Sometimes I come across items of genealogical interest.  I once bought a stack of hard-cover genealogy books for 50 cents each.  Perhaps saddest are the old portraits of unnamed and unknown ancestors that probably graced many a farm house.

Today at an estate sale I noticed a banker’s box on the top shelf marked “Genealogy.”  I thought boy oh boy, what treasures can I save from destruction and loss to hopefully find a better home.  I anxiously brought the box down, set it carefully on another box, and lifted the cover.  Oh no!  Shreds upon shreds of paper, as if the contents had been through a paper shredder.  A family of mice had at one time made their home in this box.

All of it ruined.  Hand-written notes, Xeroxed copies of records and certificates.  Nothing but strips and fragments.  Nothing salvageable.  It was obviously someone’s careful work from the time before computers, when everything was done by hand.

The lesson:  Store your genealogical paperwork in rodent-proof containers.  Avoid attics and garages.

Brick by Brick

This is another post in a series about finding the ancestors of my paternal great-grandfather Joseph Voisin¹ (1858-1916). This is a brick wall I haven’t been able to get beyond for several years. Here I chip away a few more bricks from the wall in hopes of discovering a clue.

Perhaps you can help. If you found this post while searching the Internet, chances are there’s something here that piqued your interest. That means you might know something I don’t know. If so, please post a comment. No matter how small, most any information can provide a clue.

In this installment I’ll remove four bricks from the wall.  See also Bricks 1 through 10.

Brick 11: Voisin Family from Bentinck, Grey, Ontario

A Voisin family appears in the 1861 Canadian Census in Bentinck Township, Grey County, Ontario.  John and Elizabeth Voisin had a son named Joseph about 1856.  Although this could be our Joseph Voisin, I’ve not yet researched this family.  They apparently emigrated from England and they are Congregationalists.  Our Joseph Voisin was very likely of German descent and Catholic.

Conclusion:  Additional research is necessary to prove whether or not this family is related to us.

Brick 12: A Voisin in Waterloo County, Ontario

A man with the Voisin surname appears in the 1861 Canadian Census in Wellesley Township, Waterloo County, Ontario.  He was a carpenter, 18 years old, and listed near the Joesph and Catherine (Meyer) Voisin family in the census (see Brick 6).  No first name is given and he appears with no other family members.

If this were our Joseph Voisin, he would have been born about 1843 rather than 1858, which is generally accepted.  It is curious however. Since Joseph wrote in his journal that he was in Hawkesville in 1877 (see Brick 5), perhaps he had actually been there since 1861.

Conclusion:  Additional research is necessary to determine the identity of this Voisin.

Brick 13: Photographs of Joseph Voisin

I have a hunch Joseph Voisin may have been older than he led everyone to believe.  I might be able to draw some inferences by examining photographs of him.

Joseph was a pioneer.  He was one of the first settlers in Nottawa Township, Isabella County, Michigan as that area became available to white settlers.  (Previously it was reserved for Native Americans.)  Joseph was also a farmer. As indicated in an 1880 agricultural census, only four of his forty newly-purchased acres had been improved (cleared for farming). He also was a father to eleven children.

As a pioneer, farmer, and father with all those mouths to feed, he undoubtedly worked very, very hard. Such labor would take a toll on anyone. Therefore, to evaluate how old someone looks in a photograph of that era is difficult.

I’m reminded of a television program, “Frontier House,” in which modern families attempted to live under the same circumstances as 19th century pioneers.  In one episode a man became concerned that he was malnourished.  It turned out he was dehydrated, but his body was also adapting to the meager food supply and daily labor to which he was not accustomed.

So can I distinguish whether Joseph appears older than his stated age? Or, did the rigors of pioneer and farm life cause him to look older than he was?

Joseph Voisin through Time

The first image was taken about 1907 when Joseph was 49.  The middle image was taken about 1912 when he was 54.  The last image was probably taken a year before he died, say about 1915, when he was 57.

