[WikiEN-l] Re: note David Gerard's just placed on [[WT:AFD]]

Tue Jan 31 03:52:52 UTC 2006

On Jan 30, 2006, at 7:00 PM, Oskar Sigvardsson wrote:
> Without commenting on the principle of the idea, I would like to point
> out that there are no such things as "article subpages". All pages in
> the main space are pure articles, with the exception of the shortcuts
> and [[Main Page]] (i bet some smart-ass is going to throw some minute
> exception at me that I forgot, but I'm pretty sure that those are the
> only two). For your suggestion to work, we would need a new namespace,
> as none of the current ones would be suitable for such a thing.

So put it in the talk namespace until we get a new namespace.  This is 
a good cleanup-type point, but a non-issue as far as the substance of 
the suggestion is concerned.

 > Now to comment on your suggestion: I think this is fundamentally
> unworkable. Articles can change rapidly,

Yes, some articles do change rapidly.  Some articles change *very 
slowly* (like one substantive edit per 6 months, or even slower).  This 
method is not appropriate for articles which are not stable.  But the 
vast majority of Wikipedia articles *are* stable.  When we've got all 
the stable articles fully referenced, we can work out a way to handle 
the last 2% or so that change quickly.  This is not a issue of 
feasibility, just a ~2% limit on the domain for which this applies.

> and it would be a herculean
> task to keep the citation page up to date.

We have done herculean jobs before; writing what we've written so far 
is a herculean task a number of times over.  And fundamentally, what is 
the loss if the page is not up date?  Each fact that is sourced will 
still be sourced, even if the page no longer contains it.  Every fact 
that has not yet been added to the page would not have a source whether 
the page existed or not.  We lose nothing.

> I think that it would
> simply not be possible for it to function in a meaningful way.

I'm not sure what you mean by "meaningful way" - would the facts not be 
readable?  Would writing down the source for a fact make the source 
turn into a purple teddy bear?  What exactly do you mean?

>  If we reference so heavliy that it cannot fit into the article 
> nicely, a
> separate page is going to be way longer than the article itself.

OK, I agree.  A full reference page would be longer than the page. (it 
would have to be, as it would contain nearly all the text, plus 
whatever text was needed to state the references.)  And the problem 
with this is what, exactly?  Are we running out of hard drive space?  
Do we have a limited number of alphabetical letters available?  Are we 
going to have a "e" shortage?

> It is true that every sentance in an article should be backed up by
> sources.

We agree about this.  I actually am a pretty strong eventualist on this 
subject; it's *hard* to find sources, and in nearly all cases I 
wouldn't remove something from an article merely for lacking a source. 
(Although I mean no disrespect to those who do remove such things; it's 
useful and bloody work.)

>  That does not mean that we have to actually put a superscript
> above every single fact that we use.

Note: Nowhere in my suggestion do I advocate this.  A more classic 
straw-man you are not likely to see.  In fact, my suggestion doesn't 
involve footnotes or superscripts in any way at all.

>  The fact is that when articles
> are written for wikipedia, they are not PhD thesis' (I how no idea how
> to pluralize that word) that have to be presented to a number of
> proffessors for review. No, wikipedia articles is just that,
> _articles_! They have to abide by certain criteria for style. In
> short, a good wikipedia article has to be nice looking.

I certainly agree!  I think making sure that our articles are well 
written, pleasant to the eye, and not too technical or cluttered is one 
of our most important tasks.  However, I'm a little puzzled as to why 
you brought this up, as the only change to the text of an article which 
followed my suggestion would be the addition of one line of text, 
placed in the References section (or simply at the end of the page):

"Detailed references are available here."

with "here" a wiki-link to the page with the detailed references on it.

While it's possible that the one line of text above contains some kind 
of horrific stylistic or writing error which would permanently and 
totally infect and cripple the "nice looking" quality of any article it 
was put in, I sort of doubt it.  And, if it did, it can be easily 
changed.

> As long as we can make sure that a) all sources are listed in a
> "References" section, b) very important facts are referenced directly
> and c) that we can always produce credible sources for every fact when
> queried, we should be fine!

"Fine"?  Again, I'm not sure what you mean by that.  I certainly agree 
with your points a), b) and c).  What better way to be able to "always 
produce credible sources for every fact when queried" than to write 
down those sources, in a nice list on a separate page?  That way, 
editors can work together to "produce credible sources", and if we are 
"queried" at 3 AM when no-one happens to be around, we will still be 
able to "produce" the sources?  Doesn't this seem like a good way to 
achieve your point c)?  Remember, we don't have to do it perfectly 
right now(or ever) - this is just a good direction to go in.  Every 
step we take is one less step we still have to take, at least, as long 
as we write the step down in public, where others can build on our 
work.

> As long as an article is actually sourced with sources listed, we
> don't need to specify where exactly every fact came from!
Er, you said above that we need to be able to "always produce credible 
sources for every fact when queried".  How exactly would we do that if 
we didn't know "where exactly every fact" was said in a credible 
source?  I totally agree that we don't need to include a source for 
every fact *when we include it* (as I said above, we don't need to be 
perfect), but we both agree that, when queried, we need to be able to 
provide credible sources for every fact.  I don't see how we can do 
that if we don't know where the fact we were queried about came from.  
Maybe I'm missing something here, though; maybe we could "produce 
credible sources" for a specific fact without "specify"ing where the 
fact "came from", although I can't really see how that could work in a 
"meaningful way".

> We are
> trying to build an encyclopedia, and a large part of that goal is
> readability.

As I said above, I totally agree with you.  Readability is critical, 
that's why this suggestion makes only a *one line change* to the 
article text of an article it is used in...  Readability is critical, 
but so is being able to "produce credible sources for every fact when 
queried" - which means writing down those answers in public, which is 
all I am suggesting.

Thanks for the discussion!  (And if I have offended, I apologize in 
advance.)

Jesse Weinstein