There’s a story about I.I. Rabi (the irascible physicist played by David Krumholtz in Oppenheimer) that I’ve heard from two different Nobel laureates1. Rabi was a brilliant physicist, but apparently an absolute disaster as a classroom teacher. His lectures were riddled with errors and omissions requiring significant backtracking and derivations that went down blind alleys— all the stuff that STEM faculty have nightmares about doing in class. When asked about it he claimed it was done deliberately, to show students what it was really like to do physics calculations, forcing them to think through the derivation themselves rather than just copying it down. (Both people I heard this from thought that was cope, as my kids would cringe to hear me say, and that he was just prone to failing to prepare adequately— lecturing without notes, etc…)
This is very much of a piece with the long history of successful researchers being lousy communicators, in a variety of media. I thought of this today because last night over on ex-Twitter, Tyler Austin Harper had a nice three-tweet thread on academic writing that I’ll screenshot because nobody clicks links any more:
As I said when I quote-tweeted this, there’s a lot of truth to this, especially the second point. An enormous fraction of the work product of academics is aimed squarely at people who don’t really have any choice about whether to read it or not, which reduces the incentive to actually write in a compelling manner. Some authors in some fields will even argue that this is actually a virtue— that obscure and confusing writing is a positive good because it forces the reader to grapple with ambiguities of language, or whatever. (It was cope for Rabi, it’s cope for them, too…)
To be sure, the difference of interests that Harper points to is also a big piece of it— in some cases, academic writing isn’t compelling because what matters to the author isn’t something that the reader finds interesting. If you pick up a history book expecting an epic tale about wars and battles and instead find a ton of detail about the economic practices of the merchant class in the relevant period, you’re most likely going to be disappointed, and that’s not necessarily the fault of the author. There’s also the matter of specialist jargon— unfamiliar words, or ordinary words used in ways that convey important shades of meaning to those in a particular community, but that don’t connect with outsiders.
That said, though, a lot of the problems with the writing produced by academics stem from the brute fact that a lot of academics simply aren’t good at writing, full stop.
It’s absolutely true that some of the problems with written communication from faculty and administrators stem from a difference in interests. But at the same time, the hallmark of a really good writer is the ability to make a subject interesting to reader who don’t come to the text thinking it’s intrinsically fascinating. Sometimes the gulf between the interests of the author and the reader is simply too wide to be overcome, but even then, I can usually recognize when somebody is doing a good job of writing about a subject that just doesn’t interest me. There’s a clarity of purpose and expression to good writing that stands out in a way that is largely independent of the subject matter.
I’ve read a lot of really bad student papers in my almost 23 years as a professor, and I’ve also had occasion to read a lot of really bad writing by faculty and administrators. And they tend to be bad in the same fundamental ways, which are closely related to the “MrHWM” tweet Harper is quoting in the screenshot above: they’re hopelessly muddled because the author isn’t entirely clear about what’s most important. They’re just kind of throwing stuff out there and hoping some of it will matter for the reader. There’s no organization, no structure that guides the reader through the key points that shape the argument.
And, look, I freely admit that I’m guilty of producing some of that kind of thing here: I do the occasional thinking-out-loud blog post, where I don’t really have a clear thesis and just kind of wibble about whatever for several hundred words. Most of the time, though, as I start to get into the writing process, a theme and a structure will come clear, and then I revise accordingly. I try not to hit “publish” on anything that I think is genuinely structureless and rambling.
I’ve seen some stuff from people with Ph.D.s, though, that I would be embarrassed to save in my Drafts folder, let alone send to anyone else. Way more often than I would’ve liked, and not just from colleagues in STEM, where stereotypes might lead you to expect terrible writing. Some of the very worst writing I’ve seen in a professional context has come from folks in literary disciplines.
(I would say that the mean of the writing quality distribution is probably a little lower for academics coming from the STEM fields, but the variance is also smaller. The failure mode of STEM writing is basically an unsorted list of factual bullet points that you could extract and do something with. The worst of the not-STEM fields is “I have no idea what you’re trying to say, and I’m 90% certain you don’t either.”)
This sounds harsh, but I don’t entirely mean it as a damning indictment. Writing well is as much a talent as a trainable skill, and as such is mostly orthogonal to the other talents and skills that make for a good career as a scholar, teacher, or administrator. There are brilliantly productive scientists who struggle to produce coherent writing— a hundred years on, people are still trying to figure out what the hell Niels Bohr was on about in some of his papers— and other people who excel at communicating the work of others, but don’t offer much original insight of their own. Rabi, for all his failures as a lecturer, was exceptionally good at administrative things: identifying talent, running a lab, and navigating politics.
It’s not even a problem restricted to academia— a fair fraction of the articles that generate buzz on social media after being published in professional outlets are pretty muddled. And that’s after a remarkable degree of editorial winnowing— you really don’t want to read the slush that gets turned away. Most academics aren’t very good writers because most humans aren’t very good writers, even the ones who get published by Big Name magazines and websites.
So, why does academic writing suck? Differences in interests and incentives play a big role, but at the end of the day it’s mostly just Sturgeon’s Law: 90% of academic writing is crud, because 90% of all writing is crud. Writing is hard, most people aren’t good at it, and you should cherish the 10% when you find it.
That’s somewhat cathartic to write, but hopefully vague enough to keep me out of trouble. If you’d like more of this, or want to see if I was too specific and have to awkwardly walk something back, here’s a button:
And if you want to argue, the comments will be open:
Norman Ramsey said this in an after-dinner talk at a DAMOP banquet years back, and when Martin Perl was getting an honorary degree from Union, he confirmed it. (After I repeated the Ramsey story to him, forgetting that Perl was also a Rabi student…)
I had a Rabi-like professor in grad school. He was world renowned theoretical physicist, but he was a terrible teacher. His office was across a courtyard from where he taught, and I was convinced that he dropped his notes one day on his way to class, and that they had been blown around by the wind. He picked them up, but he never sorted them back into the correct order.
The notion that good writing is a talent not a skill feels like letting people off too lightly. Anyone who can earn a PhD should be able to learn to put ideas in a coherent order, and make them clear. It may take talent to do it at a level deserving a Pulitzer, but not at the level required to spare the reader physical pain. Plus, what we tell our students really is true: good communication skills really will mean more people will read (and cite!) your paper, and you will do better at job applications, and get more grants, and on and on.