Like many colleagues, I review lots of manuscripts for journals. Probably too many, but that's another issue.
I want to think in public about the role of self-citations in double-blind articles for peer review. I'm going to assume the following about these articles, which are all ingredient in the previous sentence:
1. the author(s) are known to the editor and journal staff.
2. the author(s) unknown to the reviewer (or the reviewer is uncertain of the author's identity)
3. the reviewer(s) are unknown to the author(s)
Let's say I'm writing an article about deep brain stimulation and I want to cite previously published work in an article blinded for peer review. If I write (Author, date) as a citation, then this indicates at least two things to the reader:
1. The author has published something in this area -- it might be a peer reviewed journal article or scholarly book.
2. The author is building on his/her extant research in this area
Both of the above imply that the author is an expert in the field. Might it be that this kind of self-citation actually lends a degree of credibility to the author's paper that is unwarranted?
There is another option. One could self-cite without drawing attention to the fact. I could write, "Morrison (2014) reported" without writing, "As my previous research has demonstrated (Author, 2014)..." Isn't the former better, especially for those reviewers who do not have a great deal of familiarity with all the authors in a particular field of subfield?
Perhaps the phrase "do not have a great deal of familiarity with..." in the above paragraph is naive, but I review submissions all the time from authors who seem to be outside the US, wishing to publish in US-based journals. And so this is a real issue for me. Perhaps I'm not as well informed as I should be, but there you go.