Academic Workflow: My Writing Process

Since there are a lot of people doing PhDs out there and I would have found this useful when I started mine, I'm detailing my PhD workflow here for anyone who might find it useful. This would also work for general writing tasks. I'm more of a planner than a regular writer, meaning that if you're the kind of person who likes to write a few paragraphs regularly rather than a lot of paragraphs in a short time, this may not work for you, but feel free to read for random things that might work for you anyway. If on the other hand, you find yourself (like me) spending a lot of your time reading, constantly feeling like you haven't done enough work to begin writing, and usually find yourself spending late nights trying to meet a deadline, this has all the things that have helped me work more sustainably.

Methods

Zettelkasten

I use the Zettelkasten method for organising my thoughts and notes and use Obsidian as my primary note taking application to facilitate this. As an aside: the apps and tools that I describe and link here are either free for academic use or open source. Of course, there are other apps that facilitate zettelkasten and it's perfectly usable with old-school pen-and-paper, but of all the ones I have tried, Obsidian has been the best by far. The way I use zettelkasten is to simplify ideas and thoughts and create connections between them. So any new ideas/concepts get their own 'note' (usually a file with the concept as a title). If that concept deals with other concepts, it will refer to that other concept with a link, rather than explaining it in the same file. This works like a personal wikipedia: main topics cross-link with other topics allowing me to easily make connections with other thoughts and ideas. So my concept note "Disability" links to my note on the "Social Model" and "Disability as Oppression" among others, while "Social Model" links to my note on "Materialism". Links to files in Obsidian are automatically created to text in square brackets.

Literature Management with Zotero/Juris-M

I use Juris-M as my reference manager, it is a fork of Zotero that has additional tools for legal reference management. It also allows extraction of annotations from pdf files using Zotfile. As I read papers, I note down any ideas I have as comments in the pdf file. I colour code any annotations (eg. red comments are where I disagree with the author. Yellow comments are important things I want to keep track of. Orange comments are ideas that I have while reading the paper. Green comments are references I need to read/further reading). I'm careful to ensure to make references to concepts that already exist in my Obsidian notes as links. For instance, if I'm disagreeing with the author because the author ignores the Social Model, I will create a red comment in the pdf that says something like "The author bases this conclusion on the [[Medical Model]] and ignores the [[Social Model]] entirely". This will link the comment to my note on the Medical Model and the Social Model in Obsidian.

When I'm done reading the paper, I use Zotfile to extract annotations and then use Mdnotes to send these notes into Obsidian. This creates a file in my Obsidian app for this specific reference, which works as a literature note. Mdnotes can also create literature notes for references without pdfs, which is what I use for most books. In such cases, I create the reference note before reading the book and take notes directly in the reference note in Obsidian. I can now link to this literature note from other notes in Obsidian. So, in my concept note "Social Model", I may have a link to Michael Oliver's book, The Politics of Disablement. That link is a file with my notes on the book. I try to be diligent about note taking when I read. This makes the notes more comprehensive and makes it easy to access ideas even if I've forgotten where they originated. A quick search for the idea in Obsidian will show me what literature notes contain that idea, making it easy to see (and eventually cite) my sources.

Sketching an Outline

I use zotero and Obsidian for most of my writing work. I do eventually use MS Word for the final article, but that's in the Writing Phase of the work. To begin, I create a file in Obsidian for the project, calling it "Project-Name Outline". For papers, I find it useful to have an initial outline with 3 main points, with space to expand it to more headers if need be. Sometimes the outline follows a kind of Hegelian dialectic, in that it has a thesis-antithesis-synthesis structure. Content-wise that translates to (1) here's where we are right now (2) here's why this is bad/can be improved/should be changed (3) here's how to change it. Next, I fill in any sub-categorisations that I'm already aware of wanting to pursue in the paper. Most things in the outline will be bullet points rather than elaborated text. If I need to write elaborated text, I create a separate note for that elaborated text and link to it using Obsidian backlinks. At this stage, I have barely started research into the subject, so I don't treat these as permanent placements of ideas.

As I get into the research, I fill in ideas and concepts in the outline as and when I encounter/think of them. If I read something relevant in a book, I add it to the appropriate place in the outline with a link to the reference note and the page number. This keeps the outline succinct and easy for me to read and understand, while having links to the more detailed argument or source material for elaboration if I need it. I haven't been exacting about the note-taking, but consciously practicing this has helped make it a habit. This keeps track of the literature and sorts it into appropriate places as and when I read it, which is convenient. I don't need to go hunting for specific references to eventually cite. More importantly, it also keeps track of the amount of work I'm doing. As references and ideas collect onto the outline, the size increases. Using bullet points makes it easy to see how many ideas I'm trying to deal with and trim unnecessary bits (they often get moved to my folder entitled "possible future projects"). I usually aim for 20%-25% on word count estimates for the outline. So, if I'm trying to write a 10,000 word paper, I use reaching 2,000-2,500 words on the outline as an indicator that I need to start writing now. This is not a hard and fast rule, and if I feel I've really not got enough to be writing, I choose to trim the outline before continuing research.

Here's what my outline might look like:
List-based-Outline of "Random Topic" with Headers of different levels for Current State (Thesis), Problems (Antithesis), and How to make it Better (Synthesis). Points are sometimes tagged and colour coded. The outline includes links to books and articles along with the page number for the reference.|800

Because I'm very visually oriented, my outline is tagged and colour coded for my reading ease. Obsidian does most of the colour-coding for me so I don't have to think about it after setting it up. I've put in random links to literature notes I already have to use as examples here. I sometimes add tasks to the outline if I'm in the middle of research and don't want to switch focus, but need to follow-up on the task before finishing. I usually get around to them later in the research process but because they're on the outline I know I'll get to them eventually, even if it is while writing.

Writing

Writing becomes incredibly easy after the outline is ready. The only thing I need to do is work through the outline, elaborating on each individual bullet point as I write. Since I already place them in the most appropriate position on the outline during research, I don't often need to edit structure/rearrange text. It's obviously not perfect, but it mostly works. I will slip into research now and then during the writing phase, but for most part, I'm able to keep my momentum on writing the paper going. I use Zotero's inbuilt MS Word add-on to add references to the text as I write. Here's a quick video on how that works.

I am currently using this method to write individual chapters in my Thesis, rather than writing the entire thesis in one go. I'm doing it this way because the sheer size of it means it's easier to break it up into slightly smaller chunks. Since each chapter is between 10,000-20,000 words, it's easier to deal with chapters individually. It's also good for my stress levels to have chapters getting written as the PhD progresses. It's still one massive outline though, to maintain continuity and have space for my research to be added into other chapters, But the research-to-writing cycle happens by chapter. This workflow has helped me streamline my writing, and hopefully can help anyone else looking for ideas on how to manage academic writing.

Published