Jump to content


Photo

What is the definition of a "Master Source"?

Source Template

  • Please log in to reply
12 replies to this topic

#1 History Hunter

History Hunter

    Member

  • Members
  • PipPip
  • 14 posts

Posted 10 June 2017 - 12:10 PM

After looking at the source templates in RM7 and reading through the matching Evidence Explained sections, it seems that the definition of a source is different between the two. If I look at the EE section 6.50 and go through filling out the Census Canada (online images) form in RM7, I get the feeling that RM7 looks at a specific province, district and sub-district as the "re-usable" portion or Master Source. That is it seems to look as the Master Source as the Census Book for the area. However; EE seems to look at the website as the source and so requires far fewer Master Sources to be defined.

 

I should note that the finished citations turn out to be pretty-much the same. It's just that the RM7 template design seems to require a new template for each jurisdiction and EE does not.

 

Can someone confirm whether my suspicions are correct and maybe explain the apparent difference in philosophy a bit further?

 

While I understand how to modify the templates, I don't want to do so if there's a good reason to leave them alone and live with the status quo.



#2 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3535 posts

Posted 10 June 2017 - 03:47 PM

Let me distinguish between the source templates provided by RM and the support for source templates provided by RM. The support provided by RM allows any sourcing data to be placed either in the Master Source or in the Source Details. Different users have different needs and different preferences as to which data goes in which place. It seems to me that the source templates provided by RM differentiate the data between the two places in a moderate way - not putting too much or too little of the data in either place.

 

I may need to reread some sections of EE and my understanding of things may not be correct, but it seems to me that EE doesn't really address the distinction between Master Sources and Source Details in the sense that RM uses the terms. Rather, it seems to me that EE is describing how to construct footnote/endnote sentences as a whole. It's very similar to footnote/endnote sentences provided by Web sites such as familysearch.org and ancestry.com, where the sentences you can copy and paste are whole sentences and are not designed to be divided between RM's Master Sources and RM's Source Details.

 

I look forward to hearing additional discussion on this topic. There are surely other perspectives on how EE's sourcing recommendations relate to RM's Master Sources and Source Details.

 

In any case, if you choose to use RM's built-in source templates, I would urge you not to use them directly, but rather to make copies of the ones you want to use and to use the copies. That way, if you find any errors in a source template - a comma out of place or anything trivial like that - you can fix the error. You cannot make any changes at all in RM's built-in source templates, not even to fix an error.

 

Jerry



#3 KFN

KFN

    Advanced Member

  • Members
  • PipPipPip
  • 207 posts

Posted 11 June 2017 - 09:33 AM

c24m48,

As you described EE is about recording and the proper reporting of source information. In most cases this is (as we know it) a "footnote", or "endnote".

What EE wants everyone to understand is that different sources have (or should have) specific data points that are needed for telling readers where and what was used to assert your conclusion.

EE does not tell us how to store that information, since everyone reading that document will have different data storage mediums, including but limited to one of several computer programs. Don't forget paper and index cards as well. EE developed names for its data points independent of the recording medium. The EE specification is NOT a database structure, it does not tell readers how to normalize redundant data to make recording more efficient.

For the most part RM has also developed names for its datapoint independent of other concepts. When I see the term "Master Source" I have to relate that back to what I know more intimately since GEDCOM does not have that term anywhere. This would also be true about each individual data point.

What all this means is that while EE is a very good and reliable document on suggested datapoints to record for sources it is not a roadmap to the storing of said data into your selected personal recording medium. The templates are or should be the translation point between what EE suggests to be recorded and where RM places those datapoints into its database.

#4 History Hunter

History Hunter

    Member

  • Members
  • PipPip
  • 14 posts

Posted 11 June 2017 - 09:56 AM

Thanks for the reply, Jerry.

 

As a bit of background... I went back to the "Evidence Explained" (EE) book and note that the author does indeed say that the website is considered to be a "book" and the databases on the website are considered to be "chapters" in that "book". The author also notes that websites do not really have a "repository". Hope that I've paraphrased that accurately.

 

The RM7 template for "Census, Canada (online images)" divides the full EE-style citation into the "Master Source" and the "Source Details". It appears that RM7 uses the "Source Details" to document reference details which appear on what was actually viewed (ie. the artifact) and its access date. RM7 appears to use the "Master Source" to document how one accessed the artifact that was viewed. Admittedly; this observation is based on only the noted stock template.

