Dealing With Human Error on Genealogical Documents

An interesting point came up in a Facebook genealogy group yesterday. The clerk who hand-wrote your ancestor's birth, marriage or death record may not have been that skilled.

• He may have misspelled a word—leaving you to try to translate a word that isn't a word.
• He may have made an accidental substitution—adding the wrong sibling's birth record to a set of marriage documents.
• He may have recorded the midwife's last name as the baby's last name.

What do you do with this messed up documentation?

In an earlier article about disagreeing documentation, I spelled out my two cardinal rules:

1. The earliest recorded document is probably correct.
2. Some documents are more official than others.

Sometimes you need to:

• Make an educated guess
• Use the scientific method
• Gather as much information as you can

Then see if logic tells you what the truth must be.

Here's an example. Consider these data points:

• Michele Petruccelli was born in Baselice, Italy on March 8, 1800 to Costanzo Petruccelli and Brigida Ciusolo.
• Michele Petruccelli was born in Baselice, Italy on September 11, 1802 to the same parents.
• Two babies born to the same parents, with the same name. Logic tells us the first Michele must have died before the second Michele was born.
• No death records are available for 1800–1802, so we cannot verify the first baby's death.
• In 1828 Michele (born in 1802) married Mariarosa Mattia. She died in 1828.
• In 1830 Michele (born in 1800) married Veneranda Pozella.

The marriage records overlook the fact that Michele is a widower.

If logic says there were not two sons growing up in the same family with the same first name, then both the 1828 marriage and the 1830 marriage must belong to the younger Michele Petruccelli—the survivor who was born in 1802. This hypothesis works because the two marriages do not overlap.

This means there was a clerical error. In 1830 when widowed Michele Petruccelli married for the second time, a clerk accidentally used the birth record of the deceased Michele Petruccelli.

For further proof, I took another look at the 1829/1830 marriage records. It says that Michele was 23 when he received permission to marry. That fits 1802 Michele better than 1800 Michele.

Of course it's still wrong. He was 27!

The document does not say that Michele is a widower, but it does say his bride is a widow. In the full set of marriage documents for Michele and Veneranda, there is no mention of Michele's first wife Mariarosa Mattia.

This is also a mistake. It's really quite an oversight!

Mariarosa Mattia's death record clearly states she was the wife of Michele Petruccelli, son of Costanzo.

The death record for Michele's first wife leaves no doubt who her husband is.

So what would you do? Michele and Mariarosa were married only eight months when she died, so they had no children. Michele and Veneranda also had no children though they both lived past the year 1860. There is no more evidence.

I'm convinced the clerk made mistakes in 1829/1830. The Michele born in 1800 died before 1802. The younger Michele grew up and married twice.

So, having exhausted all resources and finding that logic is on my side, I'm going to update my family tree.

I'm going to say that the first Michele Petruccelli died before the second was born on September 11, 1802. And I'm going to give the bride, Veneranda Pozella, to the second Michele Petruccelli.

By the way, Michele Petruccelli is the brother-in-law of the sister-in-law of my fourth great uncle whose name is also Petruccelli. It's a small town.

You may also enjoy:

• How To Run Your Genealogy Research Like Clockwork
• How to Find More Great Grandfathers for Father's Day
• What Story Does Your Ancestor's Job Tell You?

How to Build a Professional-Quality Family Tree

I began writing this blog to encourage other genealogy hobbyists to take their family trees to a new level. My first several articles focus on the basics of a reliable, valuable family tree:

1. Gather multiple pieces of evidence for each fact. People make mistakes. It could be the census taker or the person providing information for someone's death certificate. Because of human error, one piece of evidence does not make proof.
2. Cite your sources. It's critical to be able to retrace your research steps. If you cite your source for a fact as the 1930 U.S. Federal Census, give the location right down to the sheet number, and provide a link to where you found it online, it's reproducible. If it's reproducible, it's more reliable.
3. Analyze your facts for discrepancies. My family tree has many people with the exact same name. I found a situation where I gave the same birth date to two different men with the same name. I needed more investigation to fix the error.
4. Ensure you're recording facts for the right person or people. For example, are you sure the census form you're looking at is for the specific family you think it is?
5. Research historical events as they pertain to your ancestor. One of my ancestor's hometown changed its name after World War II. If I hadn't discovered that, I wouldn't have been able to visit the town years ago. And I wouldn't be able to find the town's vital records today.

