How to write a great journal article – act like a fiction author

There are a number of good posts out there on how to write a good journal article and even a whole blog devoted to the topic; many of them are linked to in the comments section of my post on writing style.

Here I want to elevate above the nuts-and-bolts sentence-level-detail of my post on writing style* and even elevate above the aforementioned posts that break down different sections of a paper and zoom out to 100,000 feet and think really strategically about writing a paper.

In my experience as a student committee member and as an associate editor for three journals I must have seen many 100s if not at this point 1000s of pre-publication articles. And they are varied. But many of them are by already good writers in the sense of clear, fluid English that understands well the purpose of each of the four sections (intro, methods, results, discussion). But many (most?) of these are still missing something. Something which I think is the hardest thing to learn: to think about the paper as a cohesive unit.

Think about an artistic painting. For the artist, it is made up of 100s or 1000s of individual brush strokes, each one of which requires skill and artistry. And of course a painting typically has a few key objects – a building, a lake, a person and the strokes have to make those up convincingly. But the reason an artist makes a painting, and the reason we hang paintings in the Louvre and visit them by the millions is none of those reasons. It is the overall gestalt effect – the message, the emotional impact. The sum of the parts is MUCH greater than the whole in a great piece of art.

It is no different with a paper. A day after reading it, you don’t remember well-crafted sentences or a really clear introduction – you just have an overall gestalt. With an academic paper this gestalt usually includes a one sentence summary of the factual content of the paper (and yes it is really only one sentence). But it also includes the emotions and judgments hanging on that one sentence. Is it convincing or weak? Is it elegant? Clever? Surprising? Ultimately, much of the emotional gestalt we take from a paper is was it convincing? do I trust the author? It is my experience that first-time writers and even many more experienced writers are so caught up in the mechanics (the sentences and sections in analogy to the brush strokes and objects in the painting) that they never think about the overall gestalt. And as a result the gestalt is rather poor. Which, fairly or not, reflects on the results of the paper. This of course is what distinguishes an art school student (working on mastering the details) from a great artist. And it is what distinguishes a publishable paper from a great paper, one that is remembered, one that has impact, and, dare we dream, a paper that will achieve the analog of being hung in the Louvre (whatever that might be – and no its not getting published in Science or Nature).

My main piece of advice will sound like it is tongue-in-cheek but it is in fact straight-up serious advice. Think and work like a fiction author! Wikipedia says that the main ingredients of fiction writing are: Character, Plot, Setting, Theme and Style. I’m sure there is debate, but these sound a lot like what I learned in high school, and I’m going to go with these.Notice that these are all unifying elements – they are things that cut across the introduction, middle, and ending/resolution of a fiction story. In short they are what give the gestalt.

Let me address each of these in a little more detail as they relate to non-fiction, scholarly article writing:

  • Character – in fiction the characters need to be richly drawn to draw you into the story and make you care enough to keep reading and to remember them. The characters in a journal article are the questions you are asking. Introduce us to them. Spend a little time fleshing out their nuances. This is not achieved by a dump of literature citations, although that is a piece of it. You need to sound excited by your questions (which means you need to know what they are!). And you need to make them 3-D. And you need to dwell on them lovingly. None of this by the way means that you should write a long introduction anymore than you should spend half your book introducing the characters. Just as in the best fiction, the characters (questions) should be introduced deftly and crisply, which requires work.
  • Theme – the take home message. In fiction it is a moral, or perhaps an emotion. In a journal article it is the one sentence take home message. You may think I’m joking, but most people really will take away only a single sentence summary of the paper, so you better know what you want it to be before you start writing. “Figuring it out as you write” is a terrible approach. Your paper will sound disjointed and like you didn’t know what your theme was before you started. So figure out your one sentence BEFORE you start writing. I am known in my lab group for mercilessly asking a student who is at the writing stage of a paper “what is your one sentence?”. I ask them before they start the presentation. I ask them immediately at the end of the presentation. And I ask them several more times during the discussion with the lab. It might seem impossible, but it is actually very achievable – it just requires setting this as an explicit task and spending some time (usually interactive with other people) to achieve it. It is a sine qua non for a paper that has a good gestalt. How can a fiction writer construct plot/story arc, characters, setting to all build towards a powerful theme if they don’t know what it is? No different in non-fiction.
  • Plot – a good piece of fiction has a clear sense of movement. It starts one place, gives a sense of motion at any point you are reading, and then you end up somewhere new. It’s a big part of why people keep reading to the end. I call this the story arc. And the story arc is the thing that I find most often missing in journal articles. You need to take the reader along a very clear trajectory from question to conclusion. Just having the standard four sections is nowhere near enough. So many papers organized by the four sections still sound like a dump of everything you ever thought or did in connection to the paper. You need to work hard on story arc to make sure everything in the paper is pulling towards that one arc. This is why figuring out your one sentence before you write is so important.This lets you know what is superfluous and unnecessary and trim it away (most good writers will tell you that half the battle is knowing what to delete).
  • Setting – the place and culture in which things happen. In field experiments or observations this is pretty simple. Just as I cannot begin to fully understand or relate to a character unless I know their context, I won’t really care if p<0.05** unless I can visualize the whole experiment in my mind. Almost everybody tells me that they used a 1m x 1m quadrat (or whatever their sample unit was) but many fail to tell me if their replicates are 5m apart or 1km apart. If they’re on the same topography or randomized, surrounded by the same vegetation, etc. A well drawn, information-packed diagram of the layout is something I often find myself requesting as a reviewer or editor..
  • Style – this is a broad category that covers everything from writing dialogue to what voice is used – but it is ultimately the techniques. The brush strokes. And it is the clear writing I posted on last year in a non-fiction article.

