Monday, July 4, 2016

Frankenstrunk: Birth of a monster

A post today by Geoff Pullum at Language Log mentioned this Boston Globe column on Strunk & White's Elements of Style. Since the original is behind a paywall (and badly formatted to boot), I've republished it here in a more palatable and accessible form.)

JUST IN TIME for Halloween, it's back: Yet another edition of The Elements of Style, William Strunk and E.B. White's persistently popular guidebook for writers. And this time it's in costume, decked out with dozens of gay, whimsical illustrations by Maira Kalman (interviewed in this week's Examined Life column). But The Elements's new clothes can't hide the worsening limp and spackled complexion that plague this aging zombie of a book. 

It was never a seamless creation, to be sure; the 1959 first edition merely sandwiched Strunk's 1918 handbook for his Cornell students, lightly edited, between White's introduction and his essay on prose style. But at least you knew Strunk was Strunk, vintage 1918, and White was White, circa 1958. Succeeding revisions, instead of blending the disparate parts, have left Elements a hodgepodge, its now-antiquated pet peeves jostling for space with 1970s taboos and 1990s computer advice. 

(The illustrated Elements is essentially the 1999 edition, with a couple of small restorations from the 1918 original. Not quite a restoration, alas, in the case of Strunk's introduction: The proofreaders overlooked one of his "Words Often Misspelled," so the opening sentence now promises "to give in brief space the principle requirements of plain English style.") 

Scanning the recent editions, you sometimes wonder what could possibly have been cut, given the absurdity of what remains. Don't use claim to mean "assert"? Mark Twain did it in 1869, the year Strunk was born. Don't contact anyone? It's a "vague and self-important" verb -- or so people said in the 1920s, when it was new. Don't use they to refer to "a distributive subject" like everybody -- unless you're E.B. White: "But somebody taught you, didn't they?" says a character in Charlotte's Web

Even Strunk's signature battle cry -- "Omit needless words!" -- is question-begging bluster: Which are the needless words? Needless for whom? That's the hard decision, and S&W's editing examples do not inspire confidence. Surely "He often came late" is not always better than "He was not very often on time." And as any courtroom witness could testify, "I did not remember" doesn't equal "I forgot." 

We're told we must write "It looked more like a cormorant than like a heron" --because, I surmise, without the second like the sentence might mean "It looked more like a cormorant than a heron looks like a cormorant." Should that remote possibility loom in a stretch of actual prose, of course we can repeat the like, but how often does it happen? A Google search suggests that writers add the second like only about once in 100 such comparisons -- and not all of those likes are necessary. 

Could this messy monster have evolved into a different beast? Probably not; White was in a trap the moment he started revising the 1959 Elements. That book could stand as a quirky appreciation of White's old teacher, its dated items just amusing historical artifacts. 

White surely knew that some of Strunk's crotchets --his insistence on I shall, say, or his rule that however, meaning "nevertheless," could not begin a sentence -- were becoming untenable; but they were Strunk being Strunk. 

But if White, in his revisions, admitted that aggravate could mean irritate (as it did in 1611), or that I could care less was not a mystifying mistake, his usage notes would lose their essential Strunkiness, that bluff certainty that had hooked him in the first place. 

So rather than join the reality-based usage community, White stuck with Strunkian dogmatism. Hence a book that tells us, in 2005, that offputting and ongoing are illegitimate; that hopefully is beyond the pale; and that six people is a solecism because there's no such thing as one people. 

Why does this sort of thing send reviewers into raptures? Maybe they remember, from their college days, a reassuringly slim volume that pretended it could solve their writing problems. The "Strunkian attitude toward right-and-wrong," in White's eccentrically hyphenated phrase, may still stir in readers the eternal hope that someone, somewhere, knows what he's doing. 

If that certainty is what you liked about your old Strunk and White, you'll find it, only slightly eroded, in the newly dolled-up Elements. But the artwork is merely a colorful shroud on a corpse that's overdue for burial. May it rest in peace, someday soon.

