Think of a grammar rule. Any grammar rule. Basic, bizarre, straightforward, convoluted. Whatever.
Got it? Okay, then, consider: why do you think this counts as a grammar rule? What gives it credence?
Here are the usual reasons:
- An authority said it, like a high school English teacher or Strunk and White.
- Good, careful writers always follow this rule (which is also an authority, but by example rather than explicit statement).
- Writers who fail to follow this rule are ambiguous or otherwise hard to understand.
- It’s old.
- It’s logical. This claim is usually made with an analogy to other language rules or to math.
- It just is.
One last question: what would it take to convince you that this rule is wrong?
- Where did your authority get this rule? From usage? From long, arduous study of grammar? Or is it just a personal preference, in which case your authority is welcome to do what they want in their own writing, but what gives them the right to impose their quirks on others?
- Do the best writers really follow this rule? Have you checked? Books, especially classics, are available online. Search for instances of following or flouting this rule.
- Does this rule really avoid ambiguity? If a writer ignores this rule, does it cause difficulty? If the rule is designed to address a problem, does it adequately address that problem and only that problem? Or does attempting to follow the rule create new (possibly worse) problems?
- Does it matter? This is more common in discussions of vocabulary, actually, where it has a special name: the etymological fallacy. Does what a word meant a hundred or a thousand or two thousand years ago have any bearing on what it means today? And if you think so, would you say that Latin is silly? (I use “silly,” of course, in its Latin sense of “holy,” rather than in a way that is meaningful to modern readers.)
- Is the logic sound? Could someone construct an equally logical analogy that “proves” the opposite rule? Is there any basis for the connection to math? I thought English-lovers stereotypically despise math, anyway, so why is a sentence suddenly an equation?
- Well, I can’t argue with that one. (I suppose I could say “It just ain’t,” but that hardly seems helpful.) Okay, then, but realize you have made a grammar rule into an article of religious faith.
Note that these questions are not just here to dismantle bad grammar rules. They also validate good ones, mostly via the second question and a dash of the third. Like lexicographers and dictionaries, linguist grammarians think about grammar books in the opposite way from how many users think about grammar books, because they know that all these rules have to come from somewhere, and that somewhere is how people use language.
How exactly different people get their data, and what they do with it, depends on the point of the book. The Cambridge Grammar of the English Language uses corpora of millions of words and analyzes the crap out of them, making for a five-pound grammar behemoth. This level of detail is not what most writers are looking for – it’s meant for people studying grammar seriously in and of itself.* More handy for writers are the American Heritage Dictionary and Merriam-Webster’s Concise Dictionary of English Usage. The first surveys 200 writers, editors, and the like to find out how they use language (getting an approximation of the views modern people who use language frequently and carefully). The second lists controversial grammar points, goes through their history, and shows how they’re shaking out in modern use. Bad grammar books don’t do the work; one or two guys sit down and bang out their intuitions and peeves without checking how well their opinions are supported by others’ language use.
Linguists are often spotted in the wild doing two things: debunking false grammar rules, and showing that non-standard or informal varieties of English are equal to the standard formal (that is, they do not have the standard grammar, and are not problem-riddled offshoots of the standard, but rather have their own, equally complex grammar). This has led to the misconception that linguists are all wacky grammar anarchists (grammarchists?), out to burn The Man’s grammar to the ground.
Not so much. But linguists do want evidence. And they’re not afraid to call variation what it is – variation – instead of assuming that of two options, one is Right and the other is The Devil. Perhaps your language is different from mine – and that is okay? Perhaps a new form is springing up, and it is as useful as the current form, and may even replace it in the future – and that will also be okay? Good grammar books rarely announce that the world is ending, young people are all terrible, and civilization is soon to crumble, and this heaping dose of sanity may be why they are less famous than bad grammar books.
