Wrong Source Or Wrong Turn? (Part 2)

In Part One,  I was tempted to visualize my way into a mistake by over-interpreting a perfectly innocent piece of census data: a prime example of the source being right and the perception going astray.

But investigating this family further has produced more confusion, as further investigation often will. Makes you want to retire your Sherlock Holmes deerstalker cap for good.

By carefully looking at data on Martin and Mary Haigney and their family in the period of 1860-1870, it was possible to establish that the big gap in ages between their first and second sons resulted not from a second marriage and a second family, but because three daughters were born and died between censuses. Step by step, I re-traced my way through the evidence:

First steps: U.S. Censuses of 1860 and 1870 and an 1890 affidavit from Martin Haigney’s Civil War pension file, establishing that his first child, Joseph, was born in 1859, with the next two surviving children being  William, born in 1867, and Margaret, born in 1870.

Second step: Re-checking my handwritten family genealogy gold mine, also known as The List. It listed seven children for Martin and Mary: Joseph F.; William;  “Mary I and Margaret I — died in infancy”; Mary “II”, Margaret “II”, and Martin. It did not include birth dates.

Third step: Baptismal records for the Haigney children, transcribed from the register of St. Bridget’s Roman Catholic Church, Watervliet, now in the archives at Immaculate Heart of Mary Roman Catholic Church. St Bridget’s listed baptisms for eight children, not seven:

1. Joseph, 29 January 1859.
2. Mary, 31 March 1861.
3. Joanna, 26 July 1863.
4. Ellen, 10 September 1865.
5. William, 10 November 1867.
6. Margaret, 16 January 1870.
7. Mary Ann, 25 August 1872.
8. Martin, 11 November 1874.

I nodded tolerantly when I saw these entries. The List had been pretty accurate so far. In fact, darn near 100 percent accurate. But now it had missed a kid. And it had listed a “Margaret” as dying in infancy, when clearly that child had to be named either Ellen or Joanna.

Oh, well; even The List is entitled to an off day. And off I went to the …

Fourth step: Finding this family in the New York State census of June 1865. Obviously a very useful resource, since it provides a glimpse of the family midway between federal censuses. Here’s what it said:

1. Martin Haigney, 35, male, head of household. Born in Ireland. Parent of 3 children. Married once. Occupation: Soldier. Place of Employment: U.S. Arsenal. Currently in Army.
2. Mary Haigney, 30, female, wife. Born in Ireland. Parent of 3 children. Married once. Status: Married.  Citizenship status: Alien.
3. Joseph Haigney, 6, male, son. Born in Albany [County].
4. Mary Haigney, [age mark illegible; might be a 4, judging from other 4s on the page].  Female, daughter.  Born in Albany [County].
5. Margaret Haigney, 2, female, daughter. Born in Albany [County].

Fifth step: Huh????????

Somewhere I just know that my Aunt Catherine, compiler of The List, is crowing and saying that’s what comes of thinking you know it all.

Why is the child whose age corresponds to the baptismal register’s “Joanna” called “Margaret” by the 1865 census taker, and by Aunt Catherine’s source for her List?  Which piece of data is wrong?

I can tell you that the church archivist who is transcribing the St. Bridget’s registers mentioned that the recordkeeping can be sloppy. So maybe “Joanna” in the register is an error, plain and simple. Or maybe the little girl was called Joanna Margaret, and the family preferred to call her Margaret.

At the moment, I have compromised in my genealogy records by listing her as Joanna [Margaret]. Will I ever know her name for sure? Mysteries like this are infuriating, and addicting.

Wrong Source or Wrong Turn? (Part 1)

“Sometimes … the actual source is just fine: it’s our perception of that old document that may need a bit of work.”

“Yesss!” said I when this quote popped up as I was Google-Reader-ing the other day.

It comes from AGS fellow and author Henry Z. Jones, who gives a talk called “When the Sources are Wrong” that I’d dearly love to hear someday. [Note: If you're going to the Chula Vista (CA) Genealogical Society meeting tomorrow, you can!]

Mr. Jones’ wise words remind us how easy it can be to take a wrong turn to Genealogy Nowheresville. I’ve flirted with disaster a few times (cough), but never more temptingly than when I was trying to unravel the mystery of an Irish great-great-grandmother’s maiden name.

I’ve written before about the search for Mary Haigney’s birth name. I started with death certificates for the two eldest of Mary and her husband Martin Haigney’s surviving children: my great-grandfather Joseph (1859-1938) and his brother William (1867-1930). They differed on the mother’s maiden name. Joseph’s said it was Mary Mahon; William’s said Mary Carroll.

That earlier post focused on analyzing evidence that was very specific and personal to my family: death certificates, an obituary, a handwritten genealogy and, ultimately, my great-great-grandfather’s Civil War pension file. But there was another, less personal source to consider: census data. And here’s where perception could easily have led me astray from reality.

Conflicting death certificates in hand, I revisited the census data I had on this family, which at the time was limited to the federal censuses of 1860, 1870 and 1880.

1860: Martin, age 28, laborer, born in Ireland, listed at two addresses: by himself in a barracks at the Watervliet Arsenal, and at a dwelling in West Troy (later renamed Watervliet) with his wife, Mary, age 26, born in Ireland, and their son, Joseph, four months, born in New York state.

1870: Martin, age 40, laborer, living in West Troy with wife Mary, age 37, and children Joseph, 11; William, 2; and Margaret, 6 months.

1880: Martin Haigney, age 53, laborer, living in Watervliet with wife Mary, age 50, and four children, William, 12; Margaret, 10; Mary, 8;and Martin, 6.  Joseph F. Haigney, age 21, was living in a boardinghouse across the Hudson River in the city of Troy.

Can you see the tempting wrong turn in this data? I thought the changes in Martin’s household between 1860 and 1870 were a potential red flag. There he was in 1860 with a wife named Mary and a baby son. There he was in 1870 with Mary and what amounts to two sets of children — an 11 year old and two little ones, separated from the first birth by 8 years.

And there I was with two Marys on two death certificates — a Mahon and a Carroll. Could the first Mary (Joseph’s mother) have died, and a second Mary (William’s and Margaret’s mother) have replaced her? A lot can happen in ten years!

Well, a lot did happen — just not that. In my defense, my original theory wouldn’t have been unheard of at the time. But the real story was also sadly common. There was only one Mary, as it turned out, and her name was Mahon. The reason for the big gap between Joseph and William was that Martin and Mary had three little girls after Joseph, none of whom lived to be counted in the 1870 census.

The “two Marys” theory officially died when I obtained a copy of a handwritten family genealogy compiled by one of my aunts. It listed two of the  children who died young, bringing the total of Martin and Mary’s offspring to seven. Then I found a 1958 newspaper story about their daughter Margaret, which asserted that she was one of eight children. Finally, on a trip to Watervliet last fall, I was able to gather the baptismal dates of all of Martin and Mary’s children — and there were indeed eight. The three daughters missing from the 1870 census were born in 1861, 1863 and 1865.

When exactly they died, I don’t yet know.

But I do know that for a time there, I had some perfectly good census data in hand — and was tempted to imagine my way into a perfectly wrongheaded conclusion.

(Coming up in Part 2: Another naming mystery!)