GenealogyBank has archives for newspapers and obituaries as well as other collections. This article covers my best tips for searching the Newspaper archives on GenealogyBank.
Simple And Advanced Search On GenealogyBank
Before I get to the tips, I’ll run quickly through the search interface on the website. The basic search interface on GenealogyBank is very simple:
You may be surprised that the only input on this page is for last and first name (and the first name is optional).
Those extra filters are on what GenealogyBank calls its Advanced Search interface. Simply click the “advanced” link and you’ll see the search fields for dates, locations, and specific newspapers.
It’s reasonable to assume that you should always narrow your searches with a date range and location.
However, there is a major drawback to that approach with newspaper searches. I’ll explain this in my first tip.
Tip #1: Understand Why Newspaper Searches Are Different To Other Record Archives
If you know the difference between indexed records and OCR, then you can skip the tip. If those terms aren’t familiar, then read on.
Have you searched birth and death records on websites like FamilySearch.org, Ancestry.com, or MyHeritage.com? These sites use “indexed” record collections.
GenealogyBank.com also has indexed collections (the Social Security Index and many obituaries). But the newspaper archive is different.
Searching “indexed” records
When you search birth and death records, you are not searching through images of the original documents and certificates.
Instead, each document has been analyzed to extract the typical fields and keywords for genealogy: names, dates, and locations.
Many documents don’t have every exact field on the page. For example, death certificates may not have a year of birth. Instead, they have the date of death and the round number age of the deceased (e.g. 74).
There’s no place where “birth year” is noted. But when you’re searching through death records (not obits or notices), an approximate birth year is usually successful at narrowing down the results.
This is because the analysis process (known as indexation) has used logic to infer an approximate birth year. And the indexation adds keywords to the record that aren’t on the original document.
Because birth, marriage, and death certificates tend to be in the same format within legal jurisdictions, it’s possible to automate the indexation process.
For example, the scanning software expects that the death date is in the top left corner in Louisiana certificates. I’m making that up, but you can understand how indexing standardized documents can be automated.
Now, think about newspaper content and how your ancestors may be mentioned somewhere on the page. Scanning the page can’t target expected positions!
Instead, the scanning process works with the entire page. It uses OCR (optical character recognition) to convert the shapes of the characters on the page to digitized text (i.e. words).
If no further indexing is performed on the OCR scanning, then you are searching a wall of text.
That has implications for how to search newspapers on GenealogyBank.com. The next tip gets into this.
Tip #2: Don’t Start With Extra Filters On Your Newspaper Searches
Let’s say that you know your ancestor Anna Gamble was born in 1891 and lived in Westfield, New Jersey.
GenealogyBank.com offers a Simple Search that includes only the name. But wouldn’t it be better to use the Advanced Search to include 1891 and exclude other Anna Gambles who were born in different years?
But what if an article in a 1925 New Jersey newspaper mentions Anna Gamble as being one of several women escaping from a factory fire?
If you ask GenealogyBank to include 1891 in your search, the chances are that it won’t find the word “1891” in the text of this news article.
This is why GenealogyBank steers you to start with just a name on your newspaper searches.
Filters are better with obituaries
This doesn’t necessarily apply to Obituaries as GenealogyBank has done further indexation on the obituary scans.
If the obituary says that the deceased had reached her seventieth year when she died, then the indexed obituary record will hopefully include an inferred (and approximate) birth year.
Tip #3: Take Advantage Of The “Exclude” Feature
Let’s keep using my Gamble ancestors as an example.
Unfortunately for me, a name search brings back many interesting articles condemning the frequenting of dens of disrepute i.e. places where people went to gamble.
I also get hits that mention the venerable old institution of Proctor & Gamble.
This is where the “Exclude Keyword” feature is a massive benefit to me.
I can type “gambling OR Proctor” into this field and dismiss a swathe of search results. You don’t need to put in multiple words, but the OR operator allows a handy list.