Zinsser’s book, which I talked about so lovingly in my last post, contains not-great grammar advice. But that’s not most of his book; his advice is focused on style and craft. (How do you start an essay, go on, finish?) Mixing style, vocabulary and grammar is another common problem with underinformed books, and Zinsser doesn’t get off scot-free. Here’s one way to think about it: if you were writing in a non-standard variety, like Southern English or African-American Vernacular or Monsterrian, would the advice still apply? If yes, it’s style. If no, it’s grammar. Everyone needs to avoid ambiguous phrasing, but the mechanics of doing so can vary. Both Zinsser and Strunk & White have great style advice and misguided grammar advice. This leaves Zinsser with a mostly excellent book, S&W with a handful of usable pages.
Speaking of which, let’s get this party started. I’ve summarized S&W’s views on grammar issues, tried to explain where they go wrong, and linked to further reading on each point. This became a bit of a Language Log collection page, since the writers there often refute grammar nitpicks. In particular, a huge number of these articles were written by Geoffrey Pullum – he’s one of the two guys who put together the CGEL, and he gets cantankerous when people who do not understand grammar try to preach about it.
This is a work in progress — I have about seventy tabs open, which is embarrassing twice over, but I swear I’ll get them organized before my computer crashes. But I suppose this page will never be complete, since people write more about this every day. If you find a new article, or just something I missed, please send it in and I’ll add it.
Elementary Rules of Usage
1. Form the possessive singular of nouns by adding ‘s
2. In a series of three or more terms with a single conjunction, use a comma after each term except the last
Strunk and White come out in favor of the Oxford comma (the comma before “and” in a list). Hey, so’m I. It can help. But this is a point of variation, not ultimate rightness. Using the serial comma can clarify a sentence, as in the famous “We invited the strippers, JFK, and Stalin” versus “We invited the strippers, JFK and Stalin.” But those who don’t use it have a reason too: they feel that it makes a sentence tighter and faster, which is why it’s more common for newspapers to do without. Those who dislike it just need to double-check that they haven’t inadvertently made Stalin a stripper.
3. Enclose parenthetic expressions between commas
4. Place a comma before a conjunction introducing an independent clause
5. Do not join independent clauses by a comma
Or, in the modern English teacher vernacular, no run-ons. I’ve written about this one before myself, and looking at it again, the ending is relevant to much of this post: “Grammar serves the writer, not the other way around; if a run-on sentence helps to clarify and organize your ideas for the reader, go for it.”
6. Do not break sentences in two
No fragments. Like that one. And that. Etc. See previous quote.
7. Use a colon after an independent clause to introduce a list of particulars, an appositive, an amplification, or an illustrative quotation
8. Use a dash to set off an abrupt break or interruption and to announce a long apositive or summary
The first time Strunk uses a run-on in his “good” example. Not the last.
9. The number of the subject determines the number of the verb
In some dialects, yeah. In others it does not. Even in the dialect that Strunk was aiming for, this is not true all the time. Also: “none of us.”
10. Use the proper case of pronoun
“Whom” is a relic of a far-off time, when English had freer word order and used endings on nouns and pronouns to show their relationship in a sentence. We’ve lost the endings on nouns and seem to have survived; if we lose them on pronouns, that won’t kill us either. Can be used to indicate very, very, very formal writing but think about if it’s worth it. “It is I” has similar over-the-top formality issues, and the bad math logic that because the predicate is equated with the subject, a nominative pronoun should be used. Nah.
11. A participle phrase at the beginning of a sentence must refer to the grammatical subject
II. Elementary Principles of Composition
12. Choose a suitable design and hold to it
This is solid advice. Too bad it’s only two paragraphs long and the only example is a sonnet. Zinsser’s chapter “Unity” recasts this advice with enough detail to actually help the novice writer.
13. Make the paragraph the unit of composition
14. Use the active voice
Oh, this mess! Surely the most written-about page in the book. Strunk starts out by creating a ridiculous example of a passive and being angry at it for being ridiculous. He then rails against other sentences, which are not passives, though whether or not he knows that is unclear. The traditional concept of a passive is a verb phrase, most commonly “be-verb + past participle,” that allows a theme or a patient (nouns that are usually direct objects) to be the subject instead. Strunk and White’s muddle here may have jumpstarted the current vague idea of passive as “any weaselly or agentless sentence,” which is exactly the kind of language change they would have despised, judging by section IV. Irony!
