I recently ran across another reference to the newly-discovered sin of "self-plagiarism". I get the idea, I really do. But to equate reusing your own work with plagiarism is stupid and harmful.
Look, the primary reasons anyone with any sense hate plagiarists are (a) the wrong person gets credit for the work or idea, and (b) the plagiarist didn't get "permission" to use the work plagiarised (a very Western notion, BTW). OK, fine. I'm pretty sure that neither of these applies to taking your own work, though.
There are good reasons that you should attribute work that you copy, even if it is your own: attribution makes it possible to go back and look at the (presumably relevant) other places it has appeared, and it inhibits the sense of "deja vu" that sometimes transpires when reading separate works by the same author. There's also a value argument, that potential purchasers of your work should be protected from unwittingly buying the same work twice, I guess.
Maybe the worst reuse I ever experienced was in a Scott Adams "Dilbert" book, where he reused the same cartoons multiple times in a single book. It wasn't a good thing. That said, I would never accuse him of "self-plagiarism". Plagiarism is perhaps the most serious charge one can bring in the intellectual world. Reuse doesn't rise to nearly that level: it's a minor sin. So the term "reuse" should be sufficient. Further, individual organizations should have crystal-clear ground rules about what comprises acceptable unattributed and attributed reuse: I don't believe that there's nearly as strong an intellectual consensus here as the witch hunters seem to think.
There are enough serious problems to worry about. Let's call self-plagiarism reuse and treat it as the small problem that it is. (B)