Google Groups no longer supports new usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Should Primary Evidence be Redefined?

20 views
Skip to the first unread message

Colin B. Withers

unread,
27 Sept 2010, 05:35:0727/09/2010
to gen-me...@rootsweb.com
The recent threads discussing 'primary evidence' or 'primary sources' got me thinking.

Do we need to re-evaluate the definition of 'Primary Evidence'?

Our One-Name-Study embarked on a DNA study a couple of years back, and as is the danger with this type of study it has revealed that a small number in our group (6 of the 110 tested so far) have found that they are not of the lineage that they thought they were, and this is just over a small number of generations!

This harks back to Homer in his Odyssey when Telemachus, upon being asked by Pallas Athene whether or not he was the son of Odysseus, replied: "It is a wise child who knows its own father."

If the only person who truly knows the father of a child is the mother, and even in some cases that knowledge is in doubt, the improbability that any man is directly descended from his father, grandfather, great-grandfather, etc rises exponentially with each generation back.

If a woman has an affair, that results in a pregnancy, she might well be tempted to deceive her husband into thinking that the child is his, particularly if she was having sexual relations with her husband in the days around her conception. If she wasn't, but he was drunk around that time, then a deceit may still be contemplated. She, of course, may not be too sure herself if she was active with both her husband and her lover, and would look for traits in the baby to confirm in her own mind who the father was.

Unfortunately, if a deceit is carried off, and this was not uncommon, then 'primary evidence' of birth/baptism, wills, etc, tells us what?

Fortunately with DNA we are given a certainty, rather than a probability, of direct lineage, when it comes to "no", but still only a statistical probability of lineage when it comes to "yes".

However, given the rapid development in testing techniques over the last couple of years, and the promise of further advances to come, it would seem to me that there is a layer now forming over and above what we normally would call 'primary evidence', (which is documentary, and subject to the normal errors and fraud that is associated with such sources), and that is DNA evidence.

Given the ability of DNA testing to totally trash a lineage that is completely supported by traditional 'primary evidence', should we rename 'primary evidence' or create a new one, for DNA results?

Colin

Peter Stewart

unread,
27 Sept 2010, 07:17:5127/09/2010
to

"Colin B. Withers" <Colin....@eumetsat.int> wrote in message
news:mailman.374.12855801...@rootsweb.com...

Why should it have to be "new"? DNA results are scientific evidence; the
best documented sources are primary evidence - DNA might prove, or indicate,
biological paternity; documents prove legal paternity.

If you think it's a foregone conclusion which of them matters more, in
recent generations at least, good luck to you. Not everyone will agree, or
indeed care a fig about DNA evidence where medieval genealogy is concerned.

If a line is long enough there may be an increased probability of biological
"non-paternity" somewhere in it, perhaps even at multiple points, but there
may also a solid chain of legal (for which read also cultural and ethical)
paternity leading to a non-biological ancestor.

Genealogy to many is the study of relationships between people, not just
their genes.

Peter Stewart

Colin B. Withers

unread,
27 Sept 2010, 07:56:3527/09/2010
to gen-me...@rootsweb.com
I suggested the word 'new' because DNA is a fairly new technology available to genealogists, and has the potential to be at a higher level (of proof) than 'primary sources' and as the word 'primary' is already in use it begged the question of how to describe DNA results in conjunction with 'secondary sources', 'primary sources' etc.

I understand there will be a range of views of DNA evidence, but rather than a us & them approach I would really like to know the views of a range of people about DNA results.

If a line of descent is traced for any individual, over years sometimes, using 'primary sources', and then find that the descent is not true at all, when the DNA results come in, and there is an error or fraud somewhere in the paper trail, I cannot imagine anyone not being perturbed by that.

It would appear that we will then have three classes of top-level genealogies:

a) A line of descent proven by primary sources AND proved by DNA
b) A line of descent proven by primary sources AND disproved by DNA
c) A line of descent proven by primary sources AS YET unconfirmed by DNA

It would be interesting to know from the heraldry group, who do take bloodlines very seriously, what the legal status of the right to bear a coat of arms is if a descent is subsequently disproved by DNA.

In the case of 2 of our 6 that were surprised by the DNA test, they did find via the Ybase and YSearch databases that their results linked them to another surname altogether (the same surname), and are now busy trying to piece together where the likely error/fraud/undocumented adoption, etc., occurred. They have told me that this is as much interesting to them as their previous endeavours

In my own case, it was DNA that eventually confirmed a suspected line of descent, that had eluded documentary proof for nearly 30 years. So for me, it was a joyous result. I cannot imagine how someone who had succeeded in the paper trail over the same period, finding that the DNA just did not match, would feel. I hope more views will be forthcoming.

Colin


Why should it have to be "new"? DNA results are scientific evidence; the
best documented sources are primary evidence - DNA might prove, or indicate,
biological paternity; documents prove legal paternity.

If you think it's a foregone conclusion which of them matters more, in
recent generations at least, good luck to you. Not everyone will agree, or
indeed care a fig about DNA evidence where medieval genealogy is concerned.

If a line is long enough there may be an increased probability of biological
"non-paternity" somewhere in it, perhaps even at multiple points, but there
may also a solid chain of legal (for which read also cultural and ethical)
paternity leading to a non-biological ancestor.

Genealogy to many is the study of relationships between people, not just
their genes.

Peter Stewart


-------------------------------
To unsubscribe from the list, please send an email to GEN-MEDIEV...@rootsweb.com with the word 'unsubscribe' without the quotes in the subject and the body of the message

Peter Stewart

unread,
27 Sept 2010, 08:48:5127/09/2010
to

"Colin B. Withers" <Colin....@eumetsat.int> wrote in message
news:mailman.375.12855886...@rootsweb.com...

