I USED TO read to Jack in the library. 

‘Let’s go,’ I’d whisper. And we’d walk the cracked concrete path from the multipurpose room with its short-pile orange carpet to the school library at the back of the grounds, where a librarian sat wordlessly at the borrowing desk, eyeing us with suspicion.

I was supposed to be teaching Jack to read and write in English. Originally from China, he was eleven years old at the time, keeping in step with his other educational milestones but still unable to decipher or write more than a few basic words. We would hunch over a desk together, trying to speak over the intermittent blasting of two dozen plastic recorders, and work through exercises designed to develop, like a muscle, his automatic decoding skills.

But Jack loved stories. He could tell them. He was enthralled by them. Even now, more than a decade later, I still think of him when I see an image of the Tasmanian tiger, whose relentless hunting and eventual extermination – even in my clumsy rendering – seemed to prompt as much dismay for him as the plight of the dinosaurs. ‘What do you think extinct means?’ I asked him once. ‘I don’t know,’ he replied. ‘Did they smell very bad?’

On the mornings that dragged, when I sensed Jack’s enthusiasm beginning to fade, when we’d traversed the short path and the library door had wheezed shut behind us, I’d fold back the pale-blue cover of my childhood copy of Charlie and the Chocolate Factory and read to him aloud. Jack, pulling at the sleeves of his school jumper, would laugh as I tried to do the voices. We’d run our fingers over the spiky, gestural illustrations, fold down the corner of a yellowed page for next time.

Even the thin-lipped librarian would smile.

‘PUFFIN BOOKS AND the Dahl estate should be ashamed,’ Salman Rushdie tweeted on 19 February 2023, days after news broke that the publisher had consulted with sensitivity readers to release a range of revised classics. According to multiple outlets, seventeen of the author’s beloved books, including Matilda, The Twits and James and the Giant Peach, had been redacted or retrofitted, in places, to bring the language into line with ‘contemporary sensibilities’.

Riffle through the front pages of a brand-new copy, and there on the copyright page, in a delicate serif font, you’ll now find a publisher’s note: ‘This book was written many years ago,’ it gently explains, ‘and so we ­regularly review the language to ensure that it can continue to be enjoyed by all today.’

This was not, by any stretch, the first time an enduring classic had been tweaked to reflect changing cultural mores or a particular concern for young and impressionable readers. The original Nancy Drew books were revised as early as 1959, when the publisher, Grosset & Dunlap, attempted to modernise the series by shortening the books and purging them of racist stereotypes. Agatha Christie’s whodunnits, Enid Blyton’s middle readers and Ian Fleming’s James Bond novels have been periodically edited to remove dated and offensive references. Ray Bradbury, author of Fahrenheit 451, reluctantly acquiesced when he discovered US publishers had stripped the book of ‘damn’, ‘hell’ and other alleged vulgarities to help appease ‘classroom anxieties’ in the lucrative education market (a similar fate to befall RL Stine, who claims Scholastic went behind his back to bowdlerise dozens of books in the Goosebumps series). The Adventures of Huckleberry Finn by Mark Twain was famously republished in 2011 to replace more than 200 instances of the N-word. And Theodor Seuss Geisel – aka Dr Seuss – made headlines in 2021, exactly thirty years after he’d died, when Dr Seuss Enterprises ‘discontinued’ six of his picture books to cleanse the catalogue of strong racial overtones. The list goes on.

But as Rushdie’s tweet has come to epitomise (he famously condemned the revisions as ‘absurd censorship’), the so-called rewriting of Dahl’s novels ignited a controversy of such unprecedented intensity that it not only dominated headlines for months but also laid bare, particularly on social media, a deepening rift over questions of artistic freedom, cultural preservation and the relentless commoditisation of intellectual and creative output. The ‘desecration of Dahl’, in the words of conservative commentator Frank Haviland, amounted to ‘sterilising works of art and meddling with genius’; British comedian David Mitchell described the new editions as a ‘tin-eared and dreadful’ attempt to wring Dahl’s legacy of every last dollar. 

