Over the past few days there have been a whole bunch of interesting tweets relating to tricks that help people write better. Many of these have the #GetYourManuscriptOut hashtag that was set up by Raul Pacheco-Vega in his blogpost about the problems with finishing off papers http://www.raulpacheco.org/2014/07/getyourmanuscriptout-or-how-to-fight-procrastination-in-academic-writing-through-crowdsourcing/.
I guess all this attention on academic writing is because there’s a lot of writing going on at this time of year: many PhD students are currently finishing up their theses, as are many MSc students, academics are trying to get their paper submitted and those of us working in HCI are already focused on the CHI deadline on the 22nd September. As a result, I’ve spent most of today reading drafts and making comments on them, and pointing my students to various online resources that provide useful hints and tips about reverse-outlining http://explorationsofstyle.com/2014/06/25/reverse-outlines-from-the-archives/ and how to write good paragraphs https://medium.com/advice-and-help-in-authoring-a-phd-or-non-fiction/how-to-write-paragraphs-80781e2f3054.
So what are the secrets of writing a good paper? Rachel Cayley writes about the importance of extensive revision of a manuscript http://explorationsofstyle.com/2014/06/11/committing-to-extensive-revision-from-the-archives/. But how much revision is enough? How many drafts does it take to write a paper that’s good enough to be submitted? I discussed exactly this with one of my post-docs this summer. We’d made a rather last minute decision to submit a paper to a conference and had just one week until the deadline. In that week the paper went from verson1 to version12. After we’d submitted she asked me “Is it normal to go through that many revisions?” My immediate response was “Yes – of course!” but it prompted me to take a quick look through my folders to see how many revisions my papers usually go through before submission. Very quickly it became obvious that my papers tend to go through at least 10 versions before submission, and those that get accepted have (perhaps unsurprisingly) often had even more . It turns out that in this case 12 was enough and our paper was accepted. Of course we hadn’t been writing it all from scratch. We had various files which contained the method and analysis, and we’d spent many hours talking about the results. What we did do in that week though was quickly identify a narrative for the paper, write up the story, and iterate over the details of how best to present the findings. So, if you have done a lot of the thinking already, then in a week of concentrated effort you can iterate enough to get a paper from zero to submission.
I was surprised today by another colleague when I sat down to read version 1 of the paper we had decided to write for CHI2015. The reason I was surprised is because although it was version 1 it was so polished. It was so good it even prompted me to write to her asking how many versions it had really gone through before she sent it to me. “Just the one” she said! This isn’t the whole truth though as like the paper I wrote about above, we’d already done a lot of work on this one before she wrote the first word. We had written a fantasy abstract (an abstract for a paper that you’ve not even done the research for yet) back in January, and then after doing the study, we’d written an abstract for a talk (in May), and written and given the presentation (in July and August) so actually writing the first version of the paper was easy. The draft of the paper was started just two weeks ago with a file of notes that outline the structure of the paper, the literature that needs to be included, and the main contribution. It took just a week to write. My co-author wrote “I think a an awful lot of thinking has gone into this paper before I sat down to write it – we’ve spoken many times. […] So we’ve had a head start with this one and I’m glad we didn’t need 14 iterations to get it ready…”.
But not all papers can be written so quickly. Another paper I’m currently working on for CHI2015 is currently on version 9. We’ve already done a lot of iteration on it. Version1 was created in mid June and mainly consisted of cutting and pasting text that had already been written elsewhere into the correct format. There it sat for two weeks before more work was done on it. Through versions 2 to 6 the story changed, the paper got chopped from full 10 page paper to a short 4 page note and then back up again to a 10 page paper. During this process we changed our minds about exactly which studies were going to be included but by mid August we had settled on the story. The next two weeks saw versions 7 and 8 in which we refined the write up of the details of the studies, and worked on making the story flow. There’ll be another version or two before the deadline I’m sure. It’s clearly taken a lot longer than a week to write this one. A big part of that reason is because doing the writing has been part of the process of thinking about what our results mean. (See http://explorationsofstyle.com/2014/06/04/using-writing-to-clarify-your-own-thinking-from-the-archives/ for Rachel Cayley’s blog post about using writing to clarify your thinking.)
Having a deadline is always a good way to focus the mind and find the motivation to get something down on paper. So with the CHI deadline now three weeks away, is there enough time to get a paper together if you haven’t already got to version 8?
I’ve got one that reached version 6 today but we’ve had to strip out all the text and engage in some reverse outlining to try to identify one clear narrative for the paper. It’s been a painful day as a result but I’m still hopeful that we can craft it into a good paper before the deadline – there’s still a way to go but it seemed like we made a lot of progress today. With 3 weeks until the deadline I think this one stands a good chance. I’m even tempted to start another one – a paper I’ve been meaning to start for a year. With all the thinking that’s gone into it already perhaps it would only take a week to get into shape? The concrete deadline might be enough to make me put finger to keyboard.
And then there’s another that we’re really excited about but we’re not even half way through the data collection! If all goes to plan we’ll have finished running the study next week. It’s still tempting to think we might be able to do that one too. In my more rational moments it’s obvious that this will be hard to pull off. We won’t be able to write a polished draft in just a week as we haven’t had the luxury of months of thinking about the results and how best to communicate them. Nor will there be time for 10 to 12 versions of the paper to help us craft that story. Perhaps we should line this one up for September 2015?