Archive for March, 2009

Here are two photos from my Grandpa’s collection.  Since Grandpa’s photos are almost always of family members, it’s rare to find one with people that I can’t identify.  But these two photos are a complete mystery to me.

The photos show a bunch of young people, some of them in baseball uniforms, and one of the photos is labeled “after the game”.  So clearly it’s a group of young friends gathering after a baseball game, or maybe celebrating a victory.

Here’s the first photo.  (Click on the photo to see a larger version).


And here’s the second photo, showing some of the same people, but with labels that look like surnames.  (Again, click on photo for a larger version).


I really love these photos, and maybe more because they are such a mystery to me.  As usual with these older photos, I love just staring into the faces of the people in the photo and imagining what it would be like to have been there on that day, talking to everybody, getting to know them.

What’s great is that this isn’t your standard old photo, with everybody in the photo stiffly posed and unsmiling.  The young people in the photo look happy, almost flushed with the excitement of the day.  Every face tells a different story: pride, determination, “coolness”, or just plain happiness.

I especially like the couple who appear in the center of the first photo:


These two have wonderful faces.  We’re looking back in time 80 years or more, but with the energy and excitement in the couple’s faces, we can easily imagine them as young friends of ours.

I also like the doubly-stacked, jauntily tilted straw hats.  The silliness hints at the sort of day that everyone was having–just being crazy, goofing around, having fun.

There are other faces that are equally fascinating, like the guy and two girls in the front row:


(I’m half convinced that the girl on the left is my grandmother, but I’ve asked other family members to confirm this).

24 Mar 2009: An Uncle confirmed that this is not my Grandmother–so the mystery continues.

There is something in the faces of these three that also tells a story.  We might imagine the young man as romantically interested in the girl at the left, or the girl on the right as also a close friend of both of them.  The body language might be telling us a lot.  The man has his hand lazily draped over the girl’s shoulder, drawing close to her.  The girl on the right, in turn, drapes her arm around the man’s shoulder and holds his elbow with her other hand.  The image makes me really curious to know who these three were and to understand their connection with each other.

Here’s another face from the past that I love–wearing what looks like catcher’s gear:


What a great face!  The boy appears younger than the other players and clearly seems proud to be a part of the team.

A quick google search turns up a web page depicting what looks like exactly the catcher’s outfit that the boy is wearing:


The second photo also has some wonderful faces to stare into.  Note the two girls in the center of the photo, one of them wearing the catcher’s mitt:


If you look at the large version of the photo, you can also make out “Same gang” written on the face of the photo.

I did a little searching on the name “C. W. Fenners”, but didn’t turn anything up.  My guess is that this is some local business that sponsored the team.  The next step might be to go browse through the St. Paul, MN city directories from the mid-1920s and see if I can’t find a business with that name.  The other avenue to pursue might be to do some searching on the listed surnames to see if I could verify whether all of these people lived in a particular part of town.

In the end, maybe these sorts of photos are the most interesting–the ones that tell us a story, but which contain stories that elude us.


Read Full Post »

Searching for your ancestors in online records has become the rage in the genealogical world over the past couple of years.  And of all of the data that you can search, census records must be by far the most common place to find information on your ancestors.

It’s not all that tough to find your ancestors in the census records, by doing a quick search at ancestry.com or familysearch.org.  And it’s quite a thrill to find the entire family of one of your ancestors, with each member of the family listed, along with their age at the time the census was taken.

What gets interesting is when you try to use census records to provide a bit more information on the birth date of one of your ancestors.  If you already know the birth dates for family members that you are finding, the age information in the census record can just help ensure that you’re looking at the right person.  But if you don’t yet have a birth date, or have conflicting information, it’s possible to use the person’s age at census time as just another piece of evidence to help you arrive at the correct birth date.

Be Careful!

Be careful here.  Birth date information deduced by looking at the age at time of census is not a primary source.  In fact, it’s probably one of the least reliable pieces of evidence for the purpose of calculating a birth date.  Think about the circumstances and the setting at the time the information is gathered.  A census taker goes from house to house, asking people verbally to tell them who lives at that house and how old each resident is (among other things).  The validity of the answers, especially when it comes to age, depends on so many things—how carefully the census taker records what he/she hears, whether the resident actually knows the age of the person in question, and whether they take the time to remember the age of each person in the house.  It’s also probably pretty common for people to have lied about their age.

The quality of the data was also likely dependent on who answered the door.  If Grandpa answered and he was eager to get rid of the census taker, he may have rattled off a bunch of names and ages quickly, without worrying whether or not he got them right.  On the other hand, if the mother of a family answered the door, odds are pretty good that she’d correctly remember the ages of her kids.

Calculating a “Birth Date Range”

It’s still useful to look at census data and to use it to calculate a range of possible birth dates, based on the ages of the people listed.

So how do we calculate a range of possible birth dates, based on what we find in the census record?  It’s as simple as subtracting the listed age from the date on which the census was taken.  The date that you get becomes the “latest birth”, or the most recent possible date of birth, given that age.

For example, let’s say that my grandpa was listed as being 18 yrs old in the 1920 census, and the census sheet in question lists the census date as being Jun 1, 1920.  In other words, on Jun 1, 1920, the census taker stopped at the house and was told that Ted Sexton was 18 yrs old.  This means that the latest that Ted can have been born was Jun 1, 1902.  This would be Ted’s birth date if he had just turned 18 on the day that the census taker stopped by.  The other way to look at it is that this is the youngest that Ted could be, if the stated age was correct.