My bottom line is this. Every word, every sentence, every paragraph, every section of the paper should be working together, like a well-synchronized team of rowers all pulling towards one common goal. The introduction should introduce the questions in a way that gives them emotional pull and leaves us desperate to know the answer. The methods and results should be a page-turning path towards the answer. And the discussion should be your chance to remind the reader of the story arc you have taken them on and draw sweeping conclusions from it. Any freeloading sentence or paragraph that pulls in a different direction should be mercilessly jettisoned (or at least pushed to supplemental material). Does this sound like a novel you would want to read? Yes, it does, and it probably sounds like a journal article you would want to read too.

I wish more people saw themselves as needing to use the skills of a story teller when they write a journal article. I of course don’t mean the connotations of dissembling or making things up that the word “story” carries. But I do mean the art of story-telling that knows where it is going and does it crisply so that it sucks us in and carries us along with just the right amount of time spent on details of character and setting. Where the characters (questions), the plot (story arc), the setting, the theme (the one sentence take home message) all work together to make a cohesive whole that is greater than the sum of the parts. Like anything in writing, you can do it if you work at it, but you do have to work at it (writing is not a gift handed to you by the gods)***. So go ahead, turn your next manuscript into a cohesive whole with great characters and a compelling story arc that leaves us deeply moved.

UPDATE, 22 June 2014: Comments on this post are now closed. This post was highlighted on “Freshly Pressed“. Which is flattering, but has led to dozens of people who wouldn’t otherwise have seen our blog trying to make non-substantive comments in order to promote their own blogs. We may or may not reopen comments on this post in the future.

* (which I badly violated in this sentence by stringing 5 nouns and more connective words in a row with no verb in sight and then running on for 45+ words in one sentence! – do as I say, not as I do 🙂 )

**I probably won’t care about p<0.05 for a whole other set of statistical/philosophical reasons, but I leave that for another day!

*** just as an example of the messy, iterative process that writing is which depends as much on the bas-relief process of what is removed as what is added, I had a clear vision for this post – science writing should be more like fiction writing with the same elements as a compelling story which immediatley led to a title and intro. Then I when I started writing, I ended up with an outline that looked like

I – you have to know your main point

II – you should be like a fiction writer

IIa – character

IIb – plot

IIc – theme


Well – I clearly had lost my way. While nothing I said was untrue or unimportant, I had bifurcated and complexified off my main theme. This is something I am very prone to do (as I think are most academics). So I deleted two whole paragraphs on I – you have to know what you want to write about – and then worked a much reduced version of it into the IIc theme section. Boom – back to a single story arc, a single sentence to remember, and a tighter, stronger piece. Not every edit is this easy, and this post could certainly benefit from more, but I hope it at least makes my point that you have to edit with a mentality of “does this add or distract from my main point” and be merciless if the latter.

This entry was posted in Advice by Brian McGill. Bookmark the permalink.