The question is, do these images look like someone of the stated age, that is 49, 54 and 57, respectively?  To me he looks older.

Now to be accurate, there are mitigating circumstances.  Joseph did write two recipes for bitters in his journal in the 1890s.  He most likely used these as stomach tonics and therefore he may have suffered frequent indigestion and heartburn.  Also, he died of prostate cancer, which progressively got worse. He probably suffered for at least a couple years with that.  I suspect he may have had other, undocumented ailments besides.

Conclusion:  Joseph Voisin may have been older than generally accepted.

Brick 14: The Story of Two Brothers

One account passed down through the generations is that Joseph Voisin came to America in company with a brother.  His brother settled in the northern part of Michigan’s lower peninsula.

As with many family stories, there are few details to go on, and no substantiated evidence.  I think there may be grains of truth in family lore, but tracking such stories down is difficult.  There are Voisin families in northern Michigan, around Traverse City.  So this story is plausible.

The trouble is, Joseph Voisin died in 1916, well before any of his grandchildren came of age. Therefore no one now living ever met him and stories told by his children become blurry through time.

Conclusion:  Additional research is necessary to determine any connection to the Voisin families of northern lower Michigan.

The saga continues….

1. For source citations and images of the evidence discussed here, please see the Family Group Sheet for Joseph Voisin.

Brick by Brick

I’ve reached an impasse trying to find the ancestors of my paternal great-grandfather Joseph Voisin¹ (1858-1916). It’s a brick wall I haven’t been able to get beyond for several years. If I remove one brick from the wall at a time, I may discover a clue.

Perhaps you can help. If you found this post while searching the Internet, chances are there’s something here that piqued your interest.  That means you might know something I don’t know.  If so, please post a comment.  No matter how small, most any information can provide a clue.

In this installment I’ll remove ten bricks from the wall.  See also Bricks 11 through 14.

Brick 1: Date of Birth

There is some discrepancy as to Joseph Voisin’s birth date. No civil or church birth record has yet been found. A wedding sampler shows the date as January 10, 1858. That is perhaps the most reliable evidence to date.

His age is listed as 24 in the 1880 United States Federal Census, when he was actually 22 if born in 1858. Was he actually born in 1856? The ages listed in census records are notoriously unreliable. In the 1900 census, his birth date is listed as January 1858 and his age is 42. The 1910 census shows age 51, which is close enough to 52.

His death certificate indicates he was 65 in 1916, making his birth year 1850. However, it is apparent the birth date and age at death were originally written in lighter ink and subsequently rewritten in darker ink. This document is therefore highly suspect.

Still, I have a hunch Joseph may have been less than truthful about his date of birth for some reason. He may have actually been older.

Conclusion:  Joseph Voisin was born January 10, 1858 until it can be proven otherwise.

Brick 2: Place of Birth

The key to finding Joseph’s ancestors is locating where he was born. Most all records indicate he was born in Ontario, Canada, while his parents were probably born in Germany. That what’s listed in the 1880 United States Federal Census. The 1900 census indicates Canada, and his parents were born in Canada. The 1910 census shows Canada, and the birth place of his parents was unknown.

In an 1880 declaration of intention to become a United States citizen, Joseph renounced his allegiance to the Queen of Great Britain, who of course reigned over Canada.

His death certificate states he was born in Canada, but again, little credence is given to that document.

It is possible he was born in another Provence, like Quebec, or even in the United States, perhaps across the river in Buffalo, New York. His parents could have moved shortly after Joseph was born. However, there is no evidence of that yet.

Conclusion:  Joseph Voisin was born in Ontario, Canada, but I remain open to other possibilities.

Brick 3: Parent’s Names

On Joseph’s death certificate, his father’s name is listed also as Joseph. His mother’s name is listed as “Not Known.” Due to other errors on this document, his father’s name is considered unreliable.

Conclusion:  The names of Joseph’s parents are unknown.

