Start Authorship Discussions Early

While on our AGU Pubs trip through China last month, a good piece of advice that AGU Senior VP Brooks Hanson made in his “author advice” presentation was to start conversations about authorship early in the research process. As a researcher goes through the scientific process, the person will most likely discuss the research with colleagues or even get help and resources from colleagues. The interaction could be at the initial stage, when they see something strange in the observational or numerical data and have those initial conversations about what it could be. The interaction could occur at the literature search stage of seeing of the weird thing is already explained by some previously published study. That is, the term “literature search” could be asking a colleague down the hall about a topic and following whatever leads they suggest. The interaction could be at the time of developing the initial hypothesis of what is happening in this strange thing. It could also be in the formulation of the experiments to test that hypothesis, or in the act of conducting those experiments (whether they be data analysis tasks, numerical model runs, new lab or field data collection, or a new theoretical derivation). Or, the interaction could happen very late in the process, at the stage of writing up the study for presentation or publication.

There is a very broad spectrum for this level of interaction of the researcher with all of these different people. It could be participation in a group meeting, where the person chimed in with a few comments or suggestions. It could be a 5-minute talk in the hallway. It could be an hour of flipping through plots on a screen together. It could be writing a new chunk of analysis software or a new subroutine in a model. It could be making plots. Not only should the time involvement be considered, but the significance of the involvement should also be considered. A five-minute conversation that completely changed your thinking on the subject might be worth coauthorship, while many hours of regular participation in a group meeting at which you mentioned the work might not rise to the level of coauthorship. As a research community, we’ve been making this judgment for a long time but, even still, there are no hard and fast rules on what contribution warrants coauthorship.

Dr. Hanson’s advice: broach the question of authorship early. With the CRediT list of author contributions handy, as well as the AGU ethical guidelines of who should be an author, including AGU Council’s thoughts on this topic, researchers should have frank and honest conversations with colleagues making contributions to their work. When you think that someone’s involvement is rising to the level of coauthorship on the eventual presentation and publication of the work, then talk with them about it. Most of us wait until the paper is written before we start to have these conversations with those outside the immediate “primary author” core group (which could just be one or two people). Author role #10 on the AGU page above is worth repeating here:


All coauthors are responsible for the “quality and integrity of the submitted and published manuscript.” Which means that, to be a coauthor, you pretty much have to participate in the “writing – reviewing and editing” CRediT role. That could be your only involvement, but those that participated in the “conceptualization” of the study should also review and edit the manuscript before submission. If someone isn’t prepared to take on the reviewing and editing task, then their contribution has to be very strong in one of the other contributor roles to warrant coauthorship.

So, researchers honing in on results worthy of a manuscript:

  • Make an agreement with potential coauthors, saying something like this: “I think that what you’ve contributed so far warrants coauthorship on my future paper, but there is still more for you to do to get your name on the paper – please read the paper.” You should extract from them a promise to read and comment on the manuscript; without it they probably should decline to be a coauthor.
  • Offer coauthorship to all colleagues that significantly contributed to the work. Think about who impacted the research at each stage of the scientific process and offer them the chance to be a coauthor, with, or course, the additional work of “reviewing and editing” the manuscript.

The question still remains, what amount of interaction rises to the level of a “significant scientific contribution to the work?” I’ve addressed it here but it’s a subjective judgment call.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s