In the ensuing discussion, she was advised that some of what she had learned were what are called zombie rules (I think Geoff Pullum, on Language Log, is responsible for that turn of phrase), rules like not splitting an infinitive and other relics of the ill-begotten attempt in centuries past to make English grammar conform to Latin grammar. (Spoiler: it doesn't.)
The young editor asked how to identify these zombie rules, and in the context of the discussion, this is what I said. I thought it might be helpful to others, so I'm posting it here as well:
As several other people have already commented, it's good to follow linguists and editors trained in linguistics. The modern (i.e., since late nineteenth century) idea of "grammar" is that it's a description of the language as it has evolved, not a set of arbitrary rules endeavoring to tell people how they should speak ("grammar" in the sense that schoolteachers use the word).If you are a careful writer with a distinctive voice, you should expect an editor to preserve that voice. If an editor tears apart your perfectly fine prose to make it conform to rules she learned in fifth grade, she has done you a disservice. You deserve better than that.
That said, one of the features linguists describe is register, sometimes called diction. In natural speech communities in all cultures and language groups, speakers distinguish between formal and informal register (linguist John McWhorter commented recently on the strange flattening of that distinction by a certain elected official who shall here remain nameless). So, in a sense, the grammar of formal diction is different from the grammar of informal diction, even though together they compose the grammar of the language taken as a whole. And register can be broken down further. The informal register you use with your parents may differ from what you use with your peers or with your toddlers, and the formal register used in accepting an award from the local Rotarians may be different from what you would use to address the Nobel awards banquet in Stockholm.
Further, the way we write differs markedly from the way we speak, in any register.
We do expect schooling to teach students the dialect of the dominant culture, the idea being that if they learn that, they will be able to get a decent job and earn a living. And that's the genesis of the schoolteacher's approach to grammar: follow these rules, and you won't be thought stupid by rich white dudes.
But that's not the editor's role. We should take a more subtle approach: we know the standard dialect (the dialect of the socioeconomic elite) and can correct text to match that if we're hired to do so. But we can also respect the fact that all speech communities have their own perfectly valid dialects (each with its own various registers), and we can be flexible enough to edit in those contexts too. (This comes up mostly but not solely in the context of fiction.)
If you limit yourself to the minutiae of copyediting and proofreading and do not wish to think hard about the questions of usage and style informed by a passing familiarity with linguistic principles, then the general idea is to do no harm: just fix the little stuff, and don't "correct" the author on the basis of the kind of rule you're asking about here. If, on the other hand, you also are charged with line editing or substantive editing, then you do need to think about these questions.