Scientific writing
Why we ignore writing
Two main components
Broad picture advice
 Scientific writing: sometimes we’re not good at it, often we don’t like it, but it will always be part of our job
Writing is an incredibly important skill for your career as a scientist. For better or for worse, scientists increasingly have to be good writers to secure research grants and publish their work. Part of this tendency is justified; communicating scientific results is just as important as getting them. Researchers need to discuss and build on each other’s findings instead of being isolated to their own work. If we do not communicate properly, we end up repeating each other’s experiments or failing to understand their broader context. On the other hand, glorified writing can obscure the quality of science and transform it into a marketing contest. Regardless, here is a simple fact that graduate students should understand, especially if they are interested in a career in academia: writing is part of the job.

Sadly, this simple fact is often not emphasized enough in grad school.

Below is a set of  skills that are important for the success of early-career researchers, listed in the order of how “urgent” they feel during grad school:

  • Hard work: How many hours do you work in a week? Can you deliver on a deadline? How much data can you produce?
  • Technical expertise: What type of experiments/simulations can you perform? Which instruments can you operate? What techniques can you add to the repertoire of the research group?
  • ­­Specific scientific knowledge: Do you understand the science that is immediately relevant to your project? Do you know the landmark papers in your research field? Are you up to date with the recent literature in this field?
  • Scientific writing: Can you communicate your results?
  • Broad scientific knowledge: Do you understand the broader scientific field that you are part of? Can you come up with interesting research ideas? Do you have a solid understanding of science in general? Are you a curious scientist?

In general, students tend to 1) focus on ultra-specialized knowledge more urgently than broader scientific knowledge, and 2) nurture technical skills more urgently than communication skills. The direct consequence is that their technical abilities are not communicated effectively, and they are undervalued in the job market.

Here’s the PhDeal. Many extremely talented, hard-working, productive, excellent students finish graduate school without learning a crucial skill to take their careers to the next level: scientific writing. In this series, we delve deeper into the reasons for this problem and offer practical advice on how to overcome it.

Why we ignore writing skills, and why we should not

Early career scientists are usually better at the technical part of the job compared to the communication part. Here’s why:

  • Improving writing skills does not feel urgent. You might think: Why should I learn how to write now? My PhD advisor will edit my publications, and I will somehow magically become better with time. I have years left in grad school. Time is on my side. Oh by the way, I need to run another experiment …
  • There are no course offerings on this topic. Actually, there are, and they can be useful. The problem is that you want a direct course for scientific writing that is efficient and cuts to the main substance. But what you find are general writing courses that drag over a whole semester and seem too light to help you improve. I believe that taking a course on writing is only useful if you make it useful. It might be the best thing that ever happened to you or a waste of time depending on your attitude and how much energy you invest in it.
  • Writing sounds like it belongs in the realm of humanities. Of course, that is not true. Scientists need to communicate. Everyone needs to communicate.
  • Writing can be incredibly inefficient. You can stare at the screen and obsess about a word for hours. Why would you do this to yourself?
  • It’s hard.

Here’s the PhDeal. Your science education goes well beyond doing experiments and churning out papers. You should invest in learning how to write better, even if it first feels very hard and inefficient.

When you improve your scientific writing, you will also learn to think more about logical flow, and connect previous experiments with what to do next. You will think about the big picture and remember the reasoning behind the experiments, not simply do more for the sake of doing more. You will learn to make the most of your data, and how to turn the data into a coherent scientific story.

From a practical point of view, you will improve your ability to write manuscripts, review papers, and grant proposals, as well as job applications, CVs, and cover letters. You will get the job that you–and your excellent technical skills–deserve.

