* What is an objective definition of "Splitting"?

[Gary_G]

Nick;

Again... I have NOT misinterpreted the standard.

It is generally best if the source citation contains only information specific to the fact being cited and then points to the more general description of the source, defined in a SOURCE_RECORD

What part of this are you not understanding?
Are you saying that you have absolutely no "information specific to the fact being cited"? No names, dates, pages etc.
It would seem rather useless to have that situation.

[LornaCraig]

I sense that your point is rather important and would really like to understand it fully.
Would you mind briefly describing what you mean by "record" and "citation"?

Gary, I'm in haste now and can't spend long explaining, but if you go back to the post you mentioned earlier in this topic, which you said you had found Understanding source templates (21211) and re-read that it would be a good starting point.
Come back again after reading that and if I have time later I'll go into more detail.

[BillH]

I have to agree with NIck on this. None of those fields are required as shown in the last post from NIck. The quote you gave says "It is generally best if the source citation contains only information specific to the fact being cited ". I would interpret this to mean that is is best if the source citation ONLY contains information specific to the fact being cited and not any other information. It doesn't say that any of this information is required.

Bill

[Gary_G]

I asked about an objective definition of splitting.

As always, it seems, the thread has devolved into something that is far from what I originally asked or intended. This is not the fault of any one person, but rather that there is an inclination to dissect each post and comment on something that attracts ones attention. That, in turn, encourages a rebuttal. In total, it's not productive in addressing the original query.

I'm going try to draw this thread back on track, if possible...

What is an objective definition of "Splitting"?

If there truly isn't one, then let us close the thread and move on.

[NickWalker]

I am absolutely saying this:

Are you saying that you have absolutely no "information specific to the fact being cited"? No names, dates, pages etc. It would seem rather useless to have that situation

The source information is recorded in the source itself, the citation doesn't need any data. The source reference is in the source, the date is in the source, none of that is in the citation. The quote from your first post explains this:

Added a <<SOURCE_CITATION>> structure subordinate to the fact being cited. It is generally best if the source citation contains only information specific to the fact being cited and then points to the more general description of the source, defined in a SOURCE_RECORD. This reduces redundancy, provides a way of controlling the GEDCOM record size, and more closely represents the normalized data model.

! Systems that describe sources using the AUTHor, TITLe, PUBLication, and REPOsitory fields can and should always pass this information in GEDCOM using a SOURce record pointed to by the <<SOURCE_CITATION>>.

So the Author, Title, Publication and Repository are all recorded in the Source record, not the citation record. The Citation record points at that source. The quote tells you to store as much as possible in the source record itself to avoid redundancy and get closer to a normalized data model.

As I've said in various posts, videos, etc. if I wanted to record the "assessment" (primary, secondary, etc.) then that would be in the citation. But I don't record that (I have no need for it) and the standard doesn't require it.

As I said in my first reply, people are free to record data as they wish. I've been extremely supportive of this - I even provided "method 2" as a recording option in Ancestral Sources. I don't believe in imposing one way of doing things.

[NickWalker]

I asked about an objective definition of splitting.

As always, it seems, the thread has devolved into something that is far from what I originally asked or intended. This is not the fault of any one person, but rather that there is an inclination to dissect each post and comment on something that attracts ones attention. That, in turn, encourages a rebuttal. In total, it's not productive in addressing the original query.

I'm going try to draw this thread back on track, if possible...

What is an objective definition of "Splitting"?

If there truly isn't one, then let us close the thread and move on.

As others have said it is a continuum. There isn't a definition for lumping either. Some people lump at "UK Birth Records", others at "London, UK Birth Records", others at "Greenwich, London, UK Birth Records". A splitter would generally class one birth certificate as a source, one baptism of an individual as a source, one census household as a source but there are variations on that too, e.g. Adrian's double baptisms.

[tatewise]

It seems that much of this discussion hinges on the GEDCOM 5.5.1 quotation:

! Added a <<SOURCE_CITATION>> structure subordinate to the fact being cited. It is generally best if the source citation contains only information specific to the fact being cited and then points to the more general description of the source, defined in a SOURCE_RECORD. This reduces redundancy, provides a way of controlling the GEDCOM record size, and more closely represents the normalized data model.

IMO that is not a formal part of the GEDCOM 5.5.1 Specification.
It is in a section near the end headed:
Compatibility with Other GEDCOM Versions
Changes Introduced or Modified in Draft Version 5.4

Like other GEDCOM 'advice' it is ambiguous compared to the formal definition of the <<SOURCE_CITATION>> structure as has been mentioned earlier.

GEDCOM 5.5 and GEDCOM 7.0 have no such 'advice'.