Though a spokesperson from Roald Dahl Story Company insisted the changes were largely trivial ‘in terms of the overall percentage of text’ that had been removed or updated, British newspaper The Telegraph – which described the ‘overhaul’ as an ‘example of a growing trend in children’s publishing for content that nobody can find offensive’ – went as far as to circulate an exhaustive list of the changes, tabulating the original text side by side with text from Puffin’s reimagined editions.

Many of these changes are quite natural and seamless. Gendered terms, for example, have either been left out or replaced with neutral alternatives. (In the 2022 editions, ‘chambermaids’ are ‘cleaners’. Families ‘with a husband, a wife and several children’ are now simply ‘families’.) Derogatory or questionable language relating to appearance, class and identity has been diluted – or deleted altogether. Double chins and flabby necks are gone. Freaks, hags and ugly old cows have been recast as more sympathetic substitutes. People don’t laugh like mad; they laugh wildly. Faces turn pale instead of white. Grandmas who are fond of gin are no longer allowed to have a small nip every evening; they merely like to. Matilda reads Austen, not Kipling.

I feel disloyal for thinking it, as a writer myself, but it’s hard to imagine most readers noticing, especially first-time readers inured to what some have called ‘the hair-trigger sensitivities of children’s publishing’. At the same time, it’s difficult to defend some changes as fitting or otherwise effectual. A discerning reader – an adult, presumably, more familiar with the original text – will detect occasional inelegancies and anachronisms where the Puffin rewrites dampen Dahl’s original turn of phrase, distort his intended meaning, or simply fail to defuse the original slight. ‘Let’s not ask,’ the novelist and ­essayist Francine Prose points out, ‘if “enormous” is really less hurtful than “fat”.’

Debating their literary or even moral merit, however, seems largely beside the point. We know that Dahl was fussy with words, a tireless tinkerer who claimed that by the time he was ‘nearing the end of a story, the first part will have been re-read and altered and corrected at least one hundred and fifty times’. We know that he was openly resistant to editorial interference, arguing that a writer’s ‘only compensation is absolute freedom’, which left him frequently at loggerheads with his publishers. And we know that he took reading for pleasure seriously, invariably favouring the imaginative and the immersive over the instructive. We can safely assume, then, that no matter how well intentioned, the impulse to purify his books of their most harmful stereotypes and objectionable language would not have been met with cheerful compliance.

But the curious thing about this impulse is not the way it polarises.

It’s what it takes for granted.

I’m not the first to observe that our desire to keep literature safe, in both senses of the term, reconciles the rubrics of conservative and progressive politics alike, defying even our best attempts to position it ‘neatly and conveniently’ at only one end of the ideological spectrum. As Trisha Tucker points out in The Conversation, what harmonises these efforts is ‘a professed desire to protect young readers from dangerous content’, which seems sensible enough on the surface. Look a little deeper, though, and another consensus starts to materialise, one that elaborates the effects or results that ‘dangerous content’ only implies. ‘In all times and places’, writes Jonathan Zimmerman, a historian of education, the ‘deepest fear of the censor’ is that readers will ‘get the wrong idea’ – and this is true, experts have noted since at least the 1980s, ‘whether the call for censorship comes from the right or from the left’. 

It seems that whichever way we lean, the ascendant value of reading coagulates around some kind of connection between what we read and how we think, a collective vision that amplifies literature as cultural intervention. Our line of reasoning goes something like this: change the things people say or write, and you’ll change the way we think about the world, and so the world ends up changing, too. By this logic, books – cultural artefacts often threaded with the social values of their time – represent a particularly gnarly thorn in our sides, capable of perpetuating a world view at odds with contemporary norms and aspirations.

Yet if censorship controversies are anything to go by, it’s an intervention, funnily enough, that usually takes place at the level of vocabulary. Words matter, the new Puffin editions remind us. But how do they matter? And how much?

THE TRUE ALCHEMISTS, according to William H Gass, don’t change lead into gold. They change the world into words. In the twenty-first century, that magic is thought to move in both directions. Don’t words also render the world?

