There are two problems I see over and over in academic papers and presentations. The first is that the audience is told too much; the second is that they’re told too little. Opposite problems, it would seem, but they tend to crop up together. And they can both be avoided by observing a simple principle—what I think of as “Chekhov’s gun” for scholarly writing and presenting [1].
You’ve likely heard of “Chekhov’s gun.” The term is shorthand for a nugget of advice about writing from Anton Chekhov, maven of short-form fiction. He seems to have articulated the principle only in correspondence. In one letter, he wrote: “Remove everything that has no relevance to the story. If you say in the first chapter that there is a rifle hanging on the wall, in the second or third chapter it absolutely must go off. If it's not going to be fired, it shouldn't be hanging there.”
I’m not sure whether this is good advice for fiction writing—frankly, I suspect not (and others agree.) But I’m completely convinced that, gently adapted, it’s indispensable advice for academic articles and presentations. I’m thinking primarily of writing and presenting empirical work in the social and cognitive sciences, but I’m betting the principle has much broader relevance. Let’s zoom in on those two problems I mentioned.
The too-much problem. The first problem is that the reader is presented with too much information at the start. For simplicity I’ll just talk about “readers” and “writing” here but all also applies to presentations [2]. Introductions to empirical work are often rich with new-to-the-audience terms, distinctions, past findings, and open questions. Sometimes too rich: they can have the flavor of a general survey of a topic area, a broad literature review. What an introduction should be, in my view, is a compact motivation for what you did—an efficient and engaging setting-up of your specific question and approach.
But, hold on, you ask: “What could be the harm in providing a fuller context?” As the writer, it’s not so easy to see, but as the reader the harm is clear. Let’s say I’m reading your introduction and I encounter five new-to-me terms, four recent findings, three new questions, and two open questions. Naturally, I’ll assume all these must be important, crucial to understanding where things are eventually going. So I’ll carry all this with me (or, more often, try to carry it and fail). By the end, it may become apparent I only really needed three of those terms, that only one of those open questions was really focal. Imagine how much effort the beleaguered reader could have been spared if none of this extraneous stuff had been introduced in the first place. “But it’s so interesting!” you say. Yes, of course it’s interesting. But it’s not worth the cost of distracting and over-taxing your audience. An over-taxed reader won’t be able to engage deeply with your key contributions. That’s a shame.
So why does the “too much” problem arise? Partly because writers suffer the “curse of knowledge” [3]. They know the idea-terrain they’re writing about very well; it’s not hard for them to flit from one familiar idea to the next, and to pack in a sequence of such ideas in the span of a breath. And so they don’t see how effortful all this flitting about and idea-density is for their less-knowledgeable audience. Writers also suffer another curse, which we might call the “curse of enthusiasm.” Not only do writers know more about the topic at hand, they likely find it inherently more interesting than their audience does. This gives rise to digressions, expansiveness, and other forms of self-indulgence.
The too-little problem. The second problem is that ideas pop up out of nowhere in the middle of a paper, particularly in the methods or results. There might be a manipulation in an experiment that was never really motivated in the introduction, or an analysis that the audience could not have anticipated [4]. There will certainly be times when you need to introduce a new idea later in the exposition—either to avoid overloading the introduction or because it simply wouldn’t make sense until a later juncture. But late-breaking elements should be minimized. Perhaps the best formulation of this part of the principle is: If an idea can be introduced early, introduce it early.
So why does this second problem arise? The same sort of “curse of knowledge” dynamics are at play here. Since the late-emerging idea is well-known to the writer, it doesn’t put them off their footing to have it suddenly come into play. From the writer’s perspective, after all, it was lurking in the background all along! But there’s another source of this problem, I think, which we could call the “lure of chronology.” There’s a tendency for writers to unfold ideas in a way that mirrors how those ideas actually emerged chronologically, over the history of the project. So, for instance, if there was an analysis that didn’t occur to the researchers until late in the process of doing the work, it’s tempting to have it also emerge late in the presentation of the work. In some cases there are clear and engaging ways to do this, to sneak some storytelling into the exposition. But, generally, it’s good to keep in mind that—provided the work is accurately described—the communication of an idea need not recapitulate the development of the idea.
A perhaps-useful analogy: Think of a paper as a journey you’re sending your audience on. The intro is where you give them the equipment they need. Anything you give them is something they’ll carry with them (or try to), whether or not they really need it. Because, again, you are the only one that knows what they really will or won’t need. The reader doesn’t know this, can’t know this, because they haven’t seen the whole journey. In this analogy, the first problem is basically over-packing.
But, naturally, you can also under-pack. Your reader may arrive at a point in the journey where they really need a piece of equipment that they simply aren’t equipped with. In many such cases they can make do and get by, but why put readers through that? Why cause them that surprise and strain? Why not make their lives easier?
Zooming out a bit, it should be clear that the “Chekhov’s gun” principle I’m advocating here is just one piece of a much more general piece of advice: make your scholarly writing and presenting as easy to understand as possible. The goal is not just to make your writing possible to understand. The point is not just that you should write in a way that readers and audience members can ultimately—perhaps after mental contortions and re-reading—figure out what you’re actually saying. The point is that you should strive to spare them any unnecessary cognitive effort [5]. This is not just good for them, it’s good for your ideas. Scholarship that is easier to understand is also easier to remember, easier to engage with, easier to tell your colleagues about, and much more fun to teach.
Notes
[1] It’s tempting to give the principle a less violent shorthand. Perhaps “Chekhov’s gum”? “If someone starts chewing gum in the first chapter, they must blow a bubble or spit it out in the second or third.”
[2] Narrative economy is perhaps even more critical in a live presentation than it is in a written text. The demands on a live audience’s memory are greater, and they don’t have the luxury of revisiting parts of the presentation they didn’t quite get. This is one of the reasons I always like presenting a version of any writing I’m doing: it forces you to think extra hard about economy.
[3] This term was first introduced in 1989 by Camerer and colleagues. Steven Pinker has written about how the curse of knowledge contributes to bad prose—see also Chapter 3 of The Sense of Style.
[4] I may be going beyond Chekhov a bit. He seemed focused on the idea that if a gun is introduced early, it should go off later. Here I’m also advocating the converse: if a gun goes off at some point, we should have been told about its existence earlier on. Again, probably not great advice for fiction.
[5] On the idea of saving readers cognitive effort, see Joe Moran’s insightful book First You Write a Sentence.