A recent writeup provoked three users to send me a certain message. I expected this, actually. I've gotten this before when telling people that I study linguistics or yammering on about some little grammatical oddity that's probably far less interesting to whoever I'm talking to than it is to me. The message is this: the person tells me that they're impressed by my knowledge of grammar, which they seem to regard as a particularly arcane area of knowledge, and then they make a self-deprecating comment about their own ability with language. "Ha ha, I better watch my mouth around you," they say. "I was never any good at that stuff," they confide. "The way I talk must make you cringe," they tell me. Well, I wasn't cringing before, but I sure am now.
A minor wave of pique passes through me when I hear this. It's a bit of an insult — I didn't study this subject in order to get better at tsk-tsking at others' "misuse" of language. It also makes me feel a bit sad that a field of study that I find endlessly fascinating is so little-known to most people — do they really imagine that a linguist's job is memorizing schoolmarmish admonitions about 'split infinitives' and 'double negatives'? And finally, I become a bit depressed that all that a lot of people seem to have gotten out of years of English classes is a lasting feeling that the way they speak is somehow unacceptable.
Pardon me if this writeup is English-centric; regrettably, I don't know any other language well enough to discuss the finer points of its prescriptive tradition. By an amazing coincidence, though, you speak English too, so try not to let my parochialism get on your tits.
Describing and prescribing
There is a difference here between taking a descriptive approach to language and taking a prescriptive one. Schoolteachers usually take a prescriptive approach, teaching their students not to use ain't, or not to end sentences with prepositions, or scolding them for saying things like Joe is taller than me.
This is not what linguists do. Linguistics is a science, and it shares with every other science the basic formula. First you observe a phenomenon, and next you attempt to form a hypothesis to explain it; after that, you see what your hypothesis predicts and look further to see if those predictions are borne out. Observation is key here; from a linguist's perspective, language is an object of study. You look at the way people speak and try to understand the language's underlying rules and even the functioning of the language center of the brain itself based upon what they say. From that perspective, rules about how people speak don't make sense — after all, such rules exist to make people change the way they speak. If they didn't use some stigmatized bit of grammar in the first place, there wouldn't be a rule against it. And so these "errors" are just as legitimate as a target of study as anything else. They're just one more aspect of how people talk.
When you think about it, this is obviously the only approach that a linguist can take. Imagine Jane Goodall goes off to study the chimpanzees, and she is utterly appalled by their rampant sexual immorality. So she decides to teach them the value of abstinence. Ridiculous, right? She's there to study what the chimps do, not change it. That's what we're doing — language is an interesting human behavior, a behavior with no real parallel in any other species. It's one of the few things we can identify as quintessentially human. For linguists, it's fascinating. There's no place in trying to puzzle out this intriguing aspect of our humanity for telling people to change it. It wouldn't make sense for linguists to spend their time correcting other people's grammar.
The nature of prescriptivism
Prescriptivism takes a couple different forms; some prescriptions exist to teach the standard variety of a language, while in other cases, prescriptive rules simply have no genuine grammatical or historical basis at all in the language's usage and instead are invented out of whole cloth.
Prescribing of standard varieties
Prescribing certain ways of speaking and forbidding others is quite common. It seems to develop when a linguistic community becomes large enough and stratified enough that different varieties of language develop among different subgroups, and particularly when those subgroups differ in their power in the community. Unsurprisingly, the variety spoken by the dominant group within a population is the one that becomes the standard. It shouldn't come as a shock, for instance, that the standard variety of American English is not Black English (AAVE); the fact that Black English is stigmatized follows quite logically from the fact that black people themselves are a disadvantaged group.
Under prescriptive mandates, grammatical features present in stigmatized varieties — the habitual be of Black English, or the word ain't (common to many different varieties of English), or peculiar regionalisms like My car needs washed — are explained by teachers as incorrect. This isn't out of some genuine, logically-defensible argument — after all, what kind of logical argument can you actually make that the way a community naturally speaks is somehow inherently wrong? Instead, it takes the form of simple proscription ("ain't ain't a word . . . ") or appeals to abstract principles that are themselves not justifiable or are misunderstood and misapplied, as with the argument that so-called double negatives don't make sense because two negatives 'cancel out'. Then, of course, these rules are transmitted further, because people pursuing professional careers have to learn them in order to succeed, or else be falsely tarred as uneducated or even stupid. That doesn't mean that these rules describe "correct" English — it just means that they describe one particular variety of English.
Prescriptions that just don't make any goddamn sense
Of course, there are other prescriptive rules that get bandied about that just don't have any basis in reality whatsoever. While there is no logical basis for treating one variety of a language as "correct" and the others as inherently "incorrect", it becomes even stranger when English teachers start claiming that certain things are "bad English" even though they're routinely used even in formal writing by speakers of the standard language. One particularly egregious example is the "predicate nominative" — that is, the rule that says you should answer "Who is it?" with "It is I." This bizarre little rule doesn't describe the spontaneous usage of any native English speaker for a very good reason — it's not a rule of English grammar. It's a rule taken from Latin back in a time when Latin used as a standard that other languages should be compared to.
Some of these rules fall into a bit of a grey area between rules of formal style and rules that are just made up. For instance, students are routinely taught not to end sentences with prepositions. There's a (likely apocryphal) quote attributed to Winston Churchill upon being "corrected" with this rule: "That is the kind of pedantry up with which I will not put!" He probably never said this, but it's a nice illustration of how bad a sentence can sound when you follow rules that don't make sense. This is another one of those little rules imported from Latin — the difference is that English allows prepositions to be stranded (to use the linguistic terminology) when their objects are moved elsewhere in the sentence, and Latin simply doesn't. A similar sentence would be unintelligible in Latin, but it makes perfect sense in English.