15. Put statements in positive form
I appreciate that Strunk mostly tries to follow his own advice on this one, at least in the headings, but starting every third sentence with “avoid” or “omit” is cheating. Your audience will notice that they are still being told what not to do.
16. Use specific, definite, concrete language
Hey, high-quality advice, and a real example even! Accuracy and vigor FTW.
17. Omit needless words
More good advice! Dude’s on a roll. Seems like someone could take these two points, a few salient bits from section V, and a handful of words from IV, and come out with a ten-page document of usable style advice. I’ll call it “Don’t Be Vague: The Parts of Strunk and White that Are Not Hampered by Their Lack of Knowledge About Language.”
18. Avoid a succession of loose sentences
The point of this seems to be that the writer should pay attention to the rhythm of the sentences. Again, expanded upon to the point of usefulness in Zinsser, who gives among other things the basic advice of reading the first draft aloud.
19. Keep co-ordinate ideas in similar form
Another aspect of rhythm, which I suppose is here to keep you from getting the wrong idea after point #18 told you to vary sentence structure. This happens when the focus is not on giving writers the tools they need to understand rhythm themselves.
20. Keep related words together
Strunk has a good basic idea here – avoid ambiguous syntax – but doesn’t have the linguistic knowledge to express himself. The study of syntax was still in its early stages when he was writing this, so we’ll give him a pass, though that doesn’t excuse modern followers.
Also, this sentence makes me wonder if maybe Strunk was trolling us the whole time:
“The subject of a sentence and the principal verb should not, as a rule, be separated by a phrase or clause that can be transferred to the beginning.”
Pullum discusses this quote here.
21. In summaries, keep to one tense
Wait, this is just an example of choosing a design and holding to it, point 12. Why isn’t it there?
22. Place the emphatic words of a sentence at the end
Did Strunk miss that this can be the point of using a passive construction? Some of his examples here are passives! AGGHHHH
Pullum feels my pain: “[Elements] is often so misguided that the authors appear not to notice their own egregious flouting of its own rules. They can’t help it, because they don’t know how to identify what they condemn.”
III. A Few Matters of Form
Don’t put slang in quote marks, Strunk says, because “To do so is to put on airs, as though you were inviting the reader to join you in a select society of those who know better.” A caution which can now be applied to following most of Strunk’s advice.
Strunk and White display a brief awareness that language changes over time! “The steady evolution of the language seems to favor union: two words eventually become one, usually after a period of hyphenation.” This is real, accurate linguistics! Good for you, guys.
Not related: I misread this as “syllabifurcation” at first, and kinda wish I hadn’t noticed. Phoneticians, feel free to take that as a term for splitting a syllable into the onset and rime. Even better, somebody tell me this is already established usage.
IV. Words and Expressions Commonly Misused
Swiftly gone are the men who recognized “the steady evolution of the language.” Now changes are “the commonplaces of careless writing.” Deep breath, y’all.
Strunk and White dislike so many, many words. They’re so damned grouchy. White says in the introduction that on every line of the book he sees Strunk’s “smile shuttling to and fro under a carefully edged mustache,” but I see no smile. A sneer, maybe. Unlike the other sections, where I left in the headings even if I had nothing to say about them, here I’ve just deleted most of the words, because it was terribly repetitive. But if you come across an article that focuses on something I left out, send it to me and I’ll reinsert it.
For a little bracing reality, a quote from Peter Trudgill’s “Myth 1: The Meanings of Words Should Not be Allowed to Vary or Change,” the first essay in the anthology he co-edited, Language Myths, which I’ve cited frequently here.
Aggravate and irritate
First entry, first etymological fallacy. We do not speak Latin. We can use “aggravate” to mean “annoy.”
Alternate and alternative
Again, the distinction is based on Latin, not English usage.
How on Earth does the use of and/or in the “bad” example lead to confusion or ambiguity?
“Can I get a drink of water?” “I don’t know, can you?” Excuse me while I shudder.
