Let’s assume you are a scientist, collaborating with another scientist on some research project that will lead to exactly one article (to keep things simple). How should you go about involving further scientists, perhaps to perform even stronger research?
The answer is simple, if you work from first principles. Here there are the two key principles at work:
- Joint ownership. The research is jointly owned by existing collaborators; any decision needs unanimous consensus
- Dilution of share. Adding more collaborators leads to more authors and dilutes you share of the resulting value.
With this, how should you think about adding new collaborators?
First, asking a colleague to provide feedback does not necessarily lead to a contribution that affords them collaborator and hence co-authorship status. They may request this, of course, before they provide feedback, in case of which you need to discuss and agree on whether to let the outside person in and what their contribution should be.
When considering a colleague as a new added collaborator, you first need to have agreement by the existing collaborators (principle 1). Only then can you ask someone to join. It is probably best to have some specific potential contributions in mind already.
You should hesitate to add someone, because you are diluting your share of the work’s value (principle 2). Adding someone only makes sense if the new collaborator adds so much value that it offsets your reduced share.
Under no circumstance should a collaborator surprise their colleagues with hitherto unknown collaborators to the same work. This is why emails about the work should typically reach all involved parties.
There are some problematic situations. In a two author combo, the second author loses most if a third author is added. That’s because the citation changes from “A & B” to “A et al.”, dropping author B from the common citation form, reducing their name recognition. Given that the first author position is often fixed, a second author of two should think hard about whether they are willing to accept this. In general, the first author benefits more from extending the collaborator list than the other authors.
Another problem situation is if the contribution of new collaborators doesn’t materialize or overlaps strongly with existing work. Sorting this out can be fraught with stress. I recommend our contribution model and the associated allocation process for this. This contribution model cannot only be used for the first author’s dissertation, but also to sort out who did what and how it relates.