Asia & Pacific

  • How do I sub-edit a science article?

    Peter Wrobel

Sub-editing an article means making it readable, accurate and attractive — not putting words in the author's mouth — explains Peter Wrobel.
Scientists and science writers are always keen to communicate new ideas and developments. But they are not always the best judges of what their intended audience can absorb.

The problem is common throughout journalism, where specialist writers tend to assume that others share their knowledge. But it can be particularly acute in science journalism for two reasons: scientific specialisms often have an extremely narrow focus, and most specialist writing is about things that are, by definition, new, and therefore unfamiliar to many readers.

In the course of commissioning such an article, an editor normally learns a great deal about the story subject. Indeed, their understanding of it may rival that of the intended author, and will probably be more comprehensive than almost anyone's, apart from those involved in the research itself.

All this knowledge, however, already sets the commissioning editor apart from the intended audience, whose first acquaintance with the topic will be when they read the article. This is where the sub-editor comes in.

The sub-editor can be seen as the reader's last line of defence. A sub-editor's main concern is what the reader is likely to know before seeing the article, and whether they will be able to understand the article in the light of such knowledge. And the sub-editor's main task is to ensure that the article is comprehensible to as many readers as possible.

But comprehension is not the only challenge. During the process of making an article understandable, the sub-editor must also try to ensure that it is readable, legally sound, ethical and written in a consistent style. She or he must also make sure that it fits the space assigned to it; that it has an enticing headline and good captions on pictures and other illustrations; and that it is accurate — right down to the spelling of the author’s name.
Keeping the reader's attention

While it is only one of the sub-editor's tasks, ensuring that the article can be easily understood is perhaps the most important. Sub-editors often find it useful to start by asking how much people read of any given article.

It is often assumed that readers read everything. But this is far from the case. Sub-editors have to work on the general assumption that 80 per cent of readers will not get past the fourth paragraph of a printed article, and 90 per cent will not touch the scroll button when reading an article online.

It is easy to understand why this happens. In a newspaper, for example, every double-page spread is littered with headlines all designed to grab the reader's attention — rather like a street market, with different traders shouting out their wares. As their eye roves over the page, giving it any excuse to hesitate — a word they don't understand, an obvious mistake in spelling, grammar or fact — will break their concentration. There are plenty of other stories vying for their attention, and once their eye wanders off, it is unlikely to return.

In that sense, the sub-editor's aim is simple: that for any given story, the reader will read it all, down to the end.

To achieve this, the sub-editor has to deal with elements that can break the reader's concentration. Here we return to the sub-editor's many tasks — with making the article comprehensible at the top of the list. If a reader cannot understand what the writer is saying, there are plenty of other articles to read, or plenty of hyperlinks to click on.

So a sub-editor will assess every word (some quicker than others) to ensure that the readers will understand it. And for any publication aimed at a broad readership — SciDev.Net, for example, or New Scientist — a sub-editor will assume that the average reader of any given article is a non-specialist. After all, how much molecular biology will even the average condensed-matter physicist understand?


So what will a sub-editor identify as being incomprehensible to a non-specialist audience? The first, and most important, rule of thumb is that if the sub-editor cannot understand it, the non-specialist audience won't understand it either.

Applying this rule demands a certain level of self-confidence, however. Science is full of hierarchies, social and otherwise, but one of the most powerful is based on knowledge: the more you know, the higher your status.

In this situation, sub-editors have to be willing to assert their ignorance. For a sub-editor, saying, "I don't understand this" is not an admission of failure or inadequacy; it is a vital first step in turning the article into something that is easily understood.

At this point, writers — and even editors — can become a little frustrated, and may exclaim, "But everyone knows xxx!" The sub-editor has to remind the author that "everyone" can't know xxx, since they don't. Or that what everyone has learnt in the past is not important — it's what everyone remembers that counts.

Along with this requisite self-admitted ignorance, sub-editors also need to keep alert to changes in 'common knowledge'. Ten years ago, for example, the word 'genome' did not fall into this category, even within that section of the public closely interested in science. Today it is understood, at least up to a point — and normally that point is far enough for readers to understand what the article is about (although there are still circumstances in which the term may need to be explained).

Likewise, you may not need to explain what a protein is. But you would want to explain what the proteome is. It is all a matter of judgment, something a sub-editor requires in abundance.


The next key ambition of a sub-editor is to make an article readable, which is not the same thing as making it comprehensible.

Something may be clear, but dull; or clear, but difficult to take in at normal reading speed. Readers find it hard to maintain concentration if a text is dull or difficult, and they'll quickly move on to the next headline. Readability is what will keep them going to the last sentence.

Sometimes your word processor can help to check this. I am writing this article using Microsoft Word, which — along with other standard word processing programs — has a function for checking readability. All readability checkers do fundamentally the same thing: use algorithms based on word length (or number of syllables), numbers of words in sentences and, sometimes, numbers of sentences in paragraphs, and then tell you how difficult the article is to read.

The level of difficulty is often expressed as the number of years you'd need to have amassed within the American education system to understand the piece.

