Summary

This chapter looks at questions of style in reaching wider audiences. It starts with one of the most important aspects of communicating your research, using accessible language and writing effective plain language summaries. I consider some of the commonly agreed precepts of good writing style, drawing on work by recent academics to push the boundaries of what it means to write well as a researcher. I set out eight important features in good writing which go beyond grammar and basic style rules. This includes thinking about the title of your work; writing as you speak, avoiding the convolutions of academic thought; identifying main messages and expressing them clearly; being as true as you can to the science and avoiding over-claiming and spin; being playful; using tricks of persuasion; choosing words carefully; and finally, perhaps most importantly, finding your voice. More guidance and books on style are given in the Further reading section, and the chapter ends with three practical pointers to writing for impact. Finding the appropriate tone, language and idiom for different audiences and outputs is difficult and needs effort. The best way to learn is to take note of people and places which do this well and practise yourself.

Writing plain language summaries

This book is about how to communicate research findings to audiences who are not researchers. Perhaps the most important tool to do this is the plain language summary. This is likely to be the most read output and yet is often done as an afterthought. More time is needed to get this right, working with others – particularly people who know nothing about your research. This chapter looks at the language and format to reach wide audiences, starting with the plain language summary as a critical part of your pathway to impact.

We all know it is important to avoid jargon and technical words. This means replacing peripheral oedema with ankle swelling and avoiding terms like in vitro which are not in everyday use. The aim is to use simple, clear language. But it is very difficult to do, especially if your research is complex and nuanced. As Einstein may have said1 – ‘keep it as simple as possible, but no simpler’.

There is a useful paper by Rob Waller, who set up a charity – The Simplification Centre – from Reading University on making information more accessible (Waller 2011). He emphasises that writing simply is not dumbing down and provides some useful strategies for making work clearer. He busts some myths – for instance, to put text into plain language may make some reports longer, not shorter. Jargon works because it is a shorthand between professionals, but it is not understood by all. Opening up text for wider understanding is a political point. The Simplification Centre, like the pioneering work of Chrissie Maher and the Plain English Campaign, was started with a mission to open up the worlds of law, consumer rights, benefits and health information to more people.

In health-related research, good work has been done by the Cochrane collaboration in developing plain language summaries. It is difficult to make complex findings accessible, without distorting the science. An audit some time ago of plain language summaries by the Cochrane team found that almost a third were inaccurate in relation to the main findings in the full review or abstract and could be misleading (Soares-Weiser 2011). This included elements like presenting treatment effects in relative terms (‘the risks were doubled’) without giving a sense of the absolute numbers (‘from one in a hundred thousand to two in a hundred thousand’). Much work is being done on how to present information in a responsible and accurate way. This includes iterative work with consumers to test and refine Cochrane public summaries in a rigorous way (Synnot et al 2018).

Earlier studies showed that readers often did not understand summaries of research developed for public use by Cochrane review groups. For instance, a randomised trial of 143 members of the public from different countries compared different formats of plain language summaries, measuring reader understanding of benefits and harms of the intervention and research quality (Santesso et al 2015). Only one in five (18 per cent) readers understood the traditional format, increasing to 53 per cent for a newer version. While an improvement, it is not a great endorsement of a public-facing summary that only just over half of people reading it understood the main points. A similar proportion of the public understood findings in a smaller study, with just over half understanding the findings in a plain language summary, which was an improvement on the scientific abstract alone which only a third understood (Maguire and Clarke 2014). Cochrane review groups and staff have spent much time on this, with useful how-to guides on writing plain language summaries (Glenton 2017) and a recent consensus checklist on how to report clearly and accurately the effects of interventions (Oxman et al 2020). These present helpful advice, although geared more to clinical effectiveness studies and systematic reviews than the full range of study designs in health and social care research.

In short, even with much effort, it is still very difficult to write summaries which are accurate and accessible. But however imperfect, they tend to be much easier to understand than scientific abstracts and other research outputs. At their best, public-facing summaries can distil the key messages of your research in an economical way. But this takes time and multiple iterations to get right. You need to think about the plain language summary as an important output in its own right, not just remove a few hard words. Rob Waller points helpfully to the work of Widdowson (1979) who distinguishes between ‘simplified versions’ and ‘simple accounts’. Waller notes the importance of this distinction:

The former are translations of existing documents, while the latter are documents that are originally planned, written and designed to be easy to understand. Most documents intended for children are simple accounts: both content and form are created from the outset with that audience in mind. It is an important distinction, because it reflects a different dynamic of production. (Waller 2011)

That is, you need to start with a blank sheet of paper to construct a public-facing simple summary, rather than water down an abstract for a scientific journal changing a few words here and there. It is the difference between word for word translations or Google Translate and the work of a skilled interpreter, fluent in both languages.

