Around a month ago, I was requested to review a paper submitted to a well-known APS journal. The main idea introduced in the paper was nice and the experimental implementation had fairly good results. I therefore sent my approval with some minor revisions.
Last week, I was shocked to find (through a poster presented at an international conference) that the idea in the work I refereed had been blatantly plagiarized from another work published earlier this year. In fact, a few sentences had even been simply copy-pasted.
In the course of reviewing, I had checked all the references in the paper and not surprisingly, the original work had not been cited.
I immediately sent a correspondence via the APS referee interface to revert my earlier decision and reject the paper outrightly. The editor later sent me a detailed reply saying that the paper has been rejected by the journal and the authors have been made aware of the original work.
But anyhow, the question I wish to raise here is: whose responsibility should it be to find out if a work submitted for publication is an act of plagiarism ?
While I feel both referees and editors must work on this aspect, I think the primary onus should be on the editors/publishing team. They have far more resources to investigate plagiarism. Furthermore, in contrast to editors, referees do not get paid and are doing this job mainly because of their belief in the edifice of peer-review, to add to their CV, etc.
Perhaps the experts and experienced researchers can give their opinion here?