Two main components of writing: Language proficiency and scientific
1) Language proficiency: We signed up to do science, not to be English writers. Especially if you are a non-native speaker, it is very understandable that you struggle with writing in the early stages of your career. In fact, even if you are a native speaker, you are not expected to have a mastery of grammar and syntax. In addition to the cultural and language barriers that foreigners might experience, everyone faces an important difficulty: scientific writing is different from spoken English. It is a skill we all struggle with…

That being said, as a graduate student, you have to invest time and energy to improve your scientific writing. There is no other way. Not having excellent writing skills on day one of your PhD is okay. But doing nothing about it for years and years is a decision that will negatively affect your career sooner or later. To land your dream job in academia, you need to be as good a writer as you are a scientist. Even if you have superstar technical skills, you still need at least decent communication skills.

2) Scientific soundness: This is the more urgent and important aspect of writing. Forget the language barrier. Without scientific soundness, even writing ideas in your native language or picturing them in your mind is not good enough. Being scientifically sound means having a vision for the whole research project, focusing on the main hypothesis, thinking if the experiments truly answer the questions, scrutinizing the evidence and counter-evidence, and delivering a flow of sentences that helps the reader arrive with you at the main conclusions. A publication is not just data! It is a coherent storSome people have this talent naturally. But I believe that everyone can learn the basics, gain enough experience, and reach a very good level of scientific soundness.

Think of it this way: You just drafted a paper and will send it to your advisor. It can be good English (language proficiency) and good science (scientific soundness). It can be bad English–as long as you tried your best, tried to improve–and good science. But it should not be bad science, regardless of your language skills.

Here’s the PhDeal. Graduate students tend to invest 3 or 7 or 12 or more(!) hours a day working on an experiment or refining an ultra-specific technical skill. Sure, technical skills are important. But they should not come at the expense of zero hours a week invested in improving your scientific writing. What is the use of one more dataset if you will not communicate the results effectively? Just as you learned to operate a microscope or code a simulation, you can learn scientific writing. Not being born with this skill is out of your hands, but not investing in it is a choice.

Broad picture advice on writing a scientific manuscript

Before taking advice from anyone, remember that the most important factor in improving writing skills is your decision. Most of the following tips are just common sense and will come naturally once you invest time in writing. As cliché as it sounds, improving your writing really comes down to being proactive and going for it.

  • Do that crucial data analysis that you have been putting off for a while. Do what you should to build the story of the publication. The worst situation is the limbo phase when you have the raw data you need, but it stays half-processed forever or not in the right format or stored on multiple disk drives that you never have with you at the same time… Be proactive. Delaying the data analysis does not make it less important. You should have a clear flow of ideas–backed by data–before you start writing. Note that you do not necessarily need every single piece of data at this point. Usually, it is wise to start writing as soon as you see the complete picture even if a few of data points for completion are still pending. However, the core data should be there. You should be able to visualize the manuscript story and prepare the figure drafts before you start writing, even if one or two figures will be edited later. 
  • Write the manuscript sections in the proper order. We are tempted to start with the introduction, but that is a tricky approach. The goal of the introduction is to align previous literature with your current work. But you cannot do that effectively before actually writing up your current work. Instead, start by making proper, clear, powerful figures. This will help you visualize the whole manuscript and understand it in detail. Also, writing becomes much easier when it is built around a story with support from illustrations. While preparing the figures, jot down half-statements or ideas that you will elaborate in the manuscript text. Then proceed in the following order; I learned this strategy from my PhD advisor and have found it very effective: 
1) Figures, 2) Methods, 3) Results and Discussions, 4) Conclusion/Introduction, 5) Abstract/Title

In the next article series, we will provide practical tips for each of these manuscript sections. But here are some starting comments:

  • For your first draft, “the perfect is the enemy of the good”. The next sentence you write will later be edited at least once. It might be changed over and over beyond recognition. Keep that in mind and be practical. If you are overthinking and hunting endlessly for every perfect word, you can end up writing nothing at all. Or worse, staring at a screen for several minutes hoping for divine intervention (your advisor?). Draft and edit. Write and re-visit. Move on.

