* How many birth sources is too many?

[mnschwarz]

I've got a bit of general use question. I've been adding any source I have that proves an individuals existence to their birth fact and I think I'm probably handling this wrong or unnecessarily creating a lot of work for myself.

One of my relatives for example has his birth record as a source (obviously), but then I've copied the birth, marriage & death sources for each of his 12 children as he is also listed on those documents. My initial thought was that these documents prove his birth even though the event is unrelated to him directly. But with that thought I have found there's no real end to what I can add - ie: every source possible then adds evidence to his birth.

So should I only be adding actual birth records as a source in this example? And would his children's BDM sources be linked to him at all anywhere else since he is listed? Or they would only be specific to that individual?

Interested to hear how you guys manage this. Cheers!

[Gowermick]

I've got a bit of general use question. I've been adding any source I have that proves an individuals existence to their birth fact and I think I'm probably handling this wrong or unnecessarily creating a lot of work for myself.

Firstly, I think that the proof of a person’s existence should be set against their name , not their birth.
Therefore a child’s baptism supports their parents names, not their ages, so I wouldn’t use it as a source for their birth facts.

However, if one considers one should generally be working backwards in time, then one adds evidence to a child that proves their parentage. It is therefore redundant to add the same evidence to the parents about their child’s existence.

[tatewise]

An alternative is to add evidence of a person'e existence to the whole Individual record rather than specific details such as Name or Birth.

As Mick indicated, the Source records for the children also support the parents via the Parent-Child relationship.

Furthermore, you can use Fact Witnesses to associate child Sources with their parents.
e.g. A Baptism event for a child would be supported by a Baptism Source document than names the parents.
The parents could be added as Baptism Fact Witnesses with the Role of Parent and that relationship would be covered by the Baptism Source without any additional citations.
The same method can be used for other child BMD events

[Mark1834]

My advice would be to look at it analytically. Every individual in your database will have a birth fact, even if you have no idea of when or where. IMO, sources should only be cited against the fact they support. Does the birth of a child prove the details of the parents' birth? Probably not.

I am also wary of superfluous citations. To illustrate, if a census record lists John Smith in a particular place with approximate birth details, it would be perfectly reasonable to cite the census to support the census and birth facts. However, also citing it for name is redundant (IMO, of course) as citing the other facts implies directly that I have assumed it is the correct John Smith and does not prove in itself that the name is correct.

I will differ with Mike L. on one detail. Name is just another fact, so sources that simply demonstrate that the person existed should IMO be cited against the whole record, not the name fact.

[AdrianBruce]

As Mark implies, there is no point in circular citations. By that I mean that if I have John Smith, born Ambridge, 1872, and look for him in the 1881 census - and find him - there's no point in citing the 1881 against his name, birth, etc, because the 1881 is not additional "proof" that is his name, birth, etc. It can't be additional proof because that's how I chose that particular census record in the first place - it can't help but have those values. It's circular logic (which is bad) if I think that it's additional proof.

As for how many is enough, it's probably not worth going back to remove anything, but in future, I'd omit the circular citations where they don't add anything.

Another aspect about citations GEDCOM style is that you may need multiple citations to demonstrate the values of the different parts of the name (or whatever). One citation for the birth name, then another for that extra middle name that appeared out of nowhere years later, and a third for the nickname, etc. Probably they are all attached to one Name... Probably...

I also have a slightly different use of citations to a mother's name. Before Civil Registration, our normal method of establishing the mother's name is via the baptisms of her children. But, you don't need to go back very far to find baptisms that omit the mother's name. I will, in that case, put a citation against her name for every event that includes that name. That helps me understand how much evidence there is for her name and when - the classic error is to assume that the mother mentioned on the baptism of the 6th child is the mother of the 1st child (whose baptism doesn't mention a mother). If I don't do that, it's perfectly possible to go back to the original sources, but it can be quicker if explicit citations are there already.

[mnschwarz]

Many thanks guys. The notes about circular citations is particularly helpful. I also hadn't really considered using the 'whole record' sources so will make better use of this going forward. I don't have too many records entered yet as I'm still getting used to the software and how everything ties in together. So not a big issue to clean things up fortunately. Cheers!

[Gowermick]

I take the point that the whole record is a better place to source their existence, rather than their name.

To be honest, I rarely use the whole record, but perhaps I should start making better use of it, never too old to learn

[Jean001]

For each person mentioned in a source I consider each piece of information separately, i.e. existence, name, age, place of being ('residence'), occupation, marital status, etc.

For names, I link every Source Record that contains a person's name. I have many Individuals listed who have used/ were recorded under various names. So, for me, the more sources listed the better; I can see at a glance the number of sources recording each name. Another point of interest for me is how the Individual recorded or signed his or her own name. Nowadays, in the Citation List PaneI, I move the 'main' evidence to the top of the list (using the up/down arrows on the right).