(Jan Freeman, Boston Sunday Globe, Oct. 23, 2005)

Wednesday, March 23, 2016

Six times more ambiguous

I wonder how many people had to read this correction twice, as I did:
A driver who texts while driving is six times as likely to be involved in a crash as a driver who doesn't text. A Business News article Saturday about driver-monitoring systems incorrectly said that a driver who texts is six times more likely to be involved in a crash than one who doesn’t. 
When I got it — the Wall Street Journal had written “six times more likely,” and now was “correcting” the wording to “six times as likely” — I knew it was meant for a small band of sticklers. These are the people who claim that “six times as likely” means “multiplied by six,” but “six times more likely” really — that is, properly, mathematically — means “multiplied by seven”: It’s the original amount plus six more servings. 

But I don’t buy it. As Merriam-Webster’s Dictionary of English Usage says, “It is possible to misunderstand times more in this way, but it takes a good deal of effort.” In real life, nobody uses “six times more” to mean “seven times as much” (and if they did, how would a reader know it, without the numbers?). MWDEU concludes:
The fact is that “five times more” and “five times as much” are idiomatic phrases which have — and are understood to have — exactly the same meaning. The “ambiguity” of times more is imaginary.
The same argument is aimed at “six times less” to mean one-sixth — which, unlike “times more,” often does trigger my editorial antennae. I’d consider changing it in copy, if it were at all distracting. But I stopped worrying about it once I noticed that Mark Liberman of Language Log uses it unapologetically, even in contexts where he’s wrangling complicated statistics. If it’s OK with him, it’s OK with me.

Further reading:

Bill Walsh disagrees, firmly:

Arnold Zwicky treats “times more” and “times less” in a post on the Recency Illusion:

Eugene Volokh notes that Newton, Herschel, Darwin, and Robert Boyle used “times less”:

Tuesday, March 8, 2016

Braving the gauntlet

(Originally published in the Boston Sunday Globe, March 28, 2004)  

"MARTHA STEWART may still have to run another legal gauntlet," predicted a story in this section two weeks ago, prompting a usage dissent from reader Richard Sachs of Chelmsford: Shouldn't that be a legal gantlet? he wondered. Dan Tanner of Westborough had already issued his challenge: Run the gauntlet, he e-mailed, "is getting past Globe editors. A gauntlet is a glove. A gantlet is what one might run."

The Globe stylebook sides with the plaintiffs: "A gantlet is a military punishment in which the offender ran between two rows of men who struck him . . . A gauntlet is a medieval glove worn by knights in armor." Simple enough, but there's a catch: Most of the world disagrees.

The problem is not with gauntlet, the glove, a straightforward borrowing from French; a gauntlet may be flung down or picked up, but it's never confused with a gantlet. No, it's gantlet whose checkered past makes it hard to defend with a pure prescriptive passion, for gantlet was compromised from the beginning. On its way from Sweden to England, where it first shows up in print in 1636, it was transformed from gattlopp ("lane-course") to gantelope or gantlope.

A decade later, the Oxford English Dictionary records, the new word had already acquired a folk etymology: An imaginative commentator suggested that it derived from "Ghent Lope," a punishment "invented at Ghent . . . and therefore so called." This fanciful notion didn't catch on, though, and gantlope, under pressure from the similar gauntlet, soon gave way. Gauntlet has meant both punishment and glove since at least 1676.

In the ensuing centuries, Britons seem not to have lost any sleep over the potential confusion. In the colonies, however, some worthy watchdog must have decided we could do better, and in the 19th century gantlet was temporarily revived as the word for the ordeal. It didn't last, though -- Clint Eastwood's 1977 movie "The Gauntlet" was about the purists' gantlet. And today, editors who like gantlet have to dig deep for lexicographical support. Most dictionaries call gantlet a variant of the standard, dominant gauntlet.

Newspapers beyond US borders also use gauntlet almost exclusively -- it's the spelling of choice in 99 percent of variations on the phrase run the gauntlet. And yet, US editors haven't knuckled under. In newspapers here, the gantlet and the gauntlet now run neck and neck* -- not a bad showing for the gantlet devotees, who must be a smallish band.

