I once heard a presentation in which the speaker gave a typology of 'mixed languages', the final type of which was 'languages that should not be', and the only example of this type given was Wutun. This evaluation was based on confusing details about Wutun that had, to that time, appeared in scholarly publications. Based on the descriptions found in such publications as Chen (1982, 1988, 1989), Li (1984, 1986), Wurm (1995), and especially the analysis presented by Lee-Smith and Wurm (1996), Wutun appeared to be a fairly random mixture of Tibetan, Chinese, and Mongolian elements. Randomness in linguistic structure, of course, should not be, and a random mixture of elements from three source languages would be unusual indeed. With this background, the linguist with an interest in the outcomes of language contact will be delighted with Wutun. Janhunen et al. have done a great service to the field by giving us both a reliable outline of the structures of Wutun, and (more importantly from a contact perspective) by identifying historical sources of the various Wutun lexemes and structures that they describe. Under this well informed treatment, Wutun no longer appears to be a random mixture, nor even a particularly odd language.