And before anyone jumps in and says there are 4Rs in Science i.e. Reading, Research, Writing and Reviewing, I am including research as part of writing as without research you would have nothing to write about. This post is mainly about writing for publication as I have written about refereeing and reading in earlier posts. Almost twenty-five years ago I designed and implemented a scientific paper writing class for the undergraduate course that I used to run at Imperial College; later I re-tooled it as part of our postgraduate training programme and it was later rolled out across the university graduate school as part of the Doctoral Training Programme.
The first question I would ask students was “Why do scientists write papers?” Undergraduates usually responded that scientists wrote to tell the world and their peers about what research they had done and thus advance science and prevent duplication of effort. My response to this was that if they really wanted to publicise their research and make it accessible to the world they would publish their work in the popular press which has a lot more reach than a scientific journal. After a bit of prodding they would then decide that perhaps it was for peer recognition and subsequent scientific validation via the review process. Postgraduate students reached this stage more quickly and also understood that they needed to publish to make their cvs competitive and also of course to stake a claim to a particular research area to help with obtaining funding.
The first step in this journey is to do some good science! Before setting out on the publication trail I also think that one should ask yourself if your work is important, although of course this is pretty subjective. I am sure that all of us if asked, would consider that what we do is important enough to be published. Next ask yourself if the experimental design or methodology is sound and if the work has been done well. This will save time and remove some of the pain likely to be met during the review process. Most importantly, at least in my opinion, is to ask yourself if there is a story. There needs to be a strong narrative if you want to get people to read and cite your paper.
As a first time author you definitely need to ask advice about who does what, where you will send your paper and it is usually a good idea to get some agreement on authorship order earlier and not later. Even as an experienced author I think that this sort of discussion can be very useful. At the very least it will help you decide what particular slant your story will have.
Remember, have a clear story to tell and also remember that complexity is not the same as learning; keep your language simple, concise, precise and incisive and even at this early stage, make sure you follow the journal style!
At this point in the course I would put up this table and ask the students what each column represented.
They would quickly guess that the first column represented the traditional layout of a scientific paper. The other two columns took a bit longer, especially for the undergraduates until I asked them how they read papers when gathering material for their assignments and they were then able to identify the third column as how they, and most of us tend to read papers. If the title seems interesting then we read the abstract, zip down to the results, see what the authors said about them, then check the introduction and then check the references for follow-up literature. Methods and materials usually trail in at the end and then only if you have some doubts about what the authors have said or if you want to do something similar. Then you look at the results again and you might look at the acknowledgements to see where they got their funding and to guess how many times they had to revise the paper (how many anonymous referees they acknowledge).
The middle column represents how most of us now write papers especially in these days of cut and paste. We follow the line of least resistance, start with the title to give us a starting point, our methods should have been written already in our lab books, the results come next and then we get on to the harder bits, the Introduction and the Discussion; acknowledgements flow logically from this and then it is a matter of adding the references and perhaps the hardest bit of all, the abstract or summary. By the time you have done all this, your initial title almost certainly will no longer appeal to you so you come up with something new and more fitting.
Although this tends to be how we write papers I am not sure that it is actually the best way. In the days before personal PCs some of us had access to typing pools and even if we didn’t, we either wrote our first drafts in long-hand or at a typewriter. This meant that we got all our material together, had a long think about what we wanted to say and actually started at the beginning and worked our way through the paper in the same order as it would be printed. Some people argue that this meant that ‘flow’ of those papers was smoother and more coherent. I don’t think I know anyone who actually writes like that anymore, but I am happy to be contradicted.
Regardless of the fact that most of us live in a cut and paste world I am going to work through the various bits of a paper in the usual printed order. Remember you are telling a story and there are a lot of rival authors out there competing for space in the top journals and you have to convince the journal editor and two or three referees that your paper is the one that should see the light of day in their journal.
You need a title; ideally it should be short, snappy and very importantly informative, although perhaps not too informative. In the course I ran, I presented this to the students as a somewhat tongue in cheek example;
The effect of two lodgepole pine (Pinus contorta Douglas ex Loudon) seed origins (South coastal and Alaskan) on the growth, survival and development of larvae of the pine beauty moth, Panolis flammea (Denis & Schiffermuller) in the presence and absence of predators in a Scottish field site.