See section V #8.
Oh no. This one has been all over the news lately because some fellow who takes Elements very, very seriously decided to edit enormous swaths of Wikipedia to better fulfill this dictum. Everybody needs a hobby, I suppose.
“Do not contact anybody; get in touch with him, or look him up, or find him, or phone him, or meet him.” Why “contact” is dismissed while the equally vague but longer “get in touch with” is accepted is another of the many mysteries we find in these pages.
Again Strunk shows a flash of awareness of language change. Why here?
A word can have two meanings. Or piles of meanings – look at “run.” If you’re looking for arbitration, “disinterested” probably means “neutral.” Otherwise, it means “not interested.” Good grief, “interested” already has the two opposite meanings – an interested party, interested in cooking.
Each and every one
“Pitchman’s jargon”! Oh man, that is so my new minced oath.
Again references the Latin root as evidence of the superiority of his definition. Man, the Latin inspidus meant “without flavor,” but when I tell you this argument is insipid, I don’t mean I tried to eat it.
All that high-minded talk about greater precision in language apparently does not apply here. “Flammable” clearly means something will catch fire, “non-flammable” means it won’t, and “inflammable” is dangerously confusing, but Strunk would like for you to use “inflammable” anyways, “[u]nless you are operating such a truck [one carrying gasoline or explosives] and hence are concerned with the safety of children and illiterates[.]”
“Hopefully” is an emotional adverb (“He leapt hopefully from his seat”) and a sentence adverb (“Hopefully we won’t be stuck in traffic for six hours”). The first corresponds to “full of hope,” the second to “I hope, one hopes.” Both uses are old, but the second used to be rare; when it grew in popularity it picked up the usual detractors. But there’s no reason to single “hopefully” out, since many emotional adverbs can be used in this way — “She chopped the onions unhappily / Unhappily, they had already left when I arrived.”
No particular difficulty arises when “however” is used at the beginning of a sentence. Even in the examples he gives this is obvious, since “however” meaning “but” is followed by a comma, and “however” meaning “in whatever way” is not.
Imply and infer
These are converse terms, like “lend” and “borrow” or “teach” and “learn” (thank you, Trudgill); you can have a pair, but many dialects (and languages) have just one, and work out fine. The dialects that have only “infer” often pair it with “to” or “from.”
Look, some not bad advice! Telling your audience that something’s about to be interesting or funny is a great way to kill their interest or burgeoning laughter. A breath of fresh air, as Strunk and White again hit on something stylistic.
Also still fought occasionally, though “customize” seems to have escaped the scorn.
Less and fewer
The fellows say that “less” is only for mass nouns (quantity) and “fewer” only for count nouns (number), a common misconception and much, much less complicated or interesting than the way these two actually work. Most notably, “less” is often used with units, especially units with numbers — “less than twelve years old,” for example, or “less than 300 grams.” And yes, this extends to “ten items or less.”
This has been true for about a thousand years, so you’d think we’d be clear on this one, but around 1770 a fellow named Baker said that, well, he thought using exclusively “fewer” with number sounded pretty. Totally unassuming, not pulling a S&W and trying to blast his preferences into the public, just sayin’ what he thought. Somehow this chill guy’s concept was turned into a command, declared urgent since allowing less and fewer to mingle would create “confusion“… somehow.
Such drama. Strunk and White officially appoint themselves The Guardians of the English Language. See Trudgill quote at the beginning of this section re: self-regulating.
Literal and literally
A less vehement attack than one might guess, considering the fury of the modern antagonists of contranymic “literally.”
All these words, and this is Strunk’s sole spelling concern? Well, lucky for him Christopher Nolan brought this spelling back as a public preference.
Trudgill traces the whole history of this innocuous little word. Why do Strunk and White stop the historical journey at “delicate”? Before that, nice meant “silly”; before that, in the beloved Latin, “to be ignorant of.” Try taking it back to that point, just for consistency’s sake.
Noun used as verb
This is a way new words enter the English language. It happens.