[AdrianBruce]

...
Lorna;
FH7 has such different terminology from what I'm used to seeing outside the forum.
The program documentation uses "citation" to refer to something associated with a "source" (ie. a GEDCOM type of definition)
The rest of the world uses a "citation" as essentially equivalent to a footnote.
"Record" seems to be similarly ambiguous due to what I've seen used elsewhere.

I sense that your point is rather important and would really like to understand it fully.
Would you mind briefly describing what you mean by "record" and "citation"?

Not Lorna, but I did once get burned in another place by using "citation" differently from their terminology, so let me explain what I think those two terms mean based on that experience.

In what follows, when I use the word "we", I'm describing myself, of course, but I also have the possibly mistaken view that others have the same view as myself.

Firstly Record. When we refer to a Record, we mean the things visible in the Records Window. To be precise, one line on each tab constitutes a Record. So one Individual has an Individual Record visible on the Individuals tab, etc, etc. In particular for the purposes of this thread, one "source" has a Source Record visible on the Sources tab.

That is a Record in FH terms. It's very much an application-based definition, I guess.

To work towards Citation, please see the image below of my distant cousin Hettie Owen where I've selected her birth event (and snipped out the rest for privacy reasons - Hettie's dead but her children aren't). Down below in the yellow Sources Pane, I have highlighted one of the sources for her birth, her Funeral Home Record. To the right of the image is the resulting yellow pane that is described as the Citation consisting of two parts - the Source Record and the Citation-specific Details.

Screenshot 2023-05-07 220824.jpg (295.61 KiB) Viewed 1134 times

I should add that this is a generic, not templated source, and is a split source.

Here's where I think the confusion comes in for Citation - as you say, much of the rest of the world uses Citation to refer to the "printed" footnote / end-note / bibliography / whatever. The whole yellow pane consisting of the Source Record section plus the Citation-specific Details are processed to form the "printed" citation. It should be clear that the "printed" citation consists of data from the Source Record plus data from the Citation-specific Details. I have definitely referred in the past, and no doubt the future as well, to the Citation-specific Details as the Citation. I think others do. I think we did this because originally there wasn't a convenient name for "What goes into the printed citation apart from the Source Record."

If you look at my Citation-Specific Details, you will see that although this is a split source, I have entered the date (that the source was compiled), an assessment, and "Text From Source" - which is the text from the source relevant purely to this fact. I will tend to always enter the first two and only enter the Text if it seems to be useful.

Notice that while the Source Record is a Record, neither of the Citation (in any sense of the word) nor the Citation-specific details are Records in FH.

[tatewise]

That is a Record in FH terms. It's very much an application-based definition, I guess.

No, it is a GEDCOM definition, with which FH claims to be 100% compliant, and other products use the same term.
i.e. GEDCOM has exactly the same concept of Individual, Family, Note, Source, Repository Media, etc, Records.
Citations are subordinate fields. c.f. <<SOURCE_CITATION>> structure subordinate to the fact.

[LornaCraig]

Gary, I'm in haste now and can't spend long explaining, but if you go back to the post you mentioned earlier in this topic, which you said you had found Understanding source templates (21211) and re-read that it would be a good starting point.
Come back again after reading that and if I have time later I'll go into more detail.

Gary, it really is important that you understand the FH meaning of 'Record' and 'Citation'. Yes, the terms are used differently (and loosely) in everyday language but FH didn't invent its own use of the terms. As Mike has said, it uses the words in accordance with the Gedcom definition.

If you have read through that other topic hopefully you understand that there are 12 types of Record, each displayed on a different tab in the Records window, although not all types are displayed by default. You can display them all using the settings in Tools >Preferences > Records window. Each type of record can be opened in its Property Box.

A Source Record is one of those twelve types. A Citation is a link between a Source Record and a field (usually a fact) within another record, or another record as a whole. Certain additonal information (the citation-specific information) can be attached to that link and becomes subordinate fields of the fact. In one of my posts in that other topic I showed what this looks like in the Gedcom file. A Source record can have many citations which may or may not have the same citation-specific information. Every citation is independent of the others, so every one has to have the citation-specific details attached.

This means that if you put, for example, a transcription of a census household in the citation rather than in the Source record, the transcription is repeated in the Gedcom file every time you copy and paste the citation from one fact to another. So if you need to change the transcription for any reason you have to find all the citations and change them all. If it's in the Source record you only have to do it once. This is why the Gedcom spec says that it's best if the citation contains ONLY information specific to the fact being cited. That avoids duplication.

I believe some programs have a concept of a 'shared' citation, which would not comply with Gedcom. This may be the cause of the confusion.

[KFN]

I'm going try to draw this thread back on track, if possible...

