Don't let search engines index your "Page History" or "Recent Revisions" pages. This is important, and is included in our AntiSpamRecommendations, but in fact good wiki software already prevents this by default, by emitting "noindex nofollow" for the robots META tag, or by including a robot.txt definition. Most recent versions of wiki and blog software have the robots meta tags. But you should check that your software is doing it in case the developers don't know about the reasons yet.
If old revisions are indexed by search engines, then all AntiSpamUserSolutions are rendered ineffective. Reverting spam isn't very helpful when search engines still find the spam when they index your past versions.
Another important reason to hide your kept pages from search engine robots is that spammers use search engines to find spammable wikis. And they do this by entering spammy URLs as the search term. The idea is simple and effective: if there is a page out there that contains spam, then it could contain even more spam, so I better find it. You can read more about this here.
The majority of wiki spammers are finding good places to spam, by using google to search for existing spam. This is the SpamAttractsSpamEffect. Since this measure is hiding old spam (spam which your users have removed) from google, spammers are less likely to find your wiki this way.
In fact you get two benefits from preventing indexing of kept pages: 1) less spammers will find you and 2) spammers won't benefit from spamming your wiki.
By implementing / activating this measure, you are making a contribution to the wiki-wide anti-spam effort. If all wikis do this, then we are significantly reducing the effectiveness of the whole wiki spamming activity.
If you don't do this, then your users could become demoralised. Their spam cleaning work will not really be hurting the spammers. It really is the bare-minimum anti-spam tactic which every wiki should include.
However users will want their contributions to be accessible by search engines, so it is better to at least allow indexing of your main article pages. You could put NOFOLLOW on the article pages. This means google will list your pages, but will not follow the links…
However this will also mean google doesn't follow internal links (links between wiki pages), which will heavily impact your rankings. You could instead just put the rel=nofollow attribute on all external links…
This is getting better, in fact this is good default for wiki software to adopt (MediaWiki does this), but it is still not a recommendation, since it means legitimate links are not credited with PageRank. When links are good and relevant, we want google to boost their ranking.
All of these variations are eroding the usefulness of wikis.