What you should overthink and agonize over and mull repeatedly in the first draft:

The general science story of your manuscript. Do you picture the story in your mind? Is it coherent? Do the ideas flow? Does it make sense? Does one result lead to the next? Is it convincing data? Did you answer the hypothesis? And, of course, what did you learn from this study? Do overthink these questions. Once you have clear thoughts about them, writing itself becomes easier. Actually, this explains why it is efficient to start by preparing figures that crystallize your thoughts.

What you should NOT overthink and agonize over and mull repeatedly in the first draft:

Every single sentence! Just write the ideas as they come to your mind. When you have a mental block, write in spoken English. Do not worry about the language flow and the fancy terminology for a minute. Focus for now on the science flow. Draft the sentences in the order that makes sense as if you are reminding yourself of the rationale behind the experiments. Do not overthink every single word. Later re-visit the paragraphs you wrote and translate them into decent English.

  • For your final draft, be ruthless. Every sentence can be improved. Every word counts. This is the caveat to the previous point. Once you have a workable draft, the honeymoon period (sure, let’s call it that) is over. Now, you do have to overthink every word. This exercise can be long and tedious. But you should chop and change and optimize until you are comfortable with every single word. Own your paper. From the figure captions and reference formatting to the method details and data analysis. Write simply. Check commas and semi-colons and hyphens. Google word definitions even if you thought you knew what they mean. Skim through papers you cite. Search on Wikipedia concepts you use. Double check calculations. Edit until you are comfortable that you are telling the best possible story in the most direct, simple, and clear way.

Here’s the PhDeal.

  • The goal is to create a final manuscript draft that is a good science story written in simple, clear language.
  • To get there, you need a workable draft that presents a scientifically sound flow of ideas, even if the language is not great yet.
  • To get there, you need to crystallize the scientific story in your mind and prepare excellent figures that will guide your writing.
  • To get there, you need to be proactive in acquiring and processing the core data for your publication. Otherwise you will keep moving the goalposts as new data comes in, and you will struggle with your story and messaging.

Sounds like fun, right? You would be surprised. Writing is fun. Good coffee helps. Anyway, now let’s get into the weeds and some more practical advice.

Practical advice on writing: Setting 

There are exhaustive resources you can find for professional help to improve your writing. Here, I will not focus on common grammatical or syntax errors, but on ways to improve writing efficiency, structure, clarity, flow, story-telling, aesthetics, style, and most importantly the scientific soundness of your publication.

Again, remember that the single most important thing that will improve your writing skills is your decision to improve, and your time investment to improve. Without your personal initiative, advice and tips count for nothing.

Create the appropriate setting for productive writing sessions:

  • Disconnect. Disconnect. Disconnect. Disconnect from the outside world. Work does not get done if you are actually not working. Turn off your internet access. Put your cell phone in another room, or at least far from arm’s length. Cancel commitments and interruptions. Set aside a solid block of writing time. Exercise and sleep early the night before. Wake up early and have a good coffee. Disconnect. Attack the problem.
  • Work with a colleague, but be serious about it. If done correctly, this is very useful. Working with someone minimizes distractions and helps you stay focused. In the very least, you are forced to sit and stare at the screen for a while. Of course, you both have to be serious about the work session. Otherwise, it will be counter-productive and you will distract each other. Be clear about not wanting conversations. Maybe you can plan to have a quick bite or a coffee break at a set time, but get some the work done first.
  • Create your “zone”. It can be a clean desk in your room or a busy but not loud coffee shop or a cubicle in the school library. Figure out your preference early on and stick to it. After a couple of productive writing sessions in this “zone”, you will begin to associate it with a feeling of productivity. Good, more productivity. Be comfortable in this setting; you will come back many times. Of course, changing the scenery helps every now and then. But if you change it many times, you will be distracted more easily. Or worse, you will create excuses for why today’s writing session was not productive. The coffee was not good. The chair was too creaky. The background music was too loud. Yada yada yada. Nothing got done.
  • More details about the “zone”. Be comfortable, but minimally distracted. Again, your preferred zone can be at home if you can overcome the temptation of the fridge, TV, and Netflix. On the other hand, at a coffee shop, you do not have full control over distractions and ambience. Your office and school library are always good options. In any case, eat well beforehand and get your laptop charger with you and take care of all those niggling things that might tempt you to quit your session early. When writing, listen to classical/instrumental music or a song on repeat or no background music at all. But do not flip through a song playlist; you would be lured into changing the song every few minutes and get distracted.
  • By the way, do not put your life on hold. Although it’s a good idea to cut down on activities and focus on writing, do not get into a siege mentality and cancel your life completely. No one can write efficiently for twelve hours a day. Resting or meeting other people is productive too. Do stuff. Planning an appointment at a set time helps you create a mental deadline for the day and be more urgent, instead of being complacent and dragging through a supposed “whole day of writing”. Going to the gym is a good idea during writing season. Physically, you will feel energized. Meanwhile, your brain is still working. When you are back in front of the screen, the words come out.