>I suggested the word 'new' because DNA is a fairly new technology available
>to genealogists, and has the potential to be at a higher level (of proof)
>than 'primary sources' and as the word 'primary' is already in use it
>begged the question of how to describe DNA results in conjunction with
>'secondary sources', 'primary sources' etc.
>
> I understand there will be a range of views of DNA evidence, but rather
> than a us & them approach I would really like to know the views of a range
> of people about DNA results.
>
> If a line of descent is traced for any individual, over years sometimes,
> using 'primary sources', and then find that the descent is not true at
> all, when the DNA results come in, and there is an error or fraud
> somewhere in the paper trail, I cannot imagine anyone not being perturbed
> by that.

Presumably you mean "and then find that the descent is not biologically true
at all" - the "disproved" ancestry is still _legally_ valid. Truth is not
the same as validity in this, just as biology is the not same as legality.

For instance, if Prince William should be discovered to have DNA
incompatible with the paternity of the Prince of Wales, his succession to
the throne would not be affected at all. The legal principle, since Roman
times, has been "pater is est quem nuptiae demonstrant" - the father is
whomsever (the mother's) marriage indicates - and since Prince Charles was
married to his mother at the time of his birth, that is _legally_ that.

Biology might be different from the documentary record, but that doesn't
necessarily make it the all-determinative factor you seem to be assuming.

Peter Stewart

Colin B. Withers

unread,
27 Sept 2010, 09:30:2627/09/2010
to Peter Stewart, gen-me...@rootsweb.com
No, not at all. My only thoughts on this were this: if primary evidence is considered more valid than secondary evidence, where should DNA evidence be placed? Above primary? Below primary? Not considered at all?

The history of "pater is est quem nuptiae demonstrant" is an interesting one. The principle was originally: "Mater semper certa est, pater semper incertus est". However, this eventually gave way to a more pragmatic principle of 'presumption'. "Pater is est quem nuptiae demonstrant" does not mean that the father "is whomsever (the mother's) marriage indicates", but rather the father "is presumed to be....", in the absence of evidence to the contrary. And paternity cases did crop up, and evidence against the presumption submitted.

Even now, the bedrock of the Romanic law, of "Mater semper certa est" is now no longer valid (since 1978), with the advent of in-vitro fertilization, which will cause all kinds of problems for genealogist years from now. But even in the medieval period, it was not unknown for a woman who was being blamed for not producing an heir, to come to a mutually beneficial arrangement with a local girl who had an unwanted pregnancy, and of course in cases like this the presumption of "Mater semper certa est" is entirely wrong (and by that I mean a 'crime' is still a crime, even if undetected). It seems to me that one of the key differences between primary and secondary sources is also one of presumption.

While on the subject of primary sources, it may be interesting to the members of the list, that in the research for "Yorkshire Parish Registers", it was found that in the case of the East Riding, around 30% of all parish registers were found to be 'secondary', i.e. a copy, and it was the Bishops' Transcripts (presumed to be the copy) that were in fact the primary record.

Another interesting finding thrown up by the research for "Yorkshire Probate" was that in a sample of 100 of the wills in the East Riding Registry of Deeds 35 had not been proven at all.

Colin

Peter Stewart


WJho...@aol.com

unread,
27 Sept 2010, 10:21:4927/09/2010
to Colin....@eumetsat.int, pss...@bigpond.com, gen-me...@rootsweb.com
In a message dated 9/27/2010 6:30:35 AM Pacific Daylight Time,
Colin....@eumetsat.int writes:


> No, not at all. My only thoughts on this were this: if primary evidence
> is considered more valid than secondary evidence, where should DNA evidence
> be placed? Above primary? Below primary? Not considered at all?
>
> The history of "pater is est quem nuptiae demonstrant" is an interesting
> one. The principle was originally: "Mater semper certa est, pater semper
> incertus est". However, this eventually gave way to a more pragmatic principle
> of 'presumption'. "Pater is est quem nuptiae demonstrant" does not mean
> that the father "is whomsever (the mother's) marriage indicates", but rather
> the father "is presumed to be....", in the absence of evidence to the
> contrary. And paternity cases did crop up, and evidence against the presumption
> submitted.
>

> ....


>
> While on the subject of primary sources, it may be interesting to the
> members of the list, that in the research for "Yorkshire Parish Registers", it
> was found that in the case of the East Riding, around 30% of all parish
> registers were found to be 'secondary', i.e. a copy, and it was the Bishops'
> Transcripts (presumed to be the copy) that were in fact the primary record.
>
> Another interesting finding thrown up by the research for "Yorkshire
> Probate" was that in a sample of 100 of the wills in the East Riding Registry of
> Deeds 35 had not been proven at all.>>

Primary evidence is not considered more valid than secondary evidence. In
some cases it's the secondary evidence which tells us the "Truth" if you
will, by taking several pieces of primary evidence, and merging them into a
more realistic story of what occurred. This is the case when the primary
sources conflict with one another. Of course, if the primary sources are agree,
and come from otherwise opposing camps of view, then we can be fairly sure,
that reading them will tell us fairly accurately what occurred. There's
always the situation, that they've all been tainted, but we wouldn't be able to
know that.

A copy does not make a primary source into a secondary source. It still
remains a primary source. In fact I would submit that a high percentage, if
not all, of the sources we use from before say 1300 are in fact copies. I
might say that the sources we use from before 800 are all copies. They are
still primary sources.

A primary source is rather, a source which gives an eye-witness account, or
a "first written" account of an event. I saw it, my mother saw it, someone
told me directly who saw it, etc. In addition, the term is used for
sources which have copied otherwise no longer existant sources. We presume in
some cases that certain writers writing about events just shortly before their
own birth, had interviewees who told them directly this information.

If you copy such a source, you are merely refreshing it with ink, you are
not changing it into another kind of source.

A secondary source, changes the underlying primary source(s) by introducing
new concepts and ideas, merging multiple sources, and so on. A secondary
source is not of the same mind as a primary source. You can usually tell
immediately that it's not written as if it were an eye-witness account, but
rather as an analysis and synthesis of other existing written sources.

