On 12/18/06, jayjg <jayjg99(a)gmail.com> wrote: That's exactly right. Wikipedia cannot become the secondary source. Sarah

On 12/19/06, Steve Bennett <stevagewp(a)gmail.com> wrote: Steve, it's not a question of reaching for the "pinnacle of perfection." It isn't difficult to write articles that conform to the content policies. The articles are better for it too, and of more use to the reader, who can check whether we've used decent sources and whether we've been true to them. That's of more use than an anonymous Wikipedian adding his own opinion based on his own studies that may be correct or may be nonsense. Even very occasional editors are coming to see the importance of using good sources correctly and every day I see articles improving because of it. We're getting there. Sarah

On 12/20/06, Steve Bennett <stevagewp(a)gmail.com> wrote: I can't quantify it in any checkable way, Steve, but it's my strong impression. Every day I see editors producing good sources without being asked to, and that wasn't so common this time last year. When I click on "random article," I'm seeing a significantly higher percentage of better written, better sourced pages. When I look at new pages, more of them seem to have sources than used to be the case; usually only one source, but it's a start.

Sarah wrote: Which I understand to include such things as legal databases. In theory at least, anyone in the general public can do what is necessary to acquire access to such a database. We certainly don't require that sources are immediately available to anyone. -Rich

On 12/18/06, Sarah <slimvirgin(a)gmail.com> wrote: To reiterate, there are two entirely separate issues in this thread: 1) whether you can draw a fair conclusion from the results presented by a database search, for the purposes of encyclopedic referencing and verification. 2) whether members of the general public (whoever they might be) should be able to easily (whatever that means) access any source cited in the encyclopedia for verification purposes. My take on these questions: 1) No. Databases, in general, change and update their contents regularly; do not necessarily cover everything published in a given field (especially not a field as enormous as law); and cannot be trusted to properly index their holdings (and if something isn't indexed, you aren't finding it). On top of that, as has been pointed out the original poster's search may be incomplete or otherwise faulty. I wouldn't trust myself to do an absolutely comprehensive database search for these purposes, and I'm what passes for a trained professional these days; unless Zero's a fabulous legal researcher of some sort I'm not inclined to take his word for it. It's more than a barrier of having access to the database -- to recreate the search you also have to know *how* to search, which is not simple or easy in most legal databases. For all of these reasons, I wouldn't cite a database search in any context, except perhaps in a very specific type of academic writing that was focussed on research technique -- otherwise, it's like saying "because google said so" (which granted we use to determine lots of things around here, but hopefully not as an actual fact in an article). The search results would be a fine note to put on the talk page and should be taken as such. 2) As others have pointed out, expecting the public at large around the world to have access to all of our sources is pretty silly. There's the language barrier, for one thing; I am sure there are great sources in most all of the world's languages that I'll never be able to read, which shouldn't stop people from citing them in the relevant articles and relevant Wikipedias. There's also the expense of most major reference works and journals, the rarity of certain older and academic materials, the incomprehensiblity of certain academic topics to those who don't know the field (like Charles' math example, where I'm just going to have to take his word for it), etc. This is why libraries exist in the first place: to collect together things that may be hard to come by otherwise. Now, if you have a CHOICE between an obscure work and a better known and better accessible one, and otherwise they're the same, then by all means cite the better accessible one. Better yet, cite both. Cite the ISO Standard and the easily-available IEEE paper that summarizes it. In other words: sure, I've personally handled a 1728 [[Cyclopaedia]] and can happily tell you that the long subtitle is correct, but it's only because the work is digitized and discussed in other sources like Britannica that I expect most people in the world to ever be able to verify that. -- phoebe (brassratgirl)

On 12/19/06, David Gerard <dgerard(a)gmail.com> wrote: We use writers as sources, not databases and libraries. If Zero wants to say that all legal opinion says X, he should find a secondary source who confirms that, rather than relying on his own, possibly faulty, search of a database, and then calling the database the source (even though there are no documents in the database that say "all legal opinion says X"). What he means, in fact, is that HE is the source, and that he used the database to do his research, and that's what makes it OR. It's a separable issue, but it boils down to the same thing, viz. that we publish the published views of third parties (what they believe is in the database, for example), and not the views of Wikipedians. Sarah