Philosophers have long underscored a connection between language and perception, but a plausible theory didn’t catch on until the mid 1900s, when anthropologist Edward Sapir and his student Benjamin Lee Whorf stumbled across an intriguing phenomenon. Whorf discovered that the Native American Hopi have no way to express time, suggesting that their sense of life and the world is fundamentally at odds with that of ours, as speakers of English. The duo had originally set out to show that cultures traditionally dismissed as ‘primitive’ (usually those that hadn’t developed a written system) were just as sophisticated as those celebrated for their cosmopolitanism. But the Sapir-Whorf hypothesis – that language somehow determines or at least influences thought – rapidly flourished into something more, something that’s assumed an almost canonical status, owing, I suppose, to its deeply romantic conviction that we are what we speak.

Let me explain. Early Sapir-Whorf adherents proposed that human beings are ‘very much at the mercy’ of whatever linguistic habits they’ve come to share. Together, the vocabulary and grammar of a given language form a ‘loosely laced straightjacket’ for thinking, in which we’re bound to construct realities so different that we’re not merely occupying ‘the same world with different labels attached’.

Laboratory research seems to support this proposal, though not nearly to the extent Sapir and Whorf originally imagined. So pervasive is the so-called ‘weak’ strain of the hypothesis, however, that even if you’ve never heard of linguistic relativity, you’re probably familiar with some of the neo-Whorfian science that’s filtered its way to a receptive public. I’m not talking about the reasonable observation that we name culturally important things by inventing new words or repurposing old ones (microplastic, zoom-bombing, covexit). I’m talking about the idea that language reliably bears or bounces back a certain attunement to the world, so if we don’t have an obvious way to express something, we foreclose the opportunity to notice it.

Take colour, for starters.

It might seem an arbitrary choice, but using colour terms such as black or yellow is one of the most characteristic but variable ways we habitually describe – and apparently perceive – the world. Russian is one of the most frequently cited examples. For Russians, something is never just blue. It’s either siniy (dark blue) or golubój (light blue), just as we differentiate between red and pink in English. And native Russian speakers perform faster on tests distinguishing different shades of blue than speakers whose languages don’t make the same obligatory distinction.

In another classic experiment, speakers of languages that verbalise a difference between green and blue (like German, Korean and Icelandic) are compared with speakers of languages that don’t (like Tarahumara, a Uto-Aztecan language spoken in Mexico, and the Himba dialects from Namibia). When two groups are shown an array of colours, including various shades of bluish greens or greenish blues, something predictable happens: speakers who ‘colexify’ variants of blue and green using one catch-all term (what some linguists call grue) find it slightly more difficult to distinguish between those colours, whereas speakers who have separate words are quick to recognise shades that lean blue.

Recorded to the millisecond, these findings suggest that our perception hinges not only on what we actually see – determined by the sensitivity of the typical human eye to around one million different colours – but also on the relatively few colour terms available in whatever language we happen to use.

Other experiments reveal that if you’re a monolingual Mandarin speaker, you may think about time as moving up and down, rather than from left to right, marking a distinct conceptual difference in attitudes towards ageing. Or if you speak a language without a distinct future tense, such as Finnish, you may care more about what’s to come, including frightening realities such as climate change. Researchers have been able to isolate all sorts of subtle perceptual idiosyncrasies stemming from the way we habitually express things, leading us to surmise that how our language works makes us see the world through word-shaped glasses.

Simple enough, right? Sensible. Seductive.

But as the linguist John McWhorter points out in his book The Language Hoax, the quirky but infinitesimal differences we observe in the lab are often distorted as ‘preludes to something much grander’: a pervasive and predictable world view. And if the words we have at our disposal do afford us some unique epistemological insight – if the power to name is indeed the power to see – we’re left with an uncomfortable dilemma. Linguistic relativity is an enjoyable idea when it implies something special about the way we speak and think; it’s rather less flattering when it implies some kind of deficiency.

Not only that. If we suppose that language functions as a kind of cognitive colander, it follows that ‘gobblefunking’ around with words (as the BFG might put it) carries serious consequences for the speakers at its so-called mercy.

This deference for the constitutive or destructive capacity of language lies at the heart of George Orwell’s fictive but prophetic Newspeak, an artificial derivative of English that’s been so simplified, regularised and robbed of etymological richness that it demonstrates a ‘geometric uniformity’ capable of crushing independent thought entirely. ‘All dystopian languages technically belong to Whorf’, literary scholars have pointed out, highlighting the recurring alignment of language control with thought control in the bleak science fiction that proliferated in the second half of the twentieth century.

