There’s something about the phrase good writing that bothers me. Don’t get me wrong, I love good writing, I try to produce good writing, I write books about it, and I really want my students to get better at writing.
But when I see generic lists of tips or, worse, rules for good writing, I always want to ask “writing what?” As I said in Changing Practices in the L2 Writing Classroom: Moving Beyond the Five-Paragraph Essay: “We don’t just write. We write something to someone for some purpose.” What makes good writing in one situation would be out of place in another. That’s why a lot of Onion and McSweeney’s parodies work so well (this list of “top millennial injuries reported in urgent-care facilities” detours into a rant against the stereotyping of millennials, while using the word existential correctly, unlike some newspapers of record).
So I was rather concerned by a piece that circulated back in January (2019) with Steven Pinker’s “13 rules for writing better.” Oh good. Only 13. Now, to be clear, I have been at times a fan of Pinker: he signed my copy of The Language Instinct when he came to speak at the Cambridge Union in about 1998. He was also the first person I ever saw using a PowerPoint presentation, which looked pretty funny against the ornate wood paneling. I think he was a great writer when he wrote about linguistics, and judging by the number of books he sells, so do many others.
But what Pinker became famous for writing was a very specific genre: the erudite trade-book, the science-for-the masses (or at least first-year undergraduates) volume, the thinking person’s non-fiction. That’s not the same style he uses in his academic journal articles. Why? Because it’s not good writing in a different context. But he still wrote a book in 2014 criticizing academic writing, essentially trying to turn all journal articles into … well The Language Instinct. Rachel Cayley wrote a brilliant critique on her blog in 2015:
Academic writing isn’t laughably bad—it shouldn’t be the butt of a joke. And it isn’t monolithic. An established Harvard academic writing a book is doing something very different than a new doctoral student attempting their first article. Pinker’s critique often makes it sound as though academic writers simply appear whole cloth without any process of learning the craft.
I agree with Rachel. Writing is never monolithic, and good writing depends on the context, the audience, the purpose, and the genre. That’s what makes it interesting. And hard. Really hard.
But maybe these 13 “simple tips” will dissuade me. Let’s see … here are the tweets with my commentary in italics.
- Reverse-engineer what you read. If it feels like good writing, what makes it good? If it’s awful, why? Definitely, but can we be specific here? If you want to know how to write an academic journal article, you should “reverse-engineer” examples of effective journal articles. If you want to write op-ed columns, analyze op-ed columns. Just don’t assume that by studying one genre, you know how to write another.
- Prose is a window onto the world. Let your readers see what you are seeing by using visual, concrete language. Err, OK, Hemingway. But sometimes your topic isn’t visual and concrete; it’s abstract and theoretical. There’s still such a thing as good theoretical writing, and it’s not going to be visual and concrete.
- Don’t go meta. Minimize concepts about concepts, like “approach, assumption, concept, condition, context, framework, issue, level, model, perspective, process, range, role, strategy, tendency,” and “variable.” No, stop! This is terrible advice for academic writers. Let’s name the beast here: nominalization. Nominalization is good when used correctly. It’s a form of meta-discourse that helps writers guide readers through complex ideas while avoiding lengthy repetition. Do go meta. A lack of meta-discourse results in incoherence in academic genres. (And how do you write up quantitative research without the word “variable”?)
- Let verbs be verbs. “Appear,” not “make an appearance.” This has nothing to do with verbs. He’s talking about concision. Use the shortest form possible without compromising meaning (BTW, “appear” and “make an appearance” aren’t exactly the same: apparent synonyms never are).
- Beware of the Curse of Knowledge: When you know something, it’s hard to imagine what it’s like not to know it. Minimize acronyms and technical terms. Use “for example” liberally. Show a draft around, and prepare to learn that what’s obvious to you may not be obvious to anyone else. A better way to say this is “know thy audience and write for them.” If you are an expert writing for other experts, then of course you should use shared technical terms. I don’t want a surgeon who asks the operating room nurse for “that long sharp thingy.” There’s nothing wrong with using acronyms and technical terms when they are appropriate for your context and known to your audience (see #6). They save time and allow writers to …
- Omit needless words. Ironically, this advice repeats #4 and is therefore, um, “needless.” But of course if you keep giving examples and writing out acronyms even though your audience should know them (#5), you’ll fill the page with needless words.
- Avoid clichés like the plague. Sure (h/t George Orwell). But let’s be careful here. The word cliché means a fixed phrase (literally a preset printing plate that would save time for the typesetter), and effective academic writing is full of recurrent phrases that we use to orient the reader and make complex rhetorical moves (e.g., the Manchester Phrase Bank). Don’t throw the baby out with the … you know.