Why do they make the effort, when the rest of the English-writing world does fine without gantlet? Probably because every usage nut treasures a slightly different set of niceties. Bryan Garner (shockingly!) accepts bicep as a singular, though it should be biceps. If you keep the Latin singular, he notes, you're stuck with bicepses or bicipites for plurals -- a high price for accuracy. And yet, in his Modern American Usage, Garner argues for gantlet, distinguishes between masterly and masterful, and wants to keep both insure and ensure, though one spelling would suffice.

Similarly, syndicated columnist James Kilpatrick is ferocious on the difference between each other and one another, but indifferent to the widely censured comprised of (it should be composed) and unruffled by for free. Theodore Bernstein opposed the verb chair but was willing to give up on the farther/further (one literal, one figurative) distinction. And US journalists, it seems, have adopted gantlet as a shibboleth, a password that signals membership in a select linguistic community. It may not be the best place to make a stand, given gantlet's lack of adherents and its long-since-corrupted form. But for some usage obsessives, there's no cause like a hopeless cause.

*2016 note: The data about newspaper usage comes from Nexis searches.

Thursday, August 6, 2015

You may be a who, or you may be a that

(Originally published in the Boston Sunday Globe, August 24, 2003.) 
THE WORD / Jan Freeman: Who that?

"The man that is failing the people more than anyone is Gray Davis," said fledgling candidate Arnold Schwarzenegger of the governor he hopes to supplant. Reader Dave Furlong, a California voter himself, would have changed that sentence: He liked our recent treatment of which vs. that, he e-mailed, "but I wish you had addressed that/who also, as in 'Everyone that wants to go, line up here.' In editing, I find it's a very common error."

Very common, yes; an error, only sometimes. Arnold may not be a native speaker, but here his English is traditional, if debatable. The insistence on who for people, on the grounds that calling a man (or woman or child) that is somehow insulting, is a fairly modern prejudice.

"A woman that deliberates is lost," Joseph Addison wrote in 1713. The Oxford English Dictionary also coughs up O. Henry's "I'm no traitor to a man that's been my friend" (1910) and Ring Lardner's "Imagine being married to a woman that plays five hundred like she does" (1924).

Some stylebooks, including the Globe's, do tell writers to avoid that in referring to people. But throughout the English millennium -- from before the Wycliffe Bible's "the people that dwelt in darkness" (1382) to "The girl that I marry" (Irving Berlin, 1946), that has been a people pronoun.

That had a brief fall from grace in the 16th century, when a fad for using who or whom instead swept the English literati (including Addison, who revised his writings to reflect his new faith). But by the 20th century, that was back in favor, as the usage-edict record attests.*

In a 1906 American grammar textbook, John Wisely tells pupils that who "expresses persons or personified things," while that is for "inanimate objects, lower animals, persons." Fowler, in the 1926 edition of Modern English Usage, confesses that he'd like to see even more use of that in constructions like "the distinguished visitors that the Crawfords had."

Bergen and Cornelia Evans, in A Dictionary of Contemporary American Usage (1957), note mildly that some writers prefer who where persons are concerned: After centuries of "Our Father that art in heaven" (or "Our Father which art"), the Lord's Prayer now usually says who. But even Miss Thistlebottom, Theodore Bernstein's personification of grammar-school dogma, didn't draw a clear line on that: She would have told you, Bernstein says, that "which normally refers to things, who to persons, and that to either."

In his 1996 updating of Fowler, Robert Burchfield tries to make it simple: "Normally use who . . . following a human antecedent and that (or which) following an inanimate antecedent. Either who or that may be used when the antecedent is animate but not human, or when the antecedent is human but representative of a class."

Those guidelines (which would call Schwarzenegger's usage wrong) set forth a conceptual rationale, making the choice of who or that dependent on the abstractness of the pronoun's referent. So the barking dog that keeps you awake, two streets over, is different from the dog who goes out for a romp with you every day, an individual with a name. A cyborg that's on the assembly line becomes a who when it's programmed as a hunk. But unless it's banned by your local authority, that is a pronoun for people too -- some of the people, at least, some of the time.

*There's a detailed discussion of personal that in the indispensable Merriam-Webster Dictionary of English Usage. I can't imagine why I didn't mention it here.