Group Authors?

To start off with, it’d be good to get clear on how to understand the notion of group authorship. The key question to ask is: who has the property of being an author when several people author a paper?

The way that the properties of collectives work is a bit complicated, but one helpful distinction is between collective and distributive properties (linguists often add a third category of semi-distributive or cumulative properties, which is a fascinating topic in its own right). One simple test for the distinction is whether the following inference pattern goes through:

If the group G is F, then all of the members of G are F.

Roughly, a distributive property will pass this test, whereas a collective properties will fail it (roughly, because sometimes a sentence that involves one property can have both readings, meaning that the property can function either distributively or collectively. Ultimately, I suspect the distinction between collectivity and distributivity is a linguistic one, but I’ll leave this complication out below). Some examples: If the philosophy department evacuated the building, then every member of the philosophy department evacuated the building (test passed, meaning that evacuated is a distributive property). If the philosophy department surrounded the building, then it is not true that every member of the department surrounded the building (test failed, meaning that surrounds is a collective property). In fact the very opposite is true — only groups (and not individuals) can surround buildings.

(Sidenote: I actually suspect that the interesting notion of collective authorship is when authorship by a group is a both i) a collective property, and ii) none of the members of the team have the property of being an author. We might think of this kind of property as super-collective. Below, I’ll implicitly be working with asuper-collective notion of authorship)

What does this mean for authorship? Well, being an author is a property, and depending on whether we think about it as a collective or distributive property will have differences for how we think about the responsibilities of individual authors. If A and B write and publish a paper, we can say that the group consisting of just A and B has the property of being the author of the paper. If we think of authorship as a distributive property, this means that it will also be the case that A is an author, and B is an author. By contrast, if we think of authorship as a collective property, A and B can collectively be an author together, although neither of them are authors by themselves.

Why should we care? One might worry that this all seems like so much logical nit-picking. This distinction matters because it gives us two different ways of thinking about the responsibilities of individual authors. If being an author requires taking on certain responsibilities — say responding to response pieces, retracting the paper if it turns out to have been flawed, or claiming to know the results of the paper — then the collective/distributive distinction impacts whether these responsibilities filter down to the individual level. If authorship is distributive, then all of the responsibilities of authorship accrue to each member of the writing team, but if authorship is a collective property, then the collective of authors has the responsibilities together, and they can fulfil them as a team. This particularly matters if the results of the paper are complex, and involve interdisciplinary methods. It might well be that no author can claim to know all of the results of a paper, but that the collective taken together can claim to know the results, meaning that the authors can only fulfil the requirement to know their results as a collective.

One way to make collective authorship really salient would be to use a collective noun to pick out the research team, as the Polymath project do for papers which are outputted by their projects. This has the virtue of forcing us to ascribe authorship to a collective without knowing which individuals we might ascribe authorship to.

Interestingly, the guidelines on authorship that I’ve seen from journals seem to go precisely the opposite direction: treating authorship as a distributive property. Here’s an extract from the Journal of the American Medical Association’s guidelines to authors:

Not only do they claim that every member of an authoring team must meet the full criteria for authorship (which involves substantial contribution to design, writing, final approval of the work, and being held accountable for all of the work — sidenote: meaning that they end up with both labour and speaker related features of authorship), they have a procedure in place to exclude contributors whose work doesn’t fulfil all of these conditions, pushing them out to the acknowledgements section.

I want to close by noting a couple of worries about collective authorship.

A first concern which Torsten Wilholt raises in Collaborative Research, Scientific Communities, and the Social Diffusion of Trustworthiness is that because group authorship is typically a fleeting affair, it will be difficult to get a grip on a group’s credibility by considering their epistemic track record. The credibility of an author is important, since we will want to know how much to trust the result of a paper, and to make judgements about which papers to read. The fact that many collaborations are short-term does pose a challenge to understanding credibility via track record, but I don’t think that this is a distinctive problem for groups. Individuals also sometimes publish papers without a track record: think of a PhD student publishing the results of her doctoral work, or unknown mathematicians like Yitang Zhang proving important results. The scientific community has various other tools for assessing the credibility of papers in these cases — foremost amongst them just reading the paper in question.

Another worry is about whether a collective author can legitimately take responsibility for the results of joint work. Huebner, Kukla and Winsberg point out that large-scale work will not be susceptible to any simple methodological story, both due to scientists using methodologies from different disciplines, and contextual micro-decisions (like whether to exclude a subject from a drug trial for not complying with the protocol). They argue that in cases where no individual knows the complete methodology behind a research paper, it is not plausible for either individuals or the group to claim authorship for the paper because neither can be held accountable for the result. Although I can feel the draw of the argument that no individual should claim authorship in such cases (at least if taking responsibility for the claims made is part of what it is to be an author), I don’t see why the group shouldn’t claim authorship in such cases. Why not have a distributed notion of epistemic responsibility, whereby a group is responsible for a set of claims just in case the members of the group between them take responsibility for all of those claims? So, although no individual can defend all of the methodological choices, between them the members of the researching team could defend all of the methodological choices. (Side-note: this would involve treating responsibility as something like a cumulative property, one which a group possesses in virtue of members of the group contributing to that property).

So, we’re a long way away from the JAMA’s guidelines on authorship, and into pretty uncharted territory. I definitely worry that there are other issues for a collective account of authorship, but I haven’t seen any decisive problems. One other issue would be to work out how the collective account of authorship might interact with our account of authorship in general. It might be that certain conceptions of authorship are particularly suited to collective authors, whereas other conceptions cause problems for collective authors.

Related posts:

Islam is a spiritual ideology related to mindset. The idea is that thought is everything. Thought operates at the quantum level. And the highest thought is that thought that is aligned with the…

Conclusion. Get ready for another $10 a month drag on your credit card. It’s a rebranded, all in one Apple video and music offering in 2–3 years. An Emerging Area of Investment. It’s no secret that…

Making mistakes is a part of the learning process, and it’s completely okay to slip up! After all, how can you truly learn without experiencing some missteps along the way? I remember a time when I…