Offputting and ongoing
Strunk and White suggests “a simple test” for finding words with clear meanings: “transform the participles to verbs. It is possible to upset something. But to offput? To ongo?” Tragically, the authors were stricken with selective blindness towards phrasal verbs, explaining their ignorance of “to put off” and “to go on.” Also, guys, participles are verbs — clarify that you want verbs transformed from participles to infinitives.
“The word people is best not used with words of numbers, in place of persons. If of ‘six people’ five went away, how many people would be left? Answer: one people.”
This… this is just an irregular plural. Are Strunk and White coming out against irregular plurals?
If of “six geese” five migrated away, how many geese would be left? Answer: one geese, apparently.
Secondly, thirdly, etc.
“Modern usage prefers second, third, and so on.” Sure, but how did they do it in Latin?
Shall and will
There are two different rules given about when to use these, but fortunately it’s moot, as “shall” is gone. Despite the dire warning, drownings have been minimal.
A dictum that has gained no traction, perhaps because in the introduction White drew notice to how arbitrary this particular predilection of Strunk’s was. If only White had had another hundred and twenty anecdotes, we could have been spared the tiresome influence of this section altogether.
Thanking you in advance
Ugh. This reminds me of people who bewail modern indifference because the answer to “How are you?” is “Fine,” rather than some in-depth analysis of the psyche. These are set phrases. They add a dash of courtesy. That’s it.
That and which
Like less and fewer, we know the origin of this rule, and also like less and fewer, it started with a pleasant enough guy expressing a preference. His name was Fowler, and he noticed that non-restrictive clauses are introduced with “which.” A non-restrictive clause is a bit of information in a sentence that’s just additional, not defining — “The sofa, which was recently reupholstered, is massively comfortable.” You don’t need to know about the classy reupholstering job, I’m just throwing it in there.
A restrictive clause tells necessary information. “The sofa that the dog is sitting on has been in the family for generations.” Maybe I’m differentiating it from all the other sofas, because in this hypothetical scenario I’m fabulously wealthy and have many fancy and possibly antique sofas, but for whatever reason, this info is not just an aside. Anyway, as you’ll note, I started that clause with “that,” not “which.” But both are possible with restrictive clauses. “The sofa which my aunt gave me is truly hideous.” Little stuffy, but fine English.
Fowler saw this as an imbalance, and thought it would be, well, cleaner to use only “that” with restrictive clauses — balance out the usage, I suppose. Strunk and White (or actually just White), naturally, have turned this thought into an order. You hear the “it’s ambiguous!” argument thoughtlessly repeated with this one, but it ain’t. In speech, we say these phrases differently, pausing around non-restrictive clauses and motoring through restrictive ones, and in writing the commas or lack of commas do the job. We don’t need to double up by adding an artificial that/which distinction.
“The use of he as a pronoun for nouns embracing both genders is a simple, practical convention rooted in the beginnings of the English language. He has lost all suggestion of maleness in these circumstances.” Bullshit twice over. “He” has not at all lost connotations of maleness, and singular “they” is older and better. On which note:
“The furor recently raised about he would be more impressive if there were a handy substitute for the word.” THE FIRST WORD OF THIS ENTRY IS “THEY,” WHICH YOU TWO REJECTED WITHOUT EVEN ATTEMPTING A REASON! “They. Not to be used… Use the singular pronoun.”
AHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH THIS RIDICULOUSNESS CONTINUES TO HAVE AN EFFECT ON WRITING AHHHHHHHHHHHHH
“This showy noun, suggestive of power, hinting of sex… Use it sparingly. Save it for specific application.” Strunk and White confusingly both turned on and irked by phrases like “the thrust of his letter.”
“Many writers so [as “happen”] use it (usually when groping towards imagined elegance), but their usage finds little support in the Latin ‘breathe across or through.’” The relevance of which I’m sure will be explained at any moment.
Pro: agrees that “try and” do something is acceptable in some contexts. Con: says that “try to” is precise, implying that “try and” is not. They’re equivalent.
Forceful words tend to become generalized over time. Same concept as “literally,” and also taboos.