J Cook

unread,
27 Sept 2010, 11:11:3727/09/2010
to
On Sep 27, 5:35 am, "Colin B. Withers" <Colin.With...@eumetsat.int>
wrote:

> If the only person who truly knows the father of a child is the mother, and even in some cases that knowledge is in doubt, the improbability that any man is directly descended from his father, grandfather, great-grandfather, etc rises exponentially with each generation back.

Linearly or polynomically, but certainly it can't be exponentially.

As an aside, why is it "pater is est quem nuptiae demonstrant"? I
would have phrased it "demonstrat" instead...

JC

Colin B. Withers

unread,
27 Sept 2010, 11:25:2027/09/2010
to WJho...@aol.com, pss...@bigpond.com, gen-me...@rootsweb.com
I have always placed copies of original documents in the ‘secondary source’ category, through much painful experience, but I note your very different definition of ‘secondary source’ (as you include non-first hand accounts).

I have always used the following definitions, which although from Washington Uni are essentially the same as I was taught at Uni here in the UK:

http://www.lib.washington.edu/uwill/research101/basic03.htm

Anyone who has read “A Comedy of Errors” concerning the differences between indexes to the original records of BDMs at Superintendent Registrar Offices, and the copies of those indexes at the GRO, that exposes the missing records, wrongly copied records, wrongly indexed records, unindexed records, mistyped indexes, errors in page references, volume references and district names, will know what I mean. Transcription errors and omissions can not only change the source, it can lead to years of barking up the wrong tree.

There are differences between the Pipe Rolls and the copies on the Chancellor’s Rolls, that cannot be explained by error or omission (but I have ideas about those, which stemmed from research into the differences found between parish registers and bishops’ transcripts).

However, if the definitions of primary source and secondary source are something that are elucidated somewhere, that this list conforms to, it will be easier for me to work within those definitions. If their definitions are loose and are more a matter of opinion then I prefer the ones I have used over many years now.

Modern technical evidence, such as DNA test results, do not fit into either ‘primary’ (first hand evidence), nor ‘secondary’ (analysis or interpretation of primary evidence). It will be interesting to see where its ‘weight’ will be placed once its use is universal.

Just as ‘Guilty’ does not mean he really did do it, nor ‘Not Guilty’ means he didn’t, yet these verdicts bring legal validity. They can be a long way from the truth, which all historians, including family historians, should aspire to.

DNA has posthumously absolved people of crimes, has cleared up several mysteries (particularly about Tutankhamen) so I was amazed that Peter suggested that some medieval genealogists might not give a fig about DNA results. All evidence should be considered.

Colin

From: WJho...@aol.com [mailto:WJho...@aol.com]
Sent: Monday, September 27, 2010 4:22 PM
To: Colin B. Withers; pss...@bigpond.com; gen-me...@rootsweb.com
Subject: Re: Should Primary Evidence be Redefined?

Peter A. Kincaid

unread,
27 Sept 2010, 12:47:4827/09/2010
to gen-me...@rootsweb.com
One has to remember that DNA evidence can also
be quite interpretive. If one tests two descendants
of separate children of some person back x number
of years and two different results are obtained, which
one is the true line. You don't know. So a third person
from another child is tested and 2 now match. Do we
know the true line now? No. You are just adding
an interpretation. It is only when more samples are
added from more collateral lines that one gets more
confident in an interpretation.

Thus, IMHO, if DNA evidence is to be classified
with traditional evidence then it is only primary if
if it is first generation testing (i.e. of father-son,
mother-daughter). After that it becomes more like
secondary evidence.

Best wishes!

Peter Kincaid


----- Original Message -----
From: "Colin B. Withers" <Colin....@eumetsat.int>
To: <WJho...@aol.com>; <pss...@bigpond.com>; <gen-me...@rootsweb.com>
Sent: Monday, September 27, 2010 12:25 PM
Subject: RE: Should Primary Evidence be Redefined?

[snip]

WJho...@aol.com

unread,
27 Sept 2010, 13:01:0127/09/2010
to 7kin...@nb.sympatico.ca, gen-me...@rootsweb.com
Interesting.

Assume that in 1550 a married woman, had one male child with her legal
husband, but then three more male children with a paramour, but left no
documentary evidence that that had occurred.

In 2010, twenty seven descendants, some from each of the male lines, are
tested. Twenty three of them match each other sufficiently to conclude they
share male Y-DNA. The other four match each other, but not the 23.

And yet the actual male DNA through the surname of her legal husband, is
the four, not the 23.

Odd to think about isn't it

Ian Goddard

unread,
27 Sept 2010, 13:04:2027/09/2010
to

Maybe the problem is in being too quick to pigeonhole evidence into
primary and secondary.

Any piece of evidence needs to be treated on its individual merits. The
original medieval MS might be documenting a fabricated claim. The C18th
antiquarian book might be quoting an original, true document without
actually saying so. Your 6 DNA subjects might actually be the ones
whose lineage is what they thought, not the remaining 104 - DNA will
only tell you about common ancestry, it doesn't tell you who the common
ancestor was.

--
Ian

The Hotmail address is my spam-bin. Real mail address is iang
at austonley org uk

Neil H

unread,
27 Sept 2010, 13:51:5727/09/2010
to gen-me...@rootsweb.com

>Hi Neil,

>Coming back to the point of this whole thread - Thomas Burdett of
>Rowell, Northamptonshire. The only Thomas Burdett that I can find who
>fits your timeframe was of Arrow, Warwickshire - is this the same
>person?