There are no easy rules or prescription for writing summaries which are accessible and accurate. It is a craft and needs a combination of skills. What we found at our evidence centre was that it was best thought of as a process where many could contribute. That included science writers and journalists, those with critical appraisal skills (understanding the findings and strength of evidence, including possible risk of bias) and people with lived experience as staff or patients. This is an iterative process. The earlier chapter on patients and the public gave an example from our work with an advocacy group to co-produce an easy-read version of a research report with people with learning disabilities (Figure 5.5).

Tools are available to test how easy it is to read your text. Two common tools, measuring length of sentences and words, are the Gunning Fog index developed in the 1940s and the Flesch-Kincaid levels in the 1970s. There are free online services to calculate the readability of your text against these specified levels measured by educational level (in the US). The general aim would be to make papers accessible for students aged 13 to 15 years. These automated tools are helpful, but cannot give you all the answers. Some long words with many syllables are commonly used and understood. Some short words are obscure or formal. There also may be particular audiences you want to reach, like clinical staff or managers. For target readers, some jargon or technical terms can provide context and clarity. Using readability tools can be mechanistic and will not tell you if a summary is clear about why a study is important and what it found. Rules used by algorithms, such as use of adverbs or intensifiers being a marker of poor style, do not always hold good.

Skimming through the NIHR research portfolio, I selected two examples which I think are good plain language summaries, although their reading score age profiles are in the mid to high teens (Box 9.1). But they do the job for me in conveying complex research findings as simply as they can and explaining why the research is important. They also don’t go further than the study design allows – for instance, the second example in Box 9.1 is a feasibility study and so findings are suitably cautious. The best rule when you write a simple summary is whether a friend or neighbour can easily understand the research and relay back to you what it is all about and why it matters. Tools can’t do this for you.

Plain language summaries – two examples

‘Telephone first’ general practices (Newbould et al 2019)

Every one of us wants to be able to visit our family doctor when needed. With so many more patients, this is becoming increasingly difficult. The study looked at a new way to ask for help from one’s doctor to see if this would save overall time and NHS money. Patients were asked to speak first to a GP or doctor by telephone to see if their problem could be dealt with over the telephone or if they needed to see the doctor in person. Practices using the ‘telephone first’ approach were compared with other practices that were not using it.

In a patient survey, it was shown that just over half of patients found it easier to make appointments with the ‘telephone first’ approach than with the previous system, with nearly one quarter finding it less easy or the same. It was quicker to make an appointment in those practices using this new way, but, when asked, patients and practice staff had strong views both for and against the new system. Factors affecting patient satisfaction included the ease of getting through to the general practice on the telephone and how easy it was to wait for the callback from the GP (for example if the patient was at work).

‘Telephone first’ greatly increased the number of doctor consultations by telephone, with around half of problems managed in this way. This led to more work for practice staff, although there were big differences between practices, with some having much more work and others having much less. There was not much difference in the use of hospital services or in the costs of hospital care.

Although the study showed that, by using the ‘telephone first’ approach, many health problems can be dealt with over the telephone, this will not solve the increasing need for care by our family doctors.

Trying new approaches to sex education in schools (Mitchell et al 2020)

Young people report higher levels of unsafe sex and have higher rates of sexually transmitted infections than any other age group. Good sex education is important for later sexual health, but it needs to be done well. We tested an approach to sex education (called the Sexually Transmitted infections And Sexual Health (STASH) intervention) in which influential students – chosen by their year group – were trained to start conversations with other students about sexual health on social media and face to face. This approach has previously worked well to prevent young people taking up smoking. Working with students, teachers, health professionals and youth workers, we adapted the approach for sexual health and older students (aged 14–16 years).

We also developed a website of digital resources (memes, infographics, web links, and so on) that could be shared via social media. We tested the approach in one school, made adjustments and then delivered it in six schools in Scotland. We wanted to find out if it was practical to deliver and whether or not those taking part would like it. We observed some of the project activities, kept careful track of participation, interviewed students and teachers, and asked peer supporters to complete a brief web survey. We also asked the whole year group to fill in a questionnaire about their sexual attitudes and behaviour, and about taking part in the STASH study. We compared their answers with those of students in the year above who had completed the questionnaire the previous year.