 

I suspect, as in other genealogical software, the Master Source concept was intended to capture the invariant part of a user-defined set of references. Note that one user may define that invariant part as being a reference to an entire census year, another might define it as an enumerators book, and yet another as a census page (as was apparently done in the stock RM7 template).

 

It appears, as such, that the two concepts do not "collide". While the RM7 template does result in a citation that closely follows the EE style, it also overlays that with its own (apparently unstated) view on what parts are invariant or re-usable. In my opinion, the implied design concept of the RM7 template would likely lead to quite a few unique Source Templates for a collection such as the the Census of Canada on the Library and Archives Canada.

 

I believe that a more functional compromise (for me) would be to reallocate the contents to treat the Master Source as documenting the reference at the census year level or possibly at a enumerators book level. This is not to say that RM7 is "wrong". They have to draw the line between the Master Source and Source Details somewhere. It just doesn't lend itself to the geographical groupings I have in my ancestral line.



#5 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6218 posts

Posted 11 June 2017 - 10:13 AM

History Hunter, you have accurately described what has been named, in the vernacular, as "lumping" and "splitting", the extremes of which, in RootsMagic, would result in one versus very many "Master Sources". An extreme lumper may have one Master Source for the 1911 Canada Census, or, even more extreme, one for all Canada Censuses. An extreme splitter may have one Master Source for each page of the 1911 Canada Census. The virtue of one is a smallish and more manageable list of Master Sources but at the risk of poorly constructed footnotes when exported to third party systems if templates are used and having to edit every citation if a revision at the Source Detail level is required. The virtue of the other is being able to refine in one spot all citations, faithfully replicated footnotes exported to third party systems at the expense of a very large Source List requiring a rigorous, systematic, structured naming practice to be manageable.

 

The RM templates designer opted for a mid-range in the spectrum between those extremes.


Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.


#6 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3535 posts

Posted 11 June 2017 - 10:55 AM

The RM7 template for "Census, Canada (online images)" divides the full EE-style citation into the "Master Source" and the "Source Details". It appears that RM7 uses the "Source Details" to document reference details which appear on what was actually viewed (ie. the artifact) and its access date. RM7 appears to use the "Master Source" to document how one accessed the artifact that was viewed. Admittedly; this observation is based on only the noted stock template.

 

If you like this template except for the way it distributes the data between "Master Source" and "Source Details", simply copy the RM version of the template and change the distribution of the data in your version of the template before you start using your version. Doing so is extremely easy and it will not change the text of the footnote/endnote sentences. Also, I pre-pend the names of my templates with a special character (I presently use an *) so that they will sort to the front of the list of templates.

 

Jerry



#7 History Hunter

History Hunter

    Member

  • Members
  • PipPip
  • 14 posts

Posted 11 June 2017 - 10:57 AM

Thanks for the reply, Tom.

 

Up until recently, I'd never heard the terms you noted (i.e. lumping and splitting). I just switched from The Master Genealogist in the last few months. With that program, I didn't seem to have this decision to make. Unfortunately, it's gone and I'll have to adapt to using RM7.

 

Using the unaltered RM7 templates, I would likely spend most of my time fiddling with naming conventions for the templates and a lot less on actually filing the data. As such, I've decided to take a short time to determine the optimum amount of data t put in the Master source (for my own ancestral data locations), so that I can make the majority of anticipated changes with the minimum difficulty. Until I wrote down my thoughts in my recent post, it wasn't clear to me that the RM7 template design was pitched at a particular point on the "Lumper/Splitter" scale. There's really no discussion of the design concept in any of the documentation I've seen. Fortunately, I can easily alter the templates (without affecting EE-style compliance) by copying and doing very minor editing of the templates (something I generally prefer not to do without a lot of thought). There's always the chance I'll misjudge or conditions change and alter that optimum point, but remaining in flux is not an option. Data is coming in too fast to delay the decision any longer.



#8 History Hunter

History Hunter

    Member

  • Members
  • PipPip
  • 14 posts

Posted 11 June 2017 - 11:09 AM

Just got your reply as I posted a reply to Tom.