Brick 4: Netherby, Ontario

In a journal entry, Joseph signed his name and under it wrote, “Township of Humberstone; Netherby PO Ont 1875; 19 November.” Netherby is in Humberstone Township, in Welland County, Ontario. The area is south of Niagara Falls and west of Buffalo, New York. “PO Ont” probably means “Provence of Ontario.”

This means he was 17, and nearly 18 when he was in Netherby, Ontario.  However I can draw no conclusion that he was born or raised there.  He may have been passing through.  It is possible he had just left home and was on his own.

Conclusion:  Joseph Voisin was near Netherby, Ontario when he was about 18 years old.

Brick 5: Hawkesville, Ontario

Another entry in Joseph Voisn’s journal is, “Mister Joseph Voisin commenced to work the 4 of April 1877; H. Otterbein; Hawksville, Ont.” Hawkesville is about two miles north of St. Clements in Waterloo County, Ontario.  Another notation indicates he may have been making $12 a month.  He would have been about 19 years old.

There is evidence of a Heinrich or Henry Otterbein in Waterloo county. His wife Margaret (Steinacker) Otterbein apparently died April 17, 1877, just two weeks after Joseph started work, assuming this is the same H. Otterbein. There is no evidence to support this however.

Conclusion:  Joseph Voisin was probably working on his own as a farm hand or doing odd jobs near Hawkesville, Ontario in 1877.

Brick 6: The Voisin Family of St. Clements

Joseph Voisin was in Hawkesville very near St. Clements, Ontario in 1877 (see Brick 5). There was also a large Voisin family in the same area. This was the family of Joseph and Catherine (Meyer) Voisin. Since our Joseph Voisin was a young man, it is tempting to speculate he was related to the Voisin family of St. Clements.

Frank W. Voisin and Clifton E. Voisin from Kitchener, Ontario researched the genealogy of the St. Clements Voisin family in the 1970s and 1980s. They are convinced there is no connection between that family and our Joseph Voisin. Although Joseph and Catherine Voisin had a son named Joseph Xavier Voisin in 1854, there is clear evidence he married Anna Maria Lehnhart in 1879 and took over the family farm in 1886.

If our Joseph is not a son of Joseph and Catherine Voisin, perhaps he was a cousin or nephew to this family.

Conclusion:  Joseph Voisin’s presence in Hawkesville along with the Joseph and Catherine Voisin family is a coincidence.

Brick 7: Clements Starr

Clement or Clements Starr was born in 1852, probably in Wellesley Township, Waterloo County, Ontario, Canada near St. Clements. His parents were likely Ambrose and Anna Rosina Starr, who were early settlers of St. Clements.

According to a 1975 centennial booklet on the Beal City, Michigan area, Joseph Voisin immigrated to the United States in 1879 in company with Clements and Elizabeth (Fate) Starr and their children. This could be just family lore, but there are several circumstantial facts that support this claim.

According to his journal, Joseph Voisin was working in Hawkesville in 1877 (see Brick 5). Clements Starr was probably born and raised near Hawkesville. Clements was married (in 1875) and would have been about 25 years old while Joseph would have been about 19. Since they were both in the area, they probably met and were friends.

The 1880 United States Federal Census lists Joseph Voisin in the household of Clements and Elizabeth Starr and their children near Beal City, Michigan. After 1880 they owned two adjoining 40 acre properties and were neighbors thereafter.

Conclusion:  Joseph Voisin did immigrate to the United States in the company of Clements and Elizabeth Starr and their children.

Brick 8: Remus, Michigan

According to the 1975 centennial booklet on the Beal City, Michigan area, Joseph Voisin and the Starr family first settled in Remus, Michigan before moving to Beal City. I have not yet found any record of these families in the Remus area.

Conclusion:  Joseph Voisin may have settled first in Remus, Michigan.

