A few years ago, my local Family History Center reopened after an extensive remodel of the building which houses it. Although I find that I spend more time there working on other people’s genealogy than my own, I found myself looking forward to going back. I noticed two things during my first walk-through: 1) the bundles of wire under the desks (kick one and all the computers would lose their internet connection) were gone, and 2) it was a lot smaller.
I’ll admit, in the age of digital scans and terabyte “memory sticks”, genealogy work doesn’t need to take up a lot of space. I can fit all of my research, including photos, narratives, source documents (a few whole books), dictionaries, maps, and audio recordings in my pocket and still have room for my keys and cell phone. That said, I was still a little disappointed to find that we wouldn’t have more than a quarter of the shelf space that we had before. So, we had to get rid of a few books.
Whenever we talk about thinning out the bookshelves, one of the first items considered is usually the census indexes. Although I rarely use these volumes any more, I was sad to see them go, as they reminded me of one of my first genealogy-specific lessons: using soundex. Just like the Dewey Decimal System (which I had to learn in fourth grade), if you’ve never used soundex, you’ve missed out. It’s a relatively simple system invented by Robert C. Russell (patented in 1818) for indexing names. In the 1930s, a few changes were made (like including the letter “j”) and the Works Progress Administration used it to index several state and federal censuses. Since then, it’s been a part of just about every government record set – pretty public stuff. Learning it, though, felt like an initiation. Some kind of secret code I had to memorize or the books of genealogy would be closed to me forever – or at least the 1880 through 1920 census records. Here’s how it works:
First, except for the first letter, remove all the vowels in your name, as well as the letters y, w & h (“Younger” becomes “Yngr”). Next, separate the first letter and keep it; the rest will be encoded using the following rules:
For the letters b, p, f & v use the number 1
For the letters c, g, j, k, q, s, x & z use the number 2
For the letters d & t use the number 3
For the letter l use the number 4
For the letters m & n use the number 5, and
For the letter r use the number 6.
If any two or more letters in a row have the same number, they get encoded only once (no repeated numbers), and all soundex codes are four digits long, so you may need to add some zeros or leave off a few numbers.
So, “Younger” is now Y526. It’s a few steps to go through but, once you get used to it, it’s not so bad. There aren’t a lot of ways to spell “Younger” though, so the name is normally pretty easy to find – with or without soundex.
Others are not so simple: my fourth-great-grandfather’s last name was “Quintal” – a fairly common name in French Canada, but not so common in English-speaking Oregon. So, his kids’ last names were in my records as “Cantelle”, “Cantile”, and “Chantelle.” All of these variations (as well as some others that I wasn’t looking for) have the same soundex code: C534 – quite convenient when you’ve got thousands of names to search through. This made it very easy to find them and their families in the 1900 & 1920 census (Oregon didn’t have a soundex index for 1910).
Today, soundex appears in many databases – just about anywhere there are names to be encoded, you’ll find some version of a “phonetic algorithm.” Unfortunately, despite its awesomeness, I haven’t needed to use it outside the library (just like the Dewey Decimal System).