# The 1998 DNA papers — an honest account Compiled 24 September 2026 for the JeffersonMind vault. **Not for SillyTavern Data Bank. Not for Jefferson’s mouth.** This is a human/red-team note (`reference/` or next to `05_redteam.md`). --- ## What people mean by “the 1998 DNA papers” One short scientific correspondence in *Nature*, plus a same-issue commentary and a January 1999 clarification by the same authors. 1. **Eugene A. Foster, M. A. Jobling, P. G. Taylor, P. Donnelly, P. de Knijff, René Mieremet, T. Zerjal, C. Tyler-Smith.** “Jefferson fathered slave’s last child.” *Nature* 396 (5 November 1998): 27–28. DOI: [10.1038/23835](https://doi.org/10.1038/23835) PMID: 9817200. 2. **Eric S. Lander and Joseph J. Ellis.** “Founding Father.” *Nature* 396 (5 November 1998): 13–14. Companion commentary, not the lab paper. 3. **Foster et al., letter.** *Nature* 397 (7 January 1999): 32. Authors concede the headline overclaimed what Y-DNA can do, and that other Jefferson males (including Randolph Jefferson’s line) were not excluded. Foster, a retired pathologist in Charlottesville, gathered the blood. Typing was done in labs at Leicester, Oxford, and Leiden that already worked on Y-chromosome haplotypes. --- ## What was actually tested Y chromosomes pass father to son, almost unchanged. Thomas Jefferson had no surviving sons in a documented white male line that could be sampled. So the study used **male-line descendants of Field Jefferson**, Thomas’s paternal uncle. Those men carry the same Y haplotype Thomas would have carried, *if* the genealogy is right. Compared groups (as later summarized by the Thomas Jefferson Foundation): - Five male-line descendants of two sons of **Field Jefferson** - Three male-line descendants of three sons of **John Carr** (grandfather of Peter and Samuel Carr — Jefferson’s sister’s sons, the old “it was the nephews” theory) - Five male-line descendants of two sons of **Thomas Woodson** (a man whose family tradition said he was Sally Hemings’s first child, conceived in Paris) - One male-line descendant of **Eston Hemings** (Sally’s last known son, born 1808) They typed **nineteen markers** on the Y chromosome. --- ## What the molecules showed | Comparison | Result | |---|---| | Field Jefferson line vs Eston Hemings line | **Match.** Same rare haplotype. Chance match estimated well under 1%. | | Carr line vs Eston line | **No match.** Peter and Samuel Carr are excluded as Eston’s father. | | Jefferson line vs Woodson line | **No match.** A Jefferson male is excluded as Thomas Woodson’s father. | That is the entire genetic result. It is strong as far as it goes. The haplotype was uncommon in the reference samples then available. The Carr exclusion for **Eston** is clean. The Woodson exclusion for a Jefferson father is clean (later sampling of more Woodson descendants held). --- ## What the molecules did **not** show This is the part headlines skipped and the January 1999 letter walked back. 1. **Thomas Jefferson’s body was not sampled.** There is no “Jefferson’s DNA” in the 1998 paper. There is a **Jefferson-male-line Y**. 2. **Y-DNA cannot name which Jefferson man.** Every patrilineal Jefferson who inherited that Y — Thomas, his brother **Randolph** (1755–1815), Randolph’s sons, other male cousins in Virginia — looks the same on those markers. Contemporary counts used by later writers: on the order of **two dozen adult Jefferson males** in Virginia who could have carried it. 3. **Only Eston’s line was tested among Sally’s children.** Beverly, Harriet, and Madison Hemings were not given a Y test in 1998 (Madison’s male line, if sampled later, is a different study). A match for Eston does not, by itself, prove the father of the earlier children. 4. **It does not reconstruct a relationship.** DNA does not say when, whether consent existed, how long it lasted, or how Jefferson or Hemings described it. Those are historical questions. 5. **The title overspoke.** “Jefferson fathered slave’s last child” names Thomas. The abstract is already softer: the findings “provide evidence that he was the biological father of Eston Hemings Jefferson.” Foster later wrote that Thomas “can neither be definitely excluded nor solely implicated,” and that men of Randolph’s family could have fathered the later children. Lander and Ellis, in the companion essay, treated the case as historically closed in Thomas’s direction. That was an interpretation, not an extra experiment. --- ## How historians used it Two institutional readings still structure the argument. ### Thomas Jefferson Foundation (Monticello), January 2000 Staff research committee (chair Dianne Swann-Wright; Lucia Stanton among the historians). They accepted Foster’s method. They said DNA alone proves only “a Jefferson,” then combined it with: - Madison Hemings’s 1873 memoir (Sally was Jefferson’s “concubine”; the children were his) - James Callender’s 1802 public charge - Conception windows: Sally’s recorded children align with Jefferson’s presence at Monticello; she is not