About Brian McGill

I am a macroecologist at the University of Maine. I study how human-caused global change (especially global warming and land cover change) affect communities, biodiversity and our global ecology.

33 thoughts on “How to write a great journal article – act like a fiction author

  1. GREAT post Brian! I’m sending a link to all my colleagues and postgrad students. Only thing I’d add is that the “telling a story” aspect also applies to theses/dissertations, and especially to giving a presentation: tell the story, don’t drown your audience in data/stats/equations/minutiae.

    • Thanks. I 100% agree that this applies doubly so to presentations (and posters for that matter).

    • wow. this really looks like an advertisement. I swear I was not paid to do this, ha.

      • Yeah – I think Amazon and WordPress have worked out a deal. Its kind of nice to see the cover, but it is so big! (but all links to Amazon look like this – there are a number on Jeremy’s recent post on summer reading too). Again – thanks for the link.

    • I didn’t know about the book, but from the abstract I see a lot of parallels. Thanks for the information!

  2. Brian, thanks for this post.
    I’m going to share it with some colleagues with whom I am co-coordinating an ESA workshop on using multimedia for communicating science ( I think this post could be a great framework for our section about writing as a way of storytelling, not just reporting numbers.

    You’ve articulated well the potential there is for an academic paper or presentation (100% agree with Jeff Ollerton) to tell a story well. I love being part of the science communication process from the inside precisely because of what you described. A paper or talk really can be a pleasure to read or listen to without compromising scientific integrity.

    Bethann G. Merkle

      • Thanks for saying so! We’re crossing our fingers a bunch of people will participate. We’re also exploring the possibility of launching an ESA SciComm section, which we’ll discuss in more detail at the ESA meeting.

  3. Hi Brian,

    by and large I agree that this is good advice, but I feel that a disclaimer should be added, namely that science produces facts and not fiction. Hence, our first concern should be to find the stories that are correct, and selecting the best among all correct stories comes after that.

    I know that it’s tacitly assumed here that the results have already been established and this is only about presenting them, with no interactions between these two phases of research. However, in practice I think this is not quite so clear-cut, and there is often the situation where one might be tempted to gloss over counter-evidence to not have a great story ruined by messy data.

    • HI Florian – you are right that stretching the truth has no part of journals, a point I make myself in the last paragraph. But I don’t fully agree that people making up a few facts for a good story are common or a major threat to the field*, nor that they are the same people that my post is addressed to. I think early career researchers are on average the ones who most need to work on their storytelling skills, and in my experience most early career researchers are terrified of over interpreting their evidence and seek guidance around it. The ones who tend to “find” “convenient” facts are often more jaded and experienced I believe – certainly I think that is true of people shown to commit outright fraud.

      *NB: subconscious researcher degrees of freedom might be, but that is really a whole other issue than the intentional acts I think we’re talking about.

  4. Nice post Brian.

    I note that you’re clearly recommending a certain sort of storytelling as a model for scientists to follow. Maybe something like Elmore Leonard. As opposed to, say, Tristram Shandy. (,_Gentleman) 🙂

    Ok, joking aside, I do think that “tell a story” is good advice for scientists but that it needs to be fleshed out (which your post does). It’s not always obvious how to flesh it out.

    For instance, I sometimes see scientists advised not to structure their talks and papers like a “murder mystery”. Which at first glance seems like sound and even obvious advice. As a scientist, you don’t usually want to actively go out of your way to hide your most important lines of evidence and obscure how they fit together, so as to keep your audience from figuring out what your conclusion is going to be. On the other hand, does that mean you should always state your conclusions up front, and then afterwards explain how you arrived at them?

    Or think of ordering of results within your Results section. Is it always best for your “story” to lead with your main result? Or are other orderings of material sometimes more logical?

    Or think of Terry McGlynn’s lament that ecology journals don’t usually let authors toss in interesting natural historical asides if those are irrelevant to the story of the ms. Personally, I side with the journals on that–I don’t like my scientific stories cluttered up with lots of irrelevant asides. But I take it Terry would welcome a bit of “Tristram Shandiness” as adding color and interest.

    Or think of the distinction between papers in Ecological Monographs vs. Ecology. The former is supposed to be for stories that “need” to be longer. But it’s not entirely clear what that means, especially in this day and age of electronic supplementary material.

