Image: Responsibility, by Nathan Siemers CC BY-SA 2.0 via flickr.com
I spend a lot of time talking with students and colleagues about what authorship means, and about what criteria one might use for assigning it. That’s partly because the nature of authorship is both complex and (especially for early-career scientists) critically important. It’s also because my research has evolved in ways that mean I rarely write a single-authored paper any more. In fact, I rarely write a 2- or 3-authored paper any more.
There’s nothing unusual about me (in this respect); the lengths of author lists have been increasing in almost every field. In some fields, they’ve reached startling proportions, with author lists surpassing 5,000. It’s not universally agreed exactly what contributions merit authorship, or what responsibilities coauthors bear. However, one thing we often hear – and I’m pretty sure, one thing I’ve said – is that each coauthor should be willing to take responsibility for the entire paper. Take, for example, the recommendations on coauthorship from the International Committee of Medical Journal Editors:
The ICMJE recommends that authorship be based on the following 4 criteria:
- Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND
- Drafting the work or revising it critically for important intellectual content; AND
- Final approval of the version to be published; AND
- Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
In addition to being accountable for the parts of the work he or she has done, an author should be able to identify which co-authors are responsible for specific other parts of the work. In addition, authors should have confidence in the integrity of the contributions of their co-authors.
All those designated as authors should meet all four criteria for authorship, and all who meet the four criteria should be identified as authors. Those who do not meet all four criteria should be acknowledged—see Section II.A.3 below. These authorship criteria are intended to reserve the status of authorship for those who deserve credit and can take responsibility for the work. [emphases added]
I like these authorship criteria, and I’ve quoted them in The Scientist’s Guide to Writing and elsewhere. But they’re not particularly clear about “taking responsibility” (it’s hard to blame the ICMJE, because nobody else seems all that clear about it either). They don’t explain the difference between “accountable for” and “responsible for” (that distinction seems subtle, if there is one at all). They begin by suggesting each coauthor is accountable only for making sure some other coauthor can answer questions about the work; then they add that coauthors should have confidence in the integrity of their coauthors’ contributions; and finally, they fling caution to the wind to go all in on the idea that each coauthor should, without qualification, “take responsibility for the work”.
So which of these is it? And why?
I hear frequent recommendations that all coauthors should understand, and be able to explain and defend, all elements of the work. That, to me, sounds like the “all in” version from the ICMJE; and it’s an appealing idea. It’s especially appealing from an author’s point of view, because if it turns out that one coauthor faked data or otherwise acted unethically, it’s inevitable that all the other coauthors are going to wear some of the stain. So I used to buy into this; but I don’t any more. It’s just become impossible, as authorship teams have grown and research has become more interdisciplinary, for every coauthor to understand every aspect of the work in the depth needed to defend it if it’s challenged. Well, perhaps not “impossible”, exactly, but mindbogglingly inefficient. After all, it’s often precisely because I don’t know how to do analysis X that I recruit a coauthor who does. And sure, I’m a reasonably smart guy and I could learn analysis X (or repeat literature-review Y or master lab technique Z) – but if there’s a better way to have science slow to a crawl, I don’t know what it is. Coauthorship exists for a reason: because complementary expertise makes our work better. A year or so ago, over at Dynamic Ecology, Meg Duffy wondered whether at least each paper should have one author who understands and takes responsibility for everything (a “guarantor”). I voted “no” in her poll. Perhaps one author being guarantor is more realistic that every author being a guarantor; but I don’t want the role, and if I don’t want it, I can’t ask one of my coauthors to take it, either.
So I think I’m rejecting ICMJE’s final stab at the issue; and yet, I think their first stab fails, too. It surely isn’t enough for each coauthor, when a paper is challenged, to “ensure that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved”. That suggests that I should concern myself about my coauthors’ contributions only after publication, and only if someone makes an accusation of improper work. That sounds careless, to me. My coauthors are used to me pushing at what they’ve done, all through the writing process. For example, when an analytical result doesn’t match my intuition, I’ll suggest more analyses (and very often, analyses of made-up data. I’m not aiming to understand the surprising analysis, exactly, or at least not in detail; but I’m aiming to have confidence that it (somehow) does what my coauthor is claiming it does.
What’s stronger than “ensure that questions are investigated” but short of the pie-in-the-sky “understand and defend all aspects of the paper”? The ICMJE’s second stab at the issue is that each coauthor should have “confidence in the integrity of the contributions of their coauthors”. I think this works for me. I think I could have that confidence in two ways. Sometimes, I could have that confidence directly in the contributions themselves (I understand and can reproduce analysis X; I know the literature well enough to see that lit-review Y isn’t missing any critical papers). Other times, I could have that confidence in my coauthor instead. Perhaps I’ve assessed their statistical ability with analyses I do understand, perhaps I know them well enough to trust their judgement or their research ethics. Am I naïve to rely upon “trust”? Maybe – but ultimately, trust is at the foundation of everything we do. Despite the Royal Society’s motto Nullius in verba (“take nobody’s word for it”), science just can’t progress without each of us having some level of trust in the work of at least some of our colleagues.
So: for me, what does it mean to “take responsibility for a paper”? It means that in listing myself as an author I’m announcing to the world two things: that I’ve checked my own contributions to the best of my ability; and that I trust my coauthors’ contributions because I trust my coauthors. I know that I should allocate that trust carefully, because if I’m wrong, and my coauthor fakes a result, I’m going to wear the stain too. But I have to trust someone, don’t I? In science, as in life, there simply isn’t any other way.
© Stephen Heard July 17, 2018