It all boils down to this: You've got to do the legwork. Do not trust information that falls into your lap. Use someone else's tree, for example, as a series of leads for you to investigate.

The New York City Municipal Archives at 31 Chambers Street in downtown Manhattan. It's the Taj Mahal of research for New York City families.

How Professional Genealogists Work

Professional genealogists follow a standard of proof to deliver accurate research to their customers. Maybe you've seen references to the Genealogical Proof Standard.

Let's look at how the five elements of this standard can improve your family tree research.

1. Conduct a reasonably exhaustive research. This goes back to my first point above of gathering multiple pieces of evidence. It also involves conducting searches that may not yield any results. For example, let's say you expected to find a family at a particular address in the 1930 census, but they're not there. To be thorough, you may need to go through every page in that census. If you still haven't found them, you may need to check out the surrounding enumeration districts.
2. Maintain complete, accurate citations to the source or sources of each piece of information. This was my second point above. You want to make your steps retraceable and show the quality of the sources. For example, the 1930 U.S. Census seems more reliable than "my cousin's father". Citing your sources can also show how many sources you've used. Several excellent sources make a fact very reliable.
3. Test your information. This relates to my third point above about analyzing your family tree for discrepancies. If you find an error, you may need to test the validity of one of your sources.
4. Resolution of conflicts among evidence items. This would be the logical conclusion to your analysis of discrepancies. When you resolve a discrepancy, make a note about the steps you took, and why you made your decision.
5. A soundly reasoned, coherently written conclusion. When you resolve a discrepancy or find something in your tree that you don't have a lot of confidence in, make notes. I like to flag questionable facts in my family tree by adding a bookmark (a feature in Family Tree Maker) and a note. This alerts me to facts with a lower level of confidence so I can treat them with the proper amount of faith.

You may not have the funds to hire a professional genealogist. You may prefer to do the work yourself because the hunt is what makes it so enjoyable.

But if you can adopt these practices, you can have a professional-quality family tree.

And speaking of genealogy methods:

• Let a Few Quick Wins Reinvigorate Your Genealogy Research
• How to Visualize Your Ancestor-Finding Progress
• Find Those Stubborn Genealogy Documents with Creative Searches

How the Scientific Method Can Help You With Your Family Tree

Biology. Anthropology. Archaeology. They're all sciences, and lots of sciences end in -ology: "the study of". Genealogy is the study of lines of decent or development. Another science.

So it makes sense to use the well known Scientific Method in genealogy.

But how would that work? Well, the Scientific Method—principles and procedures for the systematic pursuit of knowledge—includes these steps:

1. Ask a Question. When were my great great grandparents married?

2. Construct a Hypothesis. Get very specific by looking at the genealogy facts you have. Do you know what year your great grandmother was born? Do you know when her parents were born? Were they old enough to have an earlier baby?

3. Test Your Hypothesis. When it comes to family trees, your research is your experiment. You can search available records for the answer to your question. I know my great grandmother was born in 1856. I know her parents' names and ages. My test is to search for other babies in 1855, 1854, 1853, etc., until the couple was too young to have children.

4. Analyze Your Data and Draw a Conclusion. Let's say you found more babies and then they stopped. You may conclude that your great great grandparents married at least nine months before the birth of their first child.

Let's see the Scientific Method in action on my family tree.

My first clues led to my question and hypothesis.

Ten years ago a friend in Italy went to the town hall in my ancestral hometown on my behalf. He requested a few birth, marriage and death records for me.

The town hall typed some of the important facts onto new forms. My friend sent copies of these forms to me.

I learned that my great grandmother, Marianna Iammucci, was born on January 1, 1856. Her parents names were not included.

A few years later I spent an extraordinary amount of time in a Family History Center poring over microfilmed vital records from this town. I went to the 1856 birth records and found Marianna Iammucci. I learned her parents were Antonio Iammucci born in 1814 and Annamaria Bozza born in 1815.

My question was, "How many brothers and sisters might she have had, and when was the first one born?"

