It is perfectly possible to reconcile strong opinions on individual points of grammar and usage—including dislike of a particular usage, or fear that a change to the language might introduce confusion—with a belief that the language on the whole is built to adapt, to minimise confusion.

What do you think of that cover? I’d read the whole book before I realized that all those letters made up the face of a cat. Probably that’s just me, though.

Lane Greene is an American journalist who writes the Johnson column on the topic of language for The Economist. This is his second book about language—the first was You Are What You Speak: Grammar Grouches, Language Laws, and the Politics of Identity (2011). That dealt with language as a political tool—about how people use language to differentiate “us” from “them” (and how governments can use minority-language suppression to force “them” to be more like “us”). This book, Talk On The Wild Side (2018), is subtitled The Untameable Nature of Language. It covers some of the same ground (there’s a big section on identity politics and minority languages), but deals mainly with how languages, like species and ecosystems, evolve and adapt. Greene’s point is that it’s foolish to imagine we can stop that process, and also that we shouldn’t worry about it—all those young folk with their annoyingly repurposed words and their disregard for the good old rules of grammar aren’t going to break the language, because no language in the history of the world has ever been broken by its speakers. The meaning of words will slowly evolve, but we’ll never end up without a word for an important concept, and we’ll always adjust the way we speak to avoid ambiguity. And if the old rules of grammar aren’t being applied, that doesn’t mean there’s no grammar—simply that new grammar has replaced the old.

In particular, Greene wants to convince us that we should ignore the self-righteous posturings of the Grammar Police. He offers several reasons: 1) They are ignorant of real grammar; 2) There is not, and has never been, only One Right Way to say anything; 3) Language will continue to evolve despite efforts to stabilize it; 4) Absolutely nothing bad will happen if you commit the Grammar Crime of ending a sentence with a preposition, or using “which” in a restrictive relative clause—these are just invented notions (and in some cases we can even name the people responsible*).

Greene zeroes in on Nevile Gwynne (author of Gwynne’s Grammar) as a particular focus for his counter-arguments. Gwynne is a useful illustrative target—he’s high profile (at least, within the severely limited demographic of those interested in grammar), and he peddles all the well-loved but baseless old nonsense about avoiding split infinitives and the singular “they”, always putting possessives before a gerund and using the nominative pronoun in a subject complement. But such is the depth and breadth of Greene’s onslaught against Gwynne, one does get the creeping feeling there might be some history between them:

Gwynne has a reasonable grasp on some elements of traditional grammar, and it’s hard to take issue with his description of proper punctuation. But on nearly every complicated, contested or even interesting point, he gets it wrong, as if he had an unerring instinct for getting it wrong, a compass with a needle that consistently pointed any way but north.^†

But there’s a lot more to this book than the debunking of fusty old grammar myths. There’s a chapter on invented languages, including Loglan and its successor Lojban—languages constructed on strict logical principles which are puzzling difficult to speak fluently. And another chapter on the problem of teaching computers how to translate between languages—the current (moderate) success of translations software didn’t happen until programmers stopped relying on intricate rule-based systems, and instead started exposing adaptive systems to large amounts of real-world data. The lesson here, Greene suggests, is that natural languages simply aren’t logical constructs, as the Grammar Police would like to believe.

And there’s a fine chapter on how languages change over time—sounds shift, meanings shift, grammar mutates. Greene talks us through the Great Vowel Shift in English, which accounts for why our spelling now so poorly matches our pronunciation. And he deals with the many incarnations of words like buxom and nice, the meanings of which are now a very long way from where they started (“flexible” and “stupid”, respectively). And finally, there’s a fascinating introduction to the theory that the grammatical structures of languages cycle continuously between synthetic and analytic forms. Modern English is largely analytic, relying on placing individual words in a particular order so as to convey meaning; Old English was synthetic, embedding complex meaning in single words by the use of noun cases and verb declensions that have now all but vanished. By gradually eliminating most of the word-endings that did grammatical duty in Old English, we’ve become more reliant on word order in Modern English to tell us (for instance) which noun is the object and which is the subject of a verb.

These changes are slow, and writing has only been around for 5000 years, so we need to look back to the most ancient texts in order to see a language pass through a complete cycle:

Old Egyptian had a complex verb structure with suffixes and prefixes, Late Egyptian lost most of these and uses multi-word phrases instead, and Coptic, Egyptian’s descendant, has developed complex verbs again.

There’s a nice chapter on code-switching—how we all switch registers between conversational and formal usages, all the time, according to our situation and our audience, and another on “framing”—using particular word choices to attempt to influence the mental metaphor a listener will adopt. So arguments are framed as being “pro-life” or “pro-choice”, because no-one wants to think of themselves as “anti-life” or “anti-choice”.

If that all sounds like a gallop through a lot of loosely related topics, that’s because it is. But it’s an entertaining and informative gallop, and at its heart is a cogent argument that we should stop trying to tame language by imposing spurious rules upon its speakers, and instead take delight in observing its behaviour in the wild.

* Greene tells us that John Dryden is responsible for the alleged rule against the “stranded preposition”, and H.W. Fowler for the idea that “which” should only be used in non-restrictive relative clauses.

^† For another damning critique of Gwynne’s Grammar, see the clever and entertaining review at Stroppy Editor.