Why are so many scientific articles difficult to read?
Linda Cooper
Wednesday, 26 March 2008 20:07 UTC
Does the fact that so few papers have made it onto the roster of “good” papers suggest that most papers are obscure and poorly written?
Over the years, I’ve searched through various journals and read several hundred papers looking for examples of well-written manuscripts. While examples of poorly written texts abound, clearly written papers are surprisingly few in number.
Perhaps, as has been suggested earlier, it will be more helpful and relevant to turn our attention to why scientific articles are so inaccessible, so poorly written.
Linda
-
Replies
Jump to resultsResults
-
The comments here suggest that several factors perpetuate poor writing in science: limited space, limited time, lack of training, a belief that it’s difficult to explain complex concepts. Whether explicit ally or implicit ally, I think that scientists do indeed believe all of the above – and therefore continue to submit unnecessarily confused manuscripts.
That being said, one wonders why journals continue to publish poorly written articles – at least high-profile journals directed to non-specialist readers. It’s easy for me to imagine that authors would work very very hard to make their articles accessible to a larger audience if journals withheld publication until such manuscripts were truly reader-ready.
And I know scientists are capable of writing clearly about highly complex research. Students in my classes do this all the time once they have the tools to transform their original confused drafts into articles that both the specialist and non- specialist can understand. As some of you have pointed out, this is hard work – but necessary and possible in my view.
-
Linda, your question is highly pertinent:
one wonders why journals continue to publish poorly written articles – at least high-profile journals directed to non-specialist readers. It’s easy for me to imagine that authors would work very very hard to make their articles accessible to a larger audience if journals withheld publication until such manuscripts were truly reader-ready.I think there are two main answers to this.
First, the editors themselves handling the manuscript, in the cell/molecular/system biological sciences certainly, are themselves increasingly specialist and, for Nature, may not have English as their own first language. Hence, over the past 20 years, the people making the decisions about publication of these areas have shifted the balance more in the direction of “interesting scientific result” (clear to them, the authors and the reviewers if not to many other people) and less in that of “well-written paper”. Our subeditors do what they can, but they are “bulk processing”— they can only spend about a day per manuscript, and given all the other things they have to do to it, language cannot be a time-consuming part — in this regard, they focus on the first paragraph, on terms being defined, grammar, and so on — they do not have time to rewrite style (which in itself can lead to debate with the author, whose manuscript has after all been accepted – and who has many coauthors, probably, who cannot all easily be contacted to approve significantly rewritten text on a fast timescale, as the journal publshes accepted papers very quickly).
Second, speed and author pressure. If you ask most authors, they would prefer a Nature publication, and fast, to having to rewrite the text. If they are in a position to do so, they will take their manuscripts to an equivalent journal elsewhere if they know they are going to have to do a lot of style work to get it accepted. One may disagree with this culture, but it exists – not everywhere, but in a lot of places, and among a lot of scientists doing world-class work. This is the real world – journals want to publish the best papers, so have to offer speed and as few impediments as possible, once the technical aspects are passed by peer-review.
One possible way forward, as is implicit in your reply above, is to have an optional editing service at the “accept in principle” stage, where the journal offers to rewrite the paper after the authors have revised in response to reviewers’ and editors’ comments, but before formal acceptance. I believe that some authors would want this service (but those under perceived “scoop” pressure would not), and I believe that a journal with the editing experience of Nature could offer it on a fast timescale. But it would mean additional costs, and who would pay? Why would the publisher be persuaded to invest in this service if the journal is already publishing the best research? Already, the subediting service offered by Nature journals has vanished or is vanishing from the bulk of main STM (science, technical and medical) jouranls. The publisher receives many, many requests from editors for resources for one thing or another, and has to weigh them all up in terms of costs and benefits (which include bottom lines!) to the journal. I doubt the authors could pay for this service, and indeed, should they?
-
I want to a make small comment regarding an implicit assumption in the previous post.
English is the standard language for science. But it is not the same language as the language spoken by mother tongue speakers. As a matter of fact native English speakers constitute a small minority of the people that read and write scientific papers.
For mother-tongue speakers it is very difficult to judge whether a particular sentence or paragraph would be difficult or easy for the average reader.
In my experience papers written by native speakers (and talks delivered by them) tend to be more difficult to follow than papers written by non-native speakers with a sublime knowledge of English. Just compare the average sentence length for the two cases (and the number of punctuation marks) .
So if Nature wants to publish well-written papers it needs to have editors whose mother tongue differ from English.
Ad Lagendijk
sciencesurvivalblog -
Ad, it is interesting that there is actually fairly recent research to prove your point:
Hayden JD. Readability in the British Journal of Surgery. Br J Surg. 2008;95:119-24
The author looked at 189 manuscripts and found that the readability score was lower (i.e. the manuscript more difficult to read) when English was the first language of the principal investigator.
-
hi Linda this is true to get good quality paper in scientific journal, is your assessment! but some other people understand very easly, so is depends on site of intrest, there is some problum with some readers how they understand paper very precisely.
so is the part of intrest. viz if you read toxicology of metal and is your field of intrest then u understand this part very easly rather than other part. -
Ad: indeed, yes, papers written in English by people who are not native English speakers are often better written. See, for example, the paper in this Good Paper Journal Club forum about the dwarf mammoths, which was selected by a Nature editor as one of the best-written manuscripts that he handled. (It is by a Russian group.)
-
I would add to Ad’s point, that there is I think a difference between understanding a paper and editing it for clarity. An editor handling papers in a field will understand the point whether or not the paper is well-written. A sub (copy) editor will ensure the paper is grammatical (something that is hard for a non-native speaker of the written language) and clear. There is a gap between the two. A considerable amount of clarity of argument is imposed by the peer-review and editing during review, but not enough, in many cases. Manuscripts accepted for publication in a journal are frequently less comprehensible than they need to be, whatever the geographical origin of the author or editor. That’s a fact of life.
-
Maxine,
Perhaps readers could pay for articles to be more scrupulously edited – through increased subscription rates. They’re the ultimate beneficiaries of well-written manuscripts – and the ones who are rarely considered in this equation.
And it’s curious to note that as mentioned in another thread, scientists who have English as a second language can often produce more straightforward articles than native English speakers. However, this same happy thought can’t be extended to editors. As you mention, editors “handling the manuscript . . . may not have English as their own first language. Hence, over the past 20 years, the people making the decisions about publication of these areas have shifted the balance more in the direction of “interesting scientific result” (clear to them, the authors and the reviewers if not to many other people) and less in that of “well-written paper”.” Some time spent training such editors in techniques for effective revising might be the way to proceed. -
Agree with your sentiiments, Linda.
At Nature, all editors go through a period of training with the subediting team: the journal is good about supporting this exercise. Over the years I have observed that some editors (irrespective of geographical origin, but it is on balance harder for the non-native English language speakers) are more interested in the format and readability of their papers than others. Those that aren’t are by no means bad editors: everyone has strengths and weaknesses. But the pace at which Nature editors (including subeditors) have to work precludes “language” training on a regular, rolling basis – and would not be welcomed by all editors.
Subscription costs and publication charges have to be balanced against ability to pay. I agree that some well-written papers are by non-native English speakers. But many are not, and it is authors in the less rich countries who suffer when charges of any kind raise their head, as they have to rely on waivers and so on.
-
Hello
Is in fact paper writing is based on good knowledge in grammer and vocabulary, so tail me any other factors which is present in a good paper. I have one question – reading or understanding these two are different things, or same ?
Results
-