Okay, you have nailed down the setting and are ready to start writing. Next, let us discuss practical strategies.

Practical advice on writing: Strategy
  • Create small achievable daily goals. If your goal for the session is “to write the paper”, you will overwhelmed. Set smaller goals for the day: prepare two good figures or edit the methods section or write up the results for Figure X or re-visit the core papers you cite. Goal-setting helps you write more urgently and win small battles. Nothing beats the feeling of crossing out items on your to-do list. Make sure the daily goals are realistic.
  • Split your long writing session into smaller 40-60 minute sessions. If your writing session is long and open-ended, you will be lured into a false mindset that you have lots and lots of time. This is a recipe for complacency. Instead, decide to focus hard for 60 minutes. Put the timer next to you. No distractions. After that mini-session, relax a bit and go again. The reason is that you can be efficient for one hour, but not for five hours. Always think in terms of smaller deliverables and smaller sessions. One strategy is to unplug your laptop and promise to write non-stop until the low-battery warning comes up.
  • Do not quit a writing session early. If you feel sluggish and unproductive, do not immediately give up. Consider taking a short break and trying again. Otherwise, you might fall into a frustrating routine where you expect too much from a single session, then fail to deliver again and again. Note that your next session might be similarly sluggish–this is the nature of writing. Fight for this session. Set a smaller deliverable, no matter how small, and achieve it. End your session on a happy note. Be bullish that next time will be even better.
  • Track your progress. One way is to email yourself the latest draft (on the same email thread every time) at the end of the writing session. This creates a daily measure of your writing progress. You will be happy to see the paper grow and also keep a timeline of how often you are working on it. Of course, this is also useful to keep a copy of the last version on hand for backup. Another approach is to set daily deliverables for the next two weeks.
  • Quickly jot down flashes of ideas. Revisit, edit, and complete them later. Sometimes you focus on a section for an hour with little progress, and then get an idea for a sentence in a completely unrelated section, which leads to another sentence, then another, then another, and that opens the floodgates. Each time you sit to write, start your session with a fresh look at what you wrote last time. Minor tweaks and edits will make the text flow better. You will also remember your thought process and build on it, instead of starting from nothing.

In later sections, we will discuss detailed strategies related to specific sections of your manuscript. But now let us look at resources you have for support.

