Over the years I’ve learned a lot from watching how other people write. Here are some things I’ve picked up that you might find useful.
Thing 1: Use a checklist
Even though scientists are great at doing research, especially at the beginning of their career they don’t always meet the basic standard for scientific writing. This is why a few organisations have published writing checklists. For example, the CONSORT statement describes how to report clinical trials. Why not make things easier for yourself and use those checklists?
Thing 2: Be consistent and structured
You’re not writing a novel, so use the same term for the same thing everywhere. If not, readers will think you mean something different.
This idea extends to structure and signposting through your manuscript. If you have two aims, then make sure the same two aims appear using the same language in the methods, and then in the results. Make it clear which results link with which methods, and which aim they are addressing. If you have results that don’t address the study aims, then those results are in the wrong paper or maybe you should reconsider your aims.
Thing 3: Don’t write the way you speak
These common phrases are perfect examples of that – as are many others:
The purpose of this section is to describe how we XYZ { Just use XYZ as a heading and start telling. }
From the literature we see that XYZ. { Just write: XYZ }.
The results are summarised in table 1. Blah blah… { Just write: Blah blah (table 1). }
You get the idea. Most manuscripts could be shortened by at least 10% without any change in meaning. In fact, they generally read better; a statement has greater authority when it isn’t sugar-coated and wrapped in these comfort phrases. Which is more convincing?
From the literature, we see that all elephants are grey [refs] OR All elephants are grey [refs]
Thing 4: Proper use of tenses
Use the present tense for stuff that is established wisdom (“By tradition, cake is baked in an oven…”)
Use the past tense for stuff you did (“In the experiment, the cake was baked in an oven…”).
In a proposal, use the future tense for stuff you will do (“The cake will be baked in an oven…”).
Thing 5: Proper use of the word ‘data’
The origin of ‘data’ is the plural of the Latin ‘datum’, but this point of view is a bit dogmatic. We (mostly) don’t speak Latin in Newcastle University, and even the Oxford English Dictionary accepts that data can be singular in common use.
Personally I’d go with plural data. But of course, you can write in 1st person and avoid the problem: “We analysed the data…“
Thing 6: UK or US spelling
Some words change their spelling when you move across the Atlantic. A “tumour” becomes a “tumor”, “aluminium” becomes “aluminum”, and verbs end in “ize” rather than “ise”.
Being English, and living in England, I write in British English (BE). Reading a nicely-written article in BE conveys a sense of quality that is hard to describe; maybe the written equivalent of a newsreader voice.
Of course, the reverse is probably also true. If you’re writing an article for an American scientific audience, you should consider writing in American English (AE). It might feel like a small thing, but can be a big deal depending on who the journal’s editor is.
Whoever you are writing for, it’s best not to mix BE and AE spelling. This simple advice can prove harder in practice than you might expect. Microsoft products gravitate towards AE spelling and it is easy to be auto-incorrected. Even when your default dictionary is British English, documents created by somebody else will likely over-ride your settings and use the AE dictionary.
Thing 7: Writing style – first person active OR third person passive
We’re getting into personal opinion here, so treat this section with caution.
Tradition has it that scientific literature is written in the third person, passive voice: the cake was baked has a ring of impartial, objective science about it. No human involvement. Presumably if there were real people involved, the cake would be over-egged or burned or something, which cannot be allowed to happen in science.
Personal view: avoiding I/we isn’t always beneficial in terms of style. I’ve never heard an argument that wasn’t a variation on ‘it’s just what you do’.
Here is my case for writing in the first person; but do let me know if you have a reasoned counter-argument and I’ll add it to this page.
Exhibit 1: 3rd person passive is usually unclear about who did what
“It was hypothesised …” is ambiguous as to who did the hypothesising.
So it’s clearer to use first person: “We hypothesised ….” { 2 words: describes the facts of the matter concisely and scientifically }
But ‘we’ is breaking the rules, so we need a different subject for our sentence: “These studies tested the hypothesis ….” { 5 words, suggested by an internet authority to address this specific example. }
Why don’t I like this sentence? It’s twice as long as it needed to be. And now ‘the studies’ are doing all the work, whereas the experimenters (you!) are nowhere to be seen. I want to know specifically WHO did the hypothesising. Avoiding we/I forces you to construct sentences like this, where inanimate objects take the place of the people who actually did the work.
Summary: You are writing up a large body of work that you and your colleagues made happen. How does it make sense that we or I cannot play any part in the action?
Exhibit 2: some verbs only work with a human subject
Very often, you will read “The study hypothesises…” or “The data suggest…“, which are nice work-arounds to avoid we or I.
I quite like “The data suggest…”, but data aren’t human: in my mind data can not speculate, or resign themselves to failure, or react violently to another author’s uninformed opinion in the way I or we can. Avoiding ‘I’ and ‘we’ restricts the verbs and other language you can use.
You could advocate very occasionally using the word ‘one’. “One can speculate as to how Kipling et al arrived at their conclusion…” sends a certain understated British passive-aggressive message that is hard to capture using any other turn of phrase.
Summary: Human-like opinions and activities need a human actor.
Exhibit 3: the past tense of most English verbs serves also as an adjective
Baked is a past tense, and also an adjective. “The cake was baked” might be a description of the cake (like: “the cake was raw”), rather than something that happened to the cake.
To clear the ambiguity, “The cake was baked” has to be rewritten as “We baked the cake” or “I baked the cake”. First person, active voice. And now I know WHO did the baking.
Summary: The 3rd person passive voice is ambiguous.
On the active voice
All these examples work better in the active voice. To write in the active voice, avoid where possible using the verb ‘to be’ . . . is, are, was, were. There is an entire variation on the English language where to be is disallowed. Not always possible and/or often results in poor English (read the criticisms on the Wiki page), but as a guideline it’s helpful.
I’ve probably thought a bit too much about this. No doubt others will come back with good counter-arguments, so have a look at this article in defence of my personal opinion (from the 4th paragraph down). This is a good reference in its own right.
Right, rant over. I await the storm of protest – so long as the protests are based on carefully constructed linguistic arguments. In the meantime, you might speak to your supervisor if you think they have a strong preference for style.