Academic writing doesn’t have to be hard to read
NB: This document is a work in progress.
Nobody sets out to write a sentence that others cannot read. Yet academic papers are full of them, and their writers rarely notice. A writer never reads their own sentences cold: they already know what each one means. The reader does not, and has to rebuild the thought from the words on the page.
Students pick up the habits honestly. They read papers written this way, mistake difficulty for the voice of scholarship, and copy it. The habits that make prose hard to read are specific, though, and so are the fixes. Here are seven.
Omit needless words
Due to the fact that is because. In order to is to. It is important to note that X is X. Delete the phrase and read the sentence again: if nothing is lost, nothing was there.
Apply the test upwards. A sentence that restates the one before it, a paragraph whose conclusion you drew two pages ago, a section that does not advance the argument: cut them rather than polish them.
Start where the reader is
Every sentence has two strong positions. The opening is read as what this is about, the close as the point. Old information belongs at the opening — usually whatever the last sentence ended on — and new information at the close, where it becomes the opening of the next.
Broken: Children’s education resembles their parents’ education. People tend to choose partners with a similar level of education. Advantage and disadvantage become concentrated in the same families.
Repaired: Children’s education resembles their parents’ education. Those parents also resemble each other, because people tend to choose partners with a similar level of education. That similarity between partners concentrates advantage and disadvantage in the same families.
The facts are the same. In the second version each sentence begins on the term the last one ended with, so the reader is never left holding a sentence in suspense, waiting to find out what it has to do with anything.
Name the actor
In the active voice the subject does the thing — the reviewer rejected the paper — and in the passive it has the thing done to it: the paper was rejected. The passive hides who did what. Mistakes were made names nobody. The sample was collected leaves the reader asking by whom. Write we collected the sample.
The passive has one defence: it lets a sentence open on old information when the actor is new. If the previous sentence ended on a questionnaire, the questionnaire was developed by a team of clinicians keeps the reader where they already are. That is the only reason to use it. Chosen for variety, or from a sense that impersonality sounds scientific, the passive voice costs clarity and readability.
Use verbs, not nouns
A nominalisation is a verb turned into a noun: analysis from analyse, measurement from measure, introduction from introduce.
A sentence is easiest to read when the subject is the actor and the verb is the action. Nominalisation moves the action into the subject, and leaves a weak verb behind — conduct, perform, carry out. Nominalisations are a key reason why academic writing is so hard to read.
So we performed an analysis of the data becomes we analysed the data. There was a reduction in the prevalence of depression following the introduction of the programme becomes depression became less common after we introduced the programme. Avoiding nominalisations also reduces the number of words, which is always a good thing.
Repeat yourself deliberately
Call the same thing a sample in one paragraph, a cohort in the next and a group in the third, and the reader must decide whether you mean one thing or three. Some will decide wrongly. Fix a term at first use and keep it. Repetition tells the reader that the referent has not changed.
Repeated structure does the same for comparisons. The first group received the drug; the second group received the placebo. Everything but the contrast is held constant, so attention falls on the one element that differs. Build the items of a list to a single grammatical pattern and the reader learns the shape once and spends the rest of their attention on the content. Break the pattern and the differences you did not intend compete with the one you did.
Repeat words and structure, not points. A point made twice is still a sentence to cut.
Formality is not obscurity
Being scientific means being honest, and honesty means writing so that a reader can check you. A claim that is hard to understand is hard to disagree with. First person, ordinary vocabulary, short sentences and the active voice are all compatible with serious scholarship. Use the long word when it is the exact one, and not otherwise.
Writing is rewriting
A first draft is rarely a good draft. A first draft is for working out what you think; the habits above apply to the second, the third and the fourth, and a plain sentence is usually the fourth version of itself.
The rest is practice, and practice means reading with attention. When a paragraph of someone else’s is easy to follow, stop and work out what makes it so; when you lose the thread, work out why. The diagnosis is what you keep — noticing that a paragraph works teaches you nothing until you can say why it works. Then do the same to your own drafts, which is harder, because you already know what you meant.
A full example
Broken: Due to the fact that the quality of student writing is of considerable importance, an investigation of the effects of feedback was undertaken by our research group. A total of 120 participants were recruited. Random allocation of the respondents to two conditions was then performed. Detailed comments were given to the first cohort, whereas only a grade was received by the other set of students. An improvement in the revision quality of the first group was observed, in contrast to the second group, where no such improvement was found. It should be noted that the finding that feedback improves revision is an important one.
Repaired: Feedback is meant to improve how students revise, but it is not clear what kind of feedback does. We recruited 120 students and assigned them at random to two groups. The first group received detailed comments; the second group received only a grade. When the students revised their work, only the first group improved.
The repair names the researchers (“we”), puts the action back in the verbs, calls the students one thing throughout, gives the two groups the same sentence shape, and ends on the result instead of on a claim that the result matters. None of these changes set out to make it shorter, yet the repair runs to 54 words against the original’s 104.
Further reading
- Steven Pinker, The Sense of Style: The Thinking Person’s Guide to Writing in the 21st Century (2014).
- Steven Pinker, Linguistics, Style and Writing in the 21st Century, a talk at the Royal Institution (2015).