About half of the students who were chosen as ‘most influential’ by their friends decided to become peer supporters. Once trained, nearly all of those chosen completed the role and many of them were active on social media and in conversations. Students and teachers generally liked the project. Using social media (closely monitored by adult trainers) was helpful and did not cause problems. Our findings suggest that it would be worth doing a larger study to find out if the STASH intervention can increase the number of young people staying safe from sexually transmitted infections (either by always using condoms or by not having sex).

What do we mean by writing well?

There is reasonable consensus on what we mean by good writing. Numerous style guides (see Further reading for some key texts) point to the importance of the following features:

  • keep it clear and simple – avoid technical jargon and keep writing as concise as possible, using plain English terms;

  • use active verbs – avoid passive forms (‘it has been demonstrated that’) and nominalisation, where verbs are turned into nouns in a formal style (‘articulation’, ‘marginalisation’);

  • shorten sentences – keep sentences short or mix longer with shorter sentences;

  • be precise – use examples and avoid generalities.

But in an arresting book by the academic Helen Sword, she notes how rarely this advice is followed in scholarly outputs (Sword 2012). Taking a thousand academic articles from different disciplines, she audited them against the checklist of good writing. Few of them satisfied these base requirements. She found that many were written in a convoluted, indirect, formal style with long sentences and abstractions. Results from this small sample confounded some preconceptions about different fields of study. Some of the humanities performed less well than fields like computer science in terms of their use of active verbs or personal pronouns.

To the general advice on good use of language, Helen Sword added other features from her scan of a hundred authors nominated by 70 colleagues as good and stylish academic writers. These additional elements included:

  • eye-catching titles;

  • arresting opening, with an engaging story or challenging question;

  • first-person anecdotes or asides that give a sense of the author;

  • numerous examples to explain abstract terms or theory;

  • visual illustrations beyond standard graphs;

  • references to broad and wide-ranging sources beyond the narrow field of enquiry;

  • humour, whether overt or implicit.

She found evidence of these traits in academic articles, as well as more public-facing books by scholars. One of her conclusions is that academics constrain themselves into a formal style at odds with the precepts of good writing. They internalise ‘rules’ which they think are required in order to get published. Helen Sword’s work shows that there is space for invention and creativity even within the constraints of academic publishing. But increasingly, researchers are looking to publish not only in academic journals. They also want to reach wider audiences beyond the library and common room. For these audiences, used to engaging and sophisticated offerings for entertainment and information, it is important that you tell your story well.

Some pointers are given in this chapter to improve the way you present your research, with some summary tips at the end. It is also important to tailor outputs to particular audiences. Issues of format and language appropriate to frontline clinical staff or community groups or government advisers were covered in earlier chapters (Chapters 4–6).

Writing for impact

Writing for impact – key features

  • start well

  • be natural – or write as you speak

  • get to the point

  • stay true to the science

  • be playful

  • be persuasive

  • choose words carefully

  • be yourself – the importance of voice

In what follows I identify eight important features when writing for impact, with some background, examples and a few practical pointers at the end so that you can develop a style which works for you.

Start well

A good title can both inform and engage the reader. While some scientific journals have minimum requirements – information about study design for indexing and transparency, there is still freedom to inject a bit of humour, surprise or liveliness. I mentioned my search for good titles on Twitter, and a number of researchers said they had had their suggestions for more playful titles rejected by journal editors, who amended them to more factual titles. This is not always true of publishers and editors (as Box 9.3 shows), so it is worth persisting in the hope of greater audience interest and reach.

I carried out my own exercise, scanning one year of six health services research journals to identify the titles which appealed to me – a subjective exercise – plus a few additional ones suggested by friends and contacts on Twitter (Box 9.3). It strikes me that qualitative research has an advantage, with many using well-selected informant quotes and data extracts to add colour to the header. There is of course a fine line between the eye-catching and the irritating, particularly when certain tropes get over-used (for instance, ‘if X is the answer, what is the question?’) A few puns can go a long way. Helen Sword (2012) has some good tips for researchers in her chapter on titles, including avoiding over-use of the colon in lengthy titles. This is particularly true when both parts of the title, separated by a colon, are purely descriptive, rather than an engaging: informative structure.