Also nowadays, I work to more of a 'family history' slant, rather than just a strict genealogy/ pedigree view which was how I started my research. I have an interest in how other family members describe a person (e.g. name/occupation soon or long after a person's death) so I record that too.

I am a splitter of sources (Method 1) and I do not add data to the Source Citation. For me, linking a Source Record to the Name, Whole Record, or Fact is very quickly done. (I enter all information 'manually' using Generic Sources.)

Choose whichever style suits you and your working with FH will sit comfortably with you. It might take a little experimentation.

[tatewise]

in the Citation List PaneI, I move the 'main' evidence to the top of the list (using the up/down arrows on the right).

That is a good piece of advice. It especially helps when using Queries to analyse the data.

[Jean001]

in the Citation List PaneI, I move the 'main' evidence to the top of the list (using the up/down arrows on the right).

That is a good piece of advice. It especially helps when using Queries to analyse the data.

That is exactly the reason why I changed from my previous system of leaving them in numerical order (i.e. oldest first, newest last).

[drewdavis]

That is exactly the reason why I changed from my previous system of leaving them in numerical order (i.e. oldest first, newest last).

yes it makes it easier to query data

[Mark1834]

In an ideal world, it shouldn't be necessary, but FH is not a "proper" relational database, and record sequence is significant, violating a basic principle of database design.

In its defence, it makes writing simple (but potentially superficial) queries much easier, but the downside is that it can be very difficult if not impossible to write fully comprehensive ones without resorting to plugins.

[KFN]

In an ideal world, it shouldn't be necessary, but FH is not a "proper" relational database, and record sequence is significant, violating a basic principle of database design.

In its defence, it makes writing simple (but potentially superficial) queries much easier, but the downside is that it can be very difficult if not impossible to write fully comprehensive ones without resorting to plugins.

Record order of the same tag within context in a GEDCOM is always important and therefore important in any database (relational or otherwise). Also in a relational database we do have a SELECT FIRST in a table so it is allowed and useful to have a an ordered set where the first one in the set has meaning.

[fhtess65]

I also do the same - move the primary evidence to the top.

in the Citation List PaneI, I move the 'main' evidence to the top of the list (using the up/down arrows on the right).

That is a good piece of advice. It especially helps when using Queries to analyse the data.

[Mark1834]

Record order of the same tag within context in a GEDCOM is always important and therefore important in any database (relational or otherwise). Also in a relational database we do have a SELECT FIRST in a table so it is allowed and useful to have a an ordered set where the first one in the set has meaning.

I didn’t say order in GEDCOM is not significant. I said record order in a relational database (a database that structures data into tables, which GEDCOM most definitely is not) is not significant. SELECT FIRST requires additional indexing to be defined, and is only available in some management systems.

The practical consequence is that sequence is highly relevant in FH, but is not in applications that are based on a relational database, such as RM, FTM and others. Neither model is fundamentally “right” or “better”, but it does create differences in how data are retrieved.

[davidf]

Record order of the same tag within context in a GEDCOM is always important and therefore important in any database (relational or otherwise). Also in a relational database we do have a SELECT FIRST in a table so it is allowed and useful to have an ordered set where the first one in the set has meaning.

Is it a requirement of Relational Database design that if order is important, there should be a specific "order" attribute - ideally separate from the internal index?

Without an order attribute in any database, there is a risk that the desired order may be accidentally disturbed. Fact/Attribute date is not an adequate proxy.

[ADC65]

This seems to have rather moved away from the original question!

[KFN]

A citation from a source should (IMHO) reflect the data asserted for the occurrence of an event or an attribute assigned to the individual. For many “facts” this data is usually the date and place, but could also include the cause, religion, witness/participant associated with the fact.

For birth, this is normally sources that contain the birth date or place, but in cases were no date or place can be cited, one or more census records can be cited. I don’t cite a census if I have a reasonable source with a date/place.

How many sources? I cite all that reasonably give me information about the fact data!

[Bobh1]

but then I've copied the birth, marriage & death sources for each of his 12 children as he is also listed on those documents
I don't add the Child's birth to the Father. I add the Father to the Child's birth (as a "Witness")

[Gowermick]

I don't add the Child's birth to the Father. I add the Father to the Child's birth (as a "Witness")

You seem to be ignoring the fact that the child has already been added to the family, so is already shown as the child of the father! You are needlessly duplicating the child’s parentage, and just stating the obvious!

[Bobh1]

I don't add the Child's birth to the Father. I add the Father to the Child's birth (as a "Witness")

You seem to be ignoring the fact that the child has already been added to the family, so is already shown as the child of the father! You are needlessly duplicating the child’s parentage, and just stating the obvious!

Yes, and the birth certificate is also a source and part of the proof argument for that father/child relationship