>Regards,

>John
Thanks John. I know of Thos of Arrow (quite an interesting life, and death!). I can't however find anything that would justify that Thomas being referred to as '_ Burdet of Rowell, Northants' so I doubt they are one and the same. The Northamptonshire Record Office has invited me to search their microfilmed poll tax receipts which I will try and find time for in due course.
Neil

WJho...@aol.com

unread,
27 Sept 2010, 13:55:5127/09/2010
to godd...@hotmail.co.uk, gen-me...@rootsweb.com
In a message dated 9/27/2010 10:05:08 AM Pacific Daylight Time,
godd...@hotmail.co.uk writes:


> Maybe the problem is in being too quick to pigeonhole evidence into
> primary and secondary.
>
> Any piece of evidence needs to be treated on its individual merits. The
> original medieval MS might be documenting a fabricated claim. The C18th
> antiquarian book might be quoting an original, true document without
> actually saying so. >>

Quotes don't make a work primary however. And the documentation of a
fabricated claim doesn't make a work secondary :)

The key question to ask any work would be: "Is this the first time
(chronologically) that such an event's occurance is recorded in any source?"

That might be a good test. It's not fool proof and it's not conclusive,
but it does address the two specific points you raised to show that
documentating a fabricated claim, or quoting an original document, doesn't alone, make
a work primary or secondary, nor does it change a work from secondary to
primary or vice versa.

Peter Stewart

unread,
27 Sept 2010, 17:22:3627/09/2010
to

"J Cook" <joe...@gmail.com> wrote in message
news:a8cdb957-a738-4414...@f6g2000yqa.googlegroups.com...

<snip>

> As an aside, why is it "pater is est quem nuptiae demonstrant"? I
> would have phrased it "demonstrat" instead...

The noun "nuptiae" (literally "nuptials") is plural, so the verb must be
too. In this case the syntax might be clearer if "nuptiae" is construed to
mean "the bonds of marriage" rather than just "marriage" as in my
translation.

Peter Stewart

Peter Stewart

unread,
27 Sept 2010, 17:36:4427/09/2010
to

"Colin B. Withers" <Colin....@eumetsat.int> wrote in message
news:mailman.376.12855942...@rootsweb.com...

> No, not at all. My only thoughts on this were this: if primary evidence is
> considered more valid than secondary evidence, where should DNA evidence
> be placed? Above primary? Below primary? Not considered at all?
>
> The history of "pater is est quem nuptiae demonstrant" is an interesting
> one. The principle was originally: "Mater semper certa est, pater semper
> incertus est". However, this eventually gave way to a more pragmatic
> principle of 'presumption'. "Pater is est quem nuptiae demonstrant" does
> not mean that the father "is whomsever (the mother's) marriage indicates",
> but rather the father "is presumed to be....", in the absence of evidence
> to the contrary. And paternity cases did crop up, and evidence against the
> presumption submitted.

This is highly debatable. For starters, the meaning of "est" is clearly just
"is" and not "is presumed to be". Secondly, there have been many recent
court cases and appeals where the mother's husband was found liable to pay
child support when DNA evidence showed he was not the biological father - in
other words, legally the "presumption" was held to trump the science.

In the case of a peerage with entail to the heirs "of the body", the law has
reached the same counter-intuitive decision - for instance, in the Russell
baby case where the mother was found to be technically a virgin, it was
accepted that her marriage had not been consummated and that she had spent
nights in bed with men other than her husband, yet the boy was nonetheless
held to be the heir to his father's peerage. There is a canonical sapect to
such outcomes: man and wife are "one flesh", so that her body is in a sense
also his.

The onus on a father attempting to disown a child is to prove that he could
not have been the father. DNA evidence doesn't necessarily meet this
standard - certainly not as well as absence from each other, as for example
when a crusader returned after five years away to find that his wife had a
child of four. There have been successful suits, but also countless
unsuccessful ones. In the medieval era where a cuckolded husband might have
been a law unto himself, there are instances where the father was generally
thought to be wrong in disinheriting a son from suspicion or hatred of the
mother.

Primary evidence is not always correct, any more than the personal
impressions or presumptions it records, but scientific evidence is a
distinct class of information.

Peter Stewart

Peter Stewart

unread,
27 Sept 2010, 17:39:2127/09/2010
to

"Peter Stewart" <pss...@bigpond.com> wrote in message
news:i7r2po$a2g$1...@news.eternal-september.org...

<snip>

> The onus on a father attempting to disown a child is to prove that he
> could not have been the father. DNA evidence doesn't necessarily meet this
> standard - certainly not as well as absence from each other, as for
> example when a crusader returned after five years away to find that his
> wife had a child of four.

I meant to write "a child of less than four".

Peter Stewart

Cherryexile

unread,
27 Sept 2010, 18:25:0527/09/2010
to
I think it may be worth remembering that in the Medieval Period we are
often interpreting genealogical constructs (or even potential
genalogical constructs) from what is actually primary evidence of
something else. A legal case about land transfer is most concerned
about who obtains title. In an adversarial legal system the best
argument wins and only the genealogical sequence that supports a
particular argument is likely to be presented. It does not have to be
entirely accurate, because that is not its purpose. Even where the
evidence presented by one side is accurate the result of the case does
not necessarily reflect it, as other factors (e.g. bias, favour and
threat) could determine the outcome.

So it would make sense to me that primary evidence has to be
evaluated. However, if primary evidence is evaluated, does the act of
evalutation render the result secondary evidence? If so is it of less,
or more, importance than the primary evidence? Who weighs the
competing arguments of what was stated in the primary evidence and
what was interpreted from it? DNA would seem to be just another facet
of this conundrum. What was the test designed to try and show; how are
we using it and what can we interpret from it? Weight and breadth of
evidence of all kinds has to be considered in trying to reach a
definitive conclusion, but in a lot of cases, if not most, we are
unlikely to ever know a real truth.

In the end a balanced debate amongst peers, is (perhaps) as good a
solution as any.