Her parents were ages 41 and 40 in 1856. I hypothesized that there could be several more children born before my great grandmother. Since women in Italy at that time went well into their 40s bearing children, there might be some brothers or sisters after her, too.

Time to test my hypothesis. I checked the indexes of each year before 1856. While looking for babies born to Antonio Iammucci and Annamaria Bozza, I found:

• Luigi Maria born in November 1852
• Giovannangelo born in July 1849
• Leonardo Antonio born in December 1846
• Leonardo Antonio born in August 1845 (who died in February 1846)
• Mariangela born in March 1843

The babies seemed to stop there.

Analyzing these births, I concluded that Leonardo and Annamaria could have gotten married as late as June 1842.

Now we've come full circle! My conclusion about their June 1842 marriage became my next hypothesis. I went straight to the 1842 marriage index.

I found them. They were married on June 12, 1842. Success!

I've spent many a weekend jumping from person to person in my tree. It's unstructured fun. But it's the Scientific Method that gets me focused and produces excellent results.

Start thinking of your genealogy research as the science that it is. My hypothesis is you'll be very pleased you did. (And my conclusion is that sentence was horribly corny.)

You may also enjoy:

• How I Gained 2 More Generations in 1 Day
• How to Visualize Your Ancestor-Finding Progress
• What To Do When Your Last Name Is So Common

What To Do When Your Last Name Is So Common

All my direct ancestors had Italian last names. I'm lucky to know in exactly which small towns they were born. And the hometown is the key to everything.

My name of Iamarino is found in only 10 towns.

Some of my Italian last names (or cognomi, in Italian) are rare. They're specific in origin to a small geographical area. The name Iamarino barely existed outside of my ancestral hometown of Colle Sannita.

But some of my Italian last names are about as rare as Smith or Brown in America. According to an Italian surname search site I like to use, you can find my family names of Leone and Caruso EVERYWHERE.

It would be impossible to identify my Leone or Caruso lines without knowing where my great grandparents were born.

There may not be a word strong enough to emphasize how important it is to know your ancestor's hometown. Critical. Crucial. Imperative. Nope—it's more important than that.

If your foreign ancestors emigrated to the U.S., Canada, Australia, the U.K., etc., cross your fingers and hope they arrived at a good time. Early ship manifests didn't capture much information about the passengers.

My name of Leone comes from a couple more places.

If your ancestor arrived at Ellis Island, which was open to immigrants from 1892 to 1954, you're in luck. If you find your ancestor, you should be able to learn the name of their hometown.

I may tell you I'm from New York City, but that's because I'm pretty sure you've never heard of the exact town where I grew up. But my grandfathers never said they were from Naples. They let us know very strongly that they were not Napolitani.

They were very proud of their small towns of Basélice and Colle Sannita. Both towns are in the middle of nowhere, about a 90-minute drive from Naples.

So before you begin chasing the wrong family, you must nail down that hometown.

Find out how. Read about other ways of finding your ancestor's hometown in Where Did Grandpa Come From?

You may also enjoy:

• Searching for the Missing Link in Your Family Tree
• Who Are These People?
• Does Independence Day Make You Think of Your Ancestors?

Work in Batches to Strengthen Your Family Tree

Do you want to make your family tree accurate, reliable, and highly credible? There are many things you can do:

• Add descriptions to your images.
• Be consistent with addresses.
• Cite your sources accurately.
• Choose a style and stick with it.

I know it can seem overwhelming—especially if you started your tree long ago or you can only work on it now and then.

But if you divide and conquer your tasks, working in batches, you'll see real and valuable progress. If you gang-up your tasks, you'll save time and gain consistency.

Here's what I mean.

Add Descriptions to Images

Step through each image in whichever family tree software you use, focusing on one type of image at a time.

In Family Tree Maker, I can sort my images by type (because I clicked a checkbox to categorize each one). Now I can go one-by-one through each census form image, for example, and include important information. I've chosen to note which lines a family is on, and everything you'd need to find the original image:

• town, county, state
• enumeration district, city ward, assembly district
• page number and image number if it's part of a set
• a URL on ancestry.com or familysearch.org.

Great annotations make your facts reproducible and verifiable.

Be Consistent with Addresses

Your family tree software may help you validate a place name when you are typing. Take advantage of that feature if you have it.