Brick 9: Land Deed

A warranty deed from John W. Hance to Joseph Voisin was filed October 3, 1879. Joseph is listed as living in Mt. Pleasant, Michigan at the time. He paid $725 for 80 acres of land near Beal City, Michigan. A deed index entry shows Joseph sold 40 acres of this land to Clements Starr, probably in 1880.

Curiously, the 1880 United States Federal Census lists Joseph Voisin in the household of Clements and Elizabeth Starr, even though Joseph originally purchased the land. Since the Starr family had children, Joseph probably assisted them in building their shelter first.  They may also have made a financial arrangement to purchase the land together.  If Joseph were 21 years old, he probably didn’t have $725, although it was probably all mortgaged.

Conclusion:  Joseph Voisin immigrated to the United States between April 1877 when he was last known to be in Hawkesville, Ontario and October 1879 when the land deed was filed.

Brick 10: Declaration of Intention

On October 11, 1880, Joseph Voisin filed a Declaration of Intention to become a naturalized citizen of the United States. In this declaration, Joseph renounced his allegiance to the Queen of Great Britain. Unfortunately, no record of a later petition has yet been found.

Conclusion:  Joseph Voisin was from Canada or England.

These are ten bricks from the brick wall.  I’m still no closer in determining who Joseph’s parents were, but I may be close.

1. For source citations and images of the evidence discussed here, please see the Family Group Sheet for Joseph Voisin.

Albert and Mary Pohl, about 1909

Who would have thought the National Oceanic and Atmospheric Administration (NOAA) could answer questions about family history?  I used NOAA to help solve a mystery about the immigration of my great-grandparents, Albert G. and Maria “Mary” (Pittner) Pohl.

According to the United States census for the years 1900, 1910 and 1920, they immigrated to America in the early 1890′s. Since each census gives a slightly different year of immigration, I could only conclude that Albert Pohl arrived first, followed a year later by his wife Mary and their children.

With that assumption, I searched for Albert Pohl in the ship manifest records at Ellis Island, New York. There I found an Albert Pohl, traveling alone, aboard the steamship SS Dresden. This ship arrived at the Port of New York on August 23, 1892. His voyage started at Bremen, Germany. An ocean voyage at that time took about 12 days. The ship arrived at New York, where some passengers disembarked, and the remaining passengers continued on to Baltimore, Maryland, where the ship arrived two days later on August 25. Albert is listed twice: First at New York among the passengers bound for Baltimore, and second at Baltimore when he arrived there. Since he was bound for Baltimore, it is doubtful he was processed at Ellis Island.

This Albert Pohl is likely my great-grandfather. If his wife and children were listed in the manifest, it would have made a more certain match. However his age and native country are consistent with the census records. According to my hypothesis, I would expect that he arranged for his wife and children to arrive as he did, at the Port of Baltimore the following year.  Unfortunately there is no record of a Maria or Mary Pohl arriving at Baltimore in 1893.

I decided to check other ports of entry.  Indeed a Maria Pohl did arrive in Philadelphia, Pennsylvania almost exactly one year later.  Her three children matched those recorded in the census. Maria, age 27, and her three children Leopold, 6, Maria, 4, and my grandmother Adelheid, age 2, were aboard the SS Scandia, which departed Hamburg, Germany on August 12, 1893. Mary states she was never in the United States before, and that she had a ticket to her final destination, for which her husband had already paid. She was going to Woodville, Pennsylvania to reside with him there. Woodville is a suburb of Pittsburgh.

But why did her ship arrive at Philadelphia rather than at Baltimore, Maryland?  In support of my hypothesis, I learned the SS Scandia was indeed originally bound for Baltimore. It was diverted to Philadelphia, where it arrived on August 29, 1893. Their voyage was very likely delayed and rerouted due to four storms in the Atlantic during this time.

According to a chart for the year 1893 at the National Oceanic and Atmospheric Administration, one of these storms was a hurricane that hit New York City on August 24 with 85 mph winds. In good weather the ship probably would have reached Baltimore on August 24. The ship’s manifest was prepared for Baltimore, but this was later crossed out and Philadelphia inserted.