Eh. Ben Franklin hated “colonize” when it first appeared. Imagine if Strunk had gotten to see “impactful.”
Cite Latin one more time, fellas.
See Section V #8.
Also good advice, though not really about “would.” Nice to end this arduous section on a positive note.
An Approach to Style (With a List of Reminders)
White’s solo section. “Here we leave solid ground,” he says. You were never on solid ground, good sir.
1. Place yourself in the background
A forebear of the “I”-hatred that permeates academia and other formal writing, affecting even Zinsser, who managed to break through and write a warm, personal book. An example of how the writer must work through this miasma of anxiety, must put aside these ill-begotten rules, in order to write humanly.
2. Write in a way that comes naturally
An absolute contradiction of #1.
3. Work from a suitable design
A rehash of “Choose a suitable design and hold to it,” with better examples.
4. Write with nouns and verbs
Odd because White spends most of this point qualifying and weakening it, and as Pullum points out, flouting it. Seems to be intended as a corollary to “Use definite, specific, concrete language.”
5. Revise and rewrite
Good advice. Fundamental advice, even. And the reference to actual scissors and glue cracks me up.
6. Do not overwrite
7. Do not overstate
8. Avoid the use of qualifiers
Also good style advice, and the source of the useful points in section IV, those rare occasions when the goal is not to clutch at words and try to keep them from the inevitable tide of language change, but to note when words dilute a sentence. Zinsser’s “Bit and Pieces” chapter is his weakest, with the influence of Elements seeping through, but his advice on qualifiers is superb.
9. Do not affect a breezy manner
Hey, screw you. I don’t “affect” a breezy manner; I have a breezy manner. I’m writing naturally, remember? Though I may not need to be offended, since what he means by “a breezy manner” is not exactly clear. Seems like use of unnatural slang.
10. Use orthodox spelling
Once again a brief understanding that language changes, and even that the point of all this style and grammar and whatnot is to get the message across to the reader. But “the reader” has to mean people like Strunk and White — others are “illiterates” and should not be kowtowed to.
11. Do not explain too much
12. Do not construct awkward adverbs
“Nobody says tangledly and not many people say tiredly. Words that are not used orally are seldom the ones to put to paper.” White, did you even read part IV?
13. Make sure the reader knows who is speaking
14. Avoid fancy words
15. Do not use dialect unless your ear is good
Another point where the old-white-guy-ness that permeates the book breaks through to the foreground. “Do not use dialect unless you are a devoted student of the tongue you hope to reproduce.” No chance a native speaker of some dialect would want to write a book, then?
16. Be clear
Relates back to the syntax concept of “Keep related words together,” and does somewhat better by not attempting grammatical analysis.
17. Do not inject opinion
This book is entirely opinion.
18. Use figures of speech sparingly
19. Do not take shortcuts at the cost of clarity
20. Avoid foreign languages
21. Prefer the standard to the offbeat
Man, I have got to stop thinking about this. I don’t even care about Strunk and White. I’d be indifferent to their grammar peeves and proclivities if they weren’t so intent on accosting the passer-by with them — they can have their rules, and understand them or not, and write with them or not. Who cares?
…a ridiculous number of people care. And they care so much. It’s such an integral part of their personalities. I don’t understand. I just don’t understand.
One more Pullum quote to cap us off: “Linguists often find themselves trying to bring reason and argument to bear on topics where standard reactions are based on dogmatic intransigence, incoherent beliefs about logic, unfounded fears of ambiguity, warped ideas of history, blind trust in Strunk and White, or panic over imagined abandonment of standards.”
Thanks for being here, Pullum.
The Language Nerd
*Like me, I recommend the CGEL for meeee, but tragically the damn thing’s three hundred bucks. Also, whoa, did you scroll all the way down here for this? Enjoy your trip back up.
Got a language question? Ask the Language Nerd! firstname.lastname@example.org
Twitter @AskTheLeague / facebook.com/asktheleagueofnerds
Note: Your next regularly scheduled Tuesday post will be by the Paramedics Nerd — the Language Nerd’s exhausted.