Retract bit-rotten publications: Aligning incentives for sustaining scientific software

How might we use our current publishing and academic reputation system to promote software sustainability? This paper seeks to provoke discussion by proposing that papers whose workflows are not kept current with the changing software ecosystem should be automatically retracted, placing the work of adjusting a purported contribution on those receiving the credit. This would be a change from the current situation where the work is placed on those seeking to use the contribution, each of which labors independently. As a provocation the proposal brings out strong reactions, and helps clarify what readers think of as worth rewarding in science. I progressively soften this proposal, eventually coming to a standpoint whereby a continuous improvement system could highlight opportunities for others to update a workflow to match changes in the software ecosystem, receiving a range of rewards from addition to a paper’s author list, to in-document acknowledgment, acknowledgement on a publisher’s or other website.