On writing

Of the many ways of and purposes for writing something, what I do most is analysing and arguing. This is a detailed walkthrough for myself about what that process entails.

As a Master’s student and later Assistant Professor, I taught Bachelor’s students and continuing education students how to write academically. And arguably there is still a lot of academics in my professional writing in my day job.

Goal: Precise and concise language in 3 major edits, for any standalone text to carry meaning and weight to an intended audience.

Wherein a major edit is printing it out to rewrite it, reading and revising the content, language, and structure.

Contents

1. Research
2. Outline
3. Write
4. Rewrite
5. Revise
6. Finalize

Research #

Before you start writing, you will have formulated a vague idea of what you’re writing about. Whilst researching the topic, take copious notes and store them in a structured format. Zotero is excellent for this^[1], but do not underestimate the cognitive impact of reading, and writing notes, on paper^[2][3][4]. The speed at which you read digital and physical materials differ, as does your ability to comprehend it.

For systematic thinkers, a long multi-level structured list works wonders, for visual thinkers, a mind-map works better. Try out different methods, but do take notes from your research – you will need them throughout.

When the idea is formed, formulate it as a question that can be answered. You don’t know what direction the answer will go yet; that is to be discovered through writing the text.

Gathering knowledge methodically

Figure 1. A simple model of how attention is selectively narrowed

When researching there are no distinct limits to where to start and where to end; you can exhaustively gather every bit of information about a subject. The amount of available information about any broad or narrow subject may be to much to digest or make sense of, and as such this is not conducive to finding the right information.

By all means, start broadly, but as you learn more you should narrow your focus and filter out information with the goal of understanding and retaining what is pertinent to the question you are trying to answer. This initial filtering helps you decide what information to further process and utilize later on, when applied in writing.

Figure 2. An enhanced model of how attention is selectively narrowed

Information in all its forms – text, images, audio, video, data, code, etc. – and formats are inputs, and your notes about them and how you’ve structured them helps you decide what is relevant to focus on further and what is superfluous. This means deciding what MUST, SHOULD, and MAY be included for analysis. Do not excessively exclude information because of your own biases, but be reflective about what data you base your text on, and be critical about the sources you choose.

The purpose of this stepwise approach is to get to the point where only necessary information remains when we start writing.

Outline #

The purpose of the outline is to describe what you will write, precisely, in the shape of a table of contents.

outline / ˈaʊtˌlaɪn / verb (used with object), out·lined, out·lin·ing.

1. to draw the outline of, or draw in outline, as a figure or object.
2. to give an outline of; sketch the main features of: On the first day, the professor just outlined the course for us.

Synonyms: draft, delineate

Express your idea of the text as an ordered list, starting with the main content. Point by point, use keywords or short sentences to describe what you will be examining, using what data, theories, and methods, and how you expect this will answer your question. This may sound rather academic, but it is the basis for how to argue rationally and compellingly. Waste no time on introductions, contextual explanations, or expected results, yet. In essence:

1. What will be examined
2. How it will be examined
3. Discussion of findings
4. Analysis of results
   • Argument A
   • Counter A
   • Argument B
   • Counter B
   • Argument C
   • Counter C
5. Discussion

Keep this idealized list at hand, and do not change it until you are ready to revise.

Outlines and reverse outlines

The outline is not static, but it is the source of truth for your plan for what your text should include. That is, the ideal of what it should contain when it is done. It is therefore a structured and exhaustive list of arguments, facts, information and points to be written out.

A reversed outline is a list of what the text actually includes, and will be compared to the ideal. When the lists match, the actual text corresponds to the ideal of the text, and we repeat this exercise with every substantial change to content or structure.

Figure 3. Outline and reverse outline

Write #

At this point you’re ready to start writing. There are several ways of doing this, but you should start by writing unhindered, at speed, without attention to detail. That means any distractions on your computer, from your phone, or elsewhere, are eliminated and whatever word-editor you are using is in fullscreen-mode. The right frame of mind^[5] and a very basic tool is all that is needed.

With your outline you have a todo-list, wherein each point is something to be written out to its full extent as one or more paragraphs. You have the idea of what you want to convey already formed, the exercise now is to remove all barriers between your idea and what ends up in the first draft. Continue writing until you feel that every point of your initial outline is covered.

