Abstract
The Jefferson–Hemings case is frequently discussed as though it were singular — either as a uniquely well-documented scandal or as a uniquely aggressive scholarly imposition. Neither framing survives comparison. This paper surveys the founding generation for structurally similar claims: allegations that a prominent figure fathered children by an enslaved or servant woman, resting on some combination of family tradition, testamentary anomaly, contemporary rumor, and genetic evidence. It proposes an evidentiary typology, places roughly a dozen cases within it, and draws three findings. First, the Jefferson–Hemings claim is the best-evidenced case of enslaver paternity among the founders — not a marginal one. Second, several claims accepted with markedly less public friction (John Wayles and Betty Hemings; Aaron Burr and Mary Emmons; the Custis–Dandridge line) rest on evidence weaker than what supports the Jefferson claim, which indicates that the resistance in Jefferson’s case was driven by the identity of the accused rather than by the quality of the evidence. Third, the ancillary claim of a negotiated Paris freedom agreement has no comparable case anywhere in the period; where enslaved persons did secure freedom by bargain or petition, the transaction generally left a paper trail, and the absence of any parallel deprives that claim of the base-rate support that the paternity claim itself enjoys.
1. Purpose and Method
1.1 The comparative question
A claim can be assessed two ways: on its own evidence, and against the treatment of like claims. The second exercise is a check on the first. If a discipline credits proposition A on evidence of type X while rejecting proposition B on evidence of type X-plus, the discrepancy is not an evidentiary judgment but a judgment about the parties. Identifying such discrepancies is a legitimate — and testable — historiographical operation.
This paper accordingly asks: within the founding generation, what other cases exist in which a prominent white man is alleged to have fathered children by an enslaved or servant woman; what evidence supports each; and how has each been received?
1.2 Criteria for comparability
A case is treated as comparable if it involves (a) a figure of national prominence in the revolutionary or early national period, (b) a specific woman of enslaved or unfree status, (c) named alleged offspring, and (d) a claim advanced or preserved by an identifiable source. Generic assertions that planters fathered children by enslaved women — true, and abundantly documented — are treated as base-rate context rather than as cases.
1.3 An evidentiary typology
Six tiers, in descending order of strength:
Tier I — Principal’s acknowledgment. The man in question admitted or openly conducted the relationship.
Tier II — Contemporary third-party documentation. Letters, diaries, or legal records generated at the time by observers with direct knowledge.
Tier III — Genetic evidence with circumstantial support. Molecular data constraining the field of possible fathers, combined with documentary correlation.
Tier IV — Testamentary or behavioral anomaly plus contemporary rumor. Conduct inexplicable under normal practice, coinciding with local report.
Tier V — Family tradition alone. Descendant testimony transmitted orally and recorded late.
Tier VI — Modern assertion without traceable tradition. Claims first appearing in the twentieth or twenty-first century without demonstrable descent from earlier memory.
2. Base-Rate Context
Before individual cases, the frequency question. Sexual coercion of enslaved women by enslavers was not an aberration in Virginia and the lower South; it was structural. William Byrd II’s coded diaries, kept a generation before the Revolution, record his own conduct toward enslaved and servant women with a candor no later planter matched in writing (Wright & Tinling, 1941). Rothman’s (2003) county-level study of Virginia demonstrates that such relationships were widely known within neighborhoods, tolerated so long as they went unnamed, and punished chiefly when publicity forced the issue. Morgan’s (1998) comparative work on the Chesapeake and lowcountry establishes the demographic footprint.
The most useful single contemporary comment comes from John Hartwell Cocke, a Virginia planter, Jefferson’s collaborator on the University of Virginia, and a temperance and colonization reformer. Writing privately in the 1850s, Cocke observed that widowed and unmarried planters commonly kept enslaved women as concubines, stated that the practice was nowhere more prevalent than in Virginia, and referred to Jefferson’s case as a well-known instance (Cocke Family Papers, 1853/1859; discussed in Gordon-Reed, 1997). Cocke was not a partisan of the allegation; he was a neighbor treating it as a familiar fact.
