[WikiEN-l] What to do about our writing quality?

[Ray Saintonge]

Abd ul-Rahman Lomax wrote:
> There is somewhere a recommendation that Talk be refactored. Right 
> now, what I see, everywhere I've looked, is that Talk pages are 
> simply archived. And then the same debates occur over and over, with 
> new participants who have not read the old debate, so more time is 
> wasted explaining everything over and over. It is *incredibly* 
> inefficient, and inefficiency is not fatal when new editors keep 
> pouring in. But it burns editors out, in the end, and that stream of 
> new users will dry up. I've called it a pyramid scheme. It works as 
> long as new blood keeps appearing.
Refactoring talk pages is an old notion that was already there when I 
became involved in early 2002.  I tried it then on a couple of 
occasions, and found it to be an incredibly difficult task.  Not 
everybody can do it.  It is even more difficult than good copyediting in 
article space.  It could also lead to complaints from purveyors of 
nonsense that their nonsense is being censored.

Archiving doesn't help, especially when those archive pages are 
accompanied by a warning that they are not to have further comments 
added.  Some of the shorter threads on a talk page might do well to be 
revived, especially when they deal with an easily refuted but popular 
misconception. If a topic is subject to constant dispute the talk page 
and its related archives become an unmanageable multitude that would 
deter anyone from looking to them for answers.  An improvement might be 
to archive by topic or question instead of by date as is currently done, 
but that would involve more work than simply using cut and past for 
everything added before a given date.

New editors that raise questions are more plentiful than new editors 
that answer them.  The answering editors can soon develop a siege 
mentality when they need to keep answering what they perceive to be the 
same questions.  The result may very well be an inability to recognize 
changes in the question.

Ec