 

Yes, what you suggest is all the editing I was going to do, unless I find an issue in a template.

So far, the templates seem to generate pretty accurate EE-style references.

In a copy, I just need to only put "Y" on the fields I want in the Master Source.

 

Given the vast number of potential data-sources, I'll end up "rolling my own" custom templates sooner or later.

However; I'd prefer that to be later and get on with my lagging data-entry.

It'll also give me time to ensure that my chosen approach to segregating Master/Detail data works.

 

I wasn't aware that an asterisk was allowed in a name.

Wouldn't that cause file-naming issues when exporting a template?

I was thinking of using an underscore to get the same sorting effect.



#9 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3535 posts

Posted 11 June 2017 - 11:11 AM

As a bit of background... I went back to the "Evidence Explained" (EE) book and note that the author does indeed say that the website is considered to be a "book" and the databases on the website are considered to be "chapters" in that "book". The author also notes that websites do not really have a "repository". Hope that I've paraphrased that accurately.

 

Thank you for this background. I have the 2007 edition of EE. Armed with your message, I found the information on pp.57-58.

 

I have never quite figured out how or if repository information is supposed to be incorporated into footnotes/endnotes. The pages you cite certainly suggest that EE does not consider repositories to be essential information. Of the 415 templates built in to RM, 134 include repository information and 281 do not. I have never figured out the rhyme or reason behind which RM templates do and which RM templates do not include repository information. I use templates of my own design, not RM's built-in ones nor copies of RM's built-in ones. For better or worse, none of my templates incorporate repository information, e.g. ancestry.com appears in my endnotes/footnotes but ancestry.com is not a repository in my templates.

 

Jerry



#10 Jerry Bryan

Jerry Bryan

    Advanced Member

  • Members
  • PipPipPip
  • 3535 posts

Posted 11 June 2017 - 11:22 AM

For the most part RM has also developed names for its datapoint independent of other concepts. When I see the term "Master Source" I have to relate that back to what I know more intimately since GEDCOM does not have that term anywhere. This would also be true about each individual data point.

 

I tend to think that the way nearly all genealogy programs handle sourcing information goes back to PAF and GEDCOM. My naive and overly simplified perception of the original PAF/GEDCOM model is that really only two fields were provided - book and page number. Well, book was sub-divided into author, title, and publishing information, and publishing information was sub-divided into publishing place, publishing date, and publishing company. Users of genealogy software were then forced to be very creative in figuring out how to map things like marriage certificates, death certificates, census records, etc. onto this book and page number model. The exact field names I'm describing aren't exactly the ones from GEDCOM, but I think the concepts I'm describing are very close to the data model from GEDCOM. To this day, RM's free form source template follows the data model I'm describing.

 

Jerry



#11 History Hunter

History Hunter

    Member

  • Members
  • PipPip
  • 14 posts

Posted 11 June 2017 - 11:47 AM

Hi Jerry.

 

First, I should note that EE doesn't say or imply that repository info is irrelevant in all cases. It is speaking specifically about web-based references in which a physical repository doesn't truly exist. In other places, it also adds that repository information is not generally included for widely available publications. So; it would appear that repository info is required for books and such that are available only at specific locations. That makes sense and is why RM7 still retains the ability to add a repository. Given that you likely don't have so many references (these days) to such items, it likely won't affect you too much. However; you might consider using repository info for such things as "Aunt Martha's personal records". :>)

 

Odd that you bring up PAF. I feel that part of the reason it was discontinued is that its design paradigm was formed in an age in which the web was in its infancy (and the need to reference a physical repositories and materials was the norm). When the web became the predominant source of information, the basic design philosophy was too entrenched to be changed to accommodate the web. The GEDCOM standard, being a less involved design to change, has evolved somewhat, but only somewhat. However; the incentive to update the GEDCOM standard and the ability to enforce adherence to even an updated standard seems to be diminishing due to the number of low-cost software alternatives and a "lack of incentive" for designers to provide any easy migration path out of their software offering. On the other hand, I've noticed that many software offering provide a relatively automated way of reading another programs native "database" files and GEDCOM support seems to often be thrown in "just in case". This makes sense for them to help you get "onto" their platform and keep you there.