What is an objective definition of "Splitting"?

If there truly isn't one, then let us close the thread and move on.

From Wikipedia:

Lumpers and splitters are opposing factions in any discipline that has to place individual examples into rigorously defined categories. The lumper–splitter problem occurs when there is the desire to create classifications and assign examples to them, for example schools of literature, biological taxa and so on. A "lumper" is a person who assigns examples broadly, assuming that differences are not as important as signature similarities. A "splitter" is one who makes precise definitions, and creates new categories to classify samples that differ in key ways.

GEDCOM is generally a “lumper” recording environment, because we usually create one SOURCE_RECORD for 1920 US Census, then add to it as a “Source_Citation” field for the location detail and text of the data found for the specific individual. A splitter would put everything in the “Source_Record”.

See the definition of “Source_Record” and “Source_Citation” from Lorna above!

[tatewise]

GEDCOM is generally a “lumper” recording environment.

Sorry, but I totally disagree. GEDCOM fully supports the entire spectrum from extreme 'lumper' to extreme 'splitter'.
That is reinforced by almost all the Citation fields and all the Source record fields being optional (i.e. {0:1} or {0:M}) as indicated by NickWalker on Sun 7th May 2023 20:43.

[AdrianBruce]

That is a Record in FH terms. It's very much an application-based definition, I guess.

No, it is a GEDCOM definition, with which FH claims to be 100% compliant, and other products use the same term. ...

I think "No" is a bit much. I'd prefer "And".

I wrote what I did because we were talking in terms of FH, so I provided an FH-based definition. The fact that GEDCOM also defines those "things" as Records is actually very important and obviously GEDCOM came first. But of course, the GEDCOM definition doesn't help with the idea that Places are records in their own right in FH. (GEDCOM does provide the framework for how Places are encoded in the GEDCOM file, but that's hard work to get back from such extension records to Places in FH, and in that instance, FH came first).

[tatewise]

Adrian, in the context of this thread regarding Sources and Citations, I maintain that FH in being compliant with GEDCOM takes its terminology from GEDCOM, as do several other products, so it is not very much an FH application-based definition.
Sorry, but IMO your guess was wrong.

[KFN]

GEDCOM is generally a “lumper” recording environment.

Sorry, but I totally disagree. GEDCOM fully supports the entire spectrum from extreme 'lumper' to extreme 'splitter'.
That is reinforced by almost all the Citation fields and all the Source record fields being optional (i.e. {0:1} or {0:M}) as indicated by NickWalker on Sun 7th May 2023 20:43.

The fact that the <<Source_Record>> fields are optional has no real bearing on the Lumper vs Splitter Debate! If you don't use a field that's fine, but you can't create a field in another place without creating a custom field in GEDCOM. Data will most likely be lost with custom fields.

I want to explain my thoughts on citations in pure GEDCOM as it pertains to my assertion that GEDCOM is generally a lumper model.

I'll start two definitions.

1) General Guidelines for Print Sources from the American Anthropology Association (AAA)

2) The current v5.5.1 definition of the data-points available in the "Sourcing Model". This "model" actually started with v5.4 when the <<Source_Record>> and <<Source_Citation>> first took their "simplified" structure.


General Guideline for cited reference.

• Author(s)
• Year of Publication
• Title of Author's work
• Title of Source
• Publishing Information
  a. Place of publication
  b. Publisher
  c. Page numbers (location) for work from journals or books

In general the citation for an article might take the form of:

Author Name (last, fore), Year of Publication, Article Title, Journal Title, Volume Number:Page


GEDCOM V5.5.1 design elements

<<Source_Record>>
The <<Source_Record>>, as described in The Standard, has five basic sections:

• data or classification,
• author,
• title,
• publication facts,
• repository.

The data/classification section is metadata about the the source for analysis of the source collection as a whole. The remaining sections contain the basic building blocks of the citation/footnote/end note to be built.

<<Source_Citation>>
The <<Source_Citation>> structure is placed subordinate to the fact being cited. It is generally best if the source citation contains only information specific to the fact being cited and then points to the more general description of the source, defined in a SOURCE_RECORD.

NOTE: In GEDOM v5.4 and beyond a <<Source_Citation>> can use an alternate structure that does not link to a <<Source_Record>> where all elements are contained in an unstructured "Source_Description". Systems that do not use a <<Source_Record>> must use this structure, otherwise the <<Source_Record>> will be used.


Conclusion:
Since most modern genealogy applications provide a <<Source_Record>> in their transmission, and considering the AAA general guidelines for sourcing information, I would conclude that the citation information, if loaded into the provided GEDCOM fields, would indicate that GEDCOM is generally a "Lumper" design, where the source data is "broadly stored in the <<Source_Record>> (Author, Title, Publisher), and specifics for the Reference are stored in the <<Source_Citation>> (page, volume).

