Metadata on collective authorship

Metadata on collective authorship
Photo by Michael Dziedzic / Unsplash

While working on ResearchEquals, I have a lot of ideas that could be worthwhile for research practice. I say I have a lot of ideas and most don’t make the cut into production for a ton of reasons (a topic for another time). What does happen with ideas is that I start seeking input to evaluate my assumptions a bit more than mere armchair pondering. 🪑

Because of more (big) team science efforts, I was thinking how collective authorships can be made more feasible.

In a panel a few months ago, researchers from the Psychological Science Accelerator and the FORRT project indicated there’s a friction between collective authorships and individual authorships due to what gets recognized. Other factors are practical limitations in the submission process — like the extreme hassle of entering tens, hundreds, or dare I say thousands of authors into archaic systems like ManuscriptCentral.

Now the point of this blog goes into what has been done in metadata registration of collective authorships. It's less about what we should do; I want to understand the current situation such that I can evaluate the potential room for improvement.

Specifically, my interest was piqued when I saw that you can create “organizational authorships” in the metadata when you mint a DOI with CrossRef. Subsequently my question was: Are collective authorships registered differently from individual authorships in metadata? How would we know?

A screenshot from the CrossRef metadata schema (v5.3.1) indicating that contributors can be organizations, person names, or anonymous.
A screenshot from the CrossRef metadata schema (v5.3.1)

The Twitter hivemind helped me find the following six papers from different publishers, across twenty years, with collective authorships:

  5. 10.1016/j.physletb.2012.08.020
  6. 10.1177/2515245919900809

I ended up checking in the CrossRef API how they are registered and how they are displayed on the original website. Here’s a comparison of these five papers on the following questions:

  • Is the collective name included in the metadata at all?
  • Is the collective author registered differently from the individual?
  • Are the individual authorships included in addition to the collective author?
  • Are the individual authors displayed only in the metadata?
Collective included? Collective different? Individual authorships? Individual only in metadata?
10.1073/pnas.2111091119 🚫
10.1126/science.aac4716 🚫 🚫 -
10.1016/j.ejim.2022.06.014 🚫 -
10.1038/35057062 🚫
10.1016/j.physletb.2012.08.020 🚫 - 🚫
10.1177/2515245919900809 🚫

Based on this small sample the following key predictions might be inferred:

  1. Collective authorship is not registered differently from individual authorship in the metadata, despite the capacity to do so (6/6 cases)
  2. When there is a collective authorship, the individual authors are most likely to be registered alongside the collective in the metadata (5/6 cases)
  3. In a minority of cases, the collective is only presented on the website, not in the metadata (2/6)

However, I might have made a fundamental flaw in my understanding and would undermine prediction 1. See the screenshot below:

A screenshot for one contributor. Is it registered as an organization or not?

My working assumption is that organizations can't have affiliations, as the schema also indicates, hence, this was not registered as an organization. However, this might be a false working assumption! Other contributors have the given and family name properties, which this one does not (see only the name). Regardless, it is clear it is not clear whether this in fact is an organization.

Does anybody have more expertise on this? I asked around on the CrossRef forum but that’s still pending. I don't think I've come across a registered organization, but maybe they were all disguised so I didn't notice them!