December 19, 2003
Follow me on Twitter

Editors considered helpful

One advantage of weblogs is that weblog writers needn't have editors. Dave Winer, for example, defines the weblog as "the unedited voice of a person." Writers and editors, of course, famously resemble lambs and lions. Or vice versa.

Readers of weblogs and of edited journals know who to blame when they read something that's foolish or false. If I write something here and it's wrong, you'll find my name at the top of the page. If we publish something in TEKKA that you don't like, you know from the masthead that you disagree with Bernstein or Rau.

Academic journals often rely on peer review in place of editors. This works well in the sciences, where the most important thing -- the only really important thing -- is to get the science right. But it's a mess in the arts, and I think it's breaking down in computer science; it's just too convenient for reviewers to go along with publishing incomplete or mediocre papers, especially in Web journals. After all, nobody will blame the anonymous reviewers. If the paper is bad enough, nobody will read the paper either, so nobody will be blamed.

Of course, this quickly leads to nobody reading the journals, and shortly thereafter you don't have a field anymore, you just have a bunch of people with tenure.

Janet Walker once wrote instructions to referees that contained the core of a good idea. "Imagine that, when an accepted paper is delivered at the conference, the reviewers who recommended it will be required to stand at the side of the stage." (Some scientific journals used to require papers to have a sponsoring member who served much the same function -- I think PNAS still does this. Such systems can be conservative and corrupt, but that might be better than the building a literature that is itself corrupt.)