These are a list of writing "strategies" that can be helpful when writing (e.g., to get going or to get unstuck). A meta-strategy is to jump between strategies: Feeling stuck? Then pick another strategy. Each strategy can be applied at different granularities, e.g. for an entire book or paper, or chapter or section, or just a subsection or paragraph. Different approaches may work better for you in different phases of a project.
In all cases, remember to avoid premature optimisation; re-writing and editing comes later.
Lastly, always be writing. Writing is like a muscle. Getting comfortable with writing comes with practice.
Write anything. Anything that comes to your head about the topic. Include what you are thinking even if it is vague, unformed, expletive laden, and non-sensical, with questions and trapdoors and no particular direction or rhythm or end. Vomit it onto the page. Hey, look! Now you have a whole pile of text you can edit, and maybe you found something new as the thoughts travelled through your brain to your fingers on their way to the page.
A bit like stream of consiousness, but more structured (see Thinking Mathematically, Mason et al.)
Write down:
- All significant ideas/questions that occur
- What you are trying to do
- Your feelings about it
- I do not understand...
- I do not know what to do about... I cannot see how to...
- I cannot see why...
All you have to do is write one true sentence. Write the truest sentence that you know.
- Ernest Hemingway, A Moveable Feast
You've got the nebulous shape of some idea, paper, writing, draft, in your head, but you don't know where to start. Follow Hemingway's advice and write something. The truest thing you know about that thing. Then try to write another. And another.
Related to the 'Hemingway' strategy. Write a single sentence that fully summarises the thing you want to say. Almost like a thesis statement: this is what this is or what this does. This is the seed. Get the seed right, then water it: turn that sentence into two or three (e.g., what follows from this, or what needs to be known to understand this). From there grow it to a paragraph. From a paragraph to a set of paragraphs etc. The idea is to be building up iteratively. Maybe the seed gets destroyed in the process (you aren't necessarily growing an onion layer by layer), and that's okay.
Start with the results and work backwards.
Start with the top-level motivation for the thing. The bird's eye view. Why should the reader care? What is the problem? What is the background that is important here?
Go anywhere and start writing about that topic. A good place to jump into the middle might be with something relatively easy to write, like a technical definition where you don't have to think too hard about creatively explaining it. Put the definition. Then maybe an example. Any remarks that follows from this?
Zoom out and then work out, what is the structure I want for this book/paper/section? How does it break down into sub-parts and how are they connected. Put that structure down on paper and write some text to explain why it follows in that way. Are there any cycles that need unpicking or explaining? This gives you an outline for the overall piece of writing.
Somewhat related to both "Growing a seed" and "Structure" is the breadth-first approach. This strategy can be helpful if time is short heading towards a deadline. This generally works will with having a high-level outline / structure. Then going through this structure in a breadth-first fashion and expanding each section with a bit more information, and iterating. This avoids doing a "deep dive" on one section and then running out of time to explain anything else in important.
Dominic Orchard