TMG-L ArchivesArchiver > TMG > 2006-03 > 1143764348
From: Terry Reigel <>
Subject: RE: [TMG] Census Sentence Formats
Date: Thu, 30 Mar 2006 19:21:20 -0500
On Fri, 31 Mar 2006 05:39:19 +1100, Keith Skinner wrote:
> As I have only been using TMG for a short while, it is
> taking a bit of time to get my head around the sentence
> structures, and the use of roles. I wasn't sure on where
> one changed the sentence structure, but from reading your
> tips completely, I now know it is in the Master Tag Type
> List. I should have read your tips more thoroughly before
> posting my original question.
> I have been looking at your expanded version, as the
> output seems to fit with what I am trying to do.
OK, that helps. <g> I like the expanded version because it lists all
the members of the household in the census listing for the head of
household. As I said before, it deals automatically with your basic
parents and children combinations by just selecting the correct roles.
The method is a bit indirect and the selection of roles may seem
counter-intuitive at a first. You choose a role for the parents that
describes the number of children, and a role for the children that
describes the parents that are present. Once this is setup in the Tag
Type, no changes are required in individual tags - just select the
When there others in the household though it gets a bit more
complicated. People you don't want to enter into your data base, such
as servants and boarders, you can just mention in the Memo.
But for those you have in your database you have to consider what
result you want. You can get the statements for their own narratives
pretty easily - enter them as Witnesses with Roles of "Witness" or
"Other" and they get a statement that they were in the household of
so-and-so. If you want the statement to reflect how they are related
to the heads of household I find you can generally get satisfactory
results by adding text to that effect to the Witness Memo for that
person - something like "his sister and her husband" for example.
It is possible to get more complex if you chose to. Say you have a
John, wife, and their two children in the household of his brother
Pete and wife. If you want you can have in the output for John that he
was in Pete's household along with his wife and children:
- Enter John with Role "Other"
- Enter John's wife with Role "Other2"
- Enter their children with Role "Other3"
Then, in John's Witness memo, put something like:
his brother and sister-in-law. With him were his wife [R:Other2] and
their children [RF:Other3].
The result will be something like:
He was enumerated in the census of... in the household of Pete Smith
and Sally West, his brother and sister-in-law. With him were his wife
Mary White and their children Sam and Joe.
You could get more clever by editing the sentence for John if you
want, but it's not really necessary.
Now... to the narrative for the head of the household. The roles I
define deal automatically with the children of the head(s) of
household, listing them in the narrative as present in the household.
But of course there are all sorts of other folks who show up in
households. Some users try to create predefined roles for all of them.
I think that's a waste time - there are too many kinds of relatives
that appear, in too many combinations, to predefine workable sentences
for all of them.
Instead, I created the non-descript roles of Witness (because it's a
standard Role and you can't get rid of it) and Other, Other2,
Other3... as many as needed. With those roles you can create
narratives for the head of household's narrative that properly list
all the combinations you come across. While I occasionally edit the
sentence in the tag, most of the time I do it all in the memo. For the
example I set up above, the main tag Memo might look like this:
reporting real estate valued at $1000||along with his brother,
[RF:Other], his wife, [R:Other2], and their children [R:Other3]
This would result in a narrative like:
He and Sally West were enumerated in the census of... reporting real
estate valued at $1000. Their children, Bobby, Billie, and Lester,
were listed with as living with them, along with his brother, John,
his wife, Mary White and their children Sam and Joe.
So, you can see, if you choose to, you can include most any grouping
of relatives with this system, with as much detail as you choose to
> Your example applies specifically to a Census carried out
> in the USA. I have several different Census locations -
> Scotland, England, Australia etc. From your example, it
> appears that I will need to create different Tags for
> each set of Census criteria (eg. CenSCT1851 - Scotland
> 1851). Is this correct?
I prefer to have a different tag for each census. That gives me a
clear signal in the tag list about which censuses are listed, and
allows me to customize a bit the sentences for each census. I put the
type of census and year in the Sentence as text, but you could get the
year with the [Y] variable. Also, there are some differences in the
information provided in different censuses - hence the slightly
different formats for pre-1850, 1850-1880, and later. I've worked very
little with census records from other countries, but I'd expect even
greater variation that might benefit from individual tag types for
However, many users prefer to use a single tag type for all years, and
that's certainly a viable choice.
> I would like to thank you also, for the excellent TMG
> Tips site and your input to the User Group. If you had
> not provided that I would be still floundering.
You're welcome - Thanks for your comments. <g>
|RE: [TMG] Census Sentence Formats by Terry Reigel <>|