The rule against "split infinitives" is another one taken from Latin. "Split infinitives" are phrases involving insertion of a word between to and an infinitive verb, as in the famous line from Star Trek, "To boldly go where no man has gone before". These are, of course, perfectly normal in English; in fact, to is not, grammatically speaking, part of the infinitive. However, in Latin, to go would have been translated with a single word; sometime in the nineteenth century, some old curmudgeon noticed, and decided that to go ought to be treated as a single word because its Latin translation is; therefore, sticking another word in the middle was "bad grammar", even though English speakers have done so for centuries.
But of course the split infinitive rule and the preposition stranding rule have gained some traction since they were invented. You'll certainly see both constructions used in formal writing, but you'll also see people avoiding them. A case could be made that these two rules now describe formal written English, even if they have their origins in nonsensical made-up rules. But both of these restrictions seem to be going by the wayside nowadays; fewer children are taught them in school and fewer people are adhering to them. Even if you accept that these rules still exist in the formal register of standard English, the alternatives are obviously not "bad grammar" except within the very tight scope of formal English; they're perfectly good grammar in any other context. Instead of teaching that one alternative is simply incorrect, a more useful approach would be to teach that there's a choice and how to make it in different circumstances.
The hazards of prescriptivism
What's important to realize is that, whether you're looking at a standard variety of a language or a non-standard one, all languages are the product of grammatical rules that children unconciously learn as they're learning the language. Non-standard varieties of language aren't the result of a failure to learn the standard variety properly — they're varieties that also obey systematic rules. It's just that the rules that define their grammar are a little bit different.
The trouble with teaching that some varieties of language are correct and others are incorrect — beyond the fact that there isn't any sort of rational or empirical justification for it — is that it means teaching a child that the way he and his community talk is wrong. On the other hand, it's not doing a kid any favors to refuse to teach him the standard variety out of some misplaced sense of cultural respect. But regardless, instruction of children in the standard variety shouldn't mean denigrating the way his community talks — particularly when there's no sensible reason to do so. You can hardly expect insulting a kid and his family and friends to be a great way to encourage him to learn.
It's also a difficulty for linguists attempting to study non-standard varieties of language, as speakers will often use only "correct" forms when they're talking with linguists. A linguist named David Gil, examining a colloquial variety of Indonesian called "Riau" wrote this:
Right away, I was struck by how different the local language was from the Standard Malay/Indonesian that I had read about in the linguistic literature. So I set out to investigate the language, by eliciting data from native speakers. But this turned out to be a virtually impossible task: the interference from the standard language was much too strong. If I asked speakers how to say something in colloquial Indonesian, they would invariably provide sentences in the formal language. If I then confronted speakers with sentences that they themselves had uttered, they would deny having produced them, and then offer to "correct" the sentences by translating them into the formal language. Similar problems occur in many or all languages; however, the extent of the phenomenon differs considerably from one speech community to another — and here it was about as difficult as it gets.
This is a normal aspect of attempting to study native speakers of any language; it creates difficulty whenever researchers want to focus on some particular poorly-documented language or some particular grammatical feature of interest that some speakers in a larger speech community use. It's a struggle sometimes to get informants to give accurate examples of the way they speak when their speech is being monitored and studied.
It's sad when you start to realize that a lot of people really feel deeply ashamed about how they speak, ashamed to the point that they'll lie and claim they don't actually speak that way when their speech is being looked at. And it's really sad that instead of the great literature they read or the essays they crafted, what people remember from their English classes is red pens scolding them for violating silly taboos. This is why linguists tend not to hold much with prescriptivism; it's obviously outside the purview of linguistic work, but further there never is much logical justification for prescriptive rules. Much of prescriptivism amounts to a spectacularly poor way of teaching the standard dialect, since it starts by instilling shame in people for the way their communities speak rather than encouraging them to develop the ability to use both varieties as needed. The rest of it amounts to silly superstitions that arose when some fool invented a rule out of thin air and started telling people off for breaking it.
When E.B. White decided to publish a new edition of The Elements of Style, originally penned by an old professor of his named William Strunk, he inserted a new rule about the use of that and which. That, according to this rule, was to be used for "restrictive relative clauses", while which was to be used for "nonrestrictive relative clauses". Restrictive clauses are normally not set off with commas; White's rule would forbid Franklin Roosevelt's famous description of December 7, 1941 as "A day which will live in infamy." It would also forbid this sentence: "Every morning, before Stuart dressed, Mrs. Little went into his room and weighed him on a small scale which was really meant for weighing letters." It's from the second paragraph of White's famous (and wonderful) children's book, Stuart Little. In fact, very few writers follow this rule; innumerable examples can be found of great authors violating it.
White obviously couldn't bring himself to follow his rule. Neither could Strunk, incidentally — when White added it to the new edition of The Elements of Style, he also took it upon himself to correct all the instances in which his beloved (but famously curmudgeonly) old professor had "misused" it. Quite some of nerve there.
That's the take home message here. The silly little invented taboos that so many people have so much trouble remembering are confusing for a reason — they don't really describe the English language. They don't describe the way people actually speak or write in most cases; the rules that do are ones that try to bolster the position of the standard variety of language with irrational proclamations that anything but the standard is "incorrect" rather than just being different.
Gil, David. "Working on Riau Indonesion." http://monolith.eva.mpg.de/~gil/riau/working.html
Pullum, Geoffrey K. "http://itre.cis.upenn.edu/~myl/languagelog/archives/001803.html." Language Log. http://itre.cis.upenn.edu/~myl/languagelog/archives/001803.html
Harper Collins's webpage for Stuart Little. http://www.harpercollinschildrens.com/HarperChildrens/Kids/BookDetail.aspx?isbn13=9780060282974&BDMode=8