Otherwise choose a style for entering place names, verify them on Google Maps, and stick with your style. I prefer to include the word County in my U.S. place names. I think it seems confusing (especially to non-Americans) to have something like "Monsey, Rockland, New York". That's why I consistently use this format: "Monsey, Rockland County, New York".

Entering the address shows me exactly where my great grandparents lived.

Cite Your Sources

This will be blasphemy to some of you, but I do not like excessively long citations for the sources of my facts. I use a short format each time:

• 1930 U.S. Federal Census
• 1915 New York State Census
• New York Passenger Lists, 1820-1957

There isn't any question which collection I'm citing. Plus I put the full description and unique link on the image's notes.

No matter what format, do include your sources. It's absolutely key to the strength and reliability of your data.

Be Consistent in Everything

Do you always use the same date format? I prefer dd-Mon-year (24 Sep 1959) because you should understand it no matter where you're from.

Do you always capitalize last names? I don't, because I think you lose something with compound names like McCartney or deBlasio. But stay consistent.

Do you always spell out every word in an address? I do because I feel it leaves no room for misinterpretation.

For example, my mother was born at 260 East 151st Street, Bronx, New York. If your native language were not English, E. 151st St. would be more challenging than East 151st Street.

Do you have a preferred style for descriptions of immigration facts? That may seem like an awfully granular thing to call out, but I like to add specific information to these descriptions. Here's my format:

For each ship manifest I record the date the ship left as an Emigration fact (for a person's first voyage) or a Departure fact (for subsequent voyages). In the description I follow this format:

"Left for [destination city] on the [ship name]."

Then I record the date the ship reaches port as an Immigration fact (for a person's first voyage) or an Arrival fact (for subsequent voyages). In that description field I follow this format:

"Arrived [with which relatives] to join [which relative] at [address], leaving his [relative] in [hometown]."

Now think about your family tree. Where do you feel your data needs the most care? Is it your sources? Images? Particular facts?

Pick one and dive in. Work your way through as many people as you can in one sitting to ensure absolute consistency. Make notes about the style decisions you've made so you can stick to them.

By picking one type of task and working hard to plow through it all, you will see overall improvement in your tree.

That accomplishment should inspire you to pick your next subject and get busy strengthening your family tree. Then you'll be ready to grow your family tree bigger and better.

How Do I Get There From Here?

Have you ever looked at a document in your family tree, let's say a census sheet, and realized you also need a family that lived on the same block? How do you get back to that census collection online (so you can see the surrounding pages) when all you have is the one document you saved?

You can look at the top of a census sheet and gather enough information to help you get to the general area where you want to look. For example, this 1930 census sheet is from Girard City, Trumbull County, Ohio, enumeration district 78-45, sheet 16B.

Top of a census sheet

With those facts I was able to use ancestry.com to easily drill down to the right link, containing 36 images. Because I knew it was sheet 16B, I was able to go right to image 32 and find the exact page I wanted. I can then go page by page to look for a related family that I believe lived nearby.

Searching for a particular census

But it isn't always that clean and easy. The top of the census sheet might be hard to read or the information incomplete.

Other documents are harder to rediscover, such as a ship manifest. This is the top of my grandfather's ship manifest from 1920.

A ship manifest with no ship name

It tells me that he arrived in New York on 29 November 1920, but what was the name of the ship? I can find a particular ship arriving on a particular day on ancestry.com, but if more than one ship arrived on that day, I may have a lot of images to look at.

To allow myself—and anyone who feels they may have a connection to my tree—to rediscover any of my saved documents, I add enough detail to the image in Family Tree Maker to make that search easy.

For a census sheet I indicate the line numbers to look at, the city, county and state, the enumeration district, sheet number, and image number, which can be a real time-saver.

Adding enough details to enable anyone to locate the original

When I decided to add this information, I spent a whole weekend updating every census sheet in my tree. Now I simply add the information the moment I add the new image. It's a practice that will pay off, and absolutely fortifies your family tree.

Case Study on "How Is That Possible?"

Here's a lesson that supports my earlier post, "How Is That Possible?" When my recent post about Italian marriage records led me to discover a mistake I had been making, I spent three solid days correcting my tons of such records in my enormous family tree.