Message has been deleted

WJho...@aol.com

unread,
27 Sept 2010, 18:45:0727/09/2010
to co...@btconnect.com, gen-me...@rootsweb.com
In a message dated 9/27/2010 3:30:19 PM Pacific Daylight Time,
co...@btconnect.com writes:


> So it would make sense to me that primary evidence has to be
> evaluated. However, if primary evidence is evaluated, does the act of
> evalutation render the result secondary evidence? If so is it of less,
> or more, importance than the primary evidence?>>

> -------------

Yes, the act, for example here, of presenting primary evidence alongwith
arguments and debate makes that secondary evidence.
If you published a book of CPR extracts with your notes on them, it would
be a secondary source. If you merely retyped the CPR, it would be a primary
source, merely "reset" in a new typeface or presentation if you will.

-----------------

>
> Who weighs the competing arguments of what was stated in the primary
> evidence and
> what was interpreted from it? >>

------------------------
The reader.
-------------------

>
> In the end a balanced debate amongst peers, is (perhaps) as good a
> solution as any. >>

-----------------------
Exactly. With the keyword on "balance" (he says smirkily)

Will

WJho...@aol.com

unread,
27 Sept 2010, 18:48:4027/09/2010
to pss...@bigpond.com, gen-me...@rootsweb.com
In a message dated 9/27/2010 3:40:08 PM Pacific Daylight Time,
pss...@bigpond.com writes:


> If "non-paternity" events are as common as supposed from DNA analysis,
> how
> does science discern a case where the husband's own brother was the
> biological father? Adultery, like charity, can begin at home. >>

-------------------

If the goal is tracing a line upwards, then the insertion of a biological
full-brother or even a paternal half-brother wouldn't make a difference to
the rest of the tree.

However the insertion of a paternal grandfather instead of a paternal
father would. That is, if the wife / mother had a child by her husband's
*father* then the entire part of the tree associated with her husband's *mother*
disappears from the tree. Y-DNA testing will not notice this alteration.
That is, it will not tell you that it occurred.


Ian Goddard

unread,
27 Sept 2010, 19:16:2327/09/2010
to

Colin's original question, as I understood it, was attempting to fit DNA
onto an axis currently defined by primary and secondary. The point I
was trying to make was that evaluating a piece of evidence is a more
complex issue than simply deciding whether it's primary or secondary.
The main question one has to ask is how much confidence one has in the
evidence.

In Colin's DNA example, for instance all it actually says is that 104
out of 110 people had a common ancestor and that the other 6 also had a
common ancestor. As you've said in another post, numbers alone aren't a
sufficient guide to picking out which set is correct in terms of the
postulated lineage. Eventually, if things work out for them, his group
may come up with an interpretation which is consistent with both
documentary & DNA evidence.

It's consistency which gives one confidence.

Christine Czarnecki

unread,
27 Sept 2010, 19:24:4727/09/2010
to WJho...@aol.com, pss...@bigpond.com, gen-me...@rootsweb.com
Regarding DNA testing, we need to be clear about which kind.  Clearly, doing a
Y-chromosome analysis does not give the same result as a full genome analysis. 
If you were trying to determine whether a particular man was the father, or if
it was one of his close male relatives, if you can analyze all his DNA, you will
see which is the case. 


The Hemings family is waiting to obtain some sample of Thomas Jefferson's DNA in
order to have his all his chromosomes analyzed.  They already know that some
male members carry the Jefferson Y DNA, but many in the family of Jefferson's
white legitimate descendants claim that Thomas' brother fathered Sally Hemings'
children.  When a sample of Jefferson's full genome can be analyzed, they (and
we) can know if it was indeed Thomas who fathered the male Hemings ancestor. 


The Monticello Association has refused the request to exhume TJ, and so hopes
are settling on the possibility of obtaining hair from one of his brushes, or
some other source.

 

________________________________
From: "WJho...@aol.com" <WJho...@aol.com>
To: pss...@bigpond.com; gen-me...@rootsweb.com
Sent: Mon, September 27, 2010 3:48:40 PM
Subject: Re: Should Primary Evidence be Redefined?

-------------------

-------------------------------

Ian Goddard

unread,
27 Sept 2010, 19:43:2327/09/2010
to
Cherryexile wrote:
> I think it may be worth remembering that in the Medieval Period we are
> often interpreting genealogical constructs (or even potential
> genalogical constructs) from what is actually primary evidence of
> something else. A legal case about land transfer is most concerned
> about who obtains title. In an adversarial legal system the best
> argument wins and only the genealogical sequence that supports a
> particular argument is likely to be presented. It does not have to be
> entirely accurate, because that is not its purpose. Even where the
> evidence presented by one side is accurate the result of the case does
> not necessarily reflect it, as other factors (e.g. bias, favour and
> threat) could determine the outcome.
>
> So it would make sense to me that primary evidence has to be
> evaluated. However, if primary evidence is evaluated, does the act of
> evalutation render the result secondary evidence? If so is it of less,
> or more, importance than the primary evidence?

That's a good point as the legal decision is the result of just such an
evaluation - and not necessarily a very rigorous one. So if one
concentrates on what might be called the legal genealogy then that
particular item of secondary evidence is all one is concerned with,
however defective it might be, and the remainder, be it DNA or primary
documentary evidence, might as well not exist.

Matt Tompkins

unread,
27 Sept 2010, 19:47:4627/09/2010
to

> "Colin B. Withers" <Colin.With...@eumetsat.int> wrote
> > The history of "pater is est quem nuptiae demonstrant" is an interesting
> > one. The principle was originally: "Mater semper certa est, pater semper
> > incertus est". However, this eventually gave way to a more pragmatic
> > principle of 'presumption'. "Pater is est quem nuptiae demonstrant" does
> > not mean that the father "is whomsever (the mother's) marriage indicates",
> > but rather the father "is presumed to be....", in the absence of evidence
> > to the contrary. And paternity cases did crop up, and evidence against the
> > presumption submitted.
>