At the other end of the spectrum, Ted’s earliest possible birth date is Jun 1, 1901—one year earlier than the latest date mentioned above.  If Ted was born late in the day on Jun 1, 1901 and the census taker visited early on Jun 1, 1920, Ted would still be 18 years old—turning 19 later that day.  So this is the earliest possible birth date for Ted and the oldest that he could be, if the stated age was correct.

So we end up with the following possible range of birth dates for Ted:

Census Date:  6/1/1920
Stated Age:  18
Birth Date Range:  Between 6/1/1901 and 6/1/1902

So Ted was born sometime between 6/1/1901 and 6/1/1902—but only if he was truly 18 years old on 6/1/1920.  Remember—although we end up with a precise range of dates, our original census data should still be considered fairly unreliable.  So this date range should just be treated as evidence—nothing more.

Factoring In a Second Census Record

Now that we have one possible range of birth dates for Ted, we can go find another census record and factor in the information that we find there.  Let’s say that we now move to the 1930 census and we find the following information:

Census Date:  10/15/1930
Stated Age: 28
Birth Date Range: Between 10/15/1901 and 10/15/1902

We use the same method as before to calculate a possible range of birth dates, given the stated age.

But now we can take this one step further.  If both ages were stated correctly (and that still might be a big “if”), we’ve further reduced the possible range of dates for Ted’s birth.  Here’s how it works:

Range #1:  Betw 6/1/1901 and 6/1/1902
Range #2:  Betw 10/15/1901 and 10/15/1902
Calculated Range:  Betw 10/15/1901 and 6/1/1902

We’ve gone from a 12-month range to an 8-month range, narrowing in on Ted’s actual birth date.  Note that all we did here was to look at where the two ranges overlapped and calculate the final range as that overlap, or the intersection of the ranges.  For example, range #2 said that Ted could not have been born before 10/15/1901.  This means that part of range #1 has just been ruled out—he could not have been born between 6/1/1901 and 10/14/1901.  In a similar way, the end of range #2 is also chopped off.  In the end, dates between 10/15/1901 and 6/1/1902 are the only dates that fall within both stated ranges.

A Tool to Help You Do These Calculations

I’ve put together a simple web-based tool to help you do birth date calculations like the ones that we just ran through.

You can find the tool by clicking on the link below (or entering it into your browser):

The Census Birth Date Calculator – www.famthings.com/census/default.aspx

Getting Started

Once you go to this site, you may see a little icon indicating that you need to first install Microsoft Silverlight:


If you see this logo, go ahead and click on it to install Silverlight, which is a browser plugin that is required in order to run the Census Birth Date Calculator.  The install should go quickly.  When it completes, just refresh the web page and you should see the calculator.

The Census Birth Date Calculator

Here’s what the main body of the calculator looks like when you start:

calcemptyThis is pretty self explanatory.  To use the calculator, just enter each of your census data records, one at a time, and click the “Add to List” button.  Enter the date on which the census was taken in the “Census Date” field and enter the reported age of the person in question in the “Age at Census” field.

If we use the calculator to enter the census data for my grandpa, from above, we see the following after entering the first date:

calc2Notice that the calculator spits out the Earliest/Latest birth range for this piece of census data: 6/1/1901 to 6/1/1902.  It also shows a “calculated birth range” below the table that is the same range (for now).

Now let’s add our second piece of census data for Ted.  Here’s what we end up with, after adding the new information to the list:

calc3Notice that the calculator now lists out all three ranges of interest—the individual range for each of the two census records, and then the calculated (combined) range.  As we showed earlier, the calculated range represents the overlap of the other two ranges.  Based on the data that we started with, this is our current best guess as to Ted’s birth date.

When Things Go Wrong

As we said, ages reported as part of census data are notoriously unreliable.  Sooner or later, we’re going to end up with a collection of dates that lead to inconsistent results.  In other words—they can’t possibly all be true, because there is no single birth date that would fit into each of the birth ranges.

Here’s an example.  Suppose we find one more piece of census data for Ted.  This time, we find him in the 1905 Minnesota state census.  The date of the census is 6/12/1905 and Ted’s stated age is 4.

Now we’ve run into problems.  Here is what the calculator tells us:

calc41Notice that the Calculated Birth Range at the bottom of the list now tells us that it’s not possible to calculate a range from these three census samples.  They cannot all be true at the same time.  Notice that the conflict arises between the age-28 and the age-4 entries.  They have ranges which don’t overlap, which means—they can’t both be true.

At this point, the calculator provides a little help for you.  If you like, you can play around with the data by including/excluding one or more of the pieces of data and seeing how that affects the calculated range.  For example, if we had reason to believe that the age-28 entry was incorrect, we could uncheck the Include column for that entry and we’d get:

calc5We have a valid range once again, which just means that the age-18 and age-4 entries do not conflict.

What Does It All Mean?

In the end, this calculator isn’t likely to give you a final answer as to the birth date of an ancestor.  It might help you narrow down the range of possible birth dates, but only if you believe in the validity of all of the census data that you find.

The other use for the birth date calculator is that it can help you zero in on which census records are suspicious, i.e. possibly inaccurate.  If you have five records that are all consistent, and then a sixth that would generate a range of dates that don’t overlap at all, then that sixth date is probably suspect.

So feel free to use this calculator to help you calculate some possible birth date ranges, based on census data.  Let me know how it works for you and if you have any suggestions for how I might improve upon it.

Read Full Post »