GEN-MEDIEVAL-L Archives

Archiver > GEN-MEDIEVAL > 1997-08 > 0870841674


From: Stewart Baldwin <>
Subject: Re: Grading Sources
Date: Tue, 5 Aug 1997 23:27:54 -0500
In-Reply-To: <33E21516.29B@worldnet.att.net>


wrote:

>It seems to me that grading sources of genealogical information the same
>way teachers grade students (A,B,C,D,F), might not be that helpful.
>After all, a student who scores less than sixty percent correct on a
>test flunks, but is a book that is only--to be extreme--twenty percent
>correct completely useless, as an F would imply? I don't think so.
>
>Let me give an example. Suppose you are looking for the mother of X and
>have been having no luck whatever. But there is a notoriously unreliable
>book, shot through with errors, that might address the question. So,
>faute de miuex, you try it, and, sure enough, it says that X's mother
>was Sarah P. Thistlethwaite. Now, obviously, you don't just enter her
>into your data base and go merrily on. Instead, you rush to the
>Thistlethwaite genealogy, or whatever, and try to confirm the
>information. With a bit of luck you'll have a eureka moment. And they,
>at least to me, are half the fun of genealogy.

I think I have covered the above point before. Many books can be used in two
ways, as a finding aid (i.e., as a source for clues toward future research),
and as a cited source (i.e., a relationship is claimed with the source given as
a reference). The example mentioned above uses a book as a finding aid. All
but the most pitiful genealogies have at least some value in this regard, but
the usefulness as a finding aid is COMPLETELY IRRELEVANT (apologies for
shouting) when grading a book on the correctness of the relationships given.
In the bext of all possible worlds you go to the original documents for every
fact claimed, but this is not always possible or practical. The grades help
people who are unfamiliar with the source to decide how likely the given source
is to be mistaken on the relationships claimed. (To give an example which has
not been discussed before, the Ancestral File database gets an "F" as a source
for genealogical relationships, because of its completely ridiculous
unreliability. However, it is still a very useful finding aid. Note that, in
this case, the "blame" for the "F" grade does not fall on the people who made
the database, because they are not responsible for the fact that an item which
should never be used as anything but a finding aid is being used by many as if
it were a citable source. On the other hand, the blame for the unreliability
of Burke rests squarely on the shoulders of the authors, as has been discussed
here before.)

>So perhaps a more useful grading system would be 1, 2, 3, 4, and 5 as
>follows:
>
>1) CANONICAL--to be accepted without question (are there any of these?).
>
>2) AUTHORITATIVE--to be accepted unless later, contrary information is
>found.
>
>3) USEFUL--to be accepted but confirmation desireable.
>
>4) SUSPECT--needs confirmation before being accepted.
>
>5) UNCRITICAL--to be used as a last resort and only with solid
>confirmation.

I really don't see any difference, other than the use of numbers instead of
letters, and the use of four grades instead of five. (Of course, #1 above does
not count, since no such source exists.) Grades 2,3,4,5 cover the same ground
as A,B,C,D,F, with perhpas slightly different borderlines because of the
different number of grades.

>Also, in my experience, many family genealogies, especially those big
>ones from the 19th century, are often reliable for people on the
>American side of the Atlantic but as soon as they cross to England they
>become totally uncritical and glom onto every title they can get their
>hands on. The Tracy family, early settlers of Norwich, Connecticut, is a
>case in point. Several 19th and early 20th century books state
>categorically that Lt. Thomas Tracy, the progenitor, was the son of Sir
>Paul Tracy, Baronet, of Toddington, Gloucestershire. But Jacobus
>demolished that theory several decades ago. So these might be rated 2/4,
>with / being the Atlantic Ocean.

Virtually all sources have their strengths and weaknesses. If a well defined
line can be drawn in such a way that one side of the line gets one grade and
the other side gets a different grade, then that would be useful. However,
most books which contain data of mixed reliability cannot be easily divided in
this way, as the unreliable data is mixed with the reliable in a way that is
difficult to disentangle.

One case like this would be sources which give data on events which occurred
within living memory of the publication date. In that case, the data can often
be accepted, even if the source is otherwise bad. For example, I would be
willing to accept a citation from Burke (or even Virkus) for a birth which took
place in the year 1900 (although a birth certificate would obviously be
preferable). On the other hand, for the medieval period, which is what we are
interedsted in here, Burke and Virkus are both extremely unreliable sources.

>So it seems to me that no genealogical source should be dismissed out of
>hand. After a poor lead is still better than no lead. What do you all
>think?
>
>John Steele Gordon

As a candidate for a genealogical source which should be dismissed out of hand,
I suggest "Holy Blood, Holy Grail".

Stewart Baldwin

This thread: