I'm in the process putting together a dissertation by publication and one of the manuscripts builds upon previously published work. As a result, the manuscript as a chapter in the dissertation has citations that can currently refer to the publications. This presents a quandary for the dissertation though. Should the chapter continue to cite the publications, or simply refer back to the chapter in the dissertation?
If you have a dissertation with structure
- Chapter m = Paper n
- Chapter m+1 = Paper n+1
where Paper n+1 builds on and cites Paper n, then the best course of action seems to have Chapter m+1 cite both Chapter m and Paper n. That way you provide both the convenience for the reader to find the relevant material by just looking at another page in the current document, and the proper original source.
The same idea applies to other expository documents too, like review articles and textbooks.