    I don’t have any answers, just some musings your post prompted.

    • I often find myself telling students that a talk should not be like a mystery novel. I agree that that’s not the kind of fiction writing we should try to emulate!

    • Following on from Jeremy’s reference to Elmore Leonard, there’s a good quote from him that is relevant to Brian’s post. Asked about his writing style, Leonard said, “I try to leave out the parts readers skip.” I always find it remarkable how often I find myself wanting to skip over whole chunks of papers. I’m sure part of that is my impatient nature (where’s that one sentence?!), but I’m also sure part of it is the style (or lack thereof) of storytelling…

      • Great quote Patrick. I think if we all approached papers with the attitude not of “can I fit in everything I thought about this subject” and replaced it with “can I write this so an impatient reader will read it?” we would all be better off. Because let’s be honest have you met an academic who is not an impatient reader these days?

  5. Brian – I LOVE this post – have sent it to all my grad students, and have bookmarked it for future use. THANK YOU!

  6. Hi Brian,

    very nice post, indeed! In the future, I will stick to this advice as much as possible!
    I would really appreciate if you or the DE-community could give some outstanding examples of great stroy-telling papers.

  7. I agree with your suggestions and I would hope that experienced scientific paper-writers follow a story arc without really thinking about it. I wonder, however, if there is an underlying tension in that I personally can’t remember when I last actually sat down and read a paper from beginning to end. Most of the time, there are so many papers to read that I take the “butterfly” approach skimming the text for information.
    There is also the issue of the perfect story arc being ruined by referees’ comments?

    • Good points Phillip,

      There is no doubt butterfly reading is increasing. I used to think abstract and title weren’t very important but I have had to revise that opinion. I guess I would argue that butterfly readers are all the more reason for the idea that the paper needs to cohere into a single hole. Butterfly readers, may not appreciate the beauty of what you achieved, but if they’re only going to look at one figure or one paragraph, one would want it to be one of the figures or paragraphs driving the story arc forward instead of some random unrelated paragraph thrown in just because you could.

      Don’t get me started on referee comments! 🙂 They wreck many things (and of course improve many things) but it doesn’t let us off the hook for doing the best job we can.

  8. I think you’re right that telling a story should be in the front of your mind when you’re writing a paper – and I’d like to add a bit of advice that’s helped me to do that. The best advice I ever got was – make sure the first sentence of every paragraph tells the story of that paragraph, so that you could take each first sentence, and have the essence of the paper right there. It keeps things clear and focused, and really makes you think about the structure and meaning of what you’re trying to say. Oh, and my drafts always come back from my supervisor with scrawls of “No logical link!!” so I guess keeping the links between sentences helps too 🙂

    • What you describe is essentially what the post on ConservationBytes (linked to in the comments on my earlier post on writing tricks) suggest (as far as writing out one sentence for each paragraph first). It’s kind of link outlining before writing – which is a good idea – but the challenge of which is then to keep the sense of flow through the subsections of the outline. Its never been a technique I use but I know it works well for many people.

  9. Pingback: Back in the field | Urban Phosphorus Ballet

  10. This article provides excellent advice which, of course, applies equally to lectures. I remember Henry Horn making the same point to graduate students years ago, after sitting through a poorly structured seminar. He said simply, “It didn’t seem to have a melody one could hum.”

    • Great quote. I was going to say that the points Brian makes apply to good song writing as well. Get too caught up with the technical specifics and you can lose the gestalt, happens all the time.

      Horn’s gap succession model (1971?) based just on the relative abundances of saplings remains one of the most amazingly simple predictive models I’ve ever seen. Seems like the man knew how to cut through the fog.

  11. Reblogged this on Damien G. Walter and commented:
    It’s really hard to overstate how much I agree with this post on using fiction techniques in journalism. Short form and blogging may not leave much scope for storytelling, but once you get up in to long form and feature writing, narrative techniques become essential. Facts nad information won’t hold a reader for thousands of words. You need characters and emotions to do that. You need story.

  12. Pingback: Morsels For The Mind – 20/06/2014 › Six Incredible Things Before Breakfast

  13. Thank you for the article. I am aware that this is framed in an academic context, but it’s really helped me understand how to frame my blogposts – somthing I am quite new to (started today!). Of course I also plan on using some of the ideas discussed to write better academic work.

Comments are closed.