On Sep 27, 10:36 pm, "Peter Stewart" <pss...@bigpond.com> wrote:
> This is highly debatable. For starters, the meaning of "est" is clearly just
> "is" and not "is presumed to be". Secondly, there have been many recent
> court cases and appeals where the mother's husband was found liable to pay
> child support when DNA evidence showed he was not the biological father - in
> other words, legally the "presumption" was held to trump the science.
>
> In the case of a peerage with entail to the heirs "of the body", the law has
> reached the same counter-intuitive decision - for instance, in the Russell
> baby case where the mother was found to be technically a virgin, it was
> accepted that her marriage had not been consummated and that she had spent
> nights in bed with men other than her husband, yet the boy was nonetheless
> held to be the heir to his father's peerage. There is a canonical sapect to
> such outcomes: man and wife are "one flesh", so that her body is in a sense
> also his.
>
> The onus on a father attempting to disown a child is to prove that he could
> not have been the father. DNA evidence doesn't necessarily meet this
> standard - certainly not as well as absence from each other, as for example
> when a crusader returned after five years away to find that his wife had a
> child of four. There have been successful suits, but also countless
> unsuccessful ones. In the medieval era where a cuckolded husband might have
> been a law unto himself, there are instances where the father was generally
> thought to be wrong in disinheriting a son from suspicion or hatred of the
> mother.


There is nothing debateable about it - in English law 'pater est quem
nuptiae demonstrant' (or its vernacular equivalent 'whoever bulls the
cow, the calf is yours') has always been only a rebuttable presumption
and not an absolute unqualified rule of law. Authorities as far back
as Bracton and Britton have described the grounds for rebuttal. There
were even technical legal terms to describe adulterine bastardy -
'special bastardy' - and the grounds for rebuttal of the presumption -
'special matter'.

However the grounds available for rebuttal have varied over time and,
until recently, were generally limited - but the bare minimum has
always been the following three grounds: the husband's impotence,
separation by sentence of divorce, or his absence overseas (the
crusader's child hypothesized by Peter Stewart would almost certainly
have been declared a bastard). The rules relating to them were
complex, and their detailed application changed over time - for a
numbingly complete discussion of their development up to 1836, google
'A treatise on the law of adulterine bastardy: with a report of the
Banbury case' (for a summary see pp. 249 et seq.).

The present law in England and Wales is quite different. I believe
the current law is contained in s. 26 of the Family Law Reform Act
1969, which states that ‘any presumption of law as to the legitimacy
or illegitimacy of any person may in any civil proceedings be rebutted
by evidence which shows that it is more probable than not that that
person is legitimate or illegitimate, as the case may be, and it shall
not be necessary to prove that fact beyond reasonable doubt in order
to rebut the presumption’. DNA evidence has been admitted, and unless
special rules apply to the succession to the throne, Prince Charles'
paternity of Prince William could certainly be rebutted by it.

Matt Tompkins

WJho...@aol.com

unread,
27 Sept 2010, 19:55:1127/09/2010
to godd...@hotmail.co.uk, gen-me...@rootsweb.com
In a message dated 9/27/2010 4:45:13 PM Pacific Daylight Time,
godd...@hotmail.co.uk writes:


> So if one
> concentrates on what might be called the legal genealogy then that
> particular item of secondary evidence is all one is concerned with,
> however defective it might be, and the remainder, be it DNA or primary
> documentary evidence, might as well not exist. >>

I think you go too far.
Secondary sources tend to quote or just cite, only that portion in which
they are interested. Ignoring perhaps another segment right before or after
that may add some kind of color to what's been quoted.

They are both important. The primary source is what the eye-witness
actually states, the secondary source can add "Well yes but... or and ...."

The trap for the non-expert is using *only* primary sources without the
critical discussion as well, which points out this or that flaw, or adds some
useful detail.

Peter Stewart

unread,
27 Sept 2010, 22:02:0827/09/2010
to

Clearly we are on different wavelengths - my point was that there have
been recent court cases and appeals where the presumption of
biological paternity has been overturned by DNA evidence and yet the
legal paternity has nonetheless been upheld (on the bais of "pater is
est...") in its cultural and ethical consequences by an order to pay
child support. If that issue is not debatable in the present context,
I don't know what is.

As for some medieval genealogists not caring a fig for DNA evidence, I
can't see what is surprising. I don't know if William the Conqueror
was really the biological bastard of his reputed father of not - the
far more important fact is that he was acknowledged as such. A test to
prove this (though of course it is impossible now) would be merely a
curiosity.

If Prince William could be shown to have DNA incompatible with the
paternity of Prince Charles this would in itself make no difference to
the operation of the Act of Succession, since he would remain legally
the son of the Prince of Wales unless the latter brought an action to
disown him. The only alternative to this course would be a new act of
parliament bastardising Prince William, which would only take effect
if his grandmother the Queen assented to it. A challenge to his
succession rights by Prince Henry would probably not even be heard in
England unless Prince Charles joined in the action disavowing
paternity. And apart from succeeding eventually under the Act, Prince
William has been in a sense "enroyalled in blood" by being accepted
from birth as a Royal Highness and grandson of the sovereign. This
could not be set aside by public opinion. It is a cultural and ethical
matter as well as a biological one.

And to go back to Colin's original point, there are other classses of
evidence apart from primary and secondary - what about circumstantial
evidence, that can justify a secondary authority in contradicting an
eye-witness account just as comprehensively as scientific evidence
might do from a DNA analysis? In medieval scholarship, primary sources
are usually taken to be the extant documentary record/s closest to an
event or person AND certain others derived from these: the problem is
where to draw the line, in time and place, with information that has
been copied or otherwise transmitted. Analysis and spinning can start
at any point in the continuum.

Peter Stewart

Richard Carruthers a.k.a. Carruthers-Zurowski

unread,
27 Sept 2010, 22:26:0927/09/2010
to Peter Stewart, gen-me...@rootsweb.com

> "J Cook" wrote in message
> news:a8cdb957-a738-4414...@f6g2000yqa.googlegroups.com...
>


>> As an aside, why is it "pater is est quem nuptiae demonstrant"? I
>> would have phrased it "demonstrat" instead...
>
> The noun "nuptiae" (literally "nuptials") is plural, so the verb must be
> too. In this case the syntax might be clearer if "nuptiae" is construed to
> mean "the bonds of marriage" rather than just "marriage" as in my
> translation.
>
> Peter Stewart

"The father is [he] whom the nuptials indicate/show/state [to be so]."

