Something that has come up recently in the group, and one of the hardest parts of writing a research paper, is the final polishing and editing of a paper before submission. When one is new at research this can be a very stressful time that leads to facing (and sometimes running away from) all sorts of personal demons. Insecurity, confusion, uncertainty…all can raise their ugly heads at this point. If you are still reading this I will assume that you have a rough first draft of your paper – it may well be your first! – and the challenge is to craft it to a state of readiness before sending it off to the journal.
So how do you get yourself to the delivery room and experience a successful birth to your baby?
One approach is to select a writer whose articles you enjoy reading and to study their writing style. Try to reverse engineer it and then simply try to copy their style as closely as possible. Since science writing is fairly algorithmic you can’t go far wrong with this approach, as long as they are writing for similar journals as the ones where you want to publish – using Terry Pratchett as a role model for writing Nature articles might not be very successful! Personally I remember as a young graduate student using the papers of Ed Witten as a model of very clear writing.
An alternative, more step-by-step, approach to polishing is to simply start at the title and then proceed sentence by sentence until the end of the paper asking yourself the following questions at each step:
(1) Is this sentence necessary? Is it adding usefully to the argument I am making? Can I simply cut it out without really losing much? This is particularly important if you are writing a letter where space is limited.
(2) Is the sentence written in good English? How would I say it a different way? Could I simplify this sentence or make it clearer? Could I make it more interesting to read? A bored reader is a dead reader!
(3) What am I assuming about the reader’s knowledge in this sentence? Is this consistent with what I have previously assumed or have I suddenly assumed they know much more or less?
(4) Does the sentence make logical sense? Does it follow logically and interestingly from the previous sentence. This is particularly important. Inexperienced writers tend to string together sentences that do not flow linguistically or logically from each other.
Now read the paper again focusing at the paragraph level. At the end of each paragraph, ask yourself:
Does this paragraph hang together as a logical unit? Is it clear? Could the argument of the paragraph, supported as it is by the sentences in the paragraph, be made clearer, stronger or more compelling? Is the paragraph’s argument necessary? Does it add real value to your paper?
If you had to visualise your paragraph, to run your fingers viscerally over it, would it be smooth and silky, or would it be jagged and rough? Does it taste sweet and moreish or a bit weird and unpleasant like licking a metal pipe or a pair of well-worn sandals? One trick to improve smoothness is to work on making the sentences of your paragraph roughly the same length. Abrupt changes are disconcerting.
Now read the paper again looking at the section level. Ask these same questions about each section…
Viewed this way, a paper is a collection of sections, each of which flows naturally from the previous one and into the next one. Each section is clear, logically complete and necessary. Each section consists in turn of paragraphs which are minimal, clear and flow logically from the previous ones and are constructed from sentences which are clear, logical and necessary as bricks and mortar are to a house. If in addition, the writing is cogent, compelling and engaging, then you are winning!
If you are a good at writing computer code, one way to think of this process is to imagine that your paper is a computer code which will be compiled and executed in the brain of the reader. Your task is to write “code” (aka the ‘paper’) as logically and efficiently as possible so that each line of the paper follows logically from the previous one, and the function of the code is to allow the reader to understand what you have done as clearly as possible and hopefully, to be excited by your work.
Finally there are a few useful things to note around expectations and emotions…
(1) Polishing is hard, and hence often slow, work. Typically you will need at least ten drafts between the first reasonable paper draft and the final polished version. Sometimes many more than ten. But the good thing is that it is very algorithmic. Each great sentence you knock off is one sentence ticked off.
(2) Don’t polish to publication quality until you are sure a section is actually going to be in the final draft. Otherwise you can waste a lot of time on stuff that won’t make the final cut. This is especially true if you have collaborators/advisors who change their minds a lot!
(3) Although your aim is to be clear, try to inject your own personality into your writing, enliven the writing so that it is interesting to read.
(4) It is natural to face some demons in polishing a paper. You may fear that your work is not good enough, that it will be rejected by the community you are trying to become a member of. You may find yourself on your 542nd draft. Or you may have a basically completed draft that has been sitting in a drawer or folder in your computer for the best part of a year. Knowing when a paper is good enough to submit is like asking how long a piece of string is, and that lack of clear guidelines can be very stressful. Allow for the stress, be gentle on yourself. I remember writing multiple drafts of papers with essentially identical results, which were therefore essentially just gauge transformations of each other.
What I didn’t do was get feedback from people I trusted and respected. Asking friends and colleagues – not just your advisor – for comments on your work is a good place to start. If they don’t have comments or criticisms your paper is either ready or utterly boring. If they have lots of comments that is good and remember the old adage that one does not first overcome fear and then act, one overcomes fear by action!