TMG-L Archives

Archiver > TMG > 2002-03 > 1015250804


From: "Paul J. Harris" <>
Subject: RE: [TMG] Generic vs specific sources (was: Newbie questions...)
Date: Mon, 4 Mar 2002 09:06:44 -0500
In-Reply-To: <5.1.0.14.0.20020302080743.00a03340@pop.sprynet.com>


-----Original Message-----
From: Darrell A. Martin [mailto:]
Sent: Saturday, March 02, 2002 9:17 AM

>>There is one thing that I do in TMG that causes me to choose to create
>>sources in TMG that are as specific as possible. That is, I cite each and
>>every specific source for every tag that it might relate to, even
>>indirectly. A Census household record often relates to birth dates for
each
>>member of the household, for one simple example. I am able to use reports
>>as described in a separate thread to check all the places where a specific
>>source is cited, and make sure I didn't inadvertently leave something out.
>>With generic sources that ability is severely limited.

>>My conclusion is that each birth certificate, each census household
record,
>>each separate letter from a correspondent, is a TMG source. It helps me,
in
>>the way *I* work. Those who work differently will have different
>>priorities, and may come to a different conclusion.

Darrell, et. al.,

I tend to agree with you, with some additional reasoning. First of all, if
one used generic sources, they could generate the report you refer to above
by focusing on "Source Number = [?] AND Citation Detail contains [?]" when
they generate the report. I don't see this as particularly limiting, unless
they have not been consistent in their CDs (see below).

Another consideration in the specific vs. generic saga, and one that is very
important to me, is consistency. If I have specific sources, i.e., a
separate source for each death certificate, then I have allowed for elements
of each specific piece of information in the citation. Therefore, I have a
NUMBER field, or CERTIFICATE NUMBER field in which I place the certificate
number for that death certificate. If my template contains, "...; Cert. no.
[NUMBER]; ...", then every citation to this source will be consistent with
that exact text. If I use generic sources, then more detail information is
shifted from the source to the CD, which has no structure for consistency.
My money has it that anyone with generic sources would have some with "Cert.
no. 11457," some with "cert. # 11457," some with "Certificate #11457," etc.!
Okay, many people could maintain better consistency than I could over the
course of many years of data entry. <g>

The other big issue for me in source management is flexibility. I recently
discovered a spelling error in a word used in every citation of a census
record. Much to my joy, the error was in my template, and not in every
citation detail that had been F3'd into my database. A one time edit of the
template fixed the problem everywhere. If it had been in the citation
details-------well, perhaps John Cardinals utility would have helped. Also,
consider the issue of citation form. Perhaps I am different from everyone
else using TMG, but the first time I cite a source type, I don't always "get
it right." I didn't know it all when I started, and I still don't, today.
Additionally, ten years ago, all we had was Lackey, now we have Mills,
tomorrow? My point is, if I have the details of my source citation broken
down into element fields, I can change the output format by simply editing
the template, or better yet, applying a different template. None of this
flexibility exists if the majority of my information is in the citation
detail.

So, for me, personally, I do not even use the [CD] element in most of my
specific sources, such as birth, death, and marriage certificates. It gets
page numbers when citing a book, and name and line numbers on census
records. But my general philosophy has become, "put as much in the source
and as little in the CD as possible." After all, I am entering this into a
genealogical database, not a word processor. <g>

Paul J. Harris

This thread: