative compound haplotypes for the human Y chromosome,” Human Molecular Genetics, 3 (1994), 115–123.
8. M. T. Seielstad, J. M. Hebert, A. A. Lin, P. A. Underhill, M. Ibrahim, D. Vollrath and L. L. Cavalli-Sforza, “Construction of human Y-chromosomal haplotypes using a new polymorphic A to G transition,” Human Molecular Genetics, 3 (1994), 2159–2161.
9. L. S. Whitfield, J. E. Sulston and P. N. Goodfellow, “Sequence variation of the human Y chromosome,” Nature, 378 (1995), 379–380.
10. C. Kwok, C. Tyler-Smith, B. B. Mendonca, I. Hughes, G. D. Berkovitz, P. N. Goodfellow and J. R. Hawkins, “Mutation analysis of the 2 kb 5’ to SRY in XY females and XY intersex subjects,” Journal of Medical Genetics, 33 (1996), 465–468.
11. M. Kayser, Caglia, M., Corach, D., Fretwell, N., Gehrig, C., Graziosi, G., Heidorn, F., Herrmann, S., Herzog, B., Hidding, M., Honda, K., Jobling, M., Krawczak, M., Leim, K., Meuser, S., Meyer, E., Oesterreich, W., Pandya, A., Parson, W., Penacino, G., Perez-Lezaun, A., Piccini, A., Prinz, M., Schmitt, C., Schneider, P. M., Szibor, R., Teifel-Greding, J., Weichhold, G., de Knijff, P., Roewer, L., “Evaluation of Y-chromosomal STRs: a multicenter study,” International Journal of Legal Medicine, 110 (1997), 125–133.
12. T. Zerjal, B. Dashnyam, A. Pandya, M. Kayser, L. Roewer, F. R. Santos, W. Schiefenhövel, N. Fretwell, M. A. Jobling, S. Harihara, K. Shimizu, D. Semjidmaa, A. Sajantila, P. Salo, M. H. Crawford, E. K. Ginter, O. E. Evgrafov and C. Tyler-Smith, “Genetic relationships of Asians and Northern Europeans, revealed by Y-chromosomal DNA analysis,” American Journal of Human Genetics, 60 (1997), 1174–1183.
13. M. A. Jobling, E. Heyer, P. Dieltjes and P. de Knijff, “Y-chromosome-specific microsatellite mutation rates reexamined using a minisatellite, MSY1,” Human Molecular Genetics, 8 (1999), 2117–2120.
14. Because interesting results occur in the microsatellite STRs and the minisatellite MSY1, only these will be defined here. Microsatellite short tandem repeats (STRs) are sequences of 2–6 bases that are repeated in tandem many times along a DNA sequence. The eleven STRs examined, by position from left to right, were: 19-388-389A-389B-389C-389D-390-391-392-dxys156y. The number at each position defines how many repeats of that STR occur. The minisatellite MSY1 is considered one site, because it contains four patterns or types of DNA sequences in one continuous sequence. In the figures these pattern types (1–4) are designated in parentheses, and the number following the parenthesis indicates how often that pattern type was repeated.
15. Supra, notes 4 and 11.
16. Supra, notes 8, 9, 10, and 11. C. A. Tilford, T. Kuroda-Kawaguchi, H. Skaletsky, S. Rozen, L. G. Brown, M. Rosenberg, J. D. McPherson, K. Wylie, M. Sekhon, T. A. Kucaba, R. H. Waterston and D. C. Page, “A physical map of the human Y chromosome,” Nature, 409 (2001), 943–945.
17. Supra, note 11.
18. E. Heyer, J. Puymirat, P. Dieltjes, E. Bakker and P. de Knijff, “Estimating Y chromosome specific microsatellite mutation frequencies using deep rooting pedigrees,” Human Molecular Genetics, 6 (1997), 799–803.
19. Cynthia H. Burton, Jefferson Vindicated, 2005, Keswick, VA: Cynthia H. Burton, p. 51.
20. E. S. Lander and J. J. Ellis, “Founding Father,” Nature, 396 (1998), 13–14.
21. Id.
22. E. A. Foster, M. A. Jobling, P. G. Taylor, P. Donnelly, P. de Knijff, R. Mieremet, T. Zerjal and C. Tyler-Smith, “The Thomas Jefferson paternity case,” Nature, 397 (1999), 32.
23. Id.
* * *
Footnotes
* Notes for this chapter begin on page 324.
Statement of
Playwright Karyn Traut
18
Thomas Jefferson: Brother’s Keeper*
* * *
A Playwright’s Progress
I first began researching Thomas Jefferson and Sally Hemings in connection with writing a
play more than three decades ago. Because my process of exploration and research into the topic was different from academic or legal conventions, I thought it best to write of my experiences in this subject matter as it evolved from a playwright’s perspective.
A playwright’s task when writing of either historical or fictional characters is to understand them so thoroughly that they come to life within the mind of the playwright. At that point the characters seemingly begin to “talk” for themselves. With an historical figure, especially one as pivotal to the very nature of the American consciousness as Thomas Jefferson, a playwright must work hard at the research, checking many sides and angles of point of view. Not every playwright sees his/her task this way. Shakespeare himself worried little about historical accuracy and his plays have not suffered for the lack. However, for myself as an American playwright who is concerned that our history and mythology are too often commingled, I felt it imperative to present an accurate portrait of this crucial American figure.
The second quality I bring to my research, which is perhaps different from a typically scholarly or legal approach to a subject, is a proficiency at working jigsaw puzzles. There are many trials and errors in the working of a jigsaw puzzle, not to mention frustration and hair pulling over near attempts at fit. Not all pieces slip perfectly into alignment. Two sides to a piece can fit yet the piece can be erroneous because a third or fourth does not. If the piece is the wrong fit, the puzzle won’t coalesce until the wrong piece is removed. When a piece does fit it is like music. Every other piece in the puzzle can now find its place in the resulting unified whole. This is what happened when I found the “piece” of the Jefferson “puzzle” that was Randolph, Thomas Jefferson’s brother.
The Jefferson-Hemings Controversy Page 56