Grammar and Usage

Back to the Syllabus


In the 1998 baseball playoffs, Chuck Knoblauch of the Yankees let a decisive run score while he argued with an umpire. Asked about the play afterwards, Knoblauch said, "I guess I should have went and got the ball." Commentators jumped on him as much for his "bad grammar" as for his mental error in play. They said snide things about how he'd obviously majored in English at Texas A&M. But was his sentence ungrammatical? Could you understand what he was saying? Or was it instead just "substandard" usage?

As late as the 1980s, language was a newsworthy, or at least column-worthy, topic. Columnists like Edwin Newman, Russell Baker, and William Safire wrote regularly about language, offering expert opinions on usage. A professor from New Jersey named Richard Mitchell became famous for skewering bad usage in memos by overpaid college administrators; he was known as "the Underground Grammarian" and published several books on language.

Somehow this popular demand for language writing has declined. Essays on grammar nowadays are likely to be much more ambivalent, much less likely to lay down the law, as a recent Star-Telegram piece by Larry Bingham illustrates. In fact, the best-known language writer of the present day is not a serious usage arbiter, but is Dave Barry in the guise of Mr. Language Person, a send-up of Safire and other "experts."

If Barry has a "take" on usage issues, it's that experts are silly, but that users of language are sometimes silly as well. Take this example from a "Mr. Language Person" column some years ago: Barry reported this as sent in by an alert reader as something she'd overheard verbatim. The context was a man in Tennessee apologizing to a woman for not having given her a ride to a church social:

"But, Lorena, if I'da knowed you'da want to went, I'da seed you'da got to get to go."

Barry used that line as a sort of "can you top this" of ungrammatical usage. But I'd like to use it as a touchstone for a brief discussion of grammar vs. usage. Let me define the two categories:

Grammar is a set of implicit rules that govern the formation of sentences. We may have no explicit knowledge of these rules, but we obey them every time we speak and use them every time we comprehend a sentence.

Usage is a set of explicit prescriptive rules that people impose on language in order to separate socially acceptable grammatical sentences from others that are not socially acceptable.


My definitions are slightly tendentious; that is, they have a bias. But they are also somewhat objective. The Barry example above is, by my definitions, a grammatical sentence. In fact its grammar is rather complex. In terms suggested by classical Latin grammar, we would call this a "past contrary-to-fact condition." If something had been the case in the past (i.e., the speaker knowing that Lorena wanted to go to the social), then a certain event would have occurred (the speaker lining up transportation). But it is clear from the phrasing of the sentence that the condition in the past was unfulfilled, and thus "contrary to fact."

If the speaker had said, "Knowed Lorena you'd got went I to seed you go I'da," we would call that "sentence" truly ungrammatical. It would not follow any pattern by which we could get meaning into or out of it.

Here's a simple example adapted from R.L. Trask's Language: The Basics. This set of sentences illustrates "Langacker's Rule," which is an actual rule of English grammar. You don't know it yet, but you understand it implicitly:

  1. After Sean got up, he had a shower.
  2. Sean had a shower after he got up.
  3. After he got up, Sean had a shower.
  4. He had a shower after Sean got up.

In which of these sentences can "he" refer to Sean?

Langacker's Rule states that "An anaphor may not both precede and command its antecedent" (Trask 30). IOW, "he" may not come before its antecedent in these sentences if that antecedent is in a subordinate clause. Since "he" precedes "Sean" in #4, and "Sean" in #4 is in a subordinate clause, "Sean" can't be the antecedent of "he" in #4. Somebody else must have taken that shower.

Note that Langacker's Rule does not represent Langacker's opinion on what #4 ought to mean; it's an objective observation. Rules of grammar are simply what we can't help abide by if we speak a certain language.

Rules of usage, OTOH, are opinions about how you should form grammatical sentences. Given the option between several grammatical ways of expressing yourself, usage prescriptions tell you which one to choose.

Let's take a famous example, the usage rule that is almost always expressed as a violation of itself:

"A preposition is something you should never end a sentence with."

The technical term for ending a sentence or clause with a preposition is "stranding"--you leave the pronoun hanging on its own, separated from the words it governs (in this case, "a preposition"). There are several different grammatical contexts for preposition stranding; our tongue-in-cheek rule would ban all of them (but make an exception for itself).

Now, we end sentences with prepositions all the time, but that can still be a grammatical habit that gets frowned upon. In the last sentence, I have stranded "upon" in an example of what's called the "prepositional passive." The object "upon" is the subject of a passive verb ("frowned").

The prepositional passive is interesting in that it is grammatical in Modern English (albeit against some usage prescriptions), but was ungrammatical in Old English. We base this assertion on finding no examples of the prepositional passive in OE (see David Denison, English Historical Syntax (Longman 1993): 124ff).

The earliest recorded citation of a prepositional passive in English is from the Early Middle English "St. Juliana" in about 1225:

"þer wes sorhe te seon hire leoflich lich faren so reowliche wið"

The construction is rare until the late 1300s, but then quite common; Chaucer uses it, so does John Mandeville, and it appears in Wycliffite writing (Denison 125-127). English speakers have used it ever since.

How we account for such a change is vexed. Even describing accurately what happens when grammatical rules change is difficult, since people typically don't notice it happening. Explaining why is nearly out of the question, unless there's a very strong and obvious substrate influence, as in some idioms in Irish English.

One thing to take away from this example is that grammar, though very precise in its rules, is subject to the same historical change as sound and meaning--and for reasons that are as irretrievable.


Compared to grammatical rules, usage rules are frequently arbitrary--in the sense that someone just makes them up. That also means that they are socially important. You can teach and explain usage logically, whereas I'll bet no-one could tell me exactly why Sean isnít taking that shower in #4 above, or formulate a rule as to when he's allowed to take it. If you speak ungrammatically, I won't understand you. If you speak against the customs of usage, I can pretend annoyingly "not to understand you"; I can humiliate and belittle you.

If you become an English teacher, for the rest of your life people will apologize to you for their bad grammar. This includes professional people who are highly educated--in fact they are more likely to apologize, as they've had more exposure to an educational system that faults them for grammatical inadequacy. But almost all people feel more or less grammatically inadequate. This is even though in the technical sense they all speak with perfect grammar. Usage, as a marker of dialect and social prestige, is what's at stake, not the technical grammatical rules of the English language.

Usage is serious business. Take a look at the Louisiana Civil Service examination--if you would like to work for the State of Louisiana (and many other employers) you ignore usage at your peril. Usage, however, is so fluid that there is often no way to study for a usage test except to study directly to that test. If the New York Times and Oxford University Press approve a split infinitive but the test you're taking doesn't, you are advised to do as the test does.

Usage is visceral--we have an emotional feeling about ways of expressing ourselves. Take these following sentences, which might be said to a little boy who has stolen a toy from his sister:

  1. Give it to her!
  2. Give it her!
  3. Give her it!
  4. It her give!

#4 is ungrammatical. So might #2 be, in most of the United States; but it's common usage in Ireland and parts of England. #1 is proper usage anywhere English is spoken, but #3, also grammatical, drives me nuts. I think of it viscerally as low, "common," trashy usage. I would only ever say #1. My imposition of that preference on you, or my censure of you for adopting another preference, can stand for the whole cultural work of usage prescriptions. I can impose my prescriptions on you because I am a professor of the English language--but should I have the right to?