BMJ No 7068 Volume 313

Education and Debate Saturday 23 November 1996

Writing in English for an international readership

Earlier this year John Kirkman, former director of the communication studies unit at the University of Wales Institute of Science and Technology, argued to BMJ staff and advisers that if they were serious about reaching an international audience they should pay more attention to the comprehensibility of the language they use for readers whose first language is not English and in particular work hard to remove a whole host of culturally specific words and phrases. Jona Heath, a British general practitioner, heard Dr Kirkman's arguments and, along with a Swedish colleague, disputed them strongly. We invited them to put their respective cases.

Confine yourself to forms of English that are easily understood

John Kirkman

Most contributors to medical journals published in English ignore the difficulties their texts present to readers for whom English is a foreign language. For example, here is an extract from a British journal, discussing how a new health minister in Australia is dealing with long waiting lists:

"His ministerial innings has begun in dour Geoffrey Boycott style, with orders from his captain Premier Bob Carr to defend one wicket at all costs."

To readers who play cricket, that passage conveys information and atmosphere. What, I wonder, does it convey to most readers in Italy or Russia, or even in the USA?

To make English-language medical journals readable by as many overseas readers as possible, we must remove allusions that are not essential to the medical discussion and write in forms of English that are easy for those readers to understand. That does not mean that we must distort the scientific content, remove necessary medical terms, or adopt nursery school vocabulary, structures, and tone. It does mean that we must confine ourselves to forms of English that are likely to be in the normal range learned by readers overseas.

Informal expressions

For example, we must avoid informal expressions such as:

"While the government's initiative represents a welcome foot in this particular door...."

"In a similar vein, the use of a placebo ...." "Two of those on which there is good agreement took a knock from scientists at....."

"...a criminal act for a doctor to accept from a patient an "under the table" fee."

We must recognise that putting a word or phrase in quotation marks to indicate that we are using it in a special way does not help overseas readers: it simply emphasises that they will probably not understand it.

We must avoid expressions that are familiar to users of one version of English but unfamiliar to users of another version. For example, I suspect few British readers would understand this extract from an American journal (the writer was explaining the source of an outbreak of enteritis):

"... prepared a jerky from cougar meat...."

According to Webster's Ninth New Collegiate Dictionary, jerked meat is meat preserved in long, sun-dried slices.

Perhaps British readers would know the meaning of this statement from Australian writers:

"All animals were fed standard laboratory chow..."

Readers in France would probably find it difficult. If they consulted a Robert and Collins Dictionnaire Anglais-Francais, they would be told that chow is equivalent to British English grub or nosh. Eventually, they would discover that the equivalent to chow in French is bouffe.

Unfamiliar words

We must avoid words that readers with a limited command of English are unlikely to have met before, especially if words in more common use are available to express our meaning adequately:

"There is, however, a surprising paucity of objective data."
[Surprisingly, however, there is not much objective data?]

"The seemingly evanescent pleasures of the drug are accompanied by....."
[short-lived, impermanent?]

"All de jure residents were included [in a medical study]."
[lawful; legally qualified?]

"Antiabortionists hijack fetal pain argument" [A headline]
[take over; take control of?]

Sentence structures

We must remove the difficulties presented by complicated sentence structures, especially by the parenthetical structures so popular with statistically correct medical writers. No doubt the information about confidence limits in the following statement is relevant, but the parentheses delivering that information are awkwardly placed even for a reader who is a native speaker of English:

"The corresponding regression coefficients defining the increase in blood pressure for a 1 kg fall in birth weight were -2.80 mm Hg (95% confidence interval -3.84 to -1.76; P<0.0001) for systolic pressure and -1.42 mm Hg (-2.14 to -0.70; P<0.0001) for diastolic pressure."

The writers could have let us absorb the primary information in a simple statement, and then told us about their confidence limits:

"The corresponding regression coefficients defining the increase in blood pressure for a 1 kg fall in birth weight were -2.80 mm Hg for systolic pressure and -1.42 mm Hg for diastolic pressure. (95% confidence interval; -3.84 to -1.76, P<0.0001, for systolic pressure, and -2.14 to-0.70, P<0.0001, for diastolic pressure.)"

Even without parentheses, long sentences are difficult language-processing tasks for readers whose command of English is limited. Think back to your school days and your struggles to translate texts from a foreign language. Imagine being confronted in that foreign language by a statement such as:

"They found that fairly short audits of the process of care based on interventions known from randomised controlled trials to influence outcome would detect relevant differences in the quality of care that would take several years to detect if mortality was being compared."

A restricted language

It is neither artificial nor patronising to restrict the language we use to address an audience. We do not use at all times the whole range of English (or other language) we have available. We choose subsets in accordance with the audience and context. For quick, accurate communication with friends during a football match, we choose a subset of vocabulary and structures that will be understood easily by those friends. We do not choose the same subset of language when we address colleagues at a cocktail party or an audience at a church ceremony. If we aim to communicate accurately and easily with an international audience of medical colleagues, many of whom have only a limited command of English, it is simply a matter of linguistic courtesy to choose a subset of vocabulary and structures that will minimise the demands we make on them.

Imagine being confronted in a foreign language by a statement such as:

The first group are nearly all bacterial exotoxins. They include exotoxins produced by staphylococci - the toxins causing food poisoning (enterotoxins A, B, C1-3, D, and E), toxic shock syndrome (TSST-1), and the scalded skin syndrome (exfoliating toxins A and B) - by group A streptococci (their pyrogenic exotoxins A, B , C, and D) and by Clostridium perfringens and Yersinia enterocolitica (their exterotoxins).

As English-language medical journals go on the worldwide web, the number of would-be readers who can access those journals but whose command of English is limited will increase dramatically. Shouldn't we try to help those readers?

All the examples in this article are genuine extracts from medical journals. However, as my purpose is to make general points, not to comment on the writing of individuals, I have given no exact references. Ten of the examples come from the BMJ, one from the JAMA, one from the Journal of Pharmacy and Pharmacology, and one from the New England Journal of Medicine.

John Kirkman,
consultant in presentation of scientific and technical communication

Communication Consultancy,
Witcha Cottage,
Ramsbury,
Marlborough,
Wiltshire SN8 2HQ