Practical advice on writing: Support
  • Ask repeatedly for (a certain kind of) feedback from your advisor. Not necessarily about the specific wording of a sentence or two, but more about the general flow of ideas in the manuscript: “I plan to discuss this result … then move on to that … then show this data … what do you think?” These conversations are both practically useful and surprisingly motivational. Get on the same page with your PI and gain more confidence. Do the same with your group colleagues. Again not necessarily in terms of the nitty gritty details of your paper. But it helps to discuss the big picture. (Nitty gritty is a funny expression, I hope I used it correctly here!)
  • (Side-note) Summon the voice of the advisor within you. (This is exactly as weird as it sounds). After the first couple of papers, you will become used to the writing style of your advisor. Remember their comments while writing and “pre-edit” before sharing the draft. Be stringy in this regard. For example, you know by now that your figures can be improved by making the axis font larger. Surely your advisor will tell you this. Do not wait for them to tell you what you already know. Be tough on yourself and plan to send perfect figures from your first draft attempt. This will raise your standards, make your writing more efficient, and cut down on the number of drafts floating back and forth between you and your advisor.
  • Find writing support at your university. During my PhD, I did not know this kind of support existed and till I kind of stumbled upon it. We had an office that helps grad students apply for grants. Some advice I got was just great and raised my game. For example, I was urged to write a more powerful and relatable introduction, I had been stuck on the tiny details of my tiny ultra-specialized science world. 
  • Search for and use words, phrases, and general scientific concepts correctly. Internets. Wikipedia is your friend. Double check every word you are not sure about. Actually, double check the meaning of some words even if you have used them many times before. Check if there are other words that better describe what you want to say. In the longer term, this will help you build a larger repertoire of phrases. Do the same for science concepts and terminology. Even if you learned these concepts in courses, make sure you understand what they mean and cite them correctly.
  • Read the author guidelines for your target journal and other boring resources. Make sure you conform to the paper format of your target journal. Also read (okay, skim through) the (incredibly boring) Style and Notation Guide of the main scientific organization in your discipline (example: American Chemical Society or American Physical Society). These guides are incredibly boring (did I say that already?), but pretty useful. They will not inspire you or change you as a human being, but you will learn how to improve your writing skills and become more professional. 
  • Get the proper software support. Make sure you have a tool for organizing references (example: endnote), preparing figures (example: Origin, Corel, Inkscape), and data processing (example: Matlab, Python). We will discuss more details on figure preparations in a later section. Also, get acquainted with useful support tools for your research field like LaTeX or software for analyzing specific types of data.

Before delving into more details, let us discuss some advice on writing style.

Practical advice on writing: Style

As we wrap up this section, we provide a broad overview on writing style.

If you are writing a scientific document, the goal is to be clear, brief, precise, simple, and direct. Each word should carry its weight. No fluff. No exaggeration. Start by observing and describing results, not by interpreting them. If a thought can be said in a simpler way, say it in a simpler way.

It is easier to write–but harder to read–lengthy sentences that drag on and on and on and on. Mark Twain once said: “If I had more time, I would write a shorter letter” (original version of this quote in French by Blaise Pascale). Simple is powerful.

  • Write similar-sized paragraphs of 150-250 words.
  • Write short, clear sentences. Avoid convoluted phrases with lots of conditional clauses. If a statement can be broken down in two, do it. Limit the use of parentheses with sideways thoughts. You do not want long sentences with interjecting ideas here and there (although that is sometimes necessary (and not a big problem (but do not overdo it))).
  • Use the active voice where appropriate. In some cases, the passive voice is just weird. “The procedure has been performed” sounds like a foot soldier delivering a message from a medieval king to his subjects. Do not be shy to write “We performed the procedure.”
  • Be consistent in your choice of tense.
  • Be clear,
    • Introduce acronyms before using them.
    • Do not use demonstrative pronouns without their subjects. “This supports our conclusion” => “This result supports our conclusion”.
    • Refer to the figures in the correct order.
    • Use proper punctuation including commas and semi-colons.

Do not get too attached to a word, phrase, or idea. Sometimes deleting is good. Save that deleted phrase for later. You might find a more appropriate use and recycle it in another section.

Here’s the PhDeal. This article wraps up our discussion on scientific writing in general. In the next section, we delve into the details of writing and publishing manuscripts.