Category talk:Works with second-hand provenance

What do you mean by proofing them against the originals ? I mean, who is going to do that ? one single contributor, or the community ?

If you rely on individual contributors, no collective proofreading is possible. I believe the only way to ensure that a text is valid is to make its scans publicly available.

As you might know, at de.wikisource new texts without scans are refused. As a result, nearly all texts on de.ws are backed with scans. I have been trying to promote the same policy at fr.wikisource, but I have encountered resistance to change.

It would be great if the English wikisource could raise its standards at the level of de.wikisource.

ThomasV 17:04, 16 December 2007 (UTC)


 * From my experience with de.wikisource, I would also not like to follow in the path of de.wikisource. It is very unfriendly to new users, and unfriendly to the other wikis.  We have many Category:Works originally in German where the original German text is available and known to be 100% accurate, but can't be placed on de: because of their stringent requirements.


 * Raising the bar doesnt have to mean rejecting works of less quality. The same result can be achieved by tagging pages and encouraging them to be improved.  We have many new contributors because they saw a problem with our text, and they fixed it using their physical copy.


 * John Vandenberg 23:00, 16 December 2007 (UTC)


 * It depends on what you mean by 'the same result'. If you simply mean text accuracy, I agree that you may achieve 100% accuracy without publishing scans. However, only the few ones who have access to a paper edition will be able to tell that the result is accurate. If you mean credibility, I believe that you need to make the scans public.
 * On wikisource, everyone can participate in the proofreading of a text. If anyone can modify a text, as a couterpart, anyone should be able to check if the modifications are valid. We cannot have a system where everyone can modify a text, but only a few ones have access to the paper version, and where all others have no choice but trust these few.
 * It is not just a question of text accuracy. It is a matter of trust.
 * ThomasV 23:27, 16 December 2007 (UTC)

I agree with John. When I started out here in '05, almost everything I posted had poor provenance (and the community expected no better). Gradually my standards have gone up (as has the community's). Judging by my own experience, I think that accepting works of poor provenance may be a price we have to pay to welcome and establish newbs.

I would be strongly opposed to a rule requiring scans for everything uploaded, as I would find that rule too inflexible, for reasons that I shall explain with an example: I have on my bookshelf a copy of Shoobert, Joanne (2006) Western Australian Exploration: Volume One: December 1826 – December 1835: The Letters, Reports & Journals of Exploration and Discovery in Western Australia, which contains scholarly transcriptions of exploration journals, the original sources of which are in some cases difficult to access, and virtually impossible to obtain permission to scan. Shoobert (2006) takes great care in transcription, and in cases where are multiple original sources, efforts have been made to establish the original source of best provenance. In all cases the provenance of the text is stated. As an abundantly footnoted copyright publication, I can't upload scans from this book. Yet the provenance of works that I upload from it are impeccable (albeit second-hand). I can't see any reason why these works should be refused, simply because I can't get a scan of the original.

Hesperian 23:39, 16 December 2007 (UTC)