In my first draft, the result was 11 years, which means, roughly, that the average 17-year-old could understand it — a little high for easy reading. Newspapers tend to aim a bit lower. The reading ease so far is 53.9 — on a scale where 100 is very easy, 60 is 'plain English', 40 is 'difficult' and anything under 20 is 'very difficult'. For comparison, a paper on molecular biology in Nature will typically rate between 20 and 25!

The principle behind all readability checkers is simple: neurological perception. However familiar a long individual word might be — take 'demonstration', which has 13 letters and four syllables — once you start filling up a sentence with long words, the eye and the brain simply cannot read them easily.

This creates particular problems for sub-editors working on science stories, since science is full of long words — 'epidemiology', for example, or 'crystallography'. The solution is to go short wherever possible: short words, short sentences and short paragraphs.

This is often easy to do. Words such as 'demonstrate' can easily be shortened to 'show', or 'consume' to 'eat'. Other words can be cut out entirely — the word 'very', for example, is almost always unnecessary — while words such as 'however' can often be changed to 'but', or just omitted, and 'furthermore' can profitably be transformed into 'and'.

The next trick is to look for 'passive' sentences — for example, "It has been shown by Swedish scientists that…". Passive sentences are notoriously common in science writing, as this is the way most scientists seem to have been trained to write papers. They slow sentences down, and put a brake on readability. One-way of keeping a check on them is with the Word grammar checker, which will tell you what percentages of your sentences are passive.

Once a sub-editor has pinned passive sentences culprits down, they need to turn them around to make them active — "Swedish scientists have shown that…". This reduces the number of words from eight to five — a reduction of 37.5 per cent!


The next responsibility of the sub-editor is to ensure that everything in an article is accurate. Every word must be spelt properly, all grammar and punctuation must be correct, and all facts should, in principle, be checked — within reason.

Spelling, grammar and punctuation are important because errors in them can alter the sense the writer intends to convey. In these matters, you can never, ever, rely on your computer's spellchecker or grammar checker.

A spellchecker will not tell you whether you have spelt the word correctly, merely that a word can be spelt in a particular way. So it will not discriminate between 'discreet' (prudent, modest, perceptive) and 'discrete' (a common word in science, meaning 'distinct'). Nor will it tell the difference between 'checker' and 'chequer'.

In addition, the rules of English grammar are relatively flexible, which means that they are not, yet, reliably susceptible to machine checking. So sub-editors must become masters of their language.

Another reason for getting these things right is that when they are not, they distract the reader. A sub-editor wants the reader who has read one group of words to go on to the next, not to stop and compose a letter of complaint to the editor about a grammatical slip.

Factual accuracy is even more important. Inaccurate facts can do quite a lot of damage: distract or mislead the reader, reduce the reader's trust in everything else, and damage the reputation of the publication. Serious errors may demand a correction in the next issue, which takes up time and space.

So sub-editors need to check everything they can, especially the names of institutions and people. It's the least they can do for scientists. After all, it's a tough job — you labour unnoticed in the lab, your family has no idea what you do, then one day your name appears in print and you have something you can show your mum.  So at this point, you want your name to be spelt correctly!

Numbers are a frequent source of error in writing about science. It's easy to make a slip of the keyboard and end up with the wrong order of magnitude, or a wrong numeral, which is why sub-editors check numbers for common sense.

The same goes for internal consistency. If a writer says a new grant scheme will spend $500,000 on six disciplines, and then lists only five disciplines, or gives sub-totals that add up to more than $500,000, then clearly something is wrong. At this point it is often only the writer who can say exactly which number is wrong, but the sub-editor has identified a mistake — normally without having to know anything about the subject!

There is much more to sub-editing, of course. When an article leaves the sub-editor, it must be legally sound (whole books have been written about the criteria that determine libel), its content must be ethical, it must fit, and it has to have a good headline and alluring captions.

And everything must be in a consistent 'house style' — a standard way of spelling, or writing down numbers, or giving dates or references, for example — not least to avoid distracting the reader by writing 'bloodstream' in one sentence and 'blood-stream' in the next. In short, when an article leaves the sub-editor, it must be entirely fit to print.

A last word

Finally, a word of warning to over-ambitious sub-editors. We are not the writers, nor the commissioning editors. It is not our job to decide what goes into the publication. And it is certainly not our job to put words into the writer's mouth, so they express views that are ours instead of theirs.

Even if we disagree with the writer, our job is to help them get their ideas over in the best way possible. We don't change things out of personal prejudice.

This is especially important in science writing, where many of authors are professional scientists first, and amateur writers second. But it is also true of science journalists. And if you think a writer has got a fact or a spelling wrong, and can't confirm it yourself, go back to the author for checking. Don't just make a guess.

Ultimately, we need to respect our profession — even if, ironically, the goal of our work is to end up with something that the writer can still call their own

Peter Wrobel is the managing editor of Nature

This article was previously part of SciDev.Net's e-guide to science communication and has been reformatted to become this practical guide.