Sunday, February 04, 2018

Should you send the scientist your draft article?

The Twitter discussion sparked by this poll was very illuminating. There’s a clear sense that scientists largely think they should be entitled to review quotes they make to a journalist (and perhaps to see the whole piece), while journalists say absolutely not, that’s not the way journalism works.

Of course (well, I say that but I’m not sure it’s obvious to everyone), the choices are not: (1) Journalist speaks to scientist, writes the piece, publishes; or (2) Journalist speaks to scientist, sends the scientist the piece so that the scientist can change it to their whim, publishes.

What more generally happens is that, after the draft is submitted to the editor, articles get fact-checked by the publication before publication. Typically this involves a fact-checker calling up the scientist and saying “Did you basically say X?” (usually with a light paraphrase). The fact-checker also typically asks the writer to send transcripts of interviews, to forward email exchanges etc, as well as to provide links or references to back up factual statements in the piece. This is, of course, time-consuming, and the extent to which, and rigour with which, it is done depends on the resources of the publication. Some science publications, like Quanta, have a great fact-checking machinery. Some smaller or more specialized journals don’t really have much of it at all, and might rely on an alert subeditor to spot things that look questionable.

This means that a scientist has no way of knowing, when he or she gives an interview, how accurately they are going to be quoted – though in some cases the writer can reassure them that a fact-checker will get in touch to check quotes. But – and this is the point many of the comments on the poll don’t quite acknowledge – it is not all about quotes! Many scientists are equally concerned about whether their work will be described accurately. If they don’t get to see any of the draft and are just asked about quotes, there is no way to ensure this.

One might say that it’s the responsibility of the writer to get that right. Of course it is. And they’ll do their best, for sure. But I don’t think I’ll be underestimating the awesomeness of my colleagues to say that we will get it wrong. We will get it wrong often. Usually this will be in little ways. We slightly misunderstood the explanation of the technique, we didn’t appreciate nuances and so our paraphrasing wasn’t quite apt, or – this is not uncommon – what the scientist wrote, and which we confidently repeated in simpler words, was not exactly what they meant. Sometimes our oversights and errors will be bigger. And if the reporter who has read the papers and talked with the scientists still didn’t quite get it right, what chance is there that even the most diligent fact-checker (and boy are they diligent) will spot that?

OK, mistakes happen. But they don’t have to, or not so often, if the scientist gets to see the text.

Now, I completely understand the arguments for why it might not be a good idea to show a draft to the people whose work is being discussed. The scientists might interfere to try to bend the text in their favour. They might insist that their critics, quoted in the piece, are talking nonsense and must be omitted. They might want to take back something they said, having got cold feet. Clearly, a practice like that couldn’t work in political writing.

Here, though, is what I don’t understand. What is to stop the writer saying No, that stays as it is? Sure, the scientist will be pissed off. But the scientist would be no less pissed off if the piece appeared without them ever having seen it.

Folks at Nature have told me, Well sometimes it’s not just a matter of scientists trying to interfere. On some sensitive subjects, they might get legal. And I can see that there are some stories, for example looking at misconduct or dodgy dealings by a pharmaceutical company, where passing round a draft is asking for trouble. Nature says that if they have a blanket policy so that the writer can just say Sorry, we don’t do that, it makes things much more clear-cut for everyone. I get that, and I respect it.

But my own personal preference is for discretion, not blanket policies. If you’re writing about, say, topological phases and it is brain-busting stuff, trying to think up paraphrases that will accurately reflect what you have said (or what the writer has said) to the interviewee while fact-checking seems a bit crazy when you could just show the researcher the way you described a Dirac fermion and ask them if it’s right. (I should say that I think Nature would buy that too in this situation.)

What’s more, there’s no reason on earth why a writer could not show a researcher a draft minus the comments that others have made on their work, so as to focus just on getting the facts right.

The real reason I feel deeply uncomfortable about the way that showing interviewees a draft is increasing frowned on, and even considered “highly unethical”, is however empirical. In decades of having done this whenever I can, and whenever I thought it advisable, I struggle to think of a single instance where a scientist came back with anything obstructive or unhelpful. Almost without exception they are incredibly generous and understanding, and any comments they made have improved the piece: by pointing out errors, offering better explanations or expanding on nuances. The accuracy of my writing has undoubtedly been enhanced as a result.

Indeed, writers of Focus articles for the American Physical Society, which report on papers generally from the Phys Rev journals, are requested to send articles to the papers’ authors before publication, and sometimes to get the authors to respond to criticisms raised by advisers. And this is done explicitly with the readers in mind: to ensure that the stories are as accurate as possible, and that they get some sense of the to-and-fro of questions raised. Now, it’s a very particular style of journalism at Focus, and wouldn’t work for everyone; but I believe it is a very defensible policy.

The New York Times explained its "no show" policy in 2012, and it made a lot of sense: it seems some political spokespeople and organizations were demanding quote approval and abusing it to exert control over what was reported. Press aides wanted to vet everything. This was clearly compromising to pen and balanced reporting.

But I have never encountered anything like that in many years of science reporting. That's not surprising, because it is (at least when we are reporting on scientific papers for the scientific press) a completely different ball game. Occasionally I have had people working at private companies needing to get their answers to my questions checked by the PR department before passing them on to me. That's tedious, but if it means that what results is something extremely anodyne, I just won't use it. I've also found some institutions - the NIH is particularly bad at this - reluctant to let their scientists speak at all, so that questions get fielded to a PR person who responds with such pathetic blandness and generality that it's a waste of everyone's time. It's a dereliction of duty for state-funded scientific research, but that's another issue.

As it happens, just recently while writing on a controversial topic in physical chemistry, I encountered the extremely rare situation where, having shown my interviewees a draft, one scientist told me that it was wrong for those in the other camp to be claiming X, because the scientific facts of the matter had been clearly established and they were not X. So I said fine, I can quote you as saying “The facts of the matter are not X” – but I will keep the others insisting that X is in fact that case. And I will retain the authorial voice implying that the matter is still being debated and is certainly not settled. And this guy was totally understanding and reasonable, and respected my position. This was no more or less than I had anticipated, given the way most scientists are.

In short, while I appreciate that an insistence that we writers not show drafts to the scientists is often made in an attempt to save us from being put in an awkward situation, in fact it can feel as though we are being treated as credulous dupes who cannot stand up to obstruction and bullying (if it should arise, which in my experience it hasn’t in this context), or resist manipulation, or make up our own minds about the right way to tell the story.

There’s another reason why I prefer to ask the scientists to review my texts, though – which is that I also write books. In non-fiction writing there simply is not this notion that you show no one except your editor the text before publication. To do so would be utter bloody madness. Because You Will Get Things Wrong – but with expert eyes seeing the draft, you will get much less wrong. I have always tried to get experts to read drafts of my books, or relevant parts of them, before publication, and I always thank God that I did and am deeply grateful that many scientists are generous enough to take on that onerous task (believe me, not all other disciplines have a tradition of being so forthcoming with help and advice). Always when I do this, I have no doubt that I am the author, and that I get the final say about what is said and how. But I have never had a single expert reader who has been anything but helpful, sympathetic and understanding. (Referees of books for academic publishers, however – now that’s another matter entirely. Don’t get me started.)

I seem to be in a minority here. And I may be misunderstanding something. Certainly, I fully understand why some science writers, writing some kinds of stories, would find it necessary to refuse to show copy to interviewees before publication. What's more, I will always respect editors’ requests not to show drafts of articles to interviewees. But I will continue to do so, when I think it is advisable, unless requested to do otherwise.