It's Not a Hypothesis - It's Wrong!

What do you do when a source says one thing, but you think that source is wrong?

All sources are not created equal. Your analyses, correlations and conclusions should show that you recognize that fact. In your personal database, you keep track of all records you view, even those that are demonstrably incorrect. How do you enter those records? How do you enter the information they contain? Or do you enter it?

• You need to track this data in a way that makes it clear that you believe a record's information is incorrect and explains why you came to that conclusion.
• You need to control this data in such a way that it is not reported in any manner that suggests it might be valid.

There are so many different ways to mark information in TMG! You can pick and choose from any of these methods, combine them, or create your own.

----------------------------------------------------------

There are three types of exclusion markers available in TMG. Consistent use of these markers will allow you to control what is printed in various reports. (Note that I have not tested these in GEDCOM file output.)

1. Any text enclosed in curly brackets, or sensitivity brackets ({}), will not appear in any report unless you override it in the Report Options > Miscellaneous tab.
2. The hyphen (-) entered first in any text field will exclude that information. No information preceded by this exclusion marker will appear in any report unless you override it in the Report Options > Miscellaneous tab.
3. Do you have information that you want to include in your database, but you never want it to appear in any report? Precede the text with a double hyphen [--]. You will be able to see the information when you open the event, but it will not be visible from the Individual Detail screen, nor will it appear in any report, regardless of the report options selected.

----------------------------------------------------------

TMG's surety values are no substitute for a written analysis of a record, its information, and the evidence it supports. They can be useful in many ways, though. The hyphen, entered as a negative surety value, marks a source's bit of information as incorrect. It may also affect the reporting of a tag. Here are some examples of the data entry and its corresponding report output - or lack thereof.

Excluded citation.

Figure 1a. Tag Entry screen showing excluded source citation.

Figure 1b. Individual Report showing tag citation in parentheses.

Tag showing incorrect information.

Figure 2a. Tag Entry screen showing incorrect information.

Figure 2b. Individual Detail report showing output.

 Same tag with negative surety value (Date).

Figure 3a. Tag entry screen showing negative surety (Date).

Figure 3b. Individual Detail report showing resulting output.

Same tag with negative surety value (Principal 1).

Figure 4. Tag Entry screen showing negative surety value (P1).

There is no corresponding report output for the entry shown in Figure 4. Entering a negative surety value for a Principal in a source citation prevents that tag from appearing in any report about that principal, assuming it's the only citation entered, of course. This technique even works for marriage tags, but it does not work for Parent-Child relationship tags.

----------------------------------------------------------

Entire genealogies are based on incorrect information, and to counter the arguments of those enamored of a fictional branch of their family tree, you will need to know exactly where those genealogies went wrong. Sometimes that incorrect information comes from original records. If so, the discrepancy must be investigated and explained. After all, you might discover that the information was correct; it just referred to a different person.

If the incorrect information comes from a derivative document, such as a poorly researched and analyzed family history, do you really need to treat it as if it were a reliable source? Consider creating a custom tag, such as Wrong_Info. Use this tag to record and discuss this "wrong information," but don't include this tag in any report you intend to publish or share.

Published genealogies on one of the families I research, the Longfellow family of Delaware and Maryland, are so full of incorrect information that I found it too confusing to include that information in my data project. I removed it to a separate data set within that project. As I find new records about these families, information from the published genealogies is handy for comparison, but it doesn't "contaminate" my project. Perhaps this is an extreme solution, but if you run into a similar problem, it might work for you, too.

Data Set Manager screen showing enabling of Longfellow data set.

View of Picklist with all data set entries visible.

In summary, track incorrect information, but treat it responsibly. It should never be reported as if it were correct - and perhaps should never be reported at all. There's enough poor genealogy going around!

It's Only a Hypothesis - Part Two (C)

What do you do when different sources provide different dates for the same event?