#12 TomH

TomH

    Advanced Member

  • Members
  • PipPipPip
  • 6218 posts

Posted 11 June 2017 - 12:12 PM

"I think the concepts I'm describing are very close to the data model from GEDCOM. To this day, RM's free form source template follows the data model I'm describing"

I disagree in that RM's Free Form is a gross simplification of the GEDCOM data model. On import, the master source tags values are concatenated and punctuated into Footnote and Short Footnote fields. On export to standard GEDCOM, the Footnote field maps to only the TITL tag. The only GEDCOM tag that survives the round trip intact is PAGE, which has a corresponding field in RM at the Source Details or Citation level. Generally, the importing software does produce its own footnotes matching those within RM, apart, possibly, for the punctuation separating the PAGE value from the rest.

Unfortunately, exports of sources that are constructed on most templates that have any number of Source Details fields produce poor footnotes in the other software because RM exports only to the TITL and PAGE tags. The former is filled with the RM footnote, sans the values and dependent text for the Source Details fields. PAGE is filled with the punctuated concatenation of the Source Details field values.

This "corruption" of footnotes has driven some to use templates only as an aid to drafting Free Form sources or to use "ultimate" splitting - no fields in Source Details.

Tom user of RM7550 FTM2017 Ancestry.ca FamilySearch.org FindMyPast.com
SQLite_Tools_For_Roots_Magic_in_PR_Celti wiki, exploiting the database in special ways >>> RMtrix-tiny.png app, a bundle of RootsMagic utilities.


#13 KFN

KFN

    Advanced Member

  • Members
  • PipPipPip
  • 207 posts

Posted 13 June 2017 - 06:49 PM

Jerry,
Your Book and Page understanding of sourcing in GEDCOM is some-what accurate in that it is very much directed toward “old school” hard copy objects such as: audio and files, books, fiche and film, magazines, manuscripts, maps, newspapers, photos, tombstones.
GEDCOM states that a PAGE (as part of a “source_citation”) is defined as: “Specific location with in the information referenced.” Which more people and software programs assume only refer to books, since a location within a book is best described with a page number, while other media requires additional information beyond a page to describe a unique location.
However, the field/tag title of “PAGE” does not best describe what GEDCOM really can expect to find in the tag PAGE. GEDCOM goes on to state:

“For a published work, this could include the volume of a multi-volume work and the page number(s). For a periodical, it could include volume, issue, and page numbers. For a newspaper, it could include a column number and page number. For an unpublished source or microfilmed works, this could be a film or sheet number, page number, frame number, etc. A census record might have an enumerating district, page number, line number, dwelling number, and family number. The data in this field should be in the form of a label and value pair, such as Label1: value, Label2: value, with each pair being separated by a comma. For example, Film: 1234567, Frame: 344, Line: 28.”

This statement therefore incorporates a whole lot of data points that many programs fail to include (or create their own tags for) in a well formed GEDCOM file.

However, this is not to say that I endorse this construct in my discussions on the topic for a successor version to v5.5.1

As it pertains to web based data, I recommend two different approaches to sourcing these data points, with the following caveat.
“Is the website REALLY a source or just a repository for the actual information?” I look at sites like Ancestry.com as repositories and treat them as such in my sourcing, footnotes/endnotes.

A source is the originator or creator of the data. They are generally responsible for the accuracy and collection of the original data. Ancestry did not go out and capture the data for a census or immigrant list. They are acting like a library who has collected books on a subject, not the author of that book. The only cases where websites in general are “Sources” is where they take other individuals work and author their own manuscript. These manuscripts generally contain the website’s conclusions and should source where they collected the data that led to their conclusions.

The two places I recommend placing website information is: (1) In the PAGE tag/field, used for specific pieces of information that collaborate your fact’s conclusion. (2) In the Source_Repository_Citation directing readers to look at a website for additional data.

Since “PAGE” is already noted in the GEDCOM standard as more than just a single page number that direct you to a fully qualified location like sheet numbers, frame numbers, volume, issue etc. A full URI (aka URL) to the source on the web you are citing would be appropriate. In cases where the site is not the actual source but the “Repository” of the data, placing the site URI in the Source_Repository_Citation structure is the appropriate location for this address. The Source_Repository_Citation structure contains a place to put notes about the data location, a CALN tag for the actual Source Call Number at the Repository and a media type.