* Recording a Family's Residence

[AdrianBruce]

I'm sure I've gone round every sort of way possible to record a family's residence details. I thought I'd got a new and decent way of doing it using witnesses, but I now realise it creates more problems than it solves. So.... any suggestions please? (This should be so obvious but it ends up so clunky).

Suppose we have a family with parents who have varying children recorded in several censuses. We also have details of the family's residences during that time from birth certificates and / or baptisms for children born during those years. I don't use Ancestral Sources ( ) and I do write extensive notes about where the house is or might be, etc., for each residence.

1. The most obvious way is to put an separate RESIdence event against each individual for each census, etc. Trouble is, I then need to repeat those descriptive notes about the house against each person. I hate repetition.

2. I then went to separate RESIdence events against each individual for each census, etc., but put the descriptive notes about the house into a Shared Note Record. Fine - no repetition in data entry but massive repetition whenever I ran off a report on the family. The same note appeared against each parent and each child.

3. My current method is to use a custom residence event for the family - to be precise, like any so-called family event, it appears against the parents. The note for the family residence event describes the house - so that's written once only. The data is entered once for the two parents. Then each child gets their own individual residence event, which just has the phrase "They were living with their parents" in the notes. No repetition in entry or printing - just a flaw in that any printed report including a child but not their parents will only say "They were living with their parents" without the description of the house. In fact, this doesn't seem to be an issue but whether that's because I take the whole database on a laptop or always start with the "brick wall" parents in producing reports, I'm not sure.

4. Having moved to v6, I thought I had the answer - use "witnesses" for the co-residents. Like the previous case, I enter a custom residence family event, with descriptive note, against the parents. But also against this event, I enter the children (and others) in the household as "witnesses" (and they don't have their own residence event any more). This seems great on a first look. The individuals' property boxes for the children show the full details, and so does a narrative report - even for individuals.

Utopia? No. I think the problems stack up like this:
- If the parents are in the same place and address in (say) 4 different censuses, I want one residence event covering that date range for them (not everyone would want that). But there is no opportunity to date the witnesses (unless you know different) so I'd have children yet unborn recorded in reports as co-resident - and worse still - they'd appear in their own property box with a witnessed residence before their birth. Or after they'd left home;
- If the residence event includes as witnesses only the children living with their parents at the time of the event, then I need a residence event for the parents for each census and each baptism - that's not what I want (though I appreciate that's what some would do);
- Queries about residences become a pain - I've not yet worked out how I write a query to pull off all people with individual residence facts for place X, family residence facts for X and witness sentences referring back to X as the children won't have their own residence fact (does that make sense?)

Is there any magic combination that I'm missing? I suspect not but who knows???? All comments gratefully accepted....

[tatewise]

It is virtually impossible to use a Query to find Witnesses. See Witness reporting (11976).

I suspect most users put Address & Place details associated with Events such as Census, Birth, Marriage, Death, etc, within the Events themselves, and thus avoid your problem. Any specific details about the residence Address can be entered once into the cited Source record.

Some purists say that it is only such Events that specify residence on a specific date, and any other inference is conjecture.

I presume you have seen how_to:create_locations_database_details|> Create a Locations Database of Place & Address Details and rejected the idea. The concept is to create Source/Repository records for each Address with details held just once about that particular Address. Then any use of that Address in Residence or any other Facts can cite the Source/Repository record.

[AdrianBruce]

If you're saying "virtually impossible", then it really must be tricky! I had glanced at the "Witness Reporting" link before, hence my concern. However, I don't think I'd appreciated how difficult it was. I hate and detest having to hard-code occurrence numbers - it only needs one big farming family with several farm servants and I've got more witnesses than columns in the query and probably don't realise it.... Given that I really need to be able to run queries that ask "Who was resident in Wardle?", then it's possibly the nail in the coffin for witnesses - at least as far as Residence goes. No idea if plug-ins would help but I fear that this is just one example and continually writing plug-ins for other questions on other aspects, instead of queries, sounds too much like hard work.

Re your para "I suspect most users put Address & Place details associated with Events such as Census, Birth, Marriage, Death, etc, within the Events themselves". Well, I do. But the issue is that I'm using the events to generate Residence facts - a baptism, for instance, gives me a residence that I record as its own fact. I appreciate it's in the text of the Source Record but I want to run off reports (narrative or otherwise) that make it clear where people live, not hide it way in source text - which I tend not to print anyway.

And I know that my merging several residence events into one residence event running over a range of years causes some purists pain, but that's where my pragmatic decision ends up. (Seriously, if they want to be purist, as a mathematician I could out-purist them if I tried!)

Yes, your location database idea is interesting but I found it difficult to see how I could produce a report that printed the location sources and repositories without printing them all.

I guess there's a theme running through this - the issue is not entering the data or accessing it from a screen or web-page - rather it's trying to produce (narrative?) reports that give "normal" people all the details they want. I appreciate real writing is infinitely superior to machine produced reprts but I know how long it took me to do just two lines of direct ancestors only.

[tatewise]

Adrian said: your location database idea is interesting but I found it difficult to see how I could produce a report that printed the location sources and repositories without printing them all.

I guess you are talking about Publish > Miscellaneous Reports > Source Summary Reports. Well the KB article says, "A new Source Type, such as Location, is used to uniquely identify these Location Source Records." So when the Select Records window is displayed, click on the Type column and select all the consecutive Location records. This works for any Type of Source Record.

If you stick with a RESIdence fact for each Individual, then there is possibly mileage in the new Sentence Template features that might help.

[AdrianBruce]

Thanks Mike. I think I'm going to revert to my previous Residence fact(s) for each individual. Certainly worth exploring the new features in sentences, though - I've hardly started that exploration yet - the witness idea was perhaps the first attempt in that direction.

