[Wikipedia-l] Deletion of Articles with No Content

Toby Bartels toby+wikipedia at math.ucr.edu
Mon Aug 26 01:44:23 UTC 2002


mirwin wrote:

>Fredbauder wrote:

>>Ok, Yes, Toby's suggestion that an article with a good title but no content
>>should be deleted is good policy.

>It seems like silly makework to me.
>The article will be recreated again anytime anyone
>clicks on one of the links leading to it.

These links will be "?" (or highlighted) if the page is deleted,
so people following them should be intending to write articles.
If they do write content, then that's great!
I don't mind contentless pages being recreated as contentful articles.

>If it is a good title then it is an invitation
>for anyone who encounters it to add to it.

For somebody interested in writeing content (which is what we want),
I think that a link of "?" is more inviting than a non"?" link.
If I know about a subject, then I'll follow "?" links but not non"?" ones.
If the non"?" link is to a contentless page, then this means
that I won't be writing articles that I otherwise would write!

>A poor title seems of no loss.  However, unless
>it is replaced with a good title which it previous
>links are edited to point at, it seems likely
>that it will reoccur quickly.  
>Somebody had to code the initial link to the poor 
>title to create the article in the first place.  
>It seems to me that the only solid way to eliminate 
>poor titles is  track the links and recode them to 
>a better title.   If the new title is a new article 
>then  unless someone undertakes a research project it
>will likely be fairly empty or stubby for a while.

If I understand you correctly, than I feel similarly.
I'm much less concerned about contentless pages
with poor titles that we'll never want in an encyclopaedia.
The way to fix ''these'' is to fix the links to them.
Only once that is done will deleting the page do much good.
So it is the good titles that we disagree on, not the bad ones.

>I think that having the Wikipedia Guard or Militia
>routinely deleting empty good titled articles may
>only slow down the growth in breadth and depth of
>the Wikipedia.   Some people may like organizing
>the link structures and establishing good initial titles
>and interconnections.   Why should this contribution
>be routinely deleted?   How much subsequent work is
>then lost from contributors who while browsing may
>choose to add an easy paragraph but who will not
>undertake an entire stub and the effort required
>to link it appropriately into an entire encyclopedia?

I really have no idea how you think that this will work.
Do you have examples on Wikipedia?
(If I need to look in page histories to see
how this worked in the past, then that's fine too.)

>Some areas of the Wikipedia already feel pretty
>circular and concise.   They have no sloppy or poor titles 
>hanging out for random fortuitous contributions from
>readers.   They have a concise complete feel to them
>that screams static encyclopedic overview with no place
>for further detail.

But I think that links to articles with no content ''adds'' to this problem.
When the link doesn't have "?", things look even more complete.
But they are not complete, and "?"s will make things look properly incomplete!
However, I do agree with you that we have this problem in some places.
People don't follow [[Wikipedia:Always leave something undone]] enough
(a sin that I've been guilty of too often in the past).

>To summarize, I am unconvinced that routine pruning
>of good article titles is useful to attaining our
>goals of depth, breadth, and reliability.  Rather, I think
>it may actually be harmful.

>P.S.  It might be an interesting experiment to build a
>detailed maze of good article titles and stubs in some
>underdeveloped subject area of the Wikipedia and toy around
>with some twikification techniques.

If they're ''good'' stubs, then I'd support this experiment.
(The page that I mentioned before is [[Wikipedia:The perfect stub article]],
and point #5 is the one that I hope people consider particularly.)

>If a couple of
>regulars cooperated in an area of common interest
>it might convince newcomers that Wikipedia is truly alive
>with sufficient utilization to keep its content dynamic
>and growing.  As it is I think the first multi Wikipedian
>contact in near realtime of many newcomers may often be in a
>negative atmosphere of panic and anger as the mailing list
>is attracted to some poor content locus for deletion sprees.

I hope that we never go on deletion sprees.
If we delete stubs for being too imperfect,
then every one should go to [[Wikipedia:Votes for deletion]] first.
I primarily just don't want things to be delisted from that page
without discussion; I'm not saying that they should all just be deleted!

>P.S.2  To address the issue of the most wanted list.  Perhaps
>its code could be enhanced to provide a weighted list or
>set of lists.   Thus two paragraph stubs referenced 26 times
>which have little or no outbound links would get some
>attention from people who prefer checking most wanted lists
>rather than subject browsing or random inspection for
>twikification efforts.  Perhaps we could identify some syntax
>factors that make a good Wikipedia article such as
>(perhaps):  length, median word size, median sentence length,
>average paragraph length, number of commas, number of inbound
>links, number of out bound links, editor rating, reader
>rating, etc. An advantage of this approach is that eventually
>various automated quality scanning tools could help people
>target material needing their particular gifts or interests.

Do you know how to write the software for this?
I think that we need to deal with how [[Special:Mostwanted]]
works now, or can be made to work easily.
I'm not against your idea; it's a longer term project, that's all.


-- Toby



More information about the Wikipedia-l mailing list