Titles which pull readers in

These were recent titles from health services research journals which made me curious and want to read more.

‘Partnership or insanity: why do health partnerships do the same thing over and over again and expect a different result?’

‘How wide is the Goldilocks Zone in your health system?’

‘“This is our liver patient …”: use of narratives during resident and nurse handoff conversations’

‘The view from nowhere? How thinktanks work to shape health policy’

‘Systematic reviews of economic evaluations: utility or futility?’

‘Health system goals: life, death … and football’

‘Variations in healthcare: the good, the bad and the inexplicable’

‘Psychotherapy Research Evidence and Reimbursement Decisions: Bambi Meets Godzilla’

Be natural – or write as you speak

Peter Elbow, a professor of English at the University of Massachusetts Amherst, wrote a powerful piece on what he calls ‘academic disability’ (Elbow 2013). It is worth quoting a passage in full, as he describes so well the way in which academic writing becomes convoluted and dense. He states:

When we academics were in graduate school, we were trained to write badly (no one put it this way of course) because every time we wrote X, our teacher always commented, ‘But have you considered Y? Don’t you see that Y completely contradicts what you write here.’ ‘Have you considered’ is the favourite knee-jerk response of academics to any idea. As a result, we learn as students to clog up our writing with added clauses and phrases to keep them from being attacked … As a result of all this training we come to internalize these written voices so that they speak to us continually from inside our own heads. So even when we talk and start to say ‘X,’ we interrupt ourselves to say ‘Y,’ but then turn around and say ‘Nevertheless X in certain respects, yet nevertheless Y in other respects.’ We end up with our minds tied in knots.

He goes on to defend this as a marker of what he calls:

a valuable habit of mind. It’s the habit of always hearing and considering a different idea or conflicting view while engaged in saying anything. Too many things seem to go on at once in our minds; we live with constant interruptions and mental invasions as we speak. We are trained as academics to look for exceptions, never to accept one idea or point of view or formulation without looking for contradictions or counter examples or opposing ideas. Yet this habit gets so internalized that we often don’t quite realize we are doing it; we just ‘talk normally’ – but this normal is fractured discourse to listeners.

What he argues is that we need to un-learn some of these good habits in order to communicate clearly and well to non-academic audiences. It is not that academics can’t write and think clearly and effectively. It is that scholars are hard-wired to qualify, justify and pose counter-arguments which leads to convoluted and stodgy prose.

His prescription is simple and effective. Read your work out loud. It forces shorter, clearer sentences and natural rhythms of speech. If it is difficult to say, chances are the sentences are too long or the syntax is too complex. Often long sentences, which become paragraphs, have a lot of words between the subject and verb. A good rule of thumb, which happens naturally when you speak, is to keep the main verb close to the object near the start of a sentence. When you don’t, it becomes difficult to say or understand. ‘Clinical staff, already enacting many of these paradoxes in their workarounds without knowingly breaching the normative rules of engagement, or at least without articulation of these transgressions in habituated practice, do not always conform to expected patterns of behaviour.’ Reading this clumsy sentence (which I made up) out loud would make you stop to take breath and try to make sense of the meaning by clearer, shorter statements. Reading out loud is a great rule for all kinds of writing, but particularly when you are trying to reach general or wider audiences.

Get to the point

Tim Radford, a former science editor for The Guardian, had useful advice for journalists writing feature articles. ‘A story will only ever say one big thing. Summarise it in one sentence’ (Radford 2011). That is a good precept for researchers preparing a version of their findings. More attention should be spent on headline findings than on anything else. Crafting a single sentence which gives the main findings while staying true to the science is difficult. But that is what the reader will take away.

In our evidence centre, we spent a lot of time drafting and redrafting the titles of our research summaries (Box 9.3). We wanted them to be bold and declarative, without over-simplifying. We were particularly aware of the risks of over-claiming for single studies and giving weight to the quality of evidence, risk of bias and level of certainty. But we wanted clear headline findings. Note in the examples in Box 9.4, the conclusions in scientific papers quite rightly need to provide accurate and detailed findings and explain the ‘workings’ to support conclusions. Nevertheless, it is sometimes difficult for readers to understand at a glance the findings of a study.

