A lesson from George Orwell for science communication

Trevor Butterworth

Executive Director, Sense About Science USA

It was announced in the U.S. as “George Orwell’s Brilliant Guide to Writing Well,” the New Republic’s editors, perhaps, judging that the author-given title, “Politics and the English Language,” was too obscure for an American audience, and missed the needling compulsion that comes with inducing self-doubt while offering the possibility of self-improvement; for who, among the literary magazine’s readers, would not want guidance or reassurance that they, too, could write as well as the brilliant Orwell?

History has, in one sense, vindicated the New Republic: Orwell’s essay—known by its original title— has been assigned to countless students; accolades from writers and academics wantonly outnumber its 5,000 words. It is “seminal,” a “classic,” and a “necessary” read. “Politics stands,” says Michael Hiltzich, “as the finest deconstruction of slovenly writing since Mark Twain’s ‘Fenimore Cooper’s Literary Offenses.’” Former London Times editor Harold Evans begins his recent New York Times bestseller, “Do I Make Myself Clear: Why Writing Well Matters,” by immediately genuflecting to ‘Politics’ in the very first sentence, and noting that Orwell has done as much as any man to “rescue” us from the corrupting power of bad language.

Occasionally, an iconoclastic voice shatters the sacerdotal consensus. It is precisely for his prose style, says the novelist Will Self, that Orwell is the “primus inter pares” of “literary mediocrities.” But whether a careful critical argument or a ‘hot take,’ the dissenter’s gripe-filled slingshot inevitably fails to take down the target. In conflicts over literary greatness, stature counts.

There are, however, two important, and somewhat overlapping, constituencies, where eye-rolling over ‘Politics’ deserves our attention: those of linguists and teachers of English composition.

Far from being a brilliant guide to writing well, ‘Politics’ turns out to make very little sense to those learning to write; in fact, as Cleo McNelly ventured in the journal College English, it is “a prime example of how not to teach composition.”

Take, for example, Orwell’s first rule: “Never use a metaphor, simile, or other figure of speech, which you are not used to seeing in print.”

Of course, that makes lots of sense, right? You want what you write to be fresh and vivid, free of the hackneyed phrasing that suggests you haven’t thought carefully and precisely about what you are trying to say. Yet, as McNelly, notes:

“Orwell presumes his readers have already been taught to write badly. This presumption may be true of the average academic, but it is certainly not true of the students we see in classes today. The problem is that they have hardly been taught to write at all.”

In this situation, clichés and formulaic phrases function as crampons and ice picks; they allow students to ascend the difficult path toward being writers. The idea that they should wince at the thought of using a stale metaphor makes little sense to those who have no experience identifying metaphors or sterility.

Or what about Orwell’s injunction to “never use a long word where a short one will do?” That presumes, says McNelly, “an active lexicon of word choices, which the student doesn’t have, and which they can only get by practice in moving from synonym to synonym both up and down the ladder of abstraction.”

There is more; much more. Inquire into the rejection of the passive voice at your own risk, but know that linguists have the statistical goods demonstrating not only that Orwell ignored his own rule but that he did so with gusto, employing passive constructions at a greater rate than other writers.  

“Those who for years have taught ‘Politics and the English Language’ as a sacred text in English composition courses will have cause to blush,” says Norman Page, summing up the extensive case against Orwell’s rules in The Journal of English and Germanic Philology. “Orwell is very far from being a profound or original thinker on the subject of language. He is uninformed… he is inaccurate… he is a dogmatic and unreliable guide.”

To borrow a useful Orwellism, sometimes one struggles to see what’s under one’s nose: so at the risk of stating the obvious, we find ourselves with a guide to writing that is lauded by professional writers as one of the best guides to writing ever written, and which has been assigned, on that basis, to those learning to write so that they may write well; and yet, to those who teach writing and who have observed students grapple with ‘Politics,’ the essay is a theoretical and practical failure, a source of incomprehension and confusion. What makes this an extraordinary and fascinating failure of communication is that it has been committed by those we, typically, judge the most skillful at communicating.

What is the message here for communicating science? In three years of sustained work promoting statistical literacy through a range of science communication projects directed at journalists and, recently, clinical researchers, we learned several important things. The first was that statisticians spent so much energy communicating with each other about what to say and how to say it to each other’s satisfaction that they had little to no bandwidth for imagining how the concepts might be communicated to the satisfaction of a non-statistical audience. In short, even when they thought they were communicating to those outside their ingroup—and inducting them into the ingroup’s discourse—they were, in fact, still talking to their own ingroup in ways that confounded access with opacity—not unlike the writers recommending Orwell’s rules for writing to those learning to write. And while this may not come as a surprise with respect to journalists, we found that scientists’ access to critical statistical concepts and problems was inhibited by a similar presumption that effective communication had already happened or was happening.

To put it another way, and with the caveat that there are exceptions, statisticians want to teach advanced Italian composition to Italian speakers; they don’t want to teach basic Italian conversation to non-Italian speakers; and, as they don’t want to this (it’s not very intellectually stimulating), and as they haven’t been trained to do this, they are not particularly good at recognizing non-Italian speakers.

This was supposed to be a blog post; it is now feature-length—and it seeks the expanse of an essay to address the theme and all its tributaries. It also may seem to be a variation on the familiar “know your audience” problem in science communication. But we would venture that this familiar recommendation is a bit too familiar, like one of Orwell’s rules: It appears simple and straightforward, but it presumes more than it instructs.

“Knowing” your audience requires thinking about scientific and statistical concepts from a user-experience (UX) perspective, to observe your intended audience with the methodological rigor you would bring to an experiment, to test multiple approaches to, in this case, explanation, and to incorporate design, visualization, and play to provide a range of access points into the concept or research claim.

Most of this is unfamiliar terrain to scientists and statisticians: but if we want to think of science communication in terms of sustainability—communication that sticks—then as much effort needs to go into communicating the research as generating the research. It may even require creating a new lexicon (as one of our collaborating partners, Accurat, did recently for IBM); it certainly requires learning and collaborating with other disciplines that have thought deeply about these problems. The lesson from Orwell’s ‘Politics’ is that we can be massively wrong even as we think we’re getting it spot-on right—and hopefully this example is so surprising it will stick with you.

There is, however, another more humbling example of not looking, studying, and understanding your audience that is, perhaps, easier to remember: it took 30 years for banks to realize that they needed to return your ATM card before they dispensed your cash.



  1. The New Republic, June 17, 1946
  2. “The disturbing new language of science under Trump, explained;” Julia Belluz and Umair Irfan, Vox, Jan 30, 2018
  3. “Orwell, Hitler, and Trump;” Robert Kuttner, Huffington Post, Jan 23, 2017
  4. Jennifer Szalai quoted in the New York Times Book Review, March 19, 2017
  5. Orwell’s Five Greatest Essays, Michael Hiltzich, Los Angeles Times, Nov 8, 2013
  6. “A point of view: Why Orwell was a literary mediocrity; Will Self, BBBC Magazine, Aug 31, 2014.
  7. A good overview is provided in “The Language of 1984: Orwell’s English and Ours,” W.F. Bolton, Knoxville, University of Tennessee Press, 1984
  8. “On not teaching Orwell,” College English, Vol 38, No 6, Feb. 1979, 553-566
  9. “Fear and loathing of the English passive, Geoffrey K. Pullum, Language and Communication July 2014, 60-74
  10. Review, Norman Page, The Journal of English and Germanic Philology, Vol 84, No 4, Oct. 1985
Kendra Kintzi