Although Arabic is usually treated as one language, in fact it is not. Arabic is a continuum of languages that share the basic morphological component (the stem), and differ in many of the affixes and some important syntactic structures and lexical content.

In this talk, I address the question of how and why the computational tools built for Arabic work well only for one of the dialects of Arabic (Modern Standard Arabic), but fail in real world situations. I also discuss some solutions to the problem.