Summary headlines – from academic to journalistic

Headline: ‘Adding a third antiplatelet drug after a stroke doesn’t reduce the risk of another stroke’ (NIHR 2018)2

Original conclusions: We aimed to compare the safety and efficacy of intensive antiplatelet therapy (combined aspirin, clopidogrel and dipyridamole) with that of guideline-based antiplatelet therapy … In this cohort of patients with acute, non-cardioembolic ischaemic stroke or TIA, a regimen of intensive antiplatelet therapy did not reduce stroke recurrence or its severity when compared with guideline antiplatelet therapy with either clopidogrel alone or combined aspirin and dipyridamole. (Bath et al 2017)

Headline: ‘Group-based interventions may help teenagers stop smoking’ (NIHR 2018)3

Original conclusions: Forty-one trials involving more than 13,000 young people met our inclusion criteria (26 individually randomised controlled trials and 15 cluster-randomised trials). Interventions were varied, with the majority adopting forms of individual or group counselling, with or without additional self-help materials to form complex interventions. Eight studies used primarily computer or messaging interventions, and four small studies used pharmacological interventions (nicotine patch or gum, or bupropion). There was evidence of an intervention effect for group counselling (nine studies, risk ratio (RR) 1.35, 95 per cent confidence interval (CI) 1.03 to 1.77), but not for individual counselling (seven studies, RR 1.07, 95 per cent CI 0.83 to 1.39), mixed delivery methods (eight studies, RR 1.26, 95 per cent CI 0.95 to 1.66) or the computer or messaging interventions (pooled RRs between 0.79 and 1.18, nine studies in total). (Fanshawe et al 2017)

Up until the end of 2017, the limit for a tweet was 140 characters. It was then doubled, but when we were creating headlines for our evidence summaries, we kept to the original limit as shorter is (almost always) better. Having a title or headline finding which can easily be tweeted is useful in promoting through social media channels. Limiting the number of words is a good discipline, which is worth adopting.

Stay true to the science

There is a fine line between simple, compelling accounts of your work and exaggerating the impact of a single study. Paying attention to language, means avoiding spin. Good work has been done by Petroc Sumner and colleagues at Cardiff University in exposing and understanding the nature of spin in science. Their work at InSciOut (sites.cardiff.ac.uk/insciout/) has shown how university press releases and press coverage can misrepresent findings. This does not always come from sloppy reporting by journalists – their study in 2016 showed that much of the exaggeration found in health and science stories came from university press releases themselves, perhaps reflecting recent drive to maximise impact in a competitive world (Sumner et al 2016).

Another study looked at how findings can be misrepresented. Analysis showed how a third of press releases and four out of five news stories contained causal claims when the research papers described correlation (Sumner et al 2014). This kind of distortion can be seen in just one recent example (Box 9.4).4

Research example – green spaces

One thing leads to another

Let’s just take one recent example of distorted accounts of research findings in the general media. An observational study published in Nature in 2019 used self-reported data from over 20,000 adults surveyed in the UK. This showed a relationship between exposure to green spaces and wellbeing. The authors were interested in looking at the ‘dose-response’ and made cautious assertions that there appeared to be a threshold of benefits at 120 minutes’ exposure a week, peaking at about 200–300 minutes after which little additional benefit was seen. They emphasised that ‘the tentative “threshold” and peak [are] discussed here more as a starting point for discussion and further investigation, than clearly established findings’ (White et al 2019).

Perhaps predictably, this was widely covered in popular and general media as well as the scientific press. In a sea of doom, the good news that spending just a couple of hours outdoors in nature could be good for you was compelling. Two errors appeared in much of the coverage. The first was the confusion of causality and association – this study was only looking at the relationship between various factors and wellbeing. For instance, one online news headline stated confidently: ‘Spending 2 hours in nature each week can make you happier and health, study says’ (Gravier 2019).

The second problem was the categorical nature of the two-hour threshold and how this was reported in some parts of the media. The New York Times for instance had a bold header, ‘How Much Nature Is Enough? 120 Minutes a Week, Doctors Say’ (Sheikh 2019). The article talked of an ideal amount of time, suggesting even that two hours might be a ceiling (‘enough’) rather than a tentative threshold level, as the authors stated.