known to have conceived in his long absences - No contemporary tradition naming another Jefferson man until after 1998 - The Carr theory dying for Eston **Conclusion:** high probability that Thomas fathered **Eston**, and that he most likely fathered **all six** children in Jefferson’s records (Harriet 1795, d. inf.; Beverly 1798; unnamed daughter 1799, d. inf.; Harriet 1801; Madison 1805; Eston 1808). Nature of the relationship left open. Woodson’s status left open (no Monticello birth record for a child before 1795). President Daniel P. Jordan: paternity cannot be established with absolute certainty; best evidence suggests a relationship over time that led to one, and perhaps all, of the known children. A minority report inside that committee dissented. ### Thomas Jefferson Heritage Society, Scholars Commission, 2001 (rev. 2011) Independent panel (Robert F. Turner, chair). Majority: the case for Thomas’s paternity was not proven; several members preferred **Randolph** or another Jefferson male for at least Eston. They stressed the headline, the size of the Jefferson-male pool, gaps in the record, and political uses of Callender. This is not a second DNA experiment. It is a different weighting of the same molecules plus documents. --- ## Older evidence the paper sat on top of The 1998 work did not create the question. - **1802** — James Thomson Callender, in the Richmond *Recorder*, alleged a long liaison and children. - **1873** — Madison Hemings, in the Pike County *Republican*, named Jefferson as father and described Paris and Monticello. - **1873** — Israel Jefferson (formerly Israel Gillette), overlapping testimony. - **Family denial** among some white Jefferson descendants, often pointing at the Carr nephews. - **Oral history** among some Hemings descendants that the father was Jefferson. DNA killed the Carr explanation **for Eston**. It did not create Madison’s memoir and did not erase Callender’s motives (he was a vicious, interested pamphleteer). Both still have to be read as sources, not as oracles. --- ## Later genetics (do not collapse into “1998”) Work after Foster is a **different literature**. As of late 2026 it includes attempts to use autosomal sharing and reconstructed genomes among living descendants. Those papers and preprints are contested (method, unpublished underlying calls, which line is informative). They are not the 1998 Y study. If a new autosomal paper survives peer review, file it here as a dated addendum. Do not rewrite 1998 to say what 2026 hoped to say. --- ## Honest one-paragraph summary In 1998, Y-chromosome typing showed that Eston Hemings’s male-line descendant carries the same rare Jefferson-family haplotype as descendants of Field Jefferson, and not the Carr haplotype. That is good evidence that **a Jefferson man** fathered Eston, and good evidence that the Carr brothers did not father Eston. It is **not** a test of Thomas Jefferson’s own tissue, **not** a test that can separate Thomas from Randolph or other patrilineal Jeffersons, and **not** a test of Sally’s other children. Whether Thomas was that Jefferson man is a historical inference. Monticello’s 2000 committee judged the combined record as a high probability that he was, including the earlier children. A Heritage Society commission judged that inference unproven and pointed at other Jefferson males. Both readings start from the same short *Nature* correspondence. --- ## Why this stays out of JeffersonMind RAG The 1826 man did not have *Nature*, Y haplotypes, or “1998.” He did not write a confession and did not write a denial that names Sally as a lover. Mode A for the character: - He may speak of people he legally freed in 1826 (Burwell, John Hemings, Joe Fossett; Madison and Eston at 21). - He may go silent or change the subject if asked to narrate a sexual life with Sally Hemings. - He must not cite Foster, Lander, Ellis, Monticello 2000, or “the DNA proved it.” Use this file when **you** red-team the card, not when he answers. --- ## Sources - Foster et al., *Nature* 396 (1998): 27–28. https://doi.org/10.1038/23835 - Lander & Ellis, *Nature* 396 (1998): 13–14. - Foster et al., *Nature* 397 (1999): 32 (clarification). - PubMed 9817200. - Thomas Jefferson Foundation, *Report of the Research Committee on Thomas Jefferson and Sally Hemings* (January 2000), esp. Assessment of DNA Study, Other Jeffersons, Conclusions. https://www.monticello.org/slavery/jefferson-slavery/thomas-jefferson-and-sally-hemings-a-brief-account/research-report-on-jefferson-and-hemings/ - TJF brief account: https://www.monticello.org/slavery/jefferson-and-sally-hemings/a-brief-account - Encyclopedia Virginia excerpt of Foster: https://encyclopediavirginia.org/ - Thomas Jefferson Heritage Society pages on the 1998 study and Scholars Commission. - Contemporary news: *New York Times*, 1 Nov 1998; *Los Angeles Times*, 1 Nov 1998. - Marshall, “Which Jefferson was the father?” *Science* 283 (8 Jan 1999): 153–54. Do not paste paywalled *Nature* PDF text into the corpus.