Google Ngram

Google Ngram View shows usage frequency of words or phrases along with a timeline. This is far better than simply checking the estimated search count of a word or phrase (as I used to rely on to sort words by frequency, for English words or Chinese characters). But there are a few things you need to know when using Google Ngram.

* Words are case-sensitive by default. So when comparing frequencies of different words (or phrases), make sure to use all lower-case, or just click the Case-Insensitive button. (The button shows on screen but is NOT highlighted by default.)
* You can only enter a string of 120 characters in the text input box.
* You can only enter a phrase or sentence of at most 5 words.

With those conditions kept in mind, if you want to check the frequencies of, say, all the 18 phrases that contain couteau ("knife" in French) on the Wiktionary page (click "show more" on the page to see all 18), you have to break them up into a few chunks. What I do is copy all the phrases and paste them into a gvim window (vim is a freeware editor), add commas to the end of each line (command: %s/$/,/), join them into one line (%j!), and manually break the long string up into chunks, each less than or equal to 120 characters. (The last phrase remuer le couteau dans la plaie has more than 5 words and can be omitted since Ngram won't accept it anyway.) Copy each chunk into Google Ngram, choose French as the language and search. Thus, I find couteau de cuisine ("kitchen knife"), couteau de chasse ("hunting knife"), and à couper au couteau ("(fog) very dense") in modern-day French have relatively high frequencies. I added the first two to my Learning French Words book as examples for headword couteau, but omitted à couper au couteau because it sounds a little too idiomatic. The actual text in my book is as follows:

couteau knife. Cognate with ... Example, couteau de cuisine / de chasse / à pain (“kitchen / hunting / bread knife”).

I added couteau à pain in spite of its slightly lower Ngram frequency, because this is a phrase common enough and I want to show the usage of à + noun in addition to the de + noun phrases.

Another use of the Ngram is in choosing the most common conjugation examples for headwords, as I talked about before. The example sentences I gave to the headword mentir ("to tell a lie") in my Learning French Words book are "je mens / tu mens / il ment (“I’m / you’re / he’s lying”); j’ai / il as / tu a menti (“I / he / you lied”)". Only two tenses are given because they're the most common, present and present perfect, and within each tense I chose the most common conjugations according to Googel Ngram.

By the way, although my Spanish words book was already published, I'm slowly revising it without a plan for publication. One type of addition to it is example phrases or sentences. For example, I added

Examples, buena estrella (“lucky star”); estrella fugaz (“shooting star”, literally “fleeting star”); estrella de mar (“starfish”, literally “star of sea”); el producto estrella (“the flagship / leading product”; note it’s not ... estrello).

to the headwords

estrellar to smash, to crash. estrella star. Cognate with stellar. ...

The Wiktionary page for estrella has 23 example phrases. I chose 3 because of their relatively high Ngram frequencies, in addition to el producto estrella, which I think is a common phrase and can illustrate the figurative usage of this word.