As you may remember, in Nineteen Eighty-Four the powers that be – Ingsoc – are destroying words, ‘scores of them, hundreds of them, every day’, cutting language ‘down to the bone’. The minor character Syme, a philologist working at the Ministry of Truth on a new edition of the Newspeak dictionary, is thrilled by the long-term implications of this work. ‘Every year fewer and fewer words,’ he gushes to our long-suffering Winston, ‘and the range of consciousness always a little smaller. Don’t you see the beauty of that?’

It’s a cautionary tale, to be sure. Orwell was convinced that if thought corrupts language, language also corrupts thought – whether by the presence of bad words or the absence of good. In the novel’s totalitarian Oceania, once Newspeak has matured, once society achieves total dominion over the universe of discourse, the conflation of language with reality will be complete. The thing that most excites Syme is the very same thing that horrifies the reader: erase the word, erase the concept, erase the phenomenon.

THE TROUBLE IS – for better or for worse – words don’t work this way. Concepts may find expression in a given language. Or they may not. Sometimes a reason is obvious. With alarming regularity, though, that reason is chance.

Common sense alone tells us that what we’re capable of thinking and what we’re capable of saying are not the same thing. When we struggle to find the right words, when a word lingers on the tip of our tongue, when words just won’t do something justice, we understand intuitively that thinking takes place independently of expression. It’s in this way that supposedly untranslatable words, for which our language has no exact or close synonym, are often so deeply pleasurable: not because those words reveal something about a worldview that’s unfamiliar or foreign to us but precisely the opposite. Schadenfreude has penetrated the English lexicon because there’s a certain universality to the experience of delighting – usually opportunistically – in the misfortune of others. Who among us has not tartled, a Scottish verb meaning to hesitate before introducing someone because their name has slipped your mind? Or felt a sense of gigil, a Tagalog term that captures the overwhelming urge to pinch or squeeze something irresistibly cute, such as a baby’s chubby thighs or a kitten’s whiskered cheeks? 

The idea that expression and thought are neatly symmetrical really only works when our noses are touching the data. The further afield we search, the higher the bird’s-eye view, the more we see that languages all manage to accomplish much the same thing, the same meanings, even without certain words, bits of words or ways of arranging them. As McWhorter puts it, ‘rather than each revealing a different take on thinking, languages – beyond having names for cultural tokens – are variations on the same take on thinking: the human one.’ 

To hope or to fear that words alone can channel or trammel thought in any profound or inevitable way is therefore feasible only if we disregard how ‘hopelessly motile’ their connotations are. And connotations, we find, demonstrate an uncanny knack for attaching to new words as old ones are forcefully decommissioned – a process sometimes called ‘the euphemism treadmill’. So, shellshock becomes battle fatigue becomes operational exhaustion becomes post-traumatic stress disorder, as the late George Carlin observes in a well-known comedy routine. In this case, forms may change while meanings or associations stay the same. Or forms stay the same as meanings move on, picking up odours and flavours from the environment around them like butter left open in the refrigerator. The real magic of language is that we manage to make any sense at all.

It’s not that words don’t matter. Of course they do. We bend them this way and that, to express love, to convey disgust, to reveal or obfuscate, to pledge, deny, comfort, wound. We venerate their beauty and recoil from their hidden barbs. Words sound. They look, taste, feel and smell. You can no doubt think of a word you instinctively hate – moist, anyone? – or a series of words you love, words one after the other that you can recite off by heart. (Here are some of mine: ‘Words can be like X-rays if you use them properly,’ Aldous Huxley writes in Brave New World, ‘they’ll go through anything.’ Or these, from David Foster Wallace’s The Pale King: ‘How odd I can have all this inside me and to you it’s just words.’)

But tell me what ugly means, tell me about home, about the particular way, when you were a child, the sky was dark. Tell me whether liberal is good and bossy is bad, if limerence and defenestration are absolutely necessary or ­irregardless definitely wrong, and we may find we can never agree.

In a sense, words matter because they are matter: pliant, impure and almost wilful in their powerlessness to either control or be controlled. Thinking of language as ‘a list of words with set meanings’, McWhorter reminds us in his New York Times column, ‘is like thinking of the ­position of the clouds right now as somehow fundamental rather than as a passing moment’. 

