Posts Tagged ‘census records’

This week’s lesson comes from the Freedman’s Bank Records that I have been recently exploring.

I once heard a lecturer say that up to 60% of the time, people are researching the wrong woman as mother of the children. This example shows the need to prove the father’s relationship to a child separately from the mother’s relationship to the child. What does that mean? Here’s the Freedman’s Bank card for “London Mathies”:

London Bankcard

London Bankcard

London’s bank card dated 8 October 1867 provides the surprise notation that his wife “Martha died in Memphis on Vance St. July 2/67.” Of course, most cards don’t typically include dates of death, so this is a lucky find. Under the section for children it says “Willy Franklin 1 yr 2 mos” which could be interpreted as one or two children until we compare it with his 1870 census household as London “Matthews”:



In this document, his wife’s name is Amanda. With one year old son “Jackson” in the household, we can probably safely conclude that London has remarried and had another son. What these records *together* show is that Amanda is not the mother of the first son “William” or as the bank card calls him, “Willy Franklin.” William’s mother was probably Martha. If we viewed this census record in isolation, we might incorrectly assume this was a man and a wife and their two children. Of course we’ll try to find London’s marriage records to confirm our hypothesis. We could also try to find church or burial records that may confirm the death of his wife and perhaps births or baptisms of the children.

We can’t assume that the wife in any household is the mother of all of the children in the household. We have to prove that relationship separately.

Here’s another one (1910):




If you know how to properly pull every clue from census records, you’ll notice that the little “M2” means that John Campbell has been married more than once, while this is his wife Harriet’s first marriage (“M1″). You’ll also see that Harriet has birthed 2 children, and 2 are living. This implies that the last two children are not Harriet’s children (Thanks for the correction, Rolanda!)

Same thing with this third example (1910):



Samuel’s marriage to Carrie has not produced any children yet; it’s her first marriage and his 2nd (or more) marriage. The two children in the household are probably his from a previous marriage.

Notice that if it’s the husband’s subsequent marriage, the children will not be noted as “step” children because the census records only state the relationship to the head of the household. If it’s the husband’s first marriage and the wife’s subsequent marriage, and she brings children, the children should be properly noted as “step” children, as they are here:

Perry Davis

Perry Davis

Here’s the rub: only the 1910 census requires an “M1” or an “M2” designator for number of marriages. And, the “M2” designator means “married more than once.” It could be a 2nd marriage or a 4th marriage, and it should still say “M2”.


–the 1900 census provides the number of years married and the number of children born and living for the women. It does not provide the number of marriages as shown in this example.

–the 1930 census provides age at first marriage. Doesn’t necessarily mean the person was married at that time to the current spouse.

–the 1900-1940 censuses all require a “D” to be written for divorced; if you see that, be sure to find the divorce record.

As you can see, all of these differences in what information each census provides is critical to understanding and interpreting the document correctly. Incorrectly interpreting the census can lead you astray in your research for years.

It goes without saying that information in census records have high degrees of error rates and should be approached with caution. The censustaker may not have recorded the information correctly or the family member may not have accurately reported their marital status. I have several examples of women marked “widowed” whose husbands were in fact not dead.

It goes without saying that information in the census records should be correlated with other records that illuminate a family.

It goes without saying that people can and did have children before and outside of marriages.

So how can you prove the relationship to the wife as mother of the children? Here are a few ways:

  • Sometimes simple age deductions can rule out the current wife as mother of the children. (ie, most women aren’t birthing children at age 13).
  • If the husband dies first, and the widowed wife now heads the census household, the stated relationship of any children in the household will be to her.
  • Marriage and death certificates of the children can name parents.
  • Estate or probate records after the father’s death may illuminate children and wives.
  • Bible records, church records, military pensions, obituaries and land records are examples of other types of records that may be used to prove a woman’s relationship to household children.

So go back and pull out some of your census records. Ask yourself, for each family unit: Is the Wife Really the Mother of all of the Children? The answer may surprise you.

Read Full Post »

I have recently realized I am utterly incapable of writing a short post. That said, I’d like to think I still have avid readers who value them and take the time to read them when they can. I thank you for that. I just had a wonderful Thanksgiving with my family & hope you all did too.

We have to continue pushing ourselves to learn more and better research methodologies. When we all start out, we’re basically doing name lookups in various record sets. What we found is what I refer to as the “low-hanging” fruit. It’s what the Genealogy Gods use to suck you into this hobby;) The tough stuff comes when the records relevant to your family/area/timeframe have been exhausted, AND, you only know how to look up names. If you don’t learn other ways of “connecting the dots” you’ll have trouble uncovering other relationships. Things will appear to be brick walls, that really aren’t. They may just require a research methodology that has not been learned. And I promise, you can learn them.

One of the critical skills to learn is how to analyze and correlate the information you have. Start spending more time practicing this. Lay out all the data you’ve gathered to answer a particular question such as: Who were Jane Smith’s parents? Who were her spouses? Having a specific question frames your research and allows you to work towards a conclusion. Learn the genealogical standards for evidence evaluation and learn how to pull out clues from each piece of evidence. There are excellent genealogy books and lectures for every conceivable type of record. Ask yourself the pertinent questions: Who said this? When did they say it? How did they know it? Who recorded it and why?  When evidence gives conflicting data (such as birthdates or death dates) learn how to address the conflict. Purchase one of the core genealogy reference books like “The Source” edited by Loretta Szucs or “The Researcher’s Guide to American Genealogy” by Val Greenwood (you can buy them used, but try to get a recent edition).

I often encounter people who have only researched in census records and maybe a few vital records. Think of census records as bookends on a shelf like this:
BookshelfIf you have discovered your ancestor in say, the 1900 and the 1910 census, that’s a great start. But the shelf itself is still empty –you still have a 10-year gap for which you don’t have any data. We must get in the habit of filling up that shelf—which represents our ancestor’s lives– with other information from other sources than just the census records. A lot can happen in 10 years.

Case in point: I have blogged before about my 4th great-grandmother Matilda and the years long odyssey to uncover her complicated roots. Her identity was hidden behind a veil of mis-transcribed records, moves between at least 4 counties and 2 states, and 4 marriages! Of the 4 men she married,  3 died within a few years of their marriage to her. The chart below illustrates her sojourn through the census  years, counties and states, with M1 through M4 representing her 4 marriages:

MatildaTimelineMatilda only appears as a married woman twice in the census (1900 and 1910 to her 2nd husband, Perry Davis). Each of her other husbands died before the other census years rolled around, so she was constantly showing up with a different name in those years as a widow. In fact, I thought there were 2 different women named Matilda.  Those other marriages were almost “hidden.” I gave more details about cracking this case in a previous post.

Had I only looked at census records, this case would never have been broken. There is just too much happening in 10 years time. I had to piece together the information I gathered from state censuses, city directories, vitals, oral history, cemetery records, deed records and more. That process allowed me see the errors in the evidence. It also led me to revisit my own assumptions. Matilda’s first marriage record –an original record—mistakenly recorded her name as “Matida Mealy”, not Matilda Nealy which was her name. The clerk probably heard it as “Mealy”. Simple enough right? But because both her first and last names were incorrect on the document, I could never find that marriage. Eventually, tracing all 4 of her marriages (not just the one to my direct ancestor) led me to the names of her parents, Charles and Lavinia Nealy in Hamilton County, FL. I added another branch to my tree.

I have never felt so proud about cracking a case as I did this one. It affirmed that I’m on the right track in terms of developing my skills.  I still have plenty more ahead of me, so right along with you, I continue to keep on learning.

Read Full Post »


Get every new post delivered to your Inbox.

Join 84 other followers