1893 Hurricane Chart

Mystery solved!  A hurricane affected the immigration of my great-grandmother Mary Pohl and her children, including my grandmother Ida (Pohl) Stewart.  Their adventure that year was probably re-told in many stories through the years, but those have been lost to history.

So develop a hypothesis, but keep an open mind as to where you can find the answer.  If you have an immigration mystery, check historical weather records.  Visit the National Oceanic and Atmospheric Administration. A would of caution:  Keep digging and be persistent in your search. There are numerous different agencies, laboratories and divisions involved here.  A search of one may not yield results, but a search at another may.

But should you use Source Templates?  I say No! The reason?  Most if not all genealogy programs have yet to get their source templates feature working properly.  Besides minor problems with formatting and punctuation, the most serious issue is:  You will not be able to use your source citations outside of your genealogy program.

I use one program as my main genealogy program.  With it, I maintain my “master” family tree by adding, editing and deleting genealogical facts and information as I discover it.  However I’ve yet to find one program that does everything I want.  One program will print a report another program does not have.  One will print a graphical family tree in a much better format than another.  Another will create a better family tree website.

There are several reasons you will want to transfer your family tree data between different programs.  Even if you do not agree now, there will probably come a time when you do.  Unfortunately if you were to use source templates, other programs will not be able to process your source citations.

When RootsMagic 4 was first released, I bought into the notion of source templates.  I spent a lot of time converting my free-form source citations using its source template facility.  I quickly found many of my citations were incomplete and lacking critical information.  Source templates are a wonderful way of filling in the blanks to construct complete, standardized citations.

I use my own program, GED-GEN, to publish my family tree website.  When I exported a GEDCOM file from earlier releases of RootsMagic 4, I discovered its output included source template information.  This resulted in unreadable source citations on my web pages.

So I modified GED-GEN to handle the proprietary way in which RootsMagic exported source citations.  It worked wonderfully.  But then in subsequent releases of RootsMagic 4, some of the source template information was removed from its GEDCOM file.  Since that time, GED-GEN or any other program for that matter, cannot process the source citations exported in a RootsMagic GEDCOM file.

Why were those template fields removed?  If source citations were exported using its template format, then only RootsMagic, or another program that understood that format, could process and understand the citations.  To “correct” this, later releases simply removed critical parts of the template format.  Now, even if another program understands that proprietary format, it cannot reconstruct the source citations.  What’s more, removing those template fields did not solve the problem.  Source citations cannot be reconstructed in either case.

What is the Problem?

It turns out RootsMagic has a fundamental flaw.  Its source citations are based on source templates.  It exports source templates to a GEDCOM file rather than source citations.  Remember source templates should only be used as a tool to generate proper source citations.  The source templates themselves should not be exported without the accompanied source citations.  RootsMagic doesn’t export source citations at all.

To understand the subtlety of this distinction, let’s look at this issue in more detail.  Here I use RootsMagic 4 as an example.  From what I read, Family Tree Maker 2011 and Legacy Family Tree 7 may have similar issues.  See for example Genea-Musings, The Seaver Source Citation Saga Compendium (February 16, 2011).

GEDCOM Source Records

Traditionally, each source record in a GEDCOM file refers to a single source, and it has a specific title that does not change. Citations (footnotes) that refer to this source all use the same source title. Thus many citations can reference one source. For example, a newspaper is cited by title and issue date:

Daily News. Washington, DC., 25 January 1977.

Introducing Templates

With source templates, the source title may contain place-holders, or fields that can be replaced by different information from different source citations (footnotes). Thus many citations can reference many variations of a single source record. The source record can now stand for many different source titles, not just one title.

For example, rather than a newspaper issue date, a variable date field is specified in a source template:

Daily News. Washington, DC., [Date].

