As a researcher and therefore a reader of papers, I often encounter poorly written papers, full of grammatical and spelling errors. Of course, one cannot fully blame the authors -- many researchers are not native english speakers (myself included), and some only speak english at a very basic level, or not at all. However, some publication platforms (journals, magazines, transactions and so on) have editors listed: a group of people that tends to change per issue or a set of issues. My impression was that editors edit the papers for publication, mainly focussing on formatting and language (with feedback from authors, of course).
However, reading papers from a Lecture Notes in Computer Science issue, I'm quite certain that language isn't part of what editors do. Some articles, especially those that come from Workshop and Conference proceedings, often lack editing, sometimes to the extent that the paper is (for me, at least) no longer understandable. However, I've even noticed this problem for Journals. In addition, my reading experience seems to indicate that the problem is increasing, which would be a rather troubling fact that could be addressed by editors or by reviewers. I expected this to be addressed by reviewers; the quality of text is important to the correct communication of information. On the other hand, it may be unfair to non-native speakers to consider language in the review process.
Thus, my question actually consists of two parts:
1. What is the 'job description' of an editor?
2. How can the research community as a whole (authors, reviewers, PCs, readers, editors, ...) improve the situation?