We extend the existing encoding of abstract argumentation frameworks in DL-PA (Dynamic Logic of Propositional Assignments) in order to capture different formalisms for arguing with qualitative forms of uncertainty. More in particular, we encode the main reasoning tasks of (rich) incomplete argumentation frameworks and control argumentation frameworks. After that, and inspired by our encoding, we define and study a new class of structures that are shown to be maximally expressive: constrained incomplete argumentation frameworks.