GEN-MEDIEVAL-L Archives

Archiver > GEN-MEDIEVAL > 1996-02 > 0823630683


From: Chris Bennett <>
Subject: Re[4]: Descents from Antiquity (was: Adultery redux)
Date: Tue, 6 Feb 1996 10:18:03 PST


___________________________________________________________________________
Subject: Re: Re[2]: Descents from Antiquity (was: Adultery redux)
Author:
Date: 2/5/96 10:59 PM

<snip>
>
Ah. Well, maybe I should rephrase and try to clarify what I said. . . .
Yes Chris, it is worth knowing how confident we can be in our conclusions.
What I object to is NOT applying probability to reconstructed or
hypothetical pedigrees, but rather to reasonably well-attested lineages
or descents. In any generation there is a 50/50 probability that the man
you think is the father *MIGHT* not be the biological father because of
female adultery--true enough. But if the mother never committed
adultery, there is a 0% probability that the father is someone else (I
leave out, of course, the Virgin Birth!). Now, with the state of the
source material at our disposal in medieval pedigrees, can we PROVE with
100% probability that in any given generation the woman DID NOT pass off
an illegitimate conception as that of her husband's? Of course not. So
why worry about this phantom 'problem' at all? We can't prove she was an
adulteress, and we can't prove she was chaste and faithful. All we have
are documents saying that So-and-so was the son of So-and-so by So-and-so.
To hedge one's conclusions with remarks like, "But MAYBE the wife had
been naughty," is rather like saying, "MAYBE Paul did not really write
the Epistle to the Galatians." Even if those statements are true, there
is no way ever to confirm it, so why bother in the first place? All we
can do is take what we've got and run with it. I hope this is a little
clearer than my previous remarks, Chris.

Jared Olar

OK, thanks for the clarification. Lets see if I can address your point.

The essence of the difficulty is your focus on individual generations.
What I described is a systemic problem, a problem of scale, a problem of
large numbers, and its inherent in the research discipline of genealogy.
That is, its not a problem of medieval (or ancient or even
biblical/mythological) genealogy per se. It is also not one which
significantly affects documentary research on individual generations.
Therefore I agree with you when talking about any individual link --
unless, of course, you have reason to believe that one of your
ancestresses may have played around! Its a problem which says that you
can never be sure your database, CONSIDERED AS A WHOLE, is error-free, and
so potentially limits the maximum useful size of a database for some
purposes.

The reason to address it is that we can estimate a level of confidence on
the accuracy of a database, or at least we can bound our level of
confidence, based on the observed statistics of adultery (such as they
are) and some fairly elementary probability theory. If your aim as a
genealogical researcher, as is the case for many people, is absolute
accuracy as far as the evidence allows, then this allows them to estimate
whether its worth even bothering to investigate beyond a certain time
horizon.

Chris

This thread: