Many people think of language as a set of rules; break them, and you're Wrong. But that's not how language works. There are different degrees of wrongness, and there's not a bright line between the degrees--and many things that people think are wrong aren't. I'm the office language-nerd at work, and also have tried to explain why so much scorn about how other people speak or write is misinformed or misguided in a book. I didn't get a chance to do this in the book, but herewith, I offer a taxonomy of language mistakes and non-mistakes, as a way of helping people think about what's right and wrong:
Rules everyone knows: These are the language rules that even a three-year-old knows: "Steve is here," not "Steve am here." These are the bedrock of the language, and there are so many thousands of them that most people don't think of them as rules in their own language because they're not what most people think of as "rules": the difficult ones that are drilled into you in school.
Standard but tricky: Many people are tripped up, for example, by "whom." "Whom" is still part of standard English, though it is so misused, even by people who are trying, that it may not survive forever. Rules in this category are also routinely ignored in speech.
Obsolescent rules: Sticklers insist on many usages that are now too late to save. I like the old philosophical-logical phrase "to beg the question," which means to try to sneak the conclusion of your argument into one of your assumptions. But the usage "to raise the question" is so much more common that I've nearly given this one up.
Disputed rules: Many sticklers insist, for example, on "None of us is leaving," but common speech often has this as "None of us are leaving." But the great English rulebook writer H.W. Fowler, among others, weighed in in favor of "none are" in his 1926 "Dictionary of Modern English Usage." Some questions are simply not settled, and you should check your pockets after talking with anyone who insists that they are.
Non-rules: A long list of peeves on the part of single individuals that somehow made it into grammar books and teaching materials. Most famously, great writers have split infinitives and ended sentences with prepositions for centuries, yet somehow bans on both usages became "rules" that have been taught to millions of speakers in English, in contravention of their own good sense for their native language. The linguist Arnold Zwicky has called the most persistent of these "zombie rules": like the two above, they've been shown as bogus in many good usage books, yet still survive thanks to many a provincial schoolmarm.
Formality differences: Speech and writing can have two different sets of rules, though many people are uncomfortable with this idea. If you knock on a door and your wife asks "Who is it?," if you're in the small category of people who say "It is I" you could use a refresher on the concept of "register": formality has its place, but so does informality, and usages like "It's me" has been part of living English forever.
Regional differences: Brits not only have different words from Americans (lift, motorway) but some subtle bits of grammar: "You should see that movie." "I will do," says the British-English speaker, using one more word than Americans do. To label regional differences "wrong" is one of the worst kinds of provincialism.
Dialect differences: This is tricker, but linguists have known forever that Black, Southern, Scots, Irish and many other kinds of English differ from the standard not randomly (because their speakers are lazy) but systematically. They are rule-bound varieties of language just like the standard is, with the main difference that they're not written down as often and have historically lacked prestige. That doesn't make them wrong; it does make them inappropriate for settings that call for standard English. But book-standard English is wrong for many other circumstances, a fact too often forgotten.
House style: "August ninth" or "August 9th?" "E-mail" or "email?" I have read the rant of a copy-editor who is convinced that there is a simple black-and-white answer to the question of "douche bag" versus "douchebag." But this is ridiculous: all these questions and many others are matters of house style, not correctness per se. It's good to keep one house style for a single publication, but for God's sake don't lose sleep over these as a matter of correctness.
Personal taste: I've heard that the New York Times bans "should" from editorials, since saying it relieves the writer of explaining why something should happen; the verb does it all. Bloomberg's business-wire service bans "but" from all copy (except direct quotations). This isn't grammar but style.
It's not easy keeping track of so many kinds of right or wrong. It'd be so much easier simply to memorize one set of rules and let that be that. But it's much more rewarding to develop a feel for the different things we mean when we say "correct," and much more interesting too.