TMG-L ArchivesArchiver > TMG > 2006-12 > 1166111177
From: Lee Hoffman <>
Subject: Re: [TMG] Source of a source of a source of a source ... HELP!
Date: Thu, 14 Dec 2006 10:46:17 -0500
Linda Kuhn wrote:
>I do prefer the Mills models. But. The problem I have with the Mills models,
>as translated into TMG, is that *some* tend to force one to be a splitter at
>times when (imho) there's really no need.
I agree. There is also the option of selecting to use the Lackey
Source Category in which the Source Types tend to be more generic and
lend themselves more easily to lumping. On the other hand, the
Mills Source Category contains many Source Types that relate to the
more modern documents and media.
>When we use the Mills model for say "Book (Authored)", we create a single
>source for the book and put the page number in the CD. The source functions
>like a Lump but reads like a Split in the final output.
You are confusing the application of Mills work to what is implement
(using Wholly Genes interpretation) that work into TMG. If you look
at Ms Mills work, you will see that it, like almost all style guides
(including the Chicago Manual of Style (CMOS)), is NOT designed for
use with computer programs. Their design is much the same now as
they have been for many decades (1906 for CMOS). This includes Mills
predecessor - Richard S. Lackey in "Cite Your Sources". That is all
these are how you would design citations when you sit at an
old-fashioned typewriter to type a report. Both Lackey and Mills
works are really no different in general from CMOS. In fact, if you
follow the precepts of CMOS, the end result will be much like you
find in Lackey and Mills. Yes, there are slight differences, but the
overall design is the same. The main thing for Lackey and Mills is
that CMOS does not address certain types of document and media
well. Thus Lackey and Mills give us a standard to go by for the more
>So why, then, does the Mills model discourage a source called (say) "Jones
>County Marriage Records, Book 2" and using that (with appropriate CD)?
>Instead, it asks us to create a separate source for every single MR
>extracted from that book. That seems inconsistent to me.
As you say in a later message, this example is not a good one. The
better example is the e-mail Source Type of even better the Source
Type that would be used for Letters (an older media still in
use). This Source Type is designed to cite a single letter from
person A to person B on a certain Date. But what if you have a file
of letters over a period of 50 years send by person A to person
B? How do you cite them? Well, according to Mills (and to Lackey,
and to CMOS), you would cite each letter separately. Somewhere along
the way, someone (with a typewriter, mind you) started creating
citations that were more generic. But the original model stayed the same.
So instead of the standard model:
Letter from [AUTHOR] ([AUTHOR ADDRESS]) to
[RECIPIENT]<, [DATE]>;< [REPOSITORY REFERENCE]> [
REPOSITORY] ([REPOSITORY ADDRESS])<.
we might make the citation more generic as:
Letter from [AUTHOR] ([AUTHOR ADDRESS]) to
[RECIPIENT]<, [DATE]>;< [REPOSITORY
REFERENCE]> [REPOSITORY] ([REPOSITORY
ADDRESS])<. [COMMENTS]><, [CD]>.
But, you say, there is no difference? And you would be right. The
difference is how we enter the data in the computer. Most entries
remain the same. But, for [DATE], we would enter a range of
dates. This assumes that all other entries remain the same. Now, if
the author had moved during that time, I would create multiple
Sources each with the date being the range of dates covering that
address. Similarly, if some letters were found in a different
Repository, then you would have separate Source(s) for those letters.
Now, if you have a file of these letters along with copies of the
replies and maybe including notes and clippings, etc. all combined in
a single collection then it comes to how to cite the entire
collection -- assuming that you want a single Source (or just a
couple) instead of a Source for each piece of paper in the
collection. The single Source for each piece of paper would tend to
be the direction in which Mills, Lackey, and CMOS would direct you --
splitting in the extreme. On the other hand, you can call that
collection of papers an unpublished manuscript and group them
together. Now we call that lumping here, but we are still able to
split by citing the specific piece of paper in the Citation Detail.
>Now I understand splitting and I understand lumping. But, once you get past
>the point of personal preference in that regard, does the technique really
>matter so long as the end result accurately portrays the source? I don't
>think so, but I may be missing something.
As others have pointed out, a Source is what prints in the
Bibliography (or the "List of Cited Works" as CMOS (and others)
delineates).But here also, you can split or lump. You can have a
Source for each piece of paper or you can have a Source for a collection.
Consider the various census citations. If you follow exactly the
guidelines of Mills (and others) and the standard Source Types in TMG
then you would have many Sources -- one for each line entry in the
census. Now if you have a paper for a single family and that family
has ten children (each born two years apart), you might have for that
one couple and their household something on the order of 80-100
Sources over say 6 to 8 censuses (assuming marriage of children, no
deaths and that each census finds the household). Most of us
would not cite each line entry in each census, but would do lumping
in some way -- probably having six to eight Sources. On the other
hand, this may lead to 80 to 1200 citations of those six to eight
Sources. But the Sources is what would go in the Bibliography.
As for printing or not printing a Bibliography, it has its
purpose. Which would you rather do, copy each citation from a book
or copy just the list of cited works? Of course, if you are only
copying a couple of Sources, the above question is moot. But if you
find a book that covers your family well and you wish to more or less
verify each document and its statements then you want more than the
one or two. In this case, the Bibliography is much easier.
So the design of Mils (et ux) is for individual citations with the
ultimate splitting in mind. Now the Lackey designs (and Wholly Genes
interpretation of them) tended to be more generic making lumping
easier. Similarly, if you compare Mills to some others, you will
find that lumping may be easier using Mills than what other
may. One has to decide which way to go, whether any lumping will be
done and then how much. The, you need to decide which Source Type is
to be used. As noted, if you decide to do lumping this may dictate
the basic Source Type to be used as well as how much modification of
the Source Templates is needed.
Hope this helps -
TMG Tips: <http://www.tmgtips.com>
My website: <http://www.tmgtips.com/lhoffman>
A user of the best genealogy program, The Master Genealogist (TMG)
|Re: [TMG] Source of a source of a source of a source ... HELP! by Lee Hoffman <>|