On my most popular post "How to write an abstract in 30 minutes" I was asked if I could come up with something similar for writing a paper.
Recently, I've noticed that I need to hack my way through getting a paper written in the least amount of time possible, as things have been getting more and more busy. Here's my method for getting a first draft of a paper to discuss with my advisors in two days (and it also involves answering questions to myself):
1. The masterplan
I never start by staring at a blank page on my computer screen. Instead, I pick up pencil and paper and start to answer the following questions to myself:
What's the target audience?
Is this paper going to a bridge engineering conference, a bridge maintenance conference, a structural engineering conference, a concrete conference or is it going to a journal? Do I have examples of proceedings of this conference from past years or do I have a paper I enjoyed reading from the journal to give me some ideas?
What's my key message?
Around which argument or idea do I want to structure my paper? This key message needs to be related to the target audience.
How do I structure my argument?
Different approaches are possible: I've been using two main schemes in the past:
1) Introduction - Experiments - Results - Conclusions
2) Introduction - Statement - Proof of Statement - QED and conclusions
What do I have?
Which material is ready, what needs to be prepared? Don't make it too detailed. It is important to know which parts of research you have that you can use here, and which parts you have written out somewhere else.
2. Filling up
Now it's time to turn to the computer again, with the masterplan as a starting point. If there is a format provided by the organizers or printers, I
immediately start writing in the right format; that helps to get an idea
of the length of the article.
I start with writing out the headings of the paper - depending on how I am planning to structure my argument. At this moment, subheadings (if necessary) are added as well. Sometimes I start with more subheadings, but then in the editing stage leave them out and turn them into the consecutive paragraphs of a section.
I like to dump all material I have ready into the document at the right section. I pull a few sections from a previous paper, or from a report, or from some word document which holds references and graphs from the literature. All this information gets pasted into the document. Next, I write the missing sections. This typically amounts to more than the double of the number of allowed pages.
Now it's time again to walk away from the computer. I like to print out this mass of half structured information and sit down with pencil and paper again. I then start to scratch out the information I don't need, condense the text, make the argument stronger and rewrite sentences. Finally, I change the images or I draw additional sketches - only the ones that are necessary to support the argument.
I like comparing this process to making (tomato) sauce: in the "masterplan" stage, I take my recipe and see what I have in my pantry or what I need from the store, in the "filling up" stage, I throw all the ingredients together; in the "editing" stage I make it simmer on the stove for a few hours so it becomes denser and richer and add some spices.