Better to think of words not as containers that shape the meanings inside them but as vessels into which we pour our own meanings, in different moments and places, according to whatever whims and exigencies make the most sense at the time. Words, we find, are always in motion. We might, from time to time, catch them in the pages of a book, or in a film or song, like an insect pinned to styrofoam. But they never stay still for long.

‘A word is dead when it is said, some say,’ goes a little poem by Emily Dickinson. ‘I say it just begins to live that day.’  

JACK AND I never finished Charlie and the Chocolate Factory. Fourth term ended. Jack finished primary school. I moved to another city to start my PhD.

In my mind’s eye, though, we’re still sitting at that desk in the multi­purpose room with its burnt-orange carpet. A textbook is open in front of us, and Jack traces his finger along a line of words, saying each one out loud. Stream. Ray. Fry. Sky. Spy. Squeal. Spy. Stray. Sky. Slay. Ray. Stream. The point is not to make sense but to get faster and faster.

Of course, this isn’t reading as we’d usually think of it. Recognising words effortlessly and instantaneously is a foundational skill that underlies comprehension, but reading is about much more than individual words. As the inimitable Stanley Fish tells us in How to Write a Sentence: And How to Read One, words are just discrete items ‘pointing everywhere and nowhere’. Only once they’re tied by ‘ligatures of relationships’ to one another do they congeal into something we can ‘contemplate, admire, reject, or refine’. Put another way, words get us halfway there. Reading is about doing the rest.

In the early 1970s, Dahl became embroiled in a war of words with one of his most vocal critics, fellow author Eleanor Cameron. In a letter addressed directly to Cameron, he rejects the assertion that his books are somehow harmful to young readers, emphasising a prickly opposition to ‘patronising’ children with didactic literature and a disdain for the idea that nasty stories mean nasty authors – or make nasty kids.

I’m inclined to agree, not in defence of Dahl himself, whose smiling face on the inside cover of my brand-new Puffin editions almost certainly exaggerates his benign appeal. Nor do I mean to say there’s nothing squeamish to discover in his stories, characters or turns of phrase. Rather, I agree because I can’t help but concede to the capricious quality of words and the ultimately incalculable ways that readers layer texts with their own meanings, the way – if I can pilfer an adage from reader-theorist Louise Rosenblatt – people also ‘happen to books’.

Wolfgang Iser, another reader-theorist, likens the words of a book to a constellation in which we see our own patterns: ‘Two people gazing at the night sky may both be looking at the same collection of stars, but one will see the image of a plough, and the other will make out a dipper. The “stars” in a text are fixed; the lines that join them are variable.’

So, what troubles me about the way we sometimes talk about books (be it George’s Marvellous Medicine or The Satanic Verses) is its aggressive narrowing. A book becomes language. Language becomes individual words. Words become determinate meanings – some acceptable (in a given time and place) and others objectionable. But when we reduce a book to words like this – when we cleave them from the materiality of the book itself, from the contexts in which we read, from the narratives in which they’re embedded, from the recalcitrant nature of words – we abrogate the possibility of those words meaning something else or something more. So convinced are we that preserving the language of a given book is to perpetuate a fixed meaning or view that we deny the reader an opportunity to draw the lines their own way.

Turning the pages of a Dahl novel now, furtively on the train to my inner-city office, I’m a different reader from the one who read aloud to Jack in the library more than a decade ago, and a different reader, it almost goes without saying, from the one who devoured The Witches and The BFG tucked up in my childhood bed. 

And so the book is different, too. 

Isn’t this the value of ‘good’ literature? To endure, even as the language and stories date, even as some good words become bad (or bad words become okay), because we’re able to make them new, to cast them in a different light? Isn’t part of the sorcery of story the chance to contemplate how things have changed and to find in words the residue of a reality we might no longer welcome, tolerate or reject? Rather than indulging in some kind of selective linguistic amnesia, couldn’t we instead marvel at the ways we’re able to happen to books, over and over?

How whoopsey-splunkers, I say. How absolutely squiffling.

I’m grateful to Waleed Aly, Scott Stephens and John Safran for participating in the Brisbane Writers Festival panel ‘War of the Words’ that sparked the idea for this essay.

_{Image credit: Getty Images}