DNA gets the final word - sort of

Jefferson scholars are as close as you come to a royal priesthood in the study of American history. The man on the nickel left behind letters and documents that take up six miles of microfilm. Other brilliant men have worn out lifetimes trying to sift through it all.

But the most important new take on the life of Thomas Jefferson comes from researchers who never made it through the canon. Their work signals the value of using DNA evidence to resolve longstanding historical puzzles. It also is an object lesson in avoiding snap conclusions.

In a Nov. 5 article in the British science journal Nature, retired pathology professor Eugene Foster and seven other scientists present genetic evidence that Jefferson likely fathered at least one child with Sally Hemings, a slave in his household.

To the classic Jefferson Establishment, such a claim has been anathema. Jefferson's political supporters denounced the charge when it first appeared in an opposition newspaper in 1802, during Jefferson's first term as president.

For nearly 200 years, most scholars agreed: Such a "vulgar liaison" is "distinctly out of character" and "virtually unthinkable," wrote Dumas Malone in his six-volume masterwork, "Jefferson and His Time." More recently, historian Joseph Ellis called the likelihood of a liaison "remote."

"Anyone who claims to have a clear answer to this most titillating question about the historical Jefferson is engaging in massive self-deception or outright lying. This is one mystery destined to remain unsolved," Dr. Ellis wrote in "American Sphinx: The Character of Thomas Jefferson," winner of the 1997 National Book Award.

To accept the scholars' account was to reject the oral histories of generations of descendants of Sally Hemings. "It's obvious why this evidence wasn't taken more seriously.... While this kind of liaison replicated itself across the South many, many times, it was simply not acceptable to society because it was Thomas Jefferson," says Beverly Gray, who has researched the Hemings family for more than 20 years.

Until recently, the only way to resolve this issue scientifically would have been to exhume the most beloved of the Founding Fathers - a prospect too ghoulish even for many who believed reports of a 38-year relationship between Jefferson and Hemings that produced at least six children.

But recent advances in DNA research opened up less-intrusive ways to study the question. Instead of exhuming bodies, geneticists can now compare the Y chromosomes from male-line descendants. Since Jefferson's only acknowledged son died in infancy, Dr. Foster collected blood samples from descendants of Field Jefferson, the president's paternal uncle. His study found a match in Y chromosome DNA with a descendant of Eston Hemings, Sally's youngest son. That means a Jefferson likely fathered at least one of the Hemings children - and the circumstantial evidence points to Thomas.

The probability of such a match arising by chance is "safely less than 1 percent," according to Eric Lander, a leading geneticist at the Massachusetts Institute of Technology, who co-authored a commentary on this study for Nature.

Reports of these findings, released early in response to pressure from US news media, set Jefferson scholars to revising their accounts. Guides at Monticello, Jefferson's home, worked the results into their tours. Some commentators opined that the revised history of the nation's third president should deflect moves to impeach its 42nd.

Foster says that he picked up the idea for his study from a casual dinner conversation in Charlottesville, Va. (Charlottesville was also the home of Anna Anderson, whose celebrated claim to be Anastasia, lost daughter of the last czar of Russia, was recently disproved by DNA analysis.)

He says he is not surprised at the intense interest in his work, but cautions that the DNA evidence not be overinterpreted. "This is not an open and shut case, even though I do think it is very likely that Thomas Jefferson was the father of Eston Hemings," he says.

"Any descendant of his uncle could conceivably have had a liaison with the wife of any descendant of Eston Hemings to account for this [match], but as far as we know there is little historical evidence to point to this possibility," he says.

Some forensic scientists caution that 1 in 100 odds are not compelling in the world of DNA evidence. The odds linking President Clinton to genetic material on Monica Lewinsky's dress are better than 1 to 7 trillion , according to the FBI.

"The news reports gave a conclusive ring to findings about Thomas Jefferson that are not conclusive," says James Starrs, a forensic scientist at George Washington University. "Those odds [1 in 100] wouldn't stand up in a court of law."

Foster responds that his study was not capable of providing a unique genetic "fingerprint," because it was based on only 19 Y chromosome markers, in contrast to the 200-plus markers used in conventional DNA fingerprinting. So the odds are in the hundreds to thousands, rather than in the millions or trillions. This kind of evidence "adds a bit of objectivity to the whole business, but it has to be taken in context," he adds.

In effect, what the DNA evidence has done is shift the terms of debate. For example, Jefferson family tradition has held that the president's nephews, Peter or Samuel Carr, fathered the Hemings children. But Foster's study found no match between the Carr and Hemings descendants. Now, new candidates are being publicly advanced, including the president's brother, Randolph, or his six sons. But there has been little evidence to support this claim. "Nobody suggested at any point that it was younger brother Randolph or that somebody else could have been involved. If we had heard other stories before, we'd be better prepared to listen to them now," says Peter Onuf, Thomas Jefferson Memorial Foundation professor of history at the University of Virginia in Charlottesville.

"The burden of proof has definitely shifted," says Ellis, who is currently revising his "American Sphinx."

New York Law School professor Annette Gordon-Reed says the new DNA evidence supports the case she developed in her 1997 book, "Thomas Jefferson and Sally Hemings: An American Controversy." "I knew that it would be hard for my book to ever carry the day, because so much of the debate has been in the realm of emotion, not reason," she says. "In my book, I said it would be settled through the medium of science."

* Send e-mail comments to chaddockg@csps.com