Then, some citations (footnotes) might refer to the 25 January 1977 issue, while others might refer to the 15 March 1995 issue. These use the same source record, but result in different dates, and thus different sources:

Daily News. Washington, DC., 25 January 1977.
Daily News. Washington, DC., 15 March 1995.

A RootsMagic GEDCOM File

In earlier releases of RootsMagic 4, a source record appeared as follows in a GEDCOM file:

0 @S3992@ SOUR
1 ABBR Newspaper, Daily News, Washington, DC
1 TITL [ItemID], <i>Daily News</i>, Washington, DC,
2 CONC [Date], [Details]. [Annotation].
...

Notice the template fields like [ItemID], [Date], and [Details].  A citation (i. e., footnote) that refers to the above source record might appear as follows. Here the [Date] field in the template above will be replaced by 25 January 1977, which is taken from the citation (footnote) here:

...
2 SOUR @S3992@
3 REFN 000651
3 PAGE John Doe; 25 January 1977; Page 2, Column 2
3 _TMPLT
4 FIELD
5 NAME ItemID
5 VALUE John Doe
4 FIELD
5 NAME Date
5 VALUE 25 January 1977
4 FIELD
5 NAME Details
5 VALUE Page 2, Column 2
4 FIELD
5 NAME Annotation
...

However, later releases of RootsMagic 4 removed the source template fields from the title (TITL) in the source record, so that now, the source record appears as:

0 @S3992@ SOUR
1 ABBR Newspaper, Daily News, Washington, DC
1 TITL , <I>Daily News</i>, Washington, DC, .
...

Notice here the empty punctuation in the title (TITL).  The [ItemID] and [Date] fields are missing, but the commas and spaces are still there.  Now there is no way to know where to substitute the date when trying to reconstruct the source title.  This is not a source record.  It is the remnant of a source template.

To correct this flaw, there would have to be multiple source records, one for each source citation that references a different source. That is, a source record that is missing the [Date] field:

Daily News. Washington, DC., [Date].

must really be exported as two different source records.  This is because, in our example, there are two source citations (footnotes) that reference the same source template, and this creates two different source titles:

Daily News. Washington, DC., 25 January 1977.
Daily News. Washington, DC., 15 March 1995.

This requires two different source records to be exported in the GEDCOM file.

(A related problem is the inability to generate a bibliography, without also processing all source citations (footnotes) that reference a given source record.  Traditionally, a genealogy program could simply read the list of source records in a GEDCOM file to build a list of sources for a bibliography.  Since the source records are templates here, the actual information can only be found in the source citations (footnotes).  But again, there’s no way to reconstruct a proper source title because the template fields are missing.)

What’s the Solution?

Problem?  What problem?  There is no problem if you continue to use RootsMagic, and only RootsMagic, for your genealogy research.  The problem occurs when you want to use your family tree data in other programs, or share it with someone who does not use RootsMagic.

One could argue that RootsMagic can read its own GEDCOM file and reconstruct the source citations just fine.  That is true, but only because it uses information it knows internally.  It uses a template identifier to look up the format of the template in another file.  Otherwise, without the template fields in the GEDCOM file, RootsMagic cannot reconstruct its own source citations.  And, hopefully, the format of a template referenced by that internal identifier will never change.  Otherwise a newer version of RootsMagic that reads an older GEDCOM file will misinterpret the older format.

In one solution, RootsMagic could correct its earlier “correction” by restoring the missing template fields.  That solves the problem for other programs, but only if those programs understand the RootsMagic source template format.  RootsMagic would still export source templates, and not source citations.  I doubt many other programs would make provision for the RootsMagic format.  And, it doesn’t solve the problem for older programs that understand only the standard GEDCOM source record format.

Secondly, regardless of whether RootsMagic corrects the flaw in exporting its own source templates, it still does not export source citations in either case.  Clearly that is more important because without properly formed source citations, your source data is useless to other programs.  The solution here is to export multiple source records, one for each citation (footnote) that would “cause” a different title.  As it is now, it simply exports the remnant of a single source template.