You might not know exactly what to enter in your TMG database when different sources provide different dates for the same event, but by this time, you should know that you don't want to say, "Eenie, meenie, minie, moe," and enter the result. If there is not enough evidence to come to a conclusion, then you cannot make that conclusion. Instead, enter the evidence, analyze it, and discuss the possible hypotheses. The date of birth for Emily August Rollins is one of my favorite every-source-says-a-different-thing situations.

Individual Detail Screen for Emily Augusta Rollins

There are five different dates of birth for Emily Augusta Rollins recorded in independently created original records. Affidavits, census records, and her tombstone do not agree with any one of these records. You cannot ignore any evidence, contradictory or not. All evidence must be analyzed individually and in correlation with all other evidence. That analysis must be written and recorded and any conclusion reached must be logically explained. Like the various -Candidate tags, the various -Alt[ernate] tags were created so long ago that they feel more like TMG standard tags rather than custom tags. My TMG database now includes the Birth-Alt, Burial-Alt, Death-Alt, and Marr-Alt tags to handle contradictory evidence relative to these life events. Each tag is part of the appropriate tag group, so each tag label can be changed if new evidence arises suggesting an -Alt tag should be "upgraded." Note the circled Comment tag. That's where my analysis is entered.

Birth tag entry screen showing customized sentence structure

Don't depend on TMG's default sentence structure to handle contradictory or qualified sentences. This screen shows the sentence resulting from the customized sentence structure. Remember, custom sentences are not changed if you edit your default sentence structure for a given tag.

Excel table summarizing birth evidence

The Excel table shown here summarizes all the birth evidence found to date on Emily Augusta Rollins. The file is linked to her Birth event tag. Although this table must be manually inserted in a report, it's at hand any time I want to review or update the evidence.

Journal report showing analysis comment and inserted table

It's important to share the results of our research with family members and other researchers, but once we've done so, we've lost control of that research. The best way to retain some measure of control over the accuracy and quality of that research, especially when it includes information of a hypothetical, contradictory, or inconclusive nature, is to share that research in narrative format only. Shown here is a first draft report with the summary table inserted. The commentary is source-cited, thanks to TMG's embedded citation feature.

Family group sheet report showing inclusion of -Alt tags

If you choose to publish abbreviated reports, like family group sheets, or send GEDCOM files (No! No! No!), make sure those reports include all -Alt tags and other qualifiers. TMG does export this information in appropriately designed GEDCOM exports, but I don't know how other genealogy programs handle the import. Also note that only the primary birth, death, and marriage tags are printed in pedigree charts.

Continue with "It's Not a Hypothesis - It's Wrong!"

It's Only a Hypothesis - Part Two (B)

"But rarely is genealogical evidence overwhelming, and frequently we have to use careful language and words beginning with the letter P, such as probably, perhaps, and possibly. You be the judge of when and how to use them in your own research." (Henry B. Hoff, “Weighing the Evidence,” New England Historic Genealogical Society, American Ancestors (http://www.americanancestors.org/weighing-the-evidence/ : accessed 31 December 2013))

How do you show that you're not sure of a place?

One method was suggested at our after-meeting lunch: add a question mark to the place name, with or without square brackets or parentheses. This works perfectly well in the database, in charts, and in GEDCOM files, but when was the last time you read something like this in a genealogy article? "Lydia Ball was born say 1802 in Groton(?), Grafton Co.(?), New Hampshire(?)."

Many of the qualifiers and qualifying statements relative to dates can be entered in TMG such that the suppositions and explanatory statements are clear in the database, as well as in narrative reports, family group sheets and pedigree charts, and GEDCOM files. I don't find that as easy to accomplish when dealing with hypothetical place information. Standard qualifiers for places don't seem to have been codified in quite the same way as they've been done for dates. My usual practice is fine for my personal database and works for journal reports, but it doesn't work as well for family group sheets, pedigree charts, or GEDCOM files. Other solutions might work for the database, but not for journal reports. Accuracy and clarity is important, though, and your TMG entry decisions might be helped by examining the meanings of these varied qualified statements. All of these statements would require additional information, whether in the text or in a note, that explained the qualification(s). Do you agree with my interpretations?

