Unfortunately, collaboration in academic writing often causes frustration. Academics are used to think that co-authoring a manuscripts means emailing back and forth Microsoft Word documents with endless “Track Changes” and “Comments” layered on top of each other. Whereas writing is dominated by Microsoft Word, citation is dominated by EndNote. These expensive solutions are probably favored by universities because they inflate budgets and staff of IT Departments by keeping people suitably occupied with resolving bugs and crashes. Whatever it is, one can’t avoid collaborating with colleagues who use different systems, so here is my experience, pointing that it is both software and other factors which determine success of collaboration.
1. After we have decided to prepare a joint paper I ask potential co-authors what software they use. Upon hearing that I use Microsoft Word sparingly because it is not an ideal writing software, my colleagues often roll up their eyes (‘oh, one of those Mac fanatics!‘) and sometimes raise the issue of ‘compatibility’. Concerns over ‘compatibility’ disguise lack of structured thinking about forms and tools of joint academic writing and thinking. So I explain my own writing workflow (OmniOutliner → Byword → Scrivener → Word, all supported by Papers). If the co-author has never heard of such software I always propose that s/he tries it for the period of our collaboration (there are good trial periods as well as PC versions for most of the apps). I also introduce them to the Dropbox and sometimes Google Drive so that we can easily share files and collaboratively edit texts, if needed. Then we agree on the distribution of roles and the stages of collaboration.
2. At the first stage of writing we always prepare a detailed outline of the future manuscript. I use OmniOutliner for this purpose. OmniOutliner is great for collective thinking if your screen can be projected on the white wall or shared across computers with Skype or WebEx. Seeing how OmniOutliner works often makes my PC-Word colleagues respectfully talk about ‘those Aleh’s outlines’. Of course, outlining can be done in Google Docs, MS Word, Scrivener, Pages or other software and supported by mind-mapping applications (I use MindNote Pro and now increasingly Tinderbox).
3. After the draft outline has been prepared, we start writing. I believe in the three drafts concept and that there are benefits in sharing and discussing all of them. Since I produce the first (‘put-it-down‘) draft in Byword or Scrivener, I need to compile it (sometimes by simple copy-and-pasting) to Word or Google Docs for sharing with colleagues who do not use this software. My most frequent co-authors, however, are all Scrivener converts so we simply share Scrivener files through the Dropbox. Scrivener has Revision and Comment tools which I find fully sufficient for discussing earlier drafts. If such drafts contain citations I simply insert Papers citekeys through their MagicManuscripts. I do not format bibliographies in early drafts.
4. After receiving comments on the first draft I usually come back to Scrivener (even if it requires copying and pasting or importing the text back from Word or Google Docs). This is not unusual: Michelle Muto, a writer and a recent MPU guest has a similar workflow. The reason is that the 2nd (‘fix-it-up‘) draft usually requires re-thinking, re-structuring and re-writing which is much easier in Scrivener.
5. The finishing touches in the 3rd (‘dental‘) draft are usually done in Microsoft Word with its ability to carefully track changes and automate dealing with captions and cross-references as well as tables. Many publishers also need submissions in Microsoft Word. This is also the stage for formatting citations and bibliographies according to publisher’s requirements as well as inserting final figures (often prepared with help of OmniGraffle). I have recently described some issues and solutions of collaborating with authors who use Papers. As of now, there are still concerns with in Microsoft Word not always properly handling Papers citations (see e.g. this post from Mekentosj), but they are solvable.
I have recently collaborated with a colleague who used EndNote on a PC. Some of the citations were inserted by him and some other added by me through Papers. He handled the final journal submission. I marked all references cited in this manuscript with a specific keyword (which I highly recommend in any case so that you can quickly refer to the list of citations in all your past manuscripts). Then I created a Smart Collection based on this keyword, exported all references in the collection as an EndNote XML library and emailed to my colleague. Voila! No glitches.