The base rate therefore does not make Jefferson’s paternity likely by itself — base rates never establish individual cases — but it disposes of the argument, once standard, that the allegation was implausible on its face for a man of his class and standing.
3. Tier I: Acknowledged Cases
These matter as calibration. They show what the record looks like when concealment is abandoned.
3.1 Thomas Bell and Mary Hemings
Mary Hemings, Sally’s eldest sister, was leased to the Charlottesville merchant Thomas Bell in the 1780s and bore him two children, Robert and Sally Bell. Bell purchased Mary at her request, lived with her openly, treated the children as his, and left them his property. Jefferson, who knew all of this, referred to Bell in warm terms in correspondence and continued to do business with him (Stanton, 2000, 2012).
This is the crucial control case. It occurred inside the same family, in the same town, in the same decade. It demonstrates that acknowledgment was possible for a merchant of middling rank — and, by contrast, what a man with national standing had to lose by it.
3.2 Benjamin Franklin and William Franklin
Franklin fathered William around 1730 by a woman never identified. He brought the child into his household, raised him, secured his appointment as royal governor of New Jersey, and never named the mother in any surviving document (Skemp, 1990).
The comparison is instructive in an uncomfortable direction. Franklin’s illegitimate paternity is universally accepted on his own acknowledgment; the mother’s identity has attracted comparatively modest investigative energy across three centuries. Where the child is claimed and the woman is not, historians have generally stopped looking.
3.3 Richard Mentor Johnson and Julia Chinn
Johnson, of the generation immediately following the founders, served as Vice President under Van Buren. He lived openly with Julia Chinn, an enslaved woman he never manumitted, treated her as mistress of his household and manager of his estate, acknowledged their daughters Imogene and Adaline, educated them, and attempted to introduce them into white society. The arrangement was politically damaging and contributed materially to his marginalization (Myers, 2023).
Johnson establishes both that open acknowledgment was survivable and that it was costly. He also establishes what the documentary residue of an acknowledged relationship looks like: correspondence, estate records, newspaper attacks answered rather than denied. The Jefferson record contains none of this, which is precisely what a concealed arrangement predicts.
3.4 Alexander Hamilton and Maria Reynolds
Not an enslavement case, but the period’s benchmark for evidentiary certainty. Hamilton published a pamphlet in 1797 confessing an extended adulterous affair in order to rebut a charge of speculation in public funds (Hamilton, 1797; Chernow, 2004). The accuser who forced the disclosure was James Callender — the same writer who five years later published the Hemings allegation.
The pairing is worth holding in view. Callender was unreliable, malicious, and correct about Hamilton. Being a bad source is not the same as being wrong, and being right once does not establish reliability. Callender’s Jefferson allegation contained a demonstrable error at its center — the existence of a son “Tom,” refuted by both Monticello’s records and by the exclusion of the Woodson line in 1998 (Foster et al., 1998).
4. Tier III: The Jefferson–Hemings Case
Treated in detail elsewhere; summarized here for comparative placement.
The 1998 Y-chromosome study established that a male of the Jefferson patriline fathered Eston Hemings, and excluded the Carr brothers whom Jefferson’s grandchildren had blamed (Foster et al., 1998). Thomas Jefferson could not be tested directly, having left no legitimate male-line descendants; the finding therefore constrains the field to roughly two dozen Jefferson men, of whom perhaps eight were plausibly proximate (Turner, 2011).
Three circumstantial elements narrow it further. Jefferson was resident at Monticello during every conception window across the 1795–1808 span, with no conception during his frequent long absences; Neiman (2000) estimated the chance coincidence at roughly one percent. Every person Jefferson freed in a lifetime of holding over six hundred was a Hemings (Stanton, 2012). And the family’s counter-testimony collapsed under examination, contradicting itself as to which Carr brother was responsible before genetics excluded both (Gordon-Reed, 1997).
