There are two problems with citing wikipedia. 1) The fluidity of it. When you reference a published book (along with the edition and year, of course), that reference will pretty much always exist. If somebody looks at your paper in 10 years, they could still find the exact text you referenced. When you try to cite something like a wikipedia article, the text you're citing could be gone the next day (as somebody mentioned before). It's like trying to build a castle in the swamp. With most websites worth citing, they archive all of their articles so, like with a book, you could likely find the exact same text in 10 years. 2) Lack of "formal" peer editing. With most books or websites you'd want to cite for a research paper, they are formally editied/reviewed by a peer in the field before they are published. This is typically what gives a source a lot of credability, because it basically means at least two people have agreed that what is being published is correct. Wikipedia has the double sin of being both an encyclopedia source (which is already a no-no for research) AND a website that can be changed pretty much on a dime.