A third solution, at the risk of being overly esoteric, is to restructure the source templates themselves.  Remove any template fields from the source titles that would be populated with information from a source citation (footnote).  This is the information that changes based on the citation (e.g., the date), and which causes the title to change.  In RootsMagic parlance, the title would be composed only of “master” source template fields.  Source “detail” template fields would appear in the footnote. That perhaps was the genesis of this design flaw.  The fields in source templates were not divided properly as to which should be in the source record and which should be in the source citation (footnote).

Unfortunately, source templates are not so easily modified.  Existing users would have to make extensive changes to their source citations within RootsMagic.  Where there is but a single source, it must now be rewritten into possibly multiple sources, which in the GEDCOM, would result in:

0 @S3992@ SOUR
1 ABBR Newspaper, Daily News, Washington, DC, 25 January 1977
...

0 @S3993@ SOUR
1 ABBR Newspaper, Daily News, Washington, DC, 15 March 1995
...

What’s my advice?

It’s been nearly a year since this problem was reported to RootsMagic.  There have been many releases since then, but as of version 4.1.1.0, none address this issue.  Clearly there are not enough people yet who want to freely move their genealogical data between programs.

I chose to use RootsMagic source templates, and my data is stuck.  I cannot use any other program without sacrificing all the effort I made in converting my free-form source citations to source templates.  I suspect you will encounter the same problem with other genealogy programs that offer source templates.

Do not use the source templates feature of your genealogy program unless you are sure you can export your family tree data and import it to other programs.  In the meantime stick with free-form (non-template) source citations, but use a separate program or website to generate your source citations.  Then copy and paste the result into your free-form source citation.

Unfortunately, your genealogy program may still not store the citation with its constituent parts, that is with separate fields for title, author and publisher. You may lose the benefit of standardized source records when exporting a GEDCOM file. But, your source citations will be readable. Although an external source template program or website is not as closely integrated with your genealogy program, you will achieve a better result in the long run.  You gotta use the right tool for the job.

Here are a couple examples:

BiblioExpress (free) is a separate program you install on your computer. It keeps a database of your sources and you can automatically generate a source citation to copy and paste into your genealogy program.

EasyBib (free) is a website where you can generate a source citation to copy and paste into your genealogy program.

One of the oldest records I’ve uncovered is a baptismal record for my seventh great-grandfather, Joannis Martin Gossé.  He was baptized March 25, 1691 at St. Remi Catholic Church in Baerendorf, Alsace, France.

What makes this record interesting is the signature (or mark) of his father, my eighth great-grandfather, Christian Gossé, seen as a figure-eight symbol at the bottom.  Years later in 1717 he used the same symbol to sign this son’s marriage record.

Of course it was not uncommon for people of the era to be illiterate.  Records from this time are filled with symbols representing the signatures of parents, brides and grooms.

See the Joannis Martin and Margaretha Gossé family group sheet.

Using Names

Genealogical research presents a problem in that names of individuals are not always known with certainty. You may not yet know an ancestor’s name. If you do, you may not yet know their full name. Names can also change. You may learn the correct or preferred spelling, or you may realize the name you have is a nickname rather than an actual name. As you continue your research, you may find their actual name is completely different from what you thought.

There are multiple ways to refer to an ancestor depending upon the focus of your research. Should the photograph of grandmother and her family be filed under her name, or that of your great-grandparents?  Should her childhood photograph be filed under her maiden name, or under her name from a second husband?

Using Family and Individual Numbers

Most genealogy database programs assign a number to each family in your family tree. Individuals are also numbered separately. One could use such numbers to refer to digital evidence files. But as with names, you may find the ancestor you thought was a child of one family actually belongs to another family. There is also the problem of changing family numbers when a child gets married and starts his or her own family.

