The message from Kevin Hardwick, in my view, illustrates in several ways why 
Hemings-TJ ought still to be discussed. I hope people who regularly delete 
these replies will give me a chance. I don't shout and I don't call names 
and I've studied the scientific evidence, and its uses and misuse, for a 
long time.

Professor Hardwick wrote:
 > Can we agree that we know these things
 > with reasonable certitude?
I'd rather change the word _certitude_ to _certainty_, but in any case: no, 
not all of these things. Please consider:

 > 5.  She accompanied Jefferson to France
Wasn't it more like the following? When TJ, already in France, sent for his 
younger daughter, he stipulated that she be chaperoned. He had no idea that 
she would be chaperoned by a mere child who seemed notably immature to both 
the transporting seacaptain and, on arrival in England, to Abigail Adams, 
who met the two children before they went on to France.

 > 6.  She had children by more than one man
I thought there was no evidence one way or the other. This is touchy 
territory, because it is often construed -- with a good deal of justice, in 
my view -- to overlap the domain of Sally Hemings's dignity, a domain that 
in my view is vitally important (and that in my view is also the grounding 
for the excesses that some paternity believers sometimes demonstrate -- just 
as it is a domain that a few paternity disbelievers too often seem to 
disrespect). A one-father assumption, sometimes called the monogamy 
assumption, underlies the Neiman statistical study of the conceptions 
coincidences, but if Professor Hardwick is right, there's a third fatal flaw 
in that study. (More below about all of that.) A one-father assumption also 
underlies Annette Gordon-Reed's and R. B. Bernstein's frequent, and 
scientifically erroneous, claim that the DNA molecular findings blasted the 
Carr paternity possibility not only for Eston's case, but for the cases of 
all of the Hemings kids. In fact the DNA says absolutely nothing about any 
of the kids but Eston. (Again, as I think I've said in this forum before, 
maybe the Carr paternity possibility is preposterous. I wouldn't know. I 
just know that the DNA says absolutely nothing about the paternity of any 
Hemings kid but Eston, and therefore says nothing about Carr paternity, or 
lack of it, for any of the other kids.)

> 7.  The father of one of her children was descended
>  from Thomas Jefferson's paternal grandfather--that
> is to say, we can narrow the list of possible fathers
>  for this child to Thomas Jefferson, his paternal uncles,
>  and his paternal cousins

Yes to all before the sentence dash, but what follows the sentence dash 
appears to exclude the possibility of paternity by some unknown carrier of 
the DNA marker within the enslaved population. We know that the DNA marker 
crossed the race line. We do not know whether it crossed the race line in a 
generation earlier than the one that produced Eston Hemings. Now, this may 
well be an outlandish possibility in terms of the historical evidence. But 
because the DNA evidence says nothing whatsoever about it, the circle of 
paternity candidates must actually -- whenever the context is what the DNA 
did and did not prove -- be defined as wider than the circle of males in the 
known, acknowledged, extended Jefferson family. As a matter of DNA science, 
we do not know the radius of the circle of paternity candidates.

> 8.  Most (all?) of Heming's pregnancies correspond with times when 
> Jefferson was at Monticello

Yes, in a qualitative sense, but in a strict biostatistical sense, no, not 
completely. This matters because Jan Lewis, R. B. Bernstein and the Thomas 
Jefferson Foundation all believe that the Neiman statistical study is a 
"pillar" of paternity proof. One of the two fatal flaws in the Neiman study 
(the first being the equating of correlation with causation) is the failure 
to account for the probabilities associated with the varying lengths of 
human gestation. Births occur, roughly speaking, nine months following 
conceptions -- plus or minus, roughly, two weeks. Dr. Neiman defined six 
kids' cases for what he called his "probabilistic evaluation." In one of 
those cases, TJ was absent for more than half of the four-week conception 
window. For that case, Dr. Neiman awarded himself a fudge factor. But TJ was 
also absent for a large portion of one other conception window, and was 
absent as well for small but not insignificant parts of the windows for two 
other cases -- including the case of Eston, the only Hemings child actually 
linked to the Jefferson DNA marker. In other words, in four of the six cases 
that Dr. Neiman defined for study, TJ was not present for the full 
conception window. Statistically, quantitatively, that is a big deal, even 
though it doesn't diminish the intriguing-ness of the qualitative evidence. 
In a letter to the Richmond Times-Dispatch, Annette Gordon-Reed once 
criticized my reasoning on all of this, but as my friend the physicist David 
R. Douglas clarified in a follow-on letter, she apparently didn't understand 
that in my op-ed I had been criticizing bogus quantification and the 
unjustifiable invoking of the authority of statistical science -- and not 
criticizing the obviously intriguing qualitative evidence that Professor 
Hardwick is sort of right in identifying here. (By the way, if he's right 
that there was more than one dad, the Neiman study has a third fatal flaw, 
for the study is built on the one-father assumption.)

> 9.  During the years when Hemings was having
> children, Jefferson was rarely at Monticello
(This is not an objection to item 9.) It seems to me that for that time 
period as Dr. Neiman chose to define it -- a choice that inevitably involved 
some assuming -- he determined that TJ was present for about half.

 > 4.  She was the half sister of Jefferson's wife
I have no opinion on #4, so I moved it out of order to last position. But I 
can report that some paternity disbelievers dispute its certainty. (I can't 
remember if one of them is Cyndi Burton, but I note that Cyndi's book often 
seems to be criticized more than it is actually read -- as for instance in 
the case of U. Richmond Professor Woody Holton, who tars her deplorably with 
an implied charge of white supremacy in an Amazon.com book review that I 
believe shows he didn't even read her book. Cyndi might be wrong about the 
paternity. I don't know. But she is a careful scholar and she deserves to be 
read -- and she certainly does not deserve to be treated as shamefully as 
Professor Holton treats her in that review.)

______________________________________
To subscribe, change options, or unsubscribe please see the instructions at
http://listlva.lib.va.us/archives/va-hist.html