Richard H.B. Carruthers(-Zurowski), B.A. (Hons.), Modern History, M.A., Oxon.



Peter Stewart

unread,
27 Sept 2010, 23:06:0127/09/2010
to
On Sep 28, 12:26 pm, "Richard Carruthers a.k.a. Carruthers-Zurowski"

<leliw...@hotmail.com> wrote:
> > "J Cook"  wrote in message
> >news:a8cdb957-a738-4414...@f6g2000yqa.googlegroups.com...
>
> >> As an aside, why is it "pater is est quem nuptiae demonstrant"? I
> >> would have phrased it "demonstrat" instead...
>
> > The noun "nuptiae" (literally "nuptials") is plural, so the verb must be
> > too. In this case the syntax might be clearer if "nuptiae" is construed to
> > mean "the bonds of marriage" rather than just "marriage" as in my
> > translation.
>
> > Peter Stewart
>
> "The father is [he] whom the nuptials indicate/show/state [to be so]."

Not quite - "nuptials" would usually mean a marriage celebration,
whereas here "nuptiae" is used to mean the condition of matrimony.

And there is no need for "he" to be in brackets - "is" means "that
man".

Hence "The father is that man whom the bonds of matrimony indicate".
Adding "to be so" makes for a somewhat circular argument with "est".

Peter Stewart

Matt Tompkins

unread,
28 Sept 2010, 10:20:3928/09/2010
to

On Sep 28, 3:02 am, Peter Stewart <p_m_stew...@msn.com> wrote:
> Clearly we are on different wavelengths - my point was that there have
> been recent court cases and appeals where the presumption of
> biological paternity has been overturned by DNA evidence and yet the
> legal paternity has nonetheless been upheld (on the bais of "pater is
> est...") in its cultural and ethical consequences by an order to pay
> child support. If that issue is not debatable in the present context,
> I don't know what is.
>

<snip>


> If Prince William could be shown to have DNA incompatible with the
> paternity of Prince Charles this would in itself make no difference to
> the operation of the Act of Succession, since he would remain legally
> the son of the Prince of Wales unless the latter brought an action to
> disown him. The only alternative to this course would be a new act of
> parliament bastardising Prince William, which would only take effect
> if his grandmother the Queen assented to it. A challenge to his
> succession rights by Prince Henry would probably not even be heard in
> England unless Prince Charles joined in the action disavowing
> paternity. And apart from succeeding eventually under the Act, Prince
> William has been in a sense "enroyalled in blood" by being accepted
> from birth as a Royal Highness and grandson of the sovereign. This
> could not be set aside by public opinion. It is a cultural and ethical
> matter as well as a biological one.


You're shifting your ground - your original assertions were that the
'pater est ...' principle is an undefeatable rule of law, not a
rebuttable presumption ("if Prince William should be discovered to


have DNA incompatible with the paternity of the Prince of Wales, his

succession to the throne would not be affected at all. ... since


Prince Charles was married to his mother at the time of his birth,

that is _legally_ that"; plus "the meaning of 'est' is clearly just
'is' and not 'is presumed to be' " etc).

The recent cases and appeals you mention are unlikely (I can't check
because you haven't named them) to have held that a non-biological
father who had rebutted the 'pater est' presumption of paternity was
nevertheless liable for child support "on the basis of 'pater is
est...' " - rather the order for child support would have been made
on the basis of some other principle or statutory provision. The
legal presumption relates to paternity only - it does not determine
other consequential matters.

As to the Succession, you're mixing up the law relating to legitimacy
with the procedural routes by which it might be enforced and the
social attitudes which might cause the law to be changed before it can
be enforced. The law is straightforward - only a legitimate child can
succeed to the throne, and the 'pater est' presumption, once rebutted,
is irrelevant and cannot preserve an illegitimate child's rights.

Incidentally, the 'pater est' presumption, now only a pale shadow of
the more draconian medieval/early modern rule, may be about to
disappear from English law entirely. The following is a comment by
Lady Butler-Sloss, then President of the High Court's Family Division,
in the judgement in a 2002 Court of Appeal case, H & A (Children):

'The judge made it plain that in the absence of scientific evidence
then the issue was to be decided on the application of ‘a very
important, well established principle … that is, the presumption of
the legitimacy of children born during the currency of the marriage’.
He went on to refer to the case of Serio v Serio [1983] FLR 756 .
Twenty years on I question the relevance of the presumption or the
justification for its application. In the nineteenth century, when
science had nothing to offer and illegitimacy was a social stigma as
well as a depriver of rights, the presumption was a necessary tool,
the use of which required no justification. That common law
presumption, only rebuttable by proof beyond reasonable doubt, was
modified by section 26 of the Family Law Reform Act 1969 by enabling
the presumption to be rebutted on the balance of probabilities. But as
science has hastened on and as more and more children are born out of
marriage it seems to me that the paternity of any child is to be
established by science and not by legal presumption or inference. Were
the judge's order to stand in the present case the consequence would
be a long and acrimonious trial of the paternity issue when, in the
absence of the only decisive evidence, each side would resort to
evidence of marginal or doubtful worth in the determination to
prevail. Such a development would be wasteful of both legal costs and
judicial time.'

Matt Tompkins

Ian Goddard

unread,
28 Sept 2010, 11:41:4128/09/2010
to

And I think you quote too much out of context.

You have chopped both the first sentence which referred back to part of
Cherryexile's post which I quoted and the quote itself.

