Written by Julie Holzman

---

When editing a manuscript, what do you do when you encounter an author’s non-hyphenated word that should be hyphenated per a style guide? As editors, our go-to sources for checking current spelling and word formatting are Merriam-Webster’s online dictionary and The Chicago Manual of Style.

If a novel or story takes place in another era, it’s best to reflect the format or spelling of that time period. This is where using an Ngram search comes in handy. The Google Books Ngram Viewer outputs a graph showing the use of a word or phrase in books through time. More details about what an Ngram is will follow after a visual and brief instructions on how to use the tool.

An example to work through for explanation

Let’s look at the term code talker from the World War II era. Under CMOS rule 7.96 section 2 titled “Compounds According to Parts of Speech,” under the “noun + noun, single function” list, the term would be hyphenated as code-talker. But what does the Ngram show?

The open form code talker is prevalent. Looking at articles and government pamphlets written during the time of the book’s storyline, we see code talker isn’t hyphenated, even when preceding another noun. To properly reflect the era, format it as open. While this goes against CMOS rules, the term is still clear to readers. If your research shows a discrepancy from the author’s usage, explain to them what is seen in the Ngram and in CMOS (see rule 7.91 regarding compound nouns not being hyphenated).

Getting started with an Ngram search

To see how the tool works, let’s repeat the search seen above.

• Go to https://books.google.com/ngrams/.
• In the search box, clear the default terms and then enter the words to compare. Note that you separate the terms using a comma with no space after it. The Google Books Ngram Viewer will then plot the terms’ relative frequency over the selected time range.
• You can adjust the time range to narrow down the years you’re looking for. I like to keep the ending year with the latest one available to see how the term is seen today.
• The default is to search both American and British English, so it’s best to select American English only for U.S. books. For this search, I’ve left both so we can see a comparison; when only American English is selected, the hyphenated form doesn’t show.
• For our purposes, leave the default setting of case-insensitive and ignore the smoothing button.
• Your graph should now look like the one above.

What exactly is an Ngram?

The term Ngram is used primarily within the research community. It’s a statistical analysis of text in order to find a number, or n, of some sort within the text. One word is a 1-gram, two words is a 2-gram, etc. Since we can be searching for multiple words, it’s an Ngram, as long as a space separates the text. The Ngram viewer makes adjustments for a hyphen.

The results come from the immense number of public domain books from the past 200-plus years that workers at Google scanned for its Google Books search engine.

What the Ngram shows

If more than one word or phrase is entered for a comparison, a color-coded line represents each phrase or word to contrast with the other search terms.

For more data and details, here’s a brief starting point. The percentages seen on the Y-axis in a Google Ngram represent the percentage of keywords in Google’s sample of books (for our purposes, those written in English and published in the United States) that are the target keyword. For example, searching for the shows that the word makes up 4.2% of modern published text.

Another tool for you

I highly recommend adding the Ngram search to your editing tools. You don’t have to be a code talker to make use of it in your editing process.

For more information about Google Books Ngrams, see https://books.google.com/ngrams/info.

---