This although informative is not necessarily going to gain you readers or publication in a high impact journal. In fact the external pre-REF (Research Excellence Framework – UK academics will know what this is) consultant employed by my university to help decide which papers should go forward for assessment was very clear that titles beginning with The effect of were very unlikely to receive high scores by the external assessors.
As it happened, the work I had done which is very clearly described by the informative title above was actually published as
Not very informative but it certainly got a readership.
The abstract is perhaps the least favourite bit of a paper for authors; I certainly find them difficult and invariably save them to last. They are however, extremely important and according to Wiley-Blackwell, publishers of the journal that I edit (Insect Conservation & Diversity) they are much more important than we as authors realise – they and the title are the ‘hook’ that gets your paper downloaded and hopefully read and then cited. You should thus not just rush it off in a couple of minutes. Think hard about what you want to say and what it is that is likely to get someone to download and read your paper.
Next is the Introduction, here you should put your work in context, remembering that it is not a literature review but make sure that you do cite some of the earlier relevant work as well as the more recent literature. State the problem clearly and indicate who else has tried and failed and why your work is special and how you have succeeded where others have failed.
Now for the Materials and methods section, which to me is the most important part of the paper. This is where you as a referee or reviewer should go first. This is the detail that matters. If the methodology is flawed then it doesn’t matter how great the writing is or how fancy the statistics, the paper should be rejected. I think it is deplorable that there are now a number of ‘high impact’ journals that have relegated the methods to a subsidiary position, almost hiding them away and placing the results at the front end of the paper. This is tantamount to telling reviewers that the methods don’t matter, just look at the results. I have heard however, that some of these journals are now reconsidering this policy after some embarrassing publicity.
My advice to students is that the methods should contain as much detail as would be required for someone else to repeat your work without having to contact you. So for example, the species involved, cultivars and phenological stage of the plants used, the sample size; for field work, the site details, the equipment used but not necessarily the supplier, unless of course it is very specialist, and the statistical treatment and assumptions.
The results section is your showcase. Decide which display method is best for the message you want to get across and then pick out the most important points from your tables and graphs and turn them into a commentary, but DO NOT discuss them. For the figures and tables do make sure that you follow the journal style. Make sure that the figure and table titles are informative and comprehensive; in the days before Japanese journals published in English, the only English bit of the papers were the figure and table legends and it was possible to get a very good idea of know what the paper was about from them. Keep symbols simple and check line thickness.
The Discussion section is where you discuss YOUR results, highlight the strengths and weaknesses of your approach, underline your most important results, compare them with similar data and interpret your results in the broader view. It is always a good idea to show how you addressed your initial hypothesis. This and the methods section are the two sections where you can try and pre-guess the reviewers and get your retaliation in first. If you can answer the reviewers before the questions are raised in their reports it increases the chances of getting your paper accepted.
Again, DO NOT use convoluted and obscure language and do AVOID jargon and pretentious statements. As scientists our job is to communicate, not just to our peers, but to a wider audience. Quite often the reason our results are misinterpreted by the popular press is not because they are doing it on purpose but because we have obscured what we have said by using over-complicated language. Be clear, use simple everyday words where possible, e.g. laid rather than oviposited and be concise.
Speaking as an Editor I like acknowledgements to be brief, but do appreciate that there are funding agencies and helpful colleagues to thank. I would advise against too much flippancy as after all you are advertising yourself and some people do read them.
Finally, the references; are they COMPLETE? Do they follow the journal style? Editors do check and if you have had the misfortune to be rejected by your first choice journal, it does not go down well with the Editor of your second choice journal if you haven’t made the effort to change the formatting! Do text citations and bibliography agree? Check and recheck!
So now are you ready to release your pride and joy into the wild to suffer “the slings and arrows of outrageous fortune” more commonly known as editors and referees? Actually no, that was just the first draft! DO NOT SUBMIT IT YET. Pass it around for comments; if you are a PhD student your supervisor definitely needs and wants to see it! Let colleagues read it too and for communication test, get a non-specialist to read it. If they can understand what you did and what your central message is then you have cracked the communication barrier. Do listen to what people say, rewrite it!! Be brutal in revision! It is better to revise before submission than to have your paper rejected without the chance to revise. Pass it around again. Then and only then, log on to the journal site and start the submission process, but do remember to read the guidelines for authors before you press the submit button!
Don't Forget the Roundabouts 