Login to participate
Register   Lost ID/password?
Louis Kessler's Behold Blog » Blog Entry           prev Prev   Next next

Recording Your Reasoning (Proof Argument) - Thu, 28 May 2015

So I was about to to update the documentation section for Source Details, and I looked at how I had implemented this in version 1.0:


Version 1.1 was unchanged from this, so I thought the update of this section would be simple. But then it struck me. I was not displaying this information correctly.

In the couple of years since the last version of Behold, I had been increasing my knowledge about dealing with sources, and last August I wrote a paper on Standardizing Sources and Citation Templates, which I submitted to FHISO.

Point number 8 in that paper was something I realized was very important: to separate out the sources from the conclusions. What I said was:

All assumptions and conclusions and assessments of the source must be placed with the source reference, not with the source

Well, take a look at S5-3 in the screenshot above. Listed under the source record is some “data” that says:

The Enumerator for this census improperly listed John J. McCarthy as being born Sep 1800, he then changed the date to 1890, which is still incorrect. John is listed as being 8-months old on the date of the census, 9 Jun 1900, if you reverse the date for 8-months John would have been born in Sep 1899, which is the proper date.

Well this information is not part of the source. This is the researcher’s assessment of the source. It should not be displayed as part of the source, but should be displayed as the analysis of the source used to arrive at the the conclusion. In this case, the conclusion is the birth event of John J. McCarthy.

Similarly, there are those “Quality” values. Those also are not part of the source. They are the researcher’s assessment of the quality of the source with respect to arriving at the conclusion. One source can be assessed differently for different conclusions, e.g. a death record may be very good for the death date, but if it only includes an age at death, then it’s not as good for the birth date.

Let’s see where this information comes up in GEDCOM. There’s conclusion information:

2 DATE SEP 1899
2 PLAC Boston, Suffolk, MA

and under that will be the source reference:

2 SOUR @S90@
3 PAGE Genealogy.com, Series: T623, …
3 QUAY 3
3 DATA The Enumerator for this census …

The SOUR line is the pointer to the source. The PAGE line describes the specific record used within the source.

Now those two lines, the QUAY and DATA, well they are not part of the source record and shouldn’t be displayed as part of the source record. They are part of the reference to the source. They describe the linkage, i.e. the analysis and reasoning used to come up with the conclusion.

As a result, the QUAY and DATA and the other information allowed with it (e.g. Notes, Objects, Date recorded, Event cited from), are all part of the linkage between the source record and the conclusion. What this means is that this information needs to be displayed in two places.

One place is with the conclusion, to describe the reasoning the source record brought to the conclusion:


The other is with the source record, to show the reasoning was used with this source for each conclusion:


Note the difference in the S5-3 listing to the 1.0 version of it shown earlier. The Quality and Data are now shown attached to the conclusion event that is supported by the source record.

Previously they were included as part of the source record. When that was done, S5-4 did not have exactly the same quality and Data values. So S5-4 previously was shown as a different record.

Now the source record can be treated as identical, and the former S5-3 and S5-4 can be put together as the new S5-3 with a combined total of 3 supported events. The reasoning based on that source record can now be attached individually to each event.

This is really major. It has opened my eyes up to the fact that GEDCOM actually has a place for a researcher’s reasoning statements. The user’s analysis/reasoning can go into the NOTE statements that are placed in source references.

Doing this can allow a step-by-step proof argument to be documented and passed on through GEDCOM to another program. You would do it like this:

2 DATE dd MMM yyyy
2 PLAC xyz
2 SOUR @S1@
3 NOTE The date and place were from the birth certificate
2 SOUR @S2@
3 NOTE Immigration record contained her age and country o 4 CONC f origin, agreeing with what I had.

Once I add editing to Behold, I’ll also add the ability to record your step-by-step proof argument, and you’ll be able to document and display all your reasoning.

I needed three days to make these changes to Behold. Now back to finishing the documentation and getting Version 1.1 out.

2 Comments           comments Leave a Comment

1. arnold (arnold)
Canada flag
Joined: Mon, 24 Nov 2014
10 blog comments, 13 forum posts
Posted: Mon, 1 Jun 2015  Permalink

Curiously, I am just now also preoccupied with SOURces in trying to ‘coerce’ my app to make it easier to add/maintain/check the source linkages.

One thing that seems a possible problem to me, based on my current knowledge of the GEDCOM standard:
with SOURce records containing NOTEs and NOTEs containing source citations, will the be a danger of (infinite) loops :-)

2. Louis Kessler (lkessler)
Canada flag
Joined: Sun, 9 Mar 2003
232 blog comments, 226 forum posts
Posted: Mon, 1 Jun 2015  Permalink

Yes, you can get infinite loops. It normally isn’t a problem for a program like Behold that may expand one level, but otherwise gives everything as links to the other structure. With links, you’ll click on one to go to the other. And then, you’ll click on the other and go back to the first.

But any program that tries to expand all those links will have to either detect that it is reached an already visited link, or it will have to put a limit on the depth it goes.

In Behold I do this sort of thing to detect ancestral loops. I have to mark visited people, and if they are encountered again, record the problem and stop.

Leave a Comment

You must login to comment.

Login to participate
Register   Lost ID/password?