The Art of Science Writing, Part 1: Getting Started

We’ve previously touched on writer’s block, and the strategies and tips you can use to get over the initial hurdle of the blank piece of paper. This week I’ve been inspired to revisit the topic of writing, in part because of my own return to science writing after a bit of a break. I greatly enjoyed writing in grad school, perhaps in part because I knew that writing before the end would help me finish my dissertation, but found that picking things up and getting in ‘writing’ mode again after almost two years of lab and computer work as a post-doc was a difficult task. Where do you start when you have nothing but a blank page? How do you go from a few figures to a draft of a manuscript?

As we touched on in our previous post, there are a few ‘blocks’ to get around in order to let the creative juices flow. Just as with presentations, there is no such thing as being held back in your writing by being a bad writer in science. You may not be a naturally prolific writer (just like I am not a naturally confident public speaker), but the great thing about writing in science is that if you stick to a plan and have a goal with what you want to write, you can always get there. In science, it’s not about how big your vocabulary is or how similar your writing is to the great novelists of the 21st century: it’s about sharing your story with clarity and enthusiasm, all laid out in a logical and progressive manner. So don’t let being a ‘bad writer’ bring you down or become a common excuse for you to avoid writing.

As with our easy* steps for a perfect** presentation series, we’ll detail a step-by-step guide to writing, focusing on how you can go from a blank piece of paper to a respectable draft. But instead of calling this the ‘Five Easy* steps for a perfect** paper’, this series will focus on the art underlying science writing. Because in reality, art isn’t only about fanning those flames of creativity, it’s also about getting your tools ready, doing some preliminary sketches, and having the technical knowledge to bring your vision to life. You can’t just be a good artist to make good art—you have to put preparation and thought into the works you create in order for them to be impactful.

Step -1: Read!
Just as with our presentation guidelines, there are things you can and should do before you begin writing a manuscript, grant, science blog, or really anything short or long related to science and to your work. Before you can begin to write and become a better writer, you should read and work towards becoming a better reader. Obviously you’ll read a lot of papers that are relevant to your work, but how many times do you actually read a paper versus just looking at a couple of relevant figures or glancing over the methods section?

If you want to see how science writing works, you need to read the results of science writing. See how manuscript authors lay out their story, how they bring together figures and results to a cohesive conclusion, and what works and what doesn’t in terms of style and structure. You’ll likely find more than a few bad or boring papers in the bunch, so when you do find a paper that sticks with you, keep it around: highlight the key points, see how they laid out their figures, and get a sense of how they developed their story. At the same time, learn how to critique a scientific paper. Focus on both the writing itself as well as the underlying logic. Do the findings they present match up with the conclusion they drew? Do the experiments they did line up with their hypothesis or project objectives? Were there any fundamental flaws in how they designed the study that weakens the conclusions they can draw? While you’re learning by seeing how others write, you can also take the time to become an evaluative, critical scientist, which is crucial at any stage of your career.

You don’t just have to turn to the scientific literature for reading inspiration, though. Your job as a writer is essentially to tell a story: a story with a beginning, a middle, and an end. Just because your story is crafted on data instead of imagination doesn’t mean that the methods used by authors from outside of science aren’t relevant. Take a break and pick up a new novel or read a non-fiction book on a topic you’re interested in. How does the author keep your attention? How do they transition between ideas or paragraphs? What words or phrases do they use that stick with you? We pick up a lot of our vocabulary and our way of phrasing ideas from listening and reading, so by enjoying more good writing you can also become a better writer.

Step 0: Make a story board
Sound familiar? That’s because it’s the same piece of advice we gave in our perfect presentations series. Before you open up that dreaded empty Word document (or the boilerplate ‘.tex’ file for those Latex nerds out there), do some ground work and set yourself up for success by drafting your outline or, to use the analogy from last time, a story board. As with the presentation guidelines, the purpose of the storyboard is to provide some structure for your ideas and to let you be creative while at the same time helping you guide your creativity in a logical manner.

So what exactly do you need to create your story board for science writing?
Figures. These should be in a 90% final form before you begin writing your paper. Maybe you’ll add something else in that was initially lacking, maybe you’ll change the label on some axes or change the color schemes, but overall they should be static the moment you begin writing. The figures should be able to tell a story on their own, the story that you’ll craft into words around these core figures.

You may think that the place to start with a manuscript is the introduction, but in reality you should focus on your figures before thinking of any other part of the paper. If you are a pen-and-paper type of person like me, print off each figure as its own separate sheet of paper. Around the sides, make notes about the figure. What do the bars show, on a very basic level (e.g. ‘Number of eggs per brood’)? Is everything labeled appropriately? Someone should be able to look at the figure, even without the caption, and have a basic understanding of what’s going on (such as ‘ok, there’s an increase in the number of eggs based on the dose of the treatment’).

Once you’ve had a thorough evaluation of the figures themselves, draft a caption for each one. Start with bullet points of the take-home messages for each figure. What does the figure show, on a more advanced level (e.g. ‘Differences in nutrient uptake in treated versus control animals’)? What should someone understand about this figure that they can’t figure out just by looking at the image itself (such as how many replicates are in each measurement)? Take these bullets as the starting point for your figure caption, and when back on your computer go ahead and write a full paragraph for each figure based on these bullet points. So now you’ve now got your figures and figure captions, which in reality what most people will turn to first in your paper-so you’re off to a good start!