- Put old information at the beginning of the sentence, new information at the end. Yes, yes, almost. This is indeed the most common pattern in English prose and is a reliable solution to a perplexing paragraph (see Grammar Choices, unit 8!). However, it’s not a hard-and-fast rule. Sometimes we put new information deliberately at the start of the sentence for emphasis, what linguists call “markedness” (see, a useful technical word).
- Save the heaviest for last: A complex phrase should go at the end of the sentence. Or to put it another way, see #8, but yes, English generally likes the weight at the end of the sentence.
- Prose must cohere: Readers must know how each sentence is related to the preceding one. If it’s not obvious, use “that is, for example, in general, on the other hand, nevertheless, as a result, because, nonetheless,” or “despite.” Yeah, but don’t overdo it, or you get some pretty odd writing. Incidentally, these transition words are just another type of meta-discourse, which he didn’t like in #3.
- Revise several times with the single goal of improving the prose. Good advice. I read and re-read this post obsessively because, frankly, it’s a bit scary to criticize Steven Pinker and I wanted to get it right! (Update: I write about this idea more in another post, with bonus music theater content.)
- Read it aloud. Maybe — I talk to myself incessantly, and I do like to hear what my writing sounds like in my head. Not great advice, though, if you’re writing in a second language and your writing is better than your speaking. Most prose isn’t written to be read aloud: maybe we need another 13 rules about good style for speaking?
- Find the best word, which is not always the fanciest word. Consult a dictionary with usage notes, and a thesaurus. I’m all in favor of dictionaries, but I’m not sure what he means by “fancy” –the “best” word is the one that most precisely creates the meanings (plural) that the author intends. This doesn’t mean the simplest word is always the best. This also contradicts tip 4 — e.g., “facilitate” is “fancier” than “make something easier,” but it’s also letting the verb do the work and is more concise.
Let’s see how well Pinker follows his own advice in a peer-reviewed journal article — a really interesting one, too, from a major journal, Cognition (Hartshorne, Tenenbaum, & Pinker, 2018).
- “This fact has influenced many areas of science, including theories about the plasticity of the young brain, the role of neural maturation in learning, and the modularity of linguistic abilities.” Oh boy, we have “meta” words like fact and theories (#3), nominalization instead of verbs (#3, 4), and lots of technical vocabulary (#5). And it’s a good sentence for an academic journal article, with new information and complex phrases at the end (#8, 9).
- “conspicuous grammatical errors” — ooh, fancy (#13). What about “gradually or precipitously“? What’s wrong with “slowly or quickly”? (Answer: they don’t work as well here.)
- “However, neither the nature nor the causes of this “critical period” for second language acquisition are well understood. ” This sentence is front-loaded (#9), and with good reason: it creates better cohesion and puts marked weight on the authors’ main idea, that these phenomena are not well understood.
- In the first two paragraphs, there is only one transition word (however): the prose coheres but through more subtle techniques than #10 suggests.
- Clichés (#7) …. “reaches a floor” (stops falling), “saddled with an accent” (speak with an accent)? The first seems fine, but the second is a clumsy piece of native-speakerism (accents are not burdens).
- Use visual and concrete language (#2): ” the modularity of linguistic abilities”; ” neural maturation” … these are complex theoretical ideas. Terms like “Type I and Type II errors” are also used without explanation or definition (#5). And rightly so. If you don’t know what they mean, skip ’em or grab a stats textbook. But if you’re reading Cognition, these phrases are very unlikely to cause you difficulties.
- “Reliability for the critical items was high across the entire dataset (Chronbach’s alpha = 0.86).” –if you read a lot of quantitative research, this might almost sound like a cliché (#7). It’s certainly technical, front-weighted, and abstract. But it’s also a very important sentence for readers who understand and care about reliability statistics. Its omission would certain draw the ire of Reviewer 2.
- “Use ‘for example’ liberally” (#5) — this paper uses it, er, twice, and for instance (the “fancier”) word another 7 times. The paper is about 11,000 words long, so these two phrases together are used less than once per 1000 words, or 4 pages of typed manuscript. Not very liberal.
I could go on, but you get my point: I don’t think there are any universal rules of good writing. In order to know what makes writing good, you have to know your genre, audience, and purpose; you have to understand the appropriate conventions; and you have to make language choices in the appropriate register. There is definitely terrible academic writing out there, but the Hartshorne et al. paper is well written despite apparently ignoring many of its co-author’s tips and tweets. This is the approach Ann Johns and I take in our textbook, Essential Actions for Academic Writing: A Genre-Based Approach (University of Michigan Press, 2022). Learn more here!
So, choose your (role) models carefully and do as they do, not as they say they do!
Update (12/12/21): About once a semester, this post suddenly sees a spike in viewers from the University of Arizona. Hi, Wildcats! I would love a UA student or instructor to comment on my post: tell me why you’re here, and what you think of it.
Nigel Caplan 