People are regularly accused of not knowing their grammar when the real issue is a possibly shaky grasp of usage.

Here’s Bryan Garner, whom I’ve invoked more often in the last week or so than in the previous 10 years, on grammar: “Grammar consists of the rules governing how words are put together into sentences” (Chicago Manual of Style, 16th ed., section 5.1).

And here he is on usage: “The great mass of linguistic issues that writers and editors wrestle with don’t really concern grammar at all — they concern usage: the collective habits of a language’s native speakers” (CMS 16, section 5.216).

Language eddies and ripples and never stops moving.

Those collective habits tend to change a lot faster than the underlying rules. Think of a river, a pond, or the ocean: the surface sparkles and ripples and can be quite turbulent, while what’s underneath moves more sedately or maybe not at all.

Usage isn’t uniform across speakers of a particular language either. Nowhere close. Much has been made of the differences between British English (BrE) and American English (AmE), but both BrE and AmE include great internal diversity, by nation, region, and other factors.

Usage that raises no eyebrows in a particular field may seem clunky, appalling, or even incomprehensible in another. Recently an editor queried the editors’ e-list we’re both on about a use of “interrogate” that raised her hackles; in her experience, suspects could be interrogated but not theories, Those of us who regularly edit in certain academic disciplines assured her that in those fields theories can be interrogated too.

The editors’ groups I’m in are not only international, they include editors from many fields, genres, and disciplines. So when we ask if a certain usage is OK or not, we mention the intended audience for whatever we’re working on: fiction or nonfiction? AmE or BrE? academic discipline? subject matter? Is the tone informal or formal?

Colloquialisms and, especially, slang can be especially tricky. Slang often arises within a particular group, and part of its purpose is to set that group off from others. A word that means one thing in the wider world may mean something else within the group. By the time the wider world catches on, it’s passé within the group. This poses a challenge for, say, novelists writing for teenagers and young adults, who in every generation come up with words and phrases that set the adults’ teeth on edge: how to come across as credible when by the time the book appears in print (usually at least a year after it’s turned in to the publisher), the dialogue may come across as ridiculously outdated to its target audience.

I’ll assure my insurance company that I’m wise enough to ensure my safety by not letting Travvy drive my car.

In English, usage gaffes often result when words sound alike; when their meanings are related, the potential for confusion grows. Consider this sentence: “I assured my friends that I’d ensured my own safety by insuring my car against theft.” “Insure” appears regularly for “ensure,” which means, more or less, “guarantee,” and to make it even more fun, my car is insured through Plymouth Rock Assurance — which works, sort of, because they’re assuring me that I’ll be covered in case of an accident.

By Googling frequently confused words I turned up lots of lists, including this one from the Oxford Dictionaries site. I see most of them pretty regularly, the exceptions being the ones that involve distinctively BrE spellings or words, like “draught,” “kerb,” and “barmy.” It’s missing one that I see a lot: reign/rein. The expression “rein in” has come adrift from its origin, which has to do with horses. “Reign in” sounds exactly the same, but written down it doesn’t make sense.

Working editors, especially copyeditors, store all these frequently confused words in our heads. We’re always adding to the collection — and discussing whether a particular word has graduated from confusable to acceptable, at least in certain quarters. These discussions can get quite heated.

The English-language dictionaries most commonly used these days are descriptive, not prescriptive. That is, they describe how speakers are actually using the words, not how they should be using the words. Take “imply” and “infer”: I can imply (suggest or hint) that something is true, but you can infer (deduce or understand) that I don’t believe it. “Infer” is used to mean “imply” often enough that this is listed as a meaning in both Merriam-Webster’s and American Heritage, though both dictionaries include a cautionary note about this usage.

My editorial mentor, circa 1980, railed against the use of “target” as a verb, which to me at the time, a generation younger, seemed unexceptional. A few years later, however, I and others were railing against the use of “impact” as a verb. What the hell’s wrong with “affect”? we asked. I’ve pretty much given up on that one, though I don’t use it myself.

I cheer loudly whenever an author uses “comprise” correctly, which isn’t very often, but mostly I’ve given up on that one too. Once upon a time the whole comprised the parts, and the list of parts was assumed to be comprehensive. If it wasn’t, you used “include.” So few people remember that distinction that if it’s important that readers know that the list of parts is comprehensive, you better not rely on “comprise” alone to get the idea across.

Similarly, I was well on in my editorial career when I learned that “dogs such as Alaskan malamutes and Siberian huskies” was assumed to include malamutes and huskies, whereas “dogs like Alaskan malamutes and Siberian huskies” did not, presumably with the rationale that malamutes are not like malamutes; they are malamutes. No way would I expect a general readership, even a literate, well-informed readership, to know this.

At the same time, I do occasionally feel a little smug because I’ve got all this esoterica stored in my head. But I do try to keep it under control when I’m editing.