Experimental protocols. Once you know the basics of what you’re going to show with your figures, start working on an outline for your experimental methods. This is generally the easiest (and also most boring) part of a paper, but from a scientific perspective is the most crucial. As you put together all of your relevant figures, dig out your lab notebooks and protocols to get all the details of your experiments. Make note of any steps of an experiment that fall outside the scope of a more standard operating procedure, or if a group of samples from one analysis was processed in a different way that the others.

Have your lab notes and protocols on hand and give them another read-through before you start writing. You can also look at methods sections from other manuscripts (even ones from your own lab) to get a feel for what information is important and what is superfluous. But be careful not to just copy-paste the methods section from another group’s manuscript, or even your own group (or your own previous manuscript). Even without any malintentions, simply reusing a section may be plagiarism or self-plagiarism.  Rewriting the methods section ensures it is current, and it may end up being more clear or concise.

A pile of papers that have already been read. This is again a spot that can trip people up in the writing process. Once you’ve got your figures and protocol in place, the next step is to think about how to craft the story around them. You used the protocols to generate the data that you’ll present in your figures. But what’s the contribution to the existing body of knowledge? What’s the context of why that work was done, and how does it fit into what other data is out there already? How does this help your field understand a problem/scientific question?

As with step -1, it’s hard to be a good writer if you don’t read. And while you may have a basic understanding of what’s going on in your field or within this topic, you need to take a closer look at the literature before you start writing in order to craft your story and lay out the logic in the most appropriate way. So before you start writing, read in detail any of the manuscripts that you’ll most likely cite: the papers with the experiments that inspired your work, the papers that did similar types of experiments but with perhaps different systems or questions, and the papers that challenge your result at some level. Even if you’re read them already, read them again and make notes on the important findings or concepts that you’ll need to construct your paper.

Once the literature review is done, you can use this pile of knowledge to construct your storyboard. Think of your introduction and conclusion not as two separate components, but instead as a continuation of one to the other. The introduction is the beginning of the story, the methods/results is the middle, and the discussion is the end.

In your introduction, you set up the coming tale. As with our presentation guidelines, you can use the following format: your paper, just like your presentation, isn’t a series of facts, but is instead a means of presenting a specific problem, its overall importance, and your approach to solving it. You can consistently keep this to anywhere from 4-5 paragraphs by using the following layout:
- Paragraph 1: What is the problem and why should the reader worry/care about it?
- Paragraph 2 (- 3): What’s been done to address/know more about the problem so far
- Paragraph 3/4: Knowledge or tools that can be used to further address the problem
- Paragraph 4/5: Aim of paper, experimental objectives, and also list any specific hypotheses

The methods and results section are pretty cut and dry, and don’t need much of an outline apart from what’s in your experimental protocols and the bullets you jotted down while working on your figures. Keep any specific or detailed interpretations of figures (such as ‘the decrease in egg production is related to an increase in temperature’) for the discussion. The results section should be very cut and dry, with one paragraph of results per figure. Focus on the basics of what each figure is telling you and save the juicy, exciting bits about what it all means for the discussion.

In the discussion, you continue the story started in the instruction, but now you have a new factor to accommodate for: the data you generated in the manuscript. How does your new data fit in with what was known already? Does anyone have data that disagrees with yours? Frame the discussion as a way of addressing the questions you presented in your introduction, how your results fit in with your hypothesis, and what the limitations/future directions of your work are.
 
As you make your outline, put as many ideas, relevant citations, and things to mention in the paper in your storyboard as you can think of. You likely won’t use half of them, but laying out any potentially relevant findings can help provide context for what you should discuss and how you should frame your writing. One way to do this is to break down each paper into a series of bullet points. List out relevant methods, rationale, hypotheses, findings, and if you think results were interpreted correctly. Another alternative is to have bullets ranked by topic, and then list papers and relevant results under that topic, and see where the similarities/differences lie. I’ve tended to use a mixture of both, and then while writing used color-coded notations to help me keep track of what sections were written where.

For my own dissertation, before I began writing the introduction and conclusion sections, I first laid out the literature and the key points I wanted to address in a very long outline. While in the end I only used about half of what I put into the outline, when I was ready to begin writing I was able to jump into it quickly, without having to go back and forth between reading and writing and disrupting the flow of my ideas and thoughts. Being ready to write means being more efficient at writing, because you can let your ideas come to life without having to jump back and forth between different tasks, thoughts, or distractions.

It may seem like a lot of work before you even start writing more than just a few bullet points. But think about it this way: How much preparation work do you put into a big experiment? How much time does it take you to code something that’s never been done before? How much washing, chopping, and reading recipes do you need to do before you cook a nice dinner? A lot of the things we do, both in lab and in life, take a lot of pre-work in order to come out at a high quality and to be done efficiently. The work you put in before you start writing will allow your work to take off once you are ready to get started-and will make the task less tedious and tiresome, since at that point all you’ll have to do is to tell the story.

Next week we’ll go more into detail of how to take the outline and figures and construct a story around them. Until then, happy storyboarding!

---