Supporting the whole is the Madison Hemings memoir of 1873 — late, secondhand, editorially mediated, but internally coherent and unrefuted in any testable particular (Hemings, 1873).
Tier III placement. No other case in the period combines genetic constraint, documentary correlation, behavioral anomaly, family tradition, and contemporary neighborhood report. Whatever one concludes, this is the strongest case of its kind in the founding generation, not the weakest.
5. Tier IV: Testamentary Anomaly and Rumor
5.1 George Wythe and Michael Brown
Wythe — signer of the Declaration, Jefferson’s law teacher, Virginia’s first professor of law — freed his housekeeper Lydia Broadnax and lived with her in Richmond in his widowerhood. His will provided for a free boy of color named Michael Brown, left him property, and named Jefferson as trustee for his education. In 1806 Wythe’s grandnephew George Wythe Sweeney, facing disinheritance, poisoned the household; Brown died, Wythe died after altering his will, and Sweeney was acquitted because Virginia law barred Broadnax, the eyewitness, from testifying against a white man (Brown, 1981; Chadwick, 2009).
The case is a near-perfect structural parallel to the Jefferson claim minus the genetics: cohabitation with a formerly enslaved woman, an extraordinary testamentary provision for a child of color, contemporary local assumption of paternity, and no documentary confirmation. Michael Brown died at roughly sixteen, leaving no descendants to test.
Wythe’s paternity of Brown is widely reported in the biographical literature with far less hedging than attends Jefferson, and on materially thinner evidence. Wythe is a minor figure whose reputation carries no national freight.
5.2 John Wayles and Betty Hemings
Jefferson’s father-in-law is held to have fathered six children by Betty Hemings, including Sally, making the Hemings children half-siblings of Jefferson’s wife Martha. The claim is essentially universal in the scholarly literature and appears in Monticello’s own interpretation without qualification.
Its evidentiary basis is Madison Hemings’ 1873 memoir, supported by circumstance: the extraordinary status of the Hemings family at Monticello, their exemption from field labor, and their monopoly on manumission (Hemings, 1873; Stanton, 2000; Gordon-Reed, 2008).
This is the single most revealing comparison in the paper. The Wayles paternity rests on one source. The Jefferson paternity rests on that same source, plus genetic constraint, plus presence correlation, plus the manumission pattern, plus the failure of every alternative account. For two centuries the first was accepted and the second was denounced as slander. The differential cannot be explained by evidence. It is explained by the fact that John Wayles has no monument.
6. Tier V: Family Tradition Alone
6.1 Aaron Burr and Mary Emmons
Burr — Revolutionary officer, senator, Vice President — is held by a documented descendant line to have fathered two children by Mary Emmons, a woman of Haitian or South Asian origin in the household of his first wife Theodosia: John Pierre Burr (1792–1864), a Philadelphia barber, abolitionist, and Underground Railroad conductor, and Louisa Charlotte Burr. The claim rests on family tradition transmitted through the descendants, with supporting circumstance in the naming, the household connection, and John Pierre Burr’s own standing in Philadelphia’s free community (Burr, 2019).
No contemporary documentation confirms it. No genetic testing has resolved it. In 2018 the Aaron Burr Association formally recognized the descendants and their claim.
The comparison is direct and unflattering to consistency. Burr’s descendants were accepted institutionally on Tier V evidence — family tradition without genetics or documentation — with essentially no public controversy. Jefferson’s descendants, holding Tier III evidence, faced three decades of contestation and were excluded from the Monticello Association burial ground by vote in 2002. The difference is Monticello and the nickel.
6.2 George Washington and West Ford
West Ford (c. 1784–1863) was born to an enslaved woman named Venus at Bushfield, the plantation of Washington’s brother John Augustine Washington. Ford was freed by Hannah Bushrod Washington’s will, later resided at Mount Vernon, and received land there. His descendants preserve a tradition of descent from George Washington (Bryant, 2001).