But it is not all bad news. Researchers from InSciOut used their rigorous methods to show, encouragingly, that science reported responsibly and well can still garner headlines and column inches (or the equivalent in social media terms). Rachel Adams and colleagues carried out a trial in which they compared regular press outputs with ‘evidence-based’ press releases where headlines were aligned with supporting evidence and conclusions were caveated, particularly being more cautious if findings were around associations not causal links. They found that the press releases which were true to the science had just as much press coverage as those which were less accurate (Adams et al 2019).

Be playful

A book came out ten years ago which was – sort of – a novel, but formed only of questions. There were no characters, no plot, just a series of questions. ‘Are you a circusgoer? Do you like to lick stamps? … Do children smell good? … Between an automobile mechanic and a psychologist, which is worth more to you per hour? … If it might be fairly said that you have hopes and fears, would you say you had more hopes than fears or more fears than hopes?’ The effect was mesmerising, juxtaposing the banal with the profound in strange and often moving ways. It was startlingly original and made me think again about what makes for a connection between writer and reader. It is worth checking out in full – The Interrogative Mood by Padgett Powell (2010).

A good test of the ideas and findings and argument running through your research is to try out different ways of presenting them. Not perhaps through creative dance or shadowplay. But it may be that you can unlock some of the threads of your research through a series of pen-portraits, vignettes, images or questions.

Be persuasive

There is still a lot to learn from the ancient masters. A recent study of rhetoric (Leith 2012) uncovered how politicians and great speechmakers use – consciously or unconsciously – the lessons of persuasion identified from Aristotle onwards. That includes combining the three branches of oratory: ethos (establishing the credibility of the speaker); logos (setting out a powerful line of argument, with reason and logic); and pathos (appealing to the emotions). A researcher telling the story of their study will need at different times to invoke all three, although on the surface perhaps only addressing the logos or main findings. To connect with the audience or reader, it helps to remind them of who you are and why you have authority to speak – as one of them or as someone with experience and expertise to share. A clinical researcher addressing healthcare audiences may start with an anecdote from their own practice or talk about the hours spent shadowing staff on the wards on this study. Injecting some emotion may mean invoking the particular – the sights, smells, sounds of those wards for you as a researcher or that of an inpatient with dementia ‘resisting’ active medical care.

For researchers, the art of persuasion is not just understanding what drives people or asserting an argument. It is also placing your work in the wider evidence base. A useful book by Graff and Birkenstein states that ‘writing well means entering a conversation summarising others (“they say”) to set up one’s own argument (“I say”)’. The authors use templates or hooks to help researchers to structure their line of debate – ‘while some argue that, I think this …’. This is a useful structuring device to make your case persuasively while recognising the wider tradition in which your research study sits (Graff and Birkenstein 2010).

An interesting tip from Larry McEnerney5 on creating text which is persuasive is to build in a sense of instability and tension in the account of your research. Words like ‘although’, ‘despite’, ‘but’, ‘inconsistent’, ‘anomaly’ provide a sense of challenge and argument in the text. As with the formulation by Graff and Birkenstein, this builds on the community and body of knowledge which went before but then injects some doubt, enriching the problem and finding a solution. This creates an onward momentum which keeps the reader engaged.

Choose words carefully

A checklist of dissemination activities for Cochrane reviews includes useful advice on language use, suggesting that terms such as ‘intervention’ and ‘outcomes’ should be avoided in plain language summaries, in favour of specific terms like ‘healthy living programme’ and ‘weight loss’ (Glenton et al 2019). A really good tip is to always talk about people involved in the research – ‘children’ or ‘overweight men’ – rather than ‘study participants’ or (worse) ‘research subjects’. We come back again to the idea of bringing the people back in to your work.

Although a general rule is to keep language simple, unusual words can sometimes be used for effect. As I was writing this section, I read a paper in a US medical journal summarising what we knew at that time from cohort studies of the rate of asymptomatic COVID-19 infection in general and particular populations (Oran and Topol 2020). I was struck by one phrase in the concluding section identifying gaps in current knowledge: ‘What individual differences might account for why two persons of the same age, sex, and health status, for example, have idiosyncratic responses to SARS-CoV-2 infection? Why does one come through with nary a symptom, while the other lies near death in intensive care?’

It was the term ‘nary a symptom’ – a strangely colloquial and perhaps slightly archaic turn of phrase in a clear and scientific paper. But it serves its purpose, providing emphasis (compared to the flatter ‘no symptoms’) and perhaps even a sense of surprise and contrast. It seems like a small thing, but the choice of words matters.

