Reading the postings on setting up an article approval system for Wikipedia, I had the following ideas (most elements of it have already been mentioned by others).

Even though experts are no guarantee for quality, the credibility of Wikipedia will grow as articles have been approved by field experts. Therefore, there should be a mechanism to nominate an article for approval by an expert (or more, if these are available); this should in a way similar to the old voting system, and should IMO only be possible by logged in users.

Experts are defined in two ways: field and nationality. This has the benefit that an article on, say, a German mathematician is not only verified by a mathematical expert, but also by somebody with knowledge of Germany, being able to check the spelling of names, cities and check local information (the mathematician's influence in Germany may be much bigger than in the rest of the world). While experts will know a great deal about their fields (mathematics, biology, philosophy), they probably have some smaller areas in which they're particularly knowledgable (logic, evolution theory, Nietzsche). When an article is submitted for approval, the submitted should indicate which general field the article is in, potentially specify the sub-field, and a country (if applicable). For example, the article on the German mathematician will be have : mathematics, sub-field trigonometry (f.e.), country Germany. This will cause the page to show up on the approval lists of the experts that have subscribed to one of these areas. An expert can than indicate if he wishes to review the article or not. If there are multiple experts willing to review, they can decide what approach to take (the country expert may f.e. let the math-man do most work and only check for names or so). After review, the experts may put the document back without approving it. In that case, they should indicate which elements are still missing in their opinion, and point out other deficiencies (probably in the talk page). Alternatively, the expert can decide to fix the current deficiencies and put the article back as approved, or may simply give the article approval without change.

After approval, the article should NOT be locked for editing, but the page should indicate that "this article is approved by an expert" or "a previous version of this article was approved by an expert", in case the article was edited since the last approval.

As for the admittance of experts, I think expert status should not just be given to people with lot of titles or the like. Actually, I think most of the current frequent editors, or at least the sysops, would already qualify as an expert in my view. Having finished a higher-level education on a specific subject, or even still studying is already some qualification. A proven record of quality edits to Wikipedia articles on the subject should be enough to get expert-status.

These are my ideas. They're still a bit of loose sand, and several problems are still not addressed (such as the issue of how we actually give people expert status), but I'd still like hear your reactions.

Jeroen Heijmans