Skip to main content

All scientists should be storytellers

No one likes reading academic papers. Deep down even scientists hate reading academic prose. Aside from the rare gems now and again, academic prose is dry, terse and quite convoluted. This state of affairs is not surprising to anyone – it is so widespread that we have grown to expect bad prose and stumbling upon good writing in science feels like a slap in the face. It energizes and wakes us up. Whatever you may think of academics like Noam Chomsky, Jerry Fodor, Steven Pinker or Douglas Hofstadter, there is no denying that they have the uncanny ability to pierce with their words. Just look at some of the examples below and tell me they don’t intrigue the hell out of you.

Fodor, for example, entices the readers of his Precis to the Modularity of Mind (1985) with the following opening paragraph:

Everybody knows that something is wrong. But it is uniquely the achievement of contemporary philosophy – indeed, it is uniquely the achievement of contemporary analytical philosophy – to have figured out just what it is. What is wrong is that not enough distinctions are being made. If only we made all the distinctions that there are, then we should all be as happy as kings. (Kings are notoriously very happy).

Similarly, you can’t help but wonder what Hofstadter’s arguments will be when he makes a statement such as:

Analogy is the very blue of the sky of cognition. It is the talent so fundamental that it fuels our minds. [..] Is it irrational, subjective and concrete? Yes indeed, but it is also the underpinning of rationality, objectivity and abstraction. Analogy is not a rare luxury of thought or an exotic, remote corner of cognition. Analogy is the entire transport system of thought, including motorways, roads and trails; it pervades thinking, from throwaway remarks to deep scientific and artistic insights. All along the spectrum, analogy lets us see the new in terms of the familiar. It guides us in learning new concepts, solving mathematical problems, dealing with interpersonal conflicts and making political decisions.
– The Forgotten Fuel of our Minds, 2013

Why are examples such as these so few and far in between? Why are academics, who live and breathe writing, so notoriously bad at it?

In his recent essay, Why Academics Stink at Writing, Steven Pinker suggests that it has nothing to do with a conscious desire to be obtuse in order to put meaningless content behind a veil of jargon. Or with the fact that academic journals and reviewers expect papers to be written in what he calls academese. Nor does academese spring from the fundamental abstractness and complexity of the subjects we study. At what then, can we point our fingers to, as the agent that poisons academic writing?

The trap of self-conscious writing.

One problem with academics is that when it comes to writing, they are like a ship-wreck survivor standing on two planks in the water that are moving in different directions. Academics are torn between satisfying two different purposes when they write – one is to inform their readers of novel data or ideas. At the same time, however, academics agonize over how they will be perceived in the eyes of their colleagues. Will they be thought of as careful, knowledgeable, creative and analytic? Or will they be dismissed as hacks who go beyond the data in making conclusions, who disregard their colleagues’ work, however remotely related, as people who are way too confident for their own good?

Pinker sees this trap of self-conscious writing as one of the fundamental causes of bad writing. It leads to metadiscourse (e.g. “This article is organized as follows”), to using vacuous qualifying phrases (e.g. almost, quite, in part), nominalizations where verbs would suffice (e.g. “The affirmation of the current hypothesis by the data…”), etc.

But even if one solves all of these local problems, academic prose might become easier to read, but wouldn’t necessarily become any more engaging. More often than not, academic papers lack spice, they lack a compelling story.

We should all be storytellers.

Why do people enjoy reading fiction? I’m sure there will be as many different answers as there are readers and writers, but I believe that the cornerstone of all is good storytelling. Even the most fascinating characters or the most beautiful prose can be empty shells when they are built around bad, misdirected, and unsatisfying stories. Lolita, for all its beauty and charm, is one unbearably boring book (don’t crucify me, please!).

Most articles in science lack a compelling story. Data by itself is meaningless, and yet, most academics, including me, often treat it as self-explanatory, as something that should enlighten readers by its sheer awesomeness. But we don’t care about data. We don’t care about numbers. Numbers are just tools to describe the world.

Scientist care, or at least they should, about what those numbers can tell us about the fundamental properties of reality, the reality of the natural world, of our minds, and of the culture our species has produced. And this is fundamentally a storytelling issue – what does the data tell us about that specific aspect of the world? How did it come to be the way it is, how does it work, and how can we fix it, if it is broken? How do I, as a scientist, share that story with my peers and with the public?

Jeffrey McDonnell, in a paper published in Nature a few days ago, deals with exactly this issue. While above I have treated it as a problem for readability, he aptly recognizes that the lack of storytelling ability is what fundamentally makes writing a challenge for many academics. He uses a storytelling approach to help him and his students write better, clearer and more engaging academic papers:

Each of my group’s papers now starts with a storyboard session at a whiteboard. I pretend to be a big-time Hollywood producer and ask the Ph.D. student or postdoc to play the role of would-be movie director pitching a new movie to me. Their pitch must answer three questions: What is the status quo? What is wrong with the status quo? How does this new paper go beyond the status quo?

This approach helps frame the story and place key figures and technical findings in context. Balancing each of the status quo elements is a great way to set up the introduction—often the toughest section for early-career scientists to write—and to lead the reader to the research questions or hypotheses.

