- plagiarism when you copy chunks of somebody else's already-published text and 'forget' to cite it accurately (or at all)
- plagiarism of a more subtle sort when you use part of something you've already published yourself elsewhere without making that clear
and then there's
- what you do when you write something and cite your own work extensively as part of it.
Not everyone would regard the second example as plagiarism at all, though many now seem to. Either way, such 'self-plagiarism' has to be unwelcome practice when done extensively, though it can be difficult to avoid completely when building an argument over several papers. Openness about what's being done is the key.
Many would see the third as a form of self-regard, immaturity or insecurity, though few seem to see it as plagiarism in the clear-cut way that the first is, or the less obvious way embodied by the second.
However, all three practices seem both avoidable, and to be avoided. I was reminded of all this as I read a UNESCO report last week where one of the papers provided a clear example of the third issue. Perhaps we all need to sign up to a self-denying ordinance: no more than 10% of papers cited should in something you write should ever be your own. Just think of all the ink and paper we'd save – and the moral high ground we'd all occupy.