Do it empirically, or don't do it at all? On the impact of 'publish or perish' on the research of ECR's | Synapse