[JoopvB]

I am trying a simpler (I hope) approach but am still working on the sentence.

The idea is to use a witness named Family and when using the standard Residence fact add all family people (spouse and children) as family. Then test in the sentence if family is single or plural and have it print "and spouse" or "and family". If needed this could be expanded to 2 witnesses (Family and Spouse) to avoid ambiguity.

However, up until now I haven't figured out how to get the sentence, or to be more precise, how to do the test for the presence of the specific witness and based on that get the associated text.

I'm probably missing something simple but help is very much appreciated.

Regards, Joop

[tatewise]

Joop, did you read and understand this posting right through from the beginning?

It explains why Witnesses are not satisfactory for this type of Fact.
The basic problem is NOT with Sentence construction, but how Facts appear in Reports and Queries.

If each family member does not have their own Residence Fact then you cannot find them with a Query.

When children grow up, at what point do they stop being Residence Witnesses for their parents, and need Residence Facts for their own children to Witness? To me, it is just is too inconsistent.

My view is that Witnesses should only be used for secondary Individuals who witness the Fact, NOT the primary Individuals who experience the Fact.

The FH V6 default Fact Set neatly solves the Sentence construction for the Residence fact with a Witness Role of Resident and saying ... lived with ..., but as I said I am not keen on using Residence Witnesses at all.

[JoopvB]

Hi Mike,

Yes, I did read it and that's why I wrote that I was trying a simpler approach. The difference being:

- I intend to use the "Family" witness to trigger a piece of text: "family" or "spouse" in the Individual's sentence and something accordingly in the one for the Family witness. I use the "witness principle" only to make the sentence also appear in their reports, not to name them as a witness.

- As I see it, the report sentence like "On/From xx He live with his spouse/family at yy" and something like that for the Family witness is the a correct sentence. When the children grow up etc. they will have their own/next residence facts with their own relevant sentences (at/from that moment in time).

- But this approach will only work if I can make text appear depending on a witness being used and I haven't figured that out yet... Is it possible?

So, in general I completely agree with the kind of misuse of the witness principle for residences but as a kind of second best solution...

About the query problem; is that a nasty thing? I mean how often do we need to query on residences?

[tatewise]

But a sentence like "On/From xx He live with his spouse/family at yy" is the problem as soon as the family members are actually mentioned, which is what a sentence usually does, otherwise it does not identify who they are.

As Adrian says, if you add all the Children as Witnesses, and the Date is from xx to yy then the Witness sentences may say they lived with their parents before they were born or after they left home, or in unfortunate circumstances after they died in childhood. This is because you cannot put a date range on the Witnesses. They are all treated as Witnesses for the whole duration of the Fact. So you finish up having to create a separate Residence Fact with a Date range for each combination of Witnesses, even if they are all at the same Address.

A Query might not be specifically on Residence Facts but on any Facts where the Place/Address fits certain criteria. If the only Fact that fits those critera for a Child is their Residence, but they are recorded only as a Witness, then they will be omitted from the Query Result even if their Parents are included.

Regarding testing the number of Witnesses, I think the only possibilities are the Template Codes:
{role(single)=____}
{role(plural)=____}
{other(single)=____}
{other(plural)=____}
They essentially differentiate between a single Witness and multiple/plural Witnesses.

You cannot combine Template Codes with Expressions.
The only other idea is to use =Exists( ) with Data Refs for %FACT._SHAR[1].ROLE% and %FACT._SHAR[2].ROLE%, et seq.

[JoopvB]

I agree with "is the problem as soon as the family members are actually mentioned"; that is why I won't mention them.

I agree also with "if you add all the Children"; that's why I don't add them all, but only the ones that were (at that time) to (be likely) to live at the family house. If I have data/sources to have some of the children to have their own residence fact, I'll remove them from the shared one and give them their own.

I value the residence query of less importance than the easily sharing of a family residence. One must make choices in life.

Indeed I found that the template codes don't combine with the functions and expressions. So I tried to find a way to access the witnesses with the data references, but couldn't find out how to. From your answer I guess it's something like =TextIf(%FACT.SHAR..... and then something. Could you help me out with this test? It should test for the presence of a family witness and print "family" (if plural) and "spouse" (if single).

[tatewise]

The Sentence Expression will go something like:
{=TextIf( Exists( %FACT._SHAR[2]% ), "family", "spouse" )}

As I think has been mentioned elsewhere, to find a Data Ref use the Data Ref Assistant in one of the other windows. In this case a Custom Fact Query would be good, and explore the Witness entry.

Following through your first two paragraphs, it sounds like your Narrative reports may often say:
In 1798 he lived with his spouse at 21 High Street in London.
In 1800 he lived with his family at 21 High Street in London.
In 1802 he lived with his family at 21 High Street in London.
In 1804 he lived with his family at 21 High Street in London.
as each Child is born in 1800, 1802 & 1804 and added as a Witness at the same address.

Another consideration is that Witnesses do NOT export in Gedcom to other genealogy products as reliably as Individual standard RESIdence facts.

[JoopvB]

Hi Mike, I had not been experimenting with the query feature enough. I think I am now getting a feeling of data references and functions; very powerful!

About the shared residences. My memory and the rich feature set of FH6 is playing tricks on me. Years ago I tried to share residence fact in TMG as well. After negative (side) results I abandoned the idea and reverted to non-shared. When I started with FH and saw a new kind of (and more powerful) tools, the old idea of optimizing the residence fact entries must have popped back up in my mind. So I tried it again.

But as you've clearly explained, the drawbacks (in FH6 also and from GECOM perspective) outweigh the benefits. So the TMG history repeats itself and again I'll revert to non-shared residence facts.

You are a experienced and patient teacher; thanks!