Two objections are heavy. Washington fathered no children in a long marriage to a demonstrably fertile woman and is generally held to have been sterile — Martha Custis had four children by her first husband. And Washington’s documented presence at Bushfield in the relevant window is thin, while John Augustine Washington and his sons Bushrod, Corbin, and William Augustine were resident or frequent (Wiencek, 2003; Thompson, 2019).
The case is the mirror image of the Randolph Jefferson objection: a patriline containing several plausible men, a tradition naming the most famous of them, and a documentary record that better fits a less famous one. Where Jefferson’s case answers the objection through presence correlation across thirteen years, Ford’s cannot. The Washington family has declined the exhumation that would permit Y-chromosome testing, so the case remains at Tier V.
The instructive point is that Ford’s tradition and Eston Hemings’ tradition are formally identical in kind, and the evidence attached to them is not. Family tradition is a real historical source and a fallible one; it is capable of both preserving truth and consolidating around the most prominent available name. Only external evidence distinguishes the two situations, and only in the Jefferson case does such evidence exist in quantity.
6.3 The Custis–Dandridge line
Ann Dandridge, an enslaved woman in the Custis household, is held to have been the daughter of Martha Washington’s father, John Dandridge, and therefore Martha’s enslaved half-sister. William Costin (1780–1842), a prominent free Black Washingtonian, is held by tradition to have been the son of Ann Dandridge and John Parke Custis, Martha’s son — which would make Costin simultaneously nephew and step-grandson within the same household (Wiencek, 2003; Thompson, 2019).
Evidence is tradition plus household circumstance plus the Washingtons’ unusual treatment of Costin. It is Tier V throughout and is generally reported in the literature with mild qualification rather than dispute.
7. Tier VI: Claims Without Traceable Tradition
Assertions of enslaver paternity attach at various times to James Madison, Patrick Henry, and others. Where such claims cannot be traced to a nineteenth-century family tradition or a contemporary document, they should be classed separately and not aggregated with Tiers III through V. The absence of a claim is also evidence: John Adams, John Jay, and the northern founders generally attract no such allegations, and in Adams’ case the family’s documentary record is dense enough that the silence carries weight.
Symmetry requires stating this plainly. If the standard is even, it excludes as well as includes.
8. Comparative Summary
| Case | Alleged offspring | Evidence class | Genetic test | Reception |
|---|---|---|---|---|
| Bell / Mary Hemings | Robert, Sally Bell | I — open acknowledgment | n/a | Uncontested |
| Franklin / unknown | William Franklin | I — acknowledgment | n/a | Uncontested; mother unresearched |
| R. M. Johnson / Chinn | Imogene, Adaline | I — open acknowledgment | n/a | Uncontested; politically punished at the time |
| Hamilton / Reynolds | none alleged | I — published confession | n/a | Uncontested |
| Jefferson / S. Hemings | six children | III — genetic + circumstantial + tradition | Yes; patriline confirmed, individual not | Contested for two centuries; majority now affirmative |
| Wythe / Broadnax | Michael Brown | IV — will anomaly + rumor | Impossible (no descendants) | Generally accepted, little scrutiny |
| Wayles / B. Hemings | six children | V — tradition + circumstance | No | Universally accepted |
| Burr / Emmons | John Pierre, Louisa | V — tradition | No | Accepted by association vote, 2018 |
| Washington / Venus | West Ford | V — tradition | Declined by family | Disputed; alternative candidates favored |
| Custis / Dandridge | William Costin | V — tradition | No | Generally accepted, mildly qualified |
9. The Paris Agreement Has No Comparable Case
The claim that Sally Hemings negotiated her return from France in exchange for privileges and a pledge to free her children at twenty-one belongs to a different analytical class from the paternity claim, and comparison exposes why.
Paternity claims have a base rate. Negotiated freedom agreements struck by enslaved women abroad do not. Searching the period for parallels yields the following:
James Hemings, 1793. Sally’s brother, likewise in Paris and likewise able to petition, entered a conditional manumission agreement with Jefferson: return to Virginia, train a successor cook, receive freedom. Jefferson honored it in 1796 (Stanton, 2000; Gordon-Reed, 2008). This agreement was written down. It is the closest analogue and it survives as a document.
