It seems to be an established fact that mixed languages are rare, and that most languages can be classified as belonging to some family. And this seems to be true; for example, in the former territories of the Roman empire, either a Romance language developed (e.g. Iberian peninsula, Italy, France, Romania), or it didn't (e.g. Britain). Though the Romance languages all bear traces of pre-Roman features, and other languages like English have Latin influences, there are no truly ambiguous cases.
So are truly mixed languages really so rare, and if so why?
Perhaps they are not so rare, but the dearth of examples is partly due to linguists shoehorning languages into classifications? E.g. if a language is roughly a 60-40 mix** of A and B, linguists would just classify it as part of family A?
Or perhaps there really are social/psychological/language-acquisition or development related reasons that prevent the emergence of mixed languages? (I.e. when two languages come into heavy contact, for whatever reasons one always ends up dominating the other) Has there been any research done into why this is?
**I don't mean, of course, merely in terms of vocabulary. I know this is hard to quantify so please think of the numbers as being used in a figurative sense.