Using a Reference Number

I decided to use an arbitrary reference number to identify each piece of evidence. At first I tried to think up a complex numbering scheme that categorized each item according to a number of criteria. For example, the number:

0100-00023-000123-01-0050

was composed of several fields, where the each field held some significance. 0100 might mean the type of item, like a birth certificate. 00023 could be an individual’s number in my genealogy database. 000123 could be a sequential serial number, and so on.

Being a perfectionist, that scheme led me to procrastinate. Not only is it time-consuming to categorize a single piece of evidence according to several criteria, but the discipline required to look up the appropriate codes and properly increment each count is onerous.

The easiest solution is a simple consecutive number, like so:

000654

There’s only one thing to remember, and that’s the last number I assigned so as not to assign a duplicate number.  I add leading zeros to make a six-digit number for consistency.  That makes it easy to know the number is an identification number.  It also assumes I won’t ever have more than a million items of evidence.  A sure bet I think.

Still, the perfectionist in me wanted to sort like items so they would be assigned consecutive numbers as a set. For instance I wanted to group all birth certificates together, then number them consecutively. Of course months later were I to add a new birth certificate, it would have to be assigned a number not contiguous with the others.

As I went through all my items of evidence I had to force myself to just assign a number, without attaching any significance to the value of that number.  Yes, an older item might be numbered after a later item. Yes, just when you thought you collected and numbered a group of related items, another one would turn up to throw off any preconceived notion of numerical order.

My advice? Just assign the next-available number. The number itself serves only to identify a particular item of evidence, nothing more.  It is much easier to refer to an item by number than by any sort of classification or description.

There are of course some subtleties involved in assigning numbers. A couple subtleties include numbering identical items that exist in different formats and deriving new items from an original one. I will address these next time.

I have already begun the process of digitizing the genealogical evidence I accumulated over the years. In the forthcoming series of blog posts entitled “Digital Evidence,” I will describe the system I use to generate, manage and display my collection.

As always, comments and feedback are welcome, especially if you have a better idea!

A “preferred” fact is a best estimate. It may include information from a combination of alternate facts. A single preferred fact is often used in genealogical reports and charts, where listing several alternate facts is infeasible.

For some facts, it is not a matter of having several conflicting facts from alternate sources. Sometimes you must draw inferences from several sources to form a conclusion. Suppose you have a hunch that a particular person is your ancestor. If you evaluate several sources, you may find proof to a reasonable degree of certainty that your hunch is correct. In other words you may be able draw a conclusion even though there are no explicit facts that prove it.

As an example, marriage records often list witnesses, their ages, and their relationship to the bride or groom. Evaluating the marriage records of two sisters may lead you to conclude that one of the witnesses who appears on both records is actually their brother. The age and hometown of this witness may lead you to conclude he is indeed your ancestor.

Therefore, besides citing individual sources for a given fact, you can also cite a conclusion. Simply document the steps that led to your conclusion, and name that as your source. Future researchers can then see your logic and verify it against your sources, and any new sources that may be discovered.

For instance someone’s date of birth taken from their death certificate has a greater chance of being incorrect since their birth happened so many years beforehand. The person filling out the death certificate may only be guessing the deceased’s birth date. The birth date is seldom verified with other official records when the death certificate is filed.

Birth certificates are considered more accurate because they are recorded soon after birth, when everyone involved is sure when it happened. However a clerk generally recorded births in a ledger book. Sometimes these ledgers were themselves re-copied several years later. So even a “birth certificate” is subject to transcription errors and recording mistakes.

Even if a birth record and a death record each point to the same birth date, that date is not necessarily accurate. Two or more inaccurate records do not make an accurate record.

That is why citing a source in genealogical research is so important. You, or a subsequent researcher may happen upon another source in the future that corroborates or refutes a given fact. All sources will again need to be weighed for accuracy before another conclusion can be drawn. Any genealogical fact has an inherent degree of accuracy that is never 100%.