French Admiralty petitions. Enslaved persons brought to France by colonial masters could and did petition for freedom under the Police des Noirs framework, and the Paris Admiralty Court registers preserve such petitions. The procedural avenue existed and generated records. Sally Hemings does not appear in them.
Anglo-American freedom suits. The Somerset framework in England and the Quock Walker and Elizabeth Freeman cases in Massachusetts all proceeded through courts and left full documentation.
Informal “time” arrangements. Common, undocumented — and this is what Sally Hemings herself actually received, from Martha Randolph around 1827, short of formal manumission (Stanton, 2000).
The pattern is consistent: where enslaved persons secured freedom by bargain, litigation, or petition, the transaction produced records, because such transactions required enforcement mechanisms and the parties knew it. The one relevant instance in Jefferson’s own household produced a signed paper.
An arrangement that would have been embarrassing to record is of course exactly the arrangement one would expect to go unrecorded. But that argument cannot be tested, and untestable arguments do not count as support. The comparative finding stands: the Paris agreement, unlike the paternity, has no class of like cases to raise its prior, and the nearest analogue within the same family cuts against it by being documented.
10. Findings
Finding 1. The Jefferson–Hemings paternity claim is the best-evidenced case of enslaver paternity in the founding generation. Every other case rests on a subset of the evidence supporting Jefferson’s, and no other case has been subjected to genetic testing that constrains the field of candidates.
Finding 2. The standard applied to Jefferson has been anomalously strict. Wayles, Burr, Wythe, and Custis paternities are accepted on Tier IV and Tier V evidence with a small fraction of the scrutiny directed at a Tier III claim. Since the evidence runs the other way, the differential measures the prominence of the accused rather than the strength of the case. This should be stated as a finding about the discipline, not as an additional argument for the claim — the discovery that others were credited too easily does not raise the evidence in Jefferson’s file.
Finding 3. The West Ford case is the necessary check on Finding 2. It shows that family tradition can attach to the most famous man in a patriline when a less famous one is the better documentary fit. This is the same objection that Randolph Jefferson poses at Monticello. The Jefferson claim answers it through thirteen years of presence correlation; the Ford claim cannot answer it at all. The two are formally alike and evidentially unlike, which is a point in favor of external evidence rather than a point against oral tradition.
Finding 4. The Paris agreement is uncorroborated within its own case and unsupported by any comparable case in the period. It should be attributed to Madison Hemings’ 1873 memoir whenever it is reported, and reported as family tradition.
11. Conclusion
Comparison neither elevates nor deflates the Jefferson–Hemings question so much as it relocates it. The case is not an outlier in what it alleges — the conduct alleged was ordinary in Virginia and is documented in the same household one generation earlier. It is an outlier in how much evidence exists and in how hard that evidence was resisted. Placed beside Wayles, Burr, Wythe, Ford, and Costin, Jefferson’s is the case with the most support and, until recently, the least credence.
The methodological lesson is that historiography should audit its own tiers. A discipline that accepts a claim about a father-in-law on one memoir while rejecting the same memoir’s claim about his son-in-law has not applied a standard; it has applied a preference. Correcting that requires holding the strong claims strongly, the weak claims weakly, and — where a single source is all there is — saying so every time.
References
Berlin, I. (1998). Many thousands gone: The first two centuries of slavery in North America. Harvard University Press.
Brown, I. (1981). American Aristides: A biography of George Wythe. Fairleigh Dickinson University Press.
Brown, K. M. (1996). Good wives, nasty wenches, and anxious patriarchs: Gender, race, and power in colonial Virginia. University of North Carolina Press.
Bryant, L. A. (2001). I cannot tell a lie: The true story of George Washington’s African American descendants. Writers Advantage.