[ColeValleyGirl]

I don't think this discussion (which will never result in agreement) is helping the OP at all. Start a new topic and move the Gedcom discussion there?

[KFN]

I don't think this discussion (which will never result in agreement) is helping the OP at all. Start a new topic and move the Gedcom discussion there?

You are probably right. My initial comment was to define Lumping vs Splitting which was directly related to the question the OP asked.

[kfunk_ia]

I asked about an objective definition of splitting.

In my world, if one follows the EE method then one is a splitter. Since you are striving for EE compliant sourcing, that would seem to make you a splitter.

[BillH]

I asked about an objective definition of splitting.

In my world, if one follows the EE method then one is a splitter.

Not sure there is any correlation. I am a splitter and I don't follow the EE method.

Bill

[kfunk_ia]

I didn't say that the EE method was the only way to be a splitter. However, I have read many of Gary's posts here and in groups for other software packages so I am somewhat familiar with what he is doing . So my reply was directed to him, not you.

[BillH]

So my reply was directed to him, not you.

A little touchy are we?

Bill

[sbell95]

I asked about an objective definition of splitting.

In my world, if one follows the EE method then one is a splitter. Since you are striving for EE compliant sourcing, that would seem to make you a splitter.

That seems to be a sentiment echoed in other forums, but Elizabeth Shown Mills herself has never endorsed either approach, and from what I can tell, doesn't rely on any software package to generate her citations.

I asked about an objective definition of splitting.

As always, it seems, the thread has devolved into something that is far from what I originally asked or intended. This is not the fault of any one person, but rather that there is an inclination to dissect each post and comment on something that attracts ones attention. That, in turn, encourages a rebuttal. In total, it's not productive in addressing the original query.

I'm going try to draw this thread back on track, if possible...

What is an objective definition of "Splitting"?

If there truly isn't one, then let us close the thread and move on.

I don't think a discussion about objectivity/subjectivity will get very far, at least not without strong disagreement as this thread has shown. The decision whether and to what degree one chooses to lump or split seems to be arbitrary. FH and GEDCOM support whatever you want to do, and I don't see how choosing to do it one way or another could affect anything more than simply your way of working.

If one wants to be as compliant with GEDCOM as possible for exporting purposes, there are a myriad of other issues that crop up because of the way that different software developers have gone about implementing the standard (as others have said). If you want to be as compliant as possible with EE, that can be done by either lumping or splitting (to whatever degree) or a combination of both, as long as the end result is a citation that conforms to her style.

So, really, doesn't it just come down to how you choose to work in FH and your end goals for producing output? And lucky for us, FH has so many different and customisable ways to deal with sources and citations. For me, it's a breeze and in conjunction with another tool (ORA) means that I produce consistent and comprehensive citations for all of my sources.

I personally would like to see this forum focusing on workflows and use of the features of FH, as opposed to philosophical questions that only invite argument and negative commentary on things that are beyond our control (like how FH implements GEDCOM).

[KFN]

Evidence Explained is not about splitting or lumping, it is a style of citation. Citation styles take many forms, APA, MLA, Chicago, Turabian, IEEE, each has a place and use. I used Society for American Archaeology (SAA) citation style back in the 1980s for my genealogy projects.

Splitting and Lumping is about how/where to store the various bits of information used in the citation. Depending on how the database is designed the various parts of a citation can either be organized into categories (Lumping) or Uncategorized (Splitters), both work but only if it creates a valid citation for the selected style!

[fhtess65]

Sarah echoes my thoughts... This may be an example of the type of conversation that intimidates those new to genealogy and/or Family Historian.

Unless Helen creates a sub-forum dedicated to philosophical discussion about gedcom etc., which most people will avoid unless it really fascinates them.

So, really, doesn't it just come down to how you choose to work in FH and your end goals for producing output? And lucky for us, FH has so many different and customisable ways to deal with sources and citations. For me, it's a breeze and in conjunction with another tool (ORA) means that I produce consistent and comprehensive citations for all of my sources.

I personally would like to see this forum focusing on workflows and use of the features of FH, as opposed to philosophical questions that only invite argument and negative commentary on things that are beyond our control (like how FH implements GEDCOM).

[jbtapscott]

Must admit I had been tempted to respond to Helen's post about the revised forum groupings to say that a separate "Gedcom" forum really needs to be included - some of the recent discussions which have degenerated into disagreements about the meaning of certain words in the Gedcom specs really are off-putting and I'm not "new to genealogy and/or Family Historian"!.