I Wrote a Master's Thesis in Law
"I must create a system or be enslaved by another man's." -William Blake
First, the finished product:
After 11 months, 53 versions (including five revisions prompted by my adviser's comments), it's done. Here's how I went about writing my thesis.
Research
Most of the resources I used for this thesis are available online. The articles were taken from the university's Westlaw and Lexis subscriptions, while the books were from Cambridge Books Online and Springer. Whenever I use an "offline" source, I took pictures of the relevant pages and converted them into a PDF. Having everything in soft copy enabled me to put all resources in Mendeley,
Using a reference manager had the following advantages:
1. Whenever I wanted to go back to a work that I need to cite, I just looked in one place that's indexed and searchable.
2. It can generate formatted footnotes and bibliographies with a few clicks. This made it easy to cite early and cite often, which helped prevent the slippery slide to plagiarism.
Writing
This is going to look comically inefficient - but every paragraph in the thesis (except minor edits and footnotes) started out as ink on paper. I wrote down the text, quoted passages, did meta stuff like questions and notes regarding the subject - all on unruled notebooks. Only when I've churned the text enough times on paper did I type it into Scrivener, adding formatting and citations (courtesy of Mendeley). As soon as I transferred the text into the computer (or decided that it shouldn't be in the draft), I'd cross it out from the notebook.
For a lot of people it would probably make sense to go all-digital and draft everything from a computer. It doesn't help my tech cred, but I stick to pen and paper because:
1. I write faster than I type. Sometimes I luck into a writing flow state, and when that happens I need to fix my thoughts as fast and as friction-free as possible. With a good pen and nice paper, there is lower impedance mismatch between my writing and my thinking.
2. Fewer distractions. Pen and paper does not have Facebook and email and YouTube. The medium allows me to keep writing for hours - no electricity or Wi-Fi required. I only use the computer when it's time to type or edit the text - something I can manage while listening to music or watching Netflix.
Encoding/Editing
I typed and edited the text using Scrivener a favorite among screenwriters and novelists, but also pretty good for legal and academic work. What I love about it is that it helps you treat the writing project as manageable modules of text. Rather than imposing a mental model of THE DOCUMENT (which could be hundreds of pages long), the metaphor it presents is that of an index card. I just need to fill in a couple of paragraphs per session, and worry about how it all fits together later.
Whenever I've encoded enough to save as an internal milestone (or to present to my adviser), I can "compile" the text (just like code) into a Word document or PDF file. I can specify document-wide formats at compile time, which means I'm not fussing with fonts and spacing while writing. With some AppleScript, you can automate the process so that you have the latest version of your work stored and emailed at the end of every work day.
Change Management
"I want to delete these paragraphs because they're irrelevant - but what if it turns out I need them later?" "What if I try this line of inquiry, and integrate it to my main draft later on if my adviser approves?" "I think I raised that point in an earlier draft, and now I think I have to emphasize that after all." The costs of change management can weigh down the creative process. Most of the writers I know try to impose order through their file system, putting date and edit information into the file names (i.e. Document - SuperFinalFinalVersion-Aprroved-July_26_2016.docx). That's good enough when you have one guy working on the canonical version of the document and you only need to coordinate one thread of edits. What if you need to branch and merge? In my case, my adviser needed at least a weekend to look over my draft and then put in his suggested revisions and comments (Prof. Pejovic worked on the document itself with track changes - which is ideal for my system). Meanwhile, I may have already made my own revisions to the document, and I need to reconcile his version with mine by Monday - and still preserve both lines of edits so I have a reliable narrative (and a backup to refer to). This problem has been solved by programmers for decades - version control systems have been developed to maintain a coherence for large complex projects. The one I used for my thesis is the same I use for programming projects - git/GitHub.
Printing
This one should have been fuss-free - I mean, you open your file and then you print it, right? We have two laser printers in the LLM Student's study room, both connected to public terminals through a LAN. The problem was, I finished my draft pretty much the same time everyone else did - and everyone was using those two printers. These were always overheating and resulted in warped/curled pages. Thankfully, since that time I learned that the Philippine Consulate can't afford to extend printing privileges to citizens and officials, I was forced to learn how to print from pretty much every conbini in Japan. After giving up on the study room printers, I went to the nearest convenience store, and printed copies of my thesis. From the cloud. In a convenience store.
And there you have it - all the tools and practices I needed to write and finish my thesis - on spec and on schedule this time.