The work was tedious, but after a while I realized that this change I was making—reclassifying certain dates as "marriage license" rather than "marriage"—provided the answer to a question that had come up a number of times.

I had quite a few men (these are small-town Italian men in the 1800s) who had gone through the process of posting their intention to marry a woman and then seemed to marry her, but went through the same process with another woman a month later.

What was going on? Divorce was not a thing, and the first wife had not died. In fact, I had proof that the first woman then went through the process with another man and married him.

Once I learned that they had not gotten married, but had merely obtained consent to marry, it became clear: The first couple intended to marry but something prevented the marriage. Each of them was then paired with someone else whom they did marry.

See When is a Marriage Not a Marriage?

I feel this corrected and more logical information makes my family tree even more solid. So I ran Family Tree Maker's built-in error report and uncovered a page full of birth date discrepancies. Some people had two birth dates from conflicting resources, while others had an original placeholder date that had been superseded by documented facts. So I was able to clean up all of the errors.

I even figured out and fixed the error I highlighted in How Is That Possible? where I had two Michele Leone's with the same birth date. The date belongs to my grandfather's first cousin; the other Michele Leone is a more distant relative born a different year.

Now I'm itching for other discrepancies to fix!

How Is That Possible?

They say the "unexamined life is not worth living". I say the "unexamined family tree is not worth publishing"!

Check Your Facts

Occasionally you need to analyze your family tree to see if anything looks illogical. Your family tree software may alert you if you’ve entered facts that are impossible, such as a woman giving birth when she’s a little girl or after she’s dead. But it won't alert you if your facts show a person living in two different states at the same time.

See why I think Family Tree Analyzer may be perfect for you.

Don't leave impossible facts in your family tree.

A common name in my family,
but 2 with the same birth date?

Recently I was adding census data to my family tree. One set of facts said the wife came to America two years before her daughter.

If this were true, it would mean she left her infant daughter in Italy, came to join her husband in America, and didn’t send for her baby for another two years.

That's highly unlikely and can only be proved or disproved by finding the mother and daughter’s immigration records.

These types of logic errors are what I frequently find in other people’s family trees—a big reason why I never accept someone else's research without seeing the documentation. Many times, these errors are difficult to spot and difficult to solve.

I like to put a bookmark on people with a logic error so I can quickly see where more investigation is needed.

See what type of reporting features your software may have. Whatever tools you can use, your tree will benefit from a logic scrubbing.

Where Did I Find This?

A Lack of Sources Can Ruin Your Tree

To give credibility to your genealogy facts and make your family tree stronger, you need good annotation.

Describe the source of each bit of information well enough that anyone can retrace your steps and find the same information. That includes:

• Name
• Birth date
• Birth place
• Marriage date
• Death date
• Death place
• and more.

For example, if you haven't found a ship manifest documenting a person’s immigration to America, but the 1920 census states that they arrived in 1905, be sure to cite the 1920 census as the source of that tidbit.

It's clear that the 1920 census is not as reliable as an actual ship manifest when it comes to immigration, but at least we know where that data point came from.

Written proof is more trustworthy than a family story passed down to you.

Myth destroyed. Not our uncle after all!

This seems like a good place to tell my passed-down family story that turned out to be 100% false. My in-laws fully believed they were descended from the brother of the captain of the Titanic.

I met Grandmother Lillian who told the story of being Captain Smith’s brother’s daughter. She was ashamed of the fact that her uncle lost so many lives at sea. It clearly pained her.

The problem is Captain Edward Smith had no brothers. He had a half-sister, but there were no other Smith boys in his family. How could Grandmother Lillian be so wrong?

I decided to see if Grandmother Lillian’s father was Captain Smith’s first cousin rather than his brother. Unfortunately, this was another dead end. No Smith boys.

This story illustrates how much you need to show exactly where your facts came from. Captain Smith’s would-be niece is no longer alive, so we can’t ask her why she believed he was her uncle. Without proof, we’ve got nothing.

Think about this: Would you want to grab someone else’s family tree and attach it to your own when a goof like this calls their entire tree into question?

Do your due diligence. Cite your sources. Here's a great reference on citing sources from FamilySearch.org.