I came across this situation that seems not immediately obvious as regards the "line" for self-plagiarism.
Suppose I have two papers on two related problems such that merging the two papers is not an option. It seems important to draw some remarks, which may just be qualitative descriptions in a novel way but which may be helpful in clarifying concerned things in the nonempty intersection. By nature of the problems the intended remarks would be similar to a certain extent. In this case, do I have to deliberately make them look sufficiently different? Otherwise, do I have to cite the remarks I made? I would love to know the predecessors's wisdom about this matter. Thank you.