Burr, S. L. (2019). Complicated lives: Free Blacks in Virginia, 1619–1865. Carolina Academic Press.
Chadwick, B. (2009). I am murdered: George Wythe, Thomas Jefferson, and the killing that shocked a new nation. Wiley.
Chernow, R. (2004). Alexander Hamilton. Penguin Press.
Cocke, J. H. (1853–1859). Diaries and papers [Cocke Family Papers, MSS 640]. Albert and Shirley Small Special Collections Library, University of Virginia.
Foster, E. A., Jobling, M. A., Taylor, P. G., Donnelly, P., de Knijff, P., Mieremet, R., Zerjal, T., & Tyler-Smith, C. (1998). Jefferson fathered slave’s last child. Nature, 396(6706), 27–28. https://doi.org/10.1038/23835
Foster, E. A., Jobling, M. A., Taylor, P. G., Donnelly, P., de Knijff, P., Mieremet, R., Zerjal, T., & Tyler-Smith, C. (1999). The Thomas Jefferson paternity case. Nature, 397(6714), 32. https://doi.org/10.1038/16266
Gordon-Reed, A. (1997). Thomas Jefferson and Sally Hemings: An American controversy. University Press of Virginia.
Gordon-Reed, A. (2008). The Hemingses of Monticello: An American family. W. W. Norton.
Hamilton, A. (1797). Observations on certain documents contained in No. V & VI of “The history of the United States for the year 1796.” Printed for John Fenno.
Hemings, M. (1873, March 13). Life among the lowly, No. 1. Pike County Republican.
Lanier, S., & Feldman, J. (2000). Jefferson’s children: The story of one American family. Random House.
Morgan, P. D. (1998). Slave counterpoint: Black culture in the eighteenth-century Chesapeake and lowcountry. University of North Carolina Press.
Myers, A. C. (2023). The Vice President’s Black wife: The untold life of Julia Chinn. University of North Carolina Press.
Neiman, F. D. (2000). Coincidence or causal connection? The relationship between Thomas Jefferson’s visits to Monticello and Sally Hemings’s conceptions. The William and Mary Quarterly, 57(1), 198–210. https://doi.org/10.2307/2674364
Onuf, P. S., & Lewis, J. E. (Eds.). (1999). Sally Hemings and Thomas Jefferson: History, memory, and civic culture. University Press of Virginia.
Peabody, S. (1996). “There are no slaves in France”: The political culture of race and slavery in the Ancien Régime. Oxford University Press.
Rothman, J. D. (2003). Notorious in the neighborhood: Sex and families across the color line in Virginia, 1787–1861. University of North Carolina Press.
Skemp, S. L. (1990). William Franklin: Son of a patriot, servant of a king. Oxford University Press.
Stanton, L. (2000). Free some day: The African American families of Monticello. Thomas Jefferson Foundation.
Stanton, L. (2012). “Those who labor for my happiness”: Slavery at Thomas Jefferson’s Monticello. University of Virginia Press.
Stanton, L., & Swann-Wright, D. (1999). Bonds of memory: Identity and the Hemings family. In P. S. Onuf & J. E. Lewis (Eds.), Sally Hemings and Thomas Jefferson: History, memory, and civic culture (pp. 161–183). University Press of Virginia.
Thomas Jefferson Memorial Foundation. (2000). Report of the research committee on Thomas Jefferson and Sally Hemings. Monticello.
Thompson, M. E. (2019). “The only unavoidable subject of regret”: George Washington, slavery, and the enslaved community at Mount Vernon. University of Virginia Press.
Turner, R. F. (Ed.). (2011). The Jefferson–Hemings controversy: Report of the Scholars Commission. Carolina Academic Press. (Original report issued 2001)
Wiencek, H. (2003). An imperfect God: George Washington, his slaves, and the creation of America. Farrar, Straus and Giroux.
Wright, L. B., & Tinling, M. (Eds.). (1941). The secret diary of William Byrd of Westover, 1709–1712. Dietz Press.