I myself, realized that this is the way to go about three years ago, when I was struggling to write one of my first papers. I believed I had something important to say, but I found that I have no idea how to stay it – I became paralyzed every time I stared at the cursor blinking on the empty page.

My breakthrough occurred, similarly to McDonnell’s, when I realized I should start asking those precise questions – what is my data telling me? Not what have I found, but what does it meanWhat is the main conflict that my data resolves? How do I build up that conflict in a way that makes the reader ache for the resolution?

I picked up those techniques not from anyone I know, or from a course on academic writing, but from a book on writing fiction, where story is king and reigns supreme. I am far from being any good at it, and I often fall back on bad habits, but I’m constantly trying to improve.

The problem is, in STEM and in the social sciences, we rarely teach writing in undergrad or grad school. Rather, it is expected that writing is something students will pick up on their own as they go, as they read more and more. Given that the papers they read suffer from all the vices of academese, this expectation is unrealistic. We need to teach ourselves, our peers, and our students, how to write better stories, because stories do not tell themselves, and data is impotent when buried in bad writing.

Like what you read? Leave a comment, share the post on social media, or subscribe to the Neuromusings blog from the sidebar.

20 thoughts on “All scientists should be storytellers

  1. Excellent article. Allow me to comment however on what look like two small slips.
    I don’t think you need the comma after ‘water’ in “One problem with academics is that when it comes to writing, they are like a ship-wreck survivor standing on two planks in the water, that are moving in different directions.”
    And “Academics are thorn between satisfying two different purposes when they write – one is to inform their readers of novel data or ideas.” Perhaps ‘torn between’?

  2. I have come to this same conclusion. The best scientists have always been great story tellers.

    Story telling in all aspects of our lives is important, yet the average human has lot the ability to tell a compelling story. This is not stricken to just science but to people and their lives. The way I have found humans communicate most effectively is through stories and the importance of a persons personal story has been lost. Stories drive community and relationships.

    In academic papers it is difficult because you want to give them more data (which took a great deal of time and effort) but this data acts similar to a novelists action of giving too much information on the setting and it gets in the way of a good story. Sometimes less data is really more because it provides focus, yet even reviewers resist this and rarely have I not had a paper where I’ve had to try and bring in some more data that I personally didn’t want to bring in because it detracts from the message, flow and quality of writing.

    Nonetheless, I keep striving (as I hope you will too) to tell stories of science because they are extremely important. I am certain that I am no good at it – but that is no excuse not to get a little bit better.

    The challenge academic writers face is daunting – to bring a subject without characters, without enticing conflict to life for the reader. Not only are we compelled to be interesting but we have to do it in a relatively short format (for sometimes very complex topics).

    This is not an excuse, but an internal search of “how do I improve” because I know I am always trying with lectures, papers, and presentation to tell a story.

    I believe the first step is to realise there is a story to be told and to seek it.

    The second step? I’m still seeking

    1. I have the same issue when it comes to data and figures – I often times try to say and show too many things, and the main message gets diluted. So I’ve been trying a new technique in which I ask “what is the minimal amount of information I need to give to make my point convincingly.”

  3. Good article. And not only academic writing suffers from it. I have sat in so many knowledge transfers that were more efficient than the most powerful sleeping pill – there was simply no story.
    One question – would you mind sharing which book on writing fiction it was that helped you?

  4. I agree that prose can be dry and a good story makes data seem more interesting and more compelling but I think you do yourself and science a disservice by talking about story-telling and not about good cognitive models. Allen Newell and Herb Simon would certainly have argued with you (and I will instead) that the data are only as good as the theory and vice-versa and that the theory needs to be sufficiently well specified that it can be implemented in a running program. The underlying theory should not have many degrees of freedom (obviously how you count them depends on the framework) and the data should guide the theorizing (changing it or modifying it if the PREdictions are not supported). The problem with stories is that they leave much underspecified such that it cannot be implemented in a running program and then we have a “just so” story that may not really work as an explanatory mechanism for the data that are found.

    I know you know that, Ven. I just do not want others to think that you believe it is okay to only hand wave an account. I find a really good mechanistic account that can be applied to many phenomena as exciting and interesting as a good novel. I don’t want scientific papers to not be held to account.


    1. That is a great point and I am glad you are bringing it up. I was planning to do a series of post on those issues in the future.

      Here, I wanted to focus on the mode of presentation, rather than on the content. I was not looking to contrast verbal explanations with detailed mechanistic models. I wholeheartedly agree that the latter is better and more satisfying. By a “story” I meant the way a finding/theory/model is introduced and discussed, regardless if it is just a small empirical observation, a simple verbal explanation, or a detailed mechanistic account. I was stressing the benefits of using fiction writing techniques such as good storytelling for the framing and writing of an academic paper.

      But no amount of storytelling will make bad science into good science, although it might make it seem better. In general, good storytelling will make both more compelling, and it is up to a critical reasoning mind to discern the quality of the content itself.

Leave a Reply

%d bloggers like this: