Prevailing winds

FOR half a century, language experts have fallen into two camps, with most lexicographers and academic linguists on one side, and traditionalist writers and editors on the other. Should language experts aim to describe the state of the language accurately? (Webster’s Third New International Dictionary, in 1961, shocked the world by including common but disparaged “ain’t” and “irregardless”.) Or should they prescribe how the language should be used (“Irregardless ain’t a word”)? Over the decades, the two sides have traded insults; prescribers are authoritarians in denial about the real world and describers are permissivists with no standards.

Two authors in the past two years have made clear that it is time to move on. Steven Pinker is a describer, a linguist and cognitive scientist. But in 2014 he published “The Sense of Style”, a guide to good writing that ended with a section of prescriptions: do this, not that. They were grounded in description, not dogma—but prescriptions they were nonetheless.

Now come two new books by Bryan Garner, a proud prescriptivist who reaches the same point from the opposite direction. Mr Garner has tangled with Mr Pinker and other descriptive linguists. His explicit aim is to tell people what they should and shouldn’t do. But he has also called himself a “descriptive prescriber”, and this is clearer than ever before in the fourth edition of his masterly usage dictionary, “Garner’s Modern English Usage”, and a new book, “The Chicago Guide to Grammar, Usage, and Punctuation”. These new books rely not on mere clippings, but on big data: millions of books scanned by Google. This lets Mr Garner compare “he pleaded guilty” with the upstart “he pled guilty”. “Pled” is gaining ground, but “pleaded” is still three times as common in books. On this basis, Mr Garner prescribes: stick with “pleaded”. But he allowed Google’s data to change his mind, too: “run the gantlet”, however traditional, has long been outnumbered by examples of “run the gauntlet”, so he has accepted the newer usage.

The conflict between description and prescription should never have become so bitter. Mr Pinker is a fine English stylist; it is no surprise that he has opinions on whether some words and formations are better than others. And Mr Garner is a deeply read man (and a lawyer), so it should come as no surprise that he marshals evidence. But both camps were ill-served by less thoughtful standard-bearers. Many clueless prescribers really did push dud rules: the ban on split infinitives, the ban on ending sentences with prepositions, the notion that “since” cannot mean “because” and so many more. These were passed down from teachers to students over generations. When academic linguists began systematically investigating English by looking at texts and listening to speakers, they found that many such “rules” were anything but, and some began taunting the rule-promoters. They also sought to defend non-standard dialects, where for example double negatives (“I ain’t got no”) are ordinary, not ignorant.

In the pushback against a history of prejudices, prescription represented authority and tradition, and description represented democracy and progress. But sensible writers on both sides have come to agree, however tacitly, that there is a variety, called standard English, with rules that can be found by looking at large volumes of the stuff. The best prescribers are becoming ever more informed, and the describers more comfortable with the idea of giving people “right” and “wrong” judgments on standard English.

A sensible consensus emerges on most usages. Linguistic liberals and conservatives may still disagree on smaller issues: Mr Garner prefers the traditional “healthful” when talking about things like diets and exercise; most people prefer “healthy”. Mr Pinker defends “more unique”, whereas many pundits still reserve “unique” to mean an unscalable “one of a kind”. On these issues, reasonable people can disagree: Johnson is one of those who will reserve “He literally exploded laughing” to refer to a bloody scene requiring a mop, even though he knows many great writers have used “literally” figuratively.

In the end, the test of a good pundit is one who will declare the methods and evidence that went into a judgment call. On that, the best descriptivists and prescriptivists should be ready to sign an armistice in the long and—not literally—bloody usage wars.