Peter Stewart

unread,
28 Sept 2010, 18:10:2528/09/2010
to

"Matt Tompkins" <ml...@le.ac.uk> wrote in message
news:ba1c6e04-a06c-4924...@l20g2000yqm.googlegroups.com...

> You're shifting your ground - your original assertions were that the
> 'pater est ...' principle is an undefeatable rule of law, not a
> rebuttable presumption ("if Prince William should be discovered to
> have DNA incompatible with the paternity of the Prince of Wales, his
> succession to the throne would not be affected at all. ... since
> Prince Charles was married to his mother at the time of his birth,
> that is _legally_ that"; plus "the meaning of 'est' is clearly just
> 'is' and not 'is presumed to be' " etc).

No shift of ground at all - for one thing, you omitted "For instance" at the
start of my sentence about Prince William, and worse the context "biology is
the not same as legality" in the prior sentence. More egregiously, you
shifted the context to one post from another: the point I called "highly
debatable" was specifically this from Colin:

"Colin B. Withers" <Colin....@eumetsat.int> wrote in message
news:mailman.376.12855942...@rootsweb.com...

> "Pater is est quem nuptiae demonstrant" does not mean that the father
> "is whomsever (the mother's) marriage indicates", but rather the father
> "is presumed to be....", in the absence of evidence to the contrary.
> And paternity cases did crop up, and evidence against the
> presumption submitted.

My point in response was that the legal presumption is extremely strong and
not based on biology alone. It can be - and has been - upheld in the face of
DNA evidence.

> The recent cases and appeals you mention are unlikely (I can't check
> because you haven't named them) to have held that a non-biological
> father who had rebutted the 'pater est' presumption of paternity was
> nevertheless liable for child support "on the basis of 'pater is
> est...' " - rather the order for child support would have been made
> on the basis of some other principle or statutory provision. The
> legal presumption relates to paternity only - it does not determine
> other consequential matters.

And my point is that paternity has cultural and ethical aspects as well as
the biological. The basis for ordering a man to pay child support is that he
is held to be responsible as the mother's husband at the time of the birth -
if his biological paternity had been impossible, it would (normally) have
been open to him to bring an action to establish this during the pregnancy.
Having neglected or failed to do so, and on the basis that his biological
non-paternity was simply unknown until subsequent testing, DNA evidence
alone does not have the probative force to overturn the "Pater est..."
principle.

> As to the Succession, you're mixing up the law relating to legitimacy
> with the procedural routes by which it might be enforced and the
> social attitudes which might cause the law to be changed before it can
> be enforced. The law is straightforward - only a legitimate child can
> succeed to the throne, and the 'pater est' presumption, once rebutted,
> is irrelevant and cannot preserve an illegitimate child's rights.

My point is that illegitimacy is not that simple to establish at law, at
least beyond the biological level that is not all-determinative. The Act of
Succession was enacted when DNA evidence could not be foreseen, but there
were plenty of cases where the phenotype of a child made the mother's
adultery and her husband's non-paternity obvious to everyone: however, this
was not enough for other putative heirs to win a legal action, just as blood
tests were not before the refinements of genotype analysis became available.
The law has not yet accounted fully for this, and is not likely to do so in
a hurry due to reservations about the absolute nature of the proof and some
known miscarriages of justice from scientists' mistakes.

> Incidentally, the 'pater est' presumption, now only a pale shadow of
> the more draconian medieval/early modern rule, may be about to
> disappear from English law entirely. The following is a comment by
> Lady Butler-Sloss, then President of the High Court's Family Division,
> in the judgement in a 2002 Court of Appeal case, H & A (Children):

The legal principle that held when the Act of Succession was enacted is not
capable of being eliminated from any issue that may arise under the
legislation. If (or when) that Act is repealed, and (assuming the monarchy
survives) replaced by one explicitly acknowledging DNA evidence, then the
"shadow" may disappear - but not yet.

Peter Stewart

Message has been deleted

Peter Stewart

unread,
29 Sept 2010, 00:43:1929/09/2010
to
On Sep 29, 8:10 am, "Peter Stewart" <pss...@bigpond.com> wrote:
> "Matt Tompkins" <ml...@le.ac.uk> wrote in message
>
> news:ba1c6e04-a06c-4924...@l20g2000yqm.googlegroups.com...
>
> > You're shifting your ground - your original assertions were that the
> > 'pater est ...' principle is an undefeatable rule of law, not a
> > rebuttable presumption ("if Prince William should be discovered to
> > have DNA incompatible with the paternity of the Prince of Wales, his
> > succession to the throne would not be affected at all. ... since
> > Prince Charles was married to his mother at the time of his birth,
> > that is _legally_ that"; plus "the meaning of 'est' is clearly just
> > 'is' and not 'is presumed to be' " etc).
>
> No shift of ground at all - for one thing, you omitted "For instance" at the
> start of my sentence about Prince William, and worse the context "biology is
> the not same as legality" in the prior sentence. More egregiously, you
> shifted the context to one post from another: the point I called "highly
> debatable" was specifically this from Colin:

On rereading this thread I can see what Matt Tomkins was getting at
when he wrote that I had shifted ground - I didn't explain myself well
enough.

By "that is _legally_ that", I meant as to the discovery of
incompatible DNA: the exposure of non-paternity would not in itself
affect the succession to the throne, i.e. there is no existing
provision of law that would bring about a change of its own accord. I
did not mean to imlpy that nothing could be done to rebut the law's
presumption in these circumstances - hence I later outlined two
processes that could have that effect as well as a third that might.
But these processes would have to be instigated by the parties
directly concerned, or by the parliament, otherwise there need be no
official reaction at all.

The example is an outlandish one, of course. But given that there are
busybodies able to collect DNA samples very readily, it could
conceivably happen to someone at some time. I expect that the law, as
it is at present, would come down on the side of "Pater is est..."
because of the incalculable risk of deciding differently. The forensic
science is not absolute for one thing, and samples can be mishandled
before it even gets to work for another.

Peter Stewart

0 new messages