Scientific shortcuts

I am currently reviewing a paper written by two chinese guys. But this doesn't really matter; they could be italian mathematicians as well. They submitted a manuscript (or better a LaTeX-script) to a rather poor journal, and I didn't expect too much from it. It is nine pages long (or short), and it deals with a nice but rather abused problem in Schrödinger equations.

After the first quick reading, I found some misprints and language mistakes, and this is common and excusable. But I found too many references to [4], a previous paper by other mathematicians. This reference was published on a high-level journal, and this too is ok. The manuscript is not completely original, but every mathematician can do this :-)

The troubles came when I downloaded [4] and checked the corresponding technical results: if a lemma of [4] proposes an alternative, "my" authors write just one possibility. If an estimate of [4] is true under some assumption, "my" authors forget this assumption. The result is that they could have cited almost any previous paper, since they probably did not read it!

I don't want to reject the manuscript right now, and I will try and see if it can be amended. But this reminds me of italian laws: they refer back to more and more laws, and nobody checks them. But if you do, you'll discover that most citations are rather random citations, therefore almost useless! We are a funny people.