There is a great book on translation by David Bellos, himself a celebrated literary translator. In an entertaining and very readable book, Is That A Fish in Your Ear? (Bellos 2012), he explores the nuance of meaning in the act of translation. The complexity of this act of cultural exchange cannot be over-stated. He looks for instance in how you translate a joke, noting the genius of the late translator Andrea Bell in the Asterix series. The skill with which she rendered style, idiom and humour – just think of the characters’ names in English, from ‘Getafix’ to ‘Dogmatix’ – is unsurpassed. It is much more than word for word translation. The relevance of this is that we move from different registers or styles without thinking. We tend to speak differently to a friend or a senior colleague. We write differently in an academic paper or a postcard. But sometimes it is helpful to switch registers and to use words and language which may be slightly out of place in a different context. It may be using homely or concrete examples or analogies when writing about your research. Or playing with humour and double meanings (although this is best done sparingly). Choose your words carefully.

Be yourself – the importance of voice

One of the hardest elements to explain or define is the ‘voice’ in someone’s writing. We know it when we see it. We could probably recognise a paragraph written by Malcolm Gladwell or Caitlin Moran without much trouble or a short extract from a documentary by Louis Theroux. The presence of the writer is partly their style and how they express themselves and partly the kinds of things that interest them. You may think that academic writing is objective and neutral and stripped of personality. But Helen Sword and others have shown that many of the best scholarly communicators bring their whole selves to the project, with all their quirks and idiosyncrasies.

A good way of practising finding your voice is by writing a blog. Sarah Chapman, knowledge broker for the UK Cochrane Centre (and former nurse), writes regular blogs, relating her personal experience to the available evidence and making sense of the research. Two of her blogs stick in my mind; her review of evidence on older people’s adherence to complex medication regimens in relation to her mother with dementia and her experience of frozen shoulder. In the latter, she bookends the short piece with her story of shoulder stiffness and pain by way of memory of her grandfather, wounded in the First World War. She interrogates evidence which may be helpful, from physical therapy to electrotherapy (not much help). She then provides detailed insight from her physiotherapist, again interpreting the evidence with her own praxis or professional wisdom. There’s a lot in here, but it is easy to read (Chapman 2017) www.evidentlycochrane.net/frozen-shoulder-2/.

Deborah Bowman, a professor of medical ethics, has written powerfully on the experience of being a cancer patient. This upended many of her beliefs and assumptions in her professional life, in areas like consent and patient choice:

Throughout my treatment, my responses surprised me. I was both rational and emotional. I both wanted to know and to not know information. I was constant and changeable. Sometimes, I was taken aback by the way the arrival of a Royal Marsden envelope made me feel – like grenades on my door mat. (Bowman 2019)

It can feel exposing to use your own voice. But it can help to connect to your readers. On that note, Helen Sword devotes a chapter to the personal voice – suggesting that there is no inviolable rule that writers of academic papers cannot speak directly to their readers. She gives example of respected scholars writing as ‘I’, ‘we’ and even calling out directly to readers with ‘you’. Writing in the third person – ‘the author found’, ‘the investigators have already asserted’ – is strangely distancing and formal. It removes you from your writing. Think about re-inserting your own voice and speak directly to your reader.

Practical pointers to strengthen your writing style

Think about writers you like

Who do you enjoy reading – for work or in your own time? What do you like about their style? Ask your friends and colleagues to name one writer, fiction or non-fiction, who they think writes well. See if you agree. Keep a journal with reflections on your research, notes on recent books or articles which have stuck in your mind. Are there any connections between them? What is it about the themes and the way in which they were expressed that resonated with you? Take one thread, perhaps linking some disparate trends or insights from different worlds and disciplines, and try writing a short blog. Keep it personal.

Sum it up

Write a short 300-word summary of your research and try it out on your neighbour. Test how accessible it is using an online readability index, aiming for a reading age of about 13 years. Read the summary out loud. This will help you express yourself as simply and clearly as possible. Remember to start and end with the problem, why it is important and how this research adds to what we know. Why should anyone be interested? Use a brief example of individuals and families and how this research might lead to better care or experience.

Play with titles

Find three titles that you like from a few journals you read. Which papers caught your eye and how did the title contribute? You can try creating five alternative titles for your research study, testing out contrasting styles and lengths, with at least one which is boldly humorous. You can market test titles as tweets and see which versions attract the most attention – and bring in readers who stay on the page.