Arguing from causality

A powerful way of building argumentation is from cause and effect^[6], wherein we show how things interact to produce an outcome. This can be abstracted to just about any method of examination, but importantly illustrates the explanatory power of thinking about how a cause leads to an outcome – in a flow chart.

Figure 4. Simple, elaborated, and complex causality graphs

Taken from Gerring (2012), figures 8.1, 9.1, 11.3^[7]. See Gerring (2005)^[8] for a shorter introduction to causation.

For something to have happened, that is, lead to an outcome or resulted in a consequence, there must be a cause of the incident. That a stone breaks a window is an example of an outcome, but the mechanism that lead to the broken window is that the stone hit it. The cause is then that the rock was thrown. Further, the mechanism is that someone threw the rock at the window, which can be expanded by explaining the motive of throwing the rock.

There might also be independent, covariate factors that lead to the outcome, such as the glass in the window already having cracks or being fragile. Finally, there may be confounding factors that affect both the cause and the outcome, but by an indeterminate amount. For example, the rock may be thrown in the general direction of a garden where it hit a shed that had a fragile window in it. But the ultimate determinate cause of breaking the window is unsure.

The purpose is not to somehow cover and estimate every possible factor under the sun that probably did lead to the outcome, but to focus in on what caused the outcome, through what mechanism.

Use subheadings to structure the text and make the todo-list manifest. Each paragraph beneath a subheading that corresponds to your outline should be a self-contained unit of thought – understandable outside of the context of your text. When and if you need to refer to a source, use the simple form Author (Year: Page) if your word-editor does not have built-in or third-party support for reference-management. Quote fastidiously, but quickly.

Rewrite #

Your second draft should in large part be done physically, ideally with pen and paper. The digital, inferior alternative is saving the first draft as a read-only PDF and annotating it with comments. The purpose of the first rewrite is to read it at normal speed, and get a feel for how it reads and adjust it accordingly. You are not editing in detail, but looking for whether the totality of the text makes sense, both its structure and arguments.

This means reading the full draft once from start to end without interruption, editing as you read through it. If it is sufficiently long – more than 5 printed pages – you would do well to write a reversed outline. Then, in a versioned document, rewrite it entirely to include your edits. That is your second draft.

Versioning

As per our stated goal, the text should be done in 3 or less major edits. For every rewrite, the updated outline is a major-version change, adapting a Semantic Versioning-scheme:

Given a version number MAJOR.MINOR.PATCH, increment the:

• MAJOR version when you make substantial or incompatible changes
• MINOR version when you add content in a backward-compatible manner
• PATCH version when you make backward-compatible fixes

We want to be able to compare the reality and ideal of the text at multiple stages, to track what was added, left out, and where the direction changed between rewrites. This ensures that we end up where we intended to.

Figure 5. Versioned outlines

Writing is an iterative process, whereby the first draft of anything is rarely to never as good as you think it is. Repeatedly rewriting the language will correct it, but won’t necessarily make it more coherent. The gained value from each repetition declines, and you end up re-correcting something you already corrected, or changing your opinion. By practice you will find the middle ground between an excessive and deficient amount of rewrites – the golden mean.

Revise #

For the third draft you will perform another rewrite, where you read and revise the text with a focus on how close the content is to your idea of it. Compare it to your outline and make a note of which arguments are not the clearest, most succinct version they can be. This is also a question of your objectivity; no arguments should be based on your personal opinion, and always be founded in relevant and solid information. Your assessments, evaluations, and conclusions should be professional.

In rewriting, be sure to look at both sides of the arguments, and include their fallacies. At this point, you should have convinced yourself of what is the seemingly right answer to your initial question. Write it out as the conclusion to your text, and include a short summary of uncertainties – again as a versioned document. That is your third draft.

Finalize #

Finally, your text should only need a once or twice going over – depending on how important its level of polish is – to correct minor details. You will have caught the most important changes in your rewrites. Read through the text and correct any mispellings and incorrect grammar, but also keep an eye on the overall structure. If it makes more sense to place one part of the text elsewhere, or rearrange your arguments, do it without hesitation.

Reading carefully and rewriting thoughtfully takes patience, and you’ll notice yourself skipping over sentences and words because you know them well. Though it takes a lot of attention and imposes heavy cognitive load, the quality of the writing improves greatly when done slowly and thoroughly.

References #