1. Lydia Ball was born say 1802 possibly in Groton, Grafton Co., New Hampshire.
2. Lydia Ball was born say 1802 probably in Groton, Grafton Co., New Hampshire.
3. Lydia Ball was born circa 21 April 1803 possibly in Groton, Grafton Co., New Hampshire.
4. Lydia Ball was born in Lisbon, Grafton Co., New Hampshire, circa 21 April 1803.
5. Lydia Ball was born possibly 21 April 1803 in Lisbon, Grafton Co., New Hampshire.
6. Lydia Ball is probably the Lydia Ball, daughter of Eleazer and Tamson [__?__] Balll, who was born 21 April 1803 in Groton, Grafton Co., New Hampshire.
7. Lydia Ball was born 1802(?) in Groton, Grafton Co., New Hampshire(?).
8. Lydia Ball was born say 1802 in Groton[?], Grafton Co. New Hampshire." 

• Statement (1) conveys the least certainty. Say dates are based on accepted life patterns. This date might be based on Lydia's marriage date and/or the birth dates or ages of her children. Possibly conveys less certainty than probably.
• Statement (2) uses the word "probably" when referring to the birthplace. The author might be basing this on accumulated evidence showing that her birth family resided in Groton at this time. It might be best to rewrite this sentence placing the least sure event last: "Lydia Ball was probably born in Groton, Grafton Co., New Hampshire, say 1802."
• Statement (3) includes a circa date of birth. The explanatory note might include a source that gives Lydia's age at death, something that might be included in the sentence: "Lydia Ball was born circa 21 April 1803, as calculated from age 41 years 3 months 12 days at her death, possibly in Groton, Grafton Co., New Hampshire." The place qualifier is possibly. Perhaps the family lived in multiple locations and the accumulated evidence cannot narrow down a location.
• Statement (4) suggests that there is a piece of evidence that states Lydia's birthplace, and there is evidence providing her exact age at some date. Note that the least certain bit of information, the date of birth, is placed last in this sentence.
• Statement (5) suggests that there is some record of a Lydia Ball birth in Lisbon, New Hampshire, but the author doesn't have enough evidence to conclude that the record refers to the subject. With no explanatory comments, this is a very ambiguous statement.
• Statement (6) suggests that a birth record was found for a Lydia Ball, daughter of Eleazer and Tamson Ball. The use of probably, rather than possibly, suggests the evidence is pointing towards identity with the subject, but the author isn't quite ready to make that conclusion. 
• I wouldn't know how to interpret Statement (7). Would you? If this came from your TMG database, would you know what was meant if you returned to this record after several years' work on a different line?
• Statement (8) suggests that there is some evidence as to Lydia's place of birth, but the source was difficult to read, and the author wanted to indicate that he or she was not sure the transcription of the place name was correct.

The longer I research, the less I want to be pushed into making conclusions when entering data in TMG, and I don't want to publish something that doesn't make any existing ambiguity clear. Just for fun - and to let you see how far my TMG database has come (and has yet to go) - here's a series of screen shots showing the changes in my data entry practice.

This screen shot shows the Lydia Ball detail screen, last edited in November 1993. Note all the question marks, abbreviations, ambiguous notes, etc.

These three screen shots show the gradual changes to Lydia's detail screen as new information was added and evaluated in a recent online research trip.

These images show screenshots of an abbreviated family group sheet report for the Lydia Ball family. Note the variety of qualifiers used. Although ambiguity is noted in this report, a journal report is a better method of explaining the reasoning behind genealogical conclusions, or lack thereof.

Continue with "It's Only a Hypothesis - Part Two (C)"