Aarhus University Seal / Aarhus Universitets segl

Line Edslev Andersen and Brad Wray present joint paper at the 11th Biennial Collective Intentionality Conference, August 22-25, 2018, Tufts University, Boston, MA, USA

"Collective authorship: What we can learn from retraction notices"


When a serious error is discovered in a collaborative scientific paper and the paper has to be retracted, the group of authors – the collective author – is put under pressure. In this talk, we examine the robustness of collective authors by examining how they react to this kind of pressure and when they dissolve. This may help uncover the nature of collective authorship.

Our approach is to study retraction notices. We have looked at retraction notices in Science from the past 35 years and made a preliminary analysis of the notices signed by the authors of the retracted papers (as opposed to editorial retractions). They constitute 85 % of the retractions.

The preliminary analysis suggests that there are three common types of retractions. First, there is the type of retraction that is due to misconduct and where the collective author lays the blame on an individual who is guilty of misconduct [Category 1]. Second, there is the type of retraction that is due to an honest mistake and where the collective author takes on the responsibility, identifies the error made, and evaluates the difference the error makes to the reported findings [Category 2]. Finally, there is the type of retraction where it is unclear whether misconduct or an honest mistake has been committed and where the collective author neither obviously takes responsibility nor places it on someone else [Category 3].

If these are three common types of retractions, what do the various types tell us about collective authors?

Category 1 cases, where a collective author does not take responsibility when a group member has committed misconduct, may suggest a certain disunity of collective authors. But it may also suggest something else. The reason why the collective author does not take responsibility in these cases may be that a group member is not acting as a group member at all when committing misconduct, since her intentions conflict with the main aim of the group, assuming the main aim of the group is to uncover the facts about something.

Category 2 cases, where a collective author takes on the responsibility when an honest error has been made, suggest a certain robustness of collective authors. The collective author will describe the error it committed and reevaluate the reported findings.

Category 3 cases are likely due to the fact that a collective author will sometimes suspect that the error causing the retraction is due to misconduct. In this case, the collective author has reason not to trust what the group member who knows what happened says, since this is the person who is suspected of having committed misconduct, and thus of having lied to the group before. At the same time, the collective author will have moral and legal reasons not to publicly state mere suspicions. This makes for retraction notices where it is not clear who is responsible and why exactly the paper is being retracted.

Link to conference