Another example is if your ancestor shares a name with a notable person. It doesn’t have to be George Washington. With a newspaper in a small town, it could be the local drunk who keeps getting fined for public order offenses.
One of my ancestors was that local drunk and one small newspaper kept yielding useful clippings!
However, if your ancestor has a namesake in frequent trouble with the law, you can try excluding keywords like “sentence OR fine OR court”.
Tip #4: Take Advantage Of The “Include” Feature
The “Include” feature adds an extra filter to your search, so you need to be judicious in your choice.
It will help if you read a few articles from the timeframe and location that you’re targeting. You may notice that when articles refer to a man, they usually mention his occupation.
This makes your ancestor’s occupation a useful keyword! But try a few variations. Shoemaker may not appear, but bootmaker or cobbler, or cordwainer might.
If you’re seeing regional newspapers covering a wide geography, then the town or city can also be a good keyword. But you may also need to vary your placenames – this leads me to the next tip.
Tip #5: Research And Vary Your Location Keywords
This comes back to the newspaper archives being based on OCR scans. That means your keywords need to be close to an exact match to the text that appears on the page.
So, if you add “New York City” as a keyword – you may not get hits on articles that mention “NY” or “NYC” or the “Big Apple”. Some newspapers outside NYC may simply refer to “the City”.
It may be easy for you to think of a list of possibilities for New York City. But what about Westfield, New Jersey? Your ancestors may have lived there, but you may have no idea how locals referred to their hometown. And the references may have changed over the years.
It’s worth running a few internet searches on nicknames for the town. And throw in different variations as a keyword.
Tip #6: Recognize Quality Issues That Hurt Search
A few years ago I was using a website with a small archive of British military journals. I knew I’d seen mention of a name on a page in an earlier research session, and I wanted to find that page again.
The search feature couldn’t find it (this wasn’t GenealogyBank). So I used the browse feature to navigate to the page. There was the name staring at me in the face (well not quite, but you know what I mean).
I pulled up the search feature in another tab and searched the specific title for that name. This company’s search technology couldn’t find it.
I think I know why. The name was in a caption beneath a photo, and this text was in a different font to the rest of the article. The human eye could read it perfectly, but the OCR scan hadn’t translated the slightly different shapes into the right word.
Newsbank OCR technology
Consider that local and regional newspapers across the United States use wide variations of fonts and print styles. And those styles may also change over time within the same title. That’s a huge challenge for OCR.
The good news for GenealogyBank customers is that the parent company, NewsBank, has a long track record in developing and improving scanning technology. You can read some of the history in our article on who owns GenealogyBank.
Recognizing challenging titles
Even the best OCR technology has problems with old newspapers where the paper has deteriorated considerably. There may be smudges and inkblots. Pages may be greatly creased or torn.
If you are striking out with newspaper results when you know your family lived in the area, it’s worth browsing a specific issue at the appropriate date range.
So, is there anything you can do if you see that the image is of poor quality? If you squint and spot your ancestor’s names visually, I recommend that you contact Customer Support.
The company can’t magically fix the original tears and breaks in the newspapers that were scanned to create the digital archive. But they may be able to source an alternative set of issues for a repeat of the digitization.
All you can do is add your request for improvement to their list – and hopefully, other customers are also pushing for improvements.
Tip #7: Read Historical Articles To Get Used To Typical Name Usage
You probably know that married women may be referred to by their husband’s names in older articles.
So, my ancestor Anna Gamble may appear as Mrs. P. Hickie in various places. Her obituary may mention her maiden name – but I’ve seen plenty of women’s obituaries that don’t.
It also took me a while to realize that my ancestor James was the “Jas. Gamble” that kept cropping up in a local newspaper. I assumed that this was a Jason. But as Jason wasn’t a common Irish name in the 19th century, I began to get suspicious.
I got clarity from this quick internet search: “genealogy what is jas. short for”
I also saw that many other people were searching the same question!
This is why an evening spent reading issues of old newspapers will help avoid getting tripped up by historic turns of phrase.