Writers should avoid unintentionally reminding readers that they are reading, says Paul Simpson 

In the 1987 Star Wars spoof Spaceballs, there’s a running joke that the characters are aware they are inside the film. The joke peaks when two characters, Dark Helmet and Colonel Sanderz, watch themselves on screen in real time. Confused, Dark Helmet exclaims: “What the hell am I looking at?!” The, admittedly thin, illusion that Spaceballs is a real story explodes as the characters interact with the usually invisible camera.

This device is called breaking the fourth wall. When characters look directly at the audience it’s thrilling and creates tension because the illusion of the story is shattered. Bill Connor used the same technique in a column about Ruth Ellis that was published on the day she was killed in 1955. Ellis was the last woman to be hanged in the UK and Connor continually reminds readers that while they are reading a woman is being murdered. By breaking the fourth wall readers are seen and therefore can’t escape the shame or responsibility for the public support of capital punishment.

When used with intention, breaking the fourth wall is a powerful way to move a reader emotionally. However, when editing journal articles, which tend to be a few thousand words or less, I often ask academic authors not to remind their readers that they are reading a text. This almost always makes an article easier to follow by improving the flow of their work. So, what’s going on?

It’s the nutgraph where problems tend to first emerge. Now, if you are wondering what the hell a nutgraph is then don’t worry, you’re probably in good company. I had no idea there was such a thing until recently, but the nutgraph is the paragraph that sums an article up in a nutshell. It is usually located at the end of an introduction, but appears in all sorts of weird and wonderful places in first drafts.

Many academic writers reflexively break the fourth wall when they get to the nutgraph as they try to summarise their article with a mini table of contents. You’ll probably recognise clunky and passively written nutgraphs that look a bit like this:

This article will draw a parallel between breaking the fourth wall in movies and the nutgraph in academic articles. In the following section a fictionalised example will be used and it will be argued that this standard way of writing nutgraphs, where the writer directly refers to the article being read, reduces the comprehension of an article’s discourse. The article concludes that there is an urgent need to stop writing like this.

You might wonder, what’s wrong? Isn’t this the scholarly way to do things? Well, yes this is certainly a motif of academese. However, that doesn’t make it good.

In his book The Sense of Style Steven Pinker describes this type of writing as metadiscourse, essentially writing about writing. He argues that these summarising lists are too long and arbitrary to stay in readers’ minds for long. He also takes aim at the thoughtless signposting that appears in academic articles. You’ll surely recognise these: “as we will discuss,” “as described above.” I’m with Pinker on this, they need to go.

Pinker suggests that writers are following the advice that comes from classical rhetoric for long speeches or lectures: say what you’re going to say, say it, and then say what you’ve said. But he argues this isn’t needed if readers can see the text in front of them and can scan back and forth when they need to. Ultimately, these continued references to the text itself break the illusion of a conversation between the writer and their readers.

What should be done instead? You can normally just delete unnecessary signposting without any loss. For nutgraphs, I like the advice given in The Science Writers’ Handbook series. Resist the urge to summarise your entire story and simply prepare the reader for the journey ahead. Most people don’t need a detailed map of the article they are going to read, they just need to understand the general direction of travel an article will take.

This advice isn’t just for critical essays, it’s useful for research articles too. Research articles have an abstract, which already summarises the article, and they tend to follow a predictable structure. Readers don’t need another summary, but they may need a preview of research findings so they know what to expect of the road ahead.

This may feel like a picky issue of style and taste but at stake is a reader’s attention and understanding. It’s far better to spend their energy on ideas and findings rather than figuring out a clumsily written manuscript. You’re also less likely to leave your readers wondering what the hell they are looking at.

Paul J Simpson is The BMJ’s international editor. 

Competing Interests:  I bought The Sense of Style, The Science Writers’ Handbook, and The Science Writers’ Essay Handbook myself and enjoyed them. I don’t have any connection with the authors or the books’ publishers.