In the early days of my research career my interest in legal reasoning led me to study logical accounts of norms as applied to (human or artificial) agents. This includes the study of deontic logic, the logic of such modalities as 'obligatory', 'permitted' and 'forbidden'. Deontic logic has important applications in computer science, since computer systems, and their interaction with other systems and with humans can be regarded as norm-governed.
In [Prakken 1994] I explained my view on formalizing defeasible deontic reasoning: just combine any deontic logic you like with any nonmonotonic logic you like ... . With Marek Sergot I studied the logic of `contrary-to-duty obligations' (e.g. "You should keep your promises, but if you don't keep them, pay for damages caused"). Our point of departure was that such obligations should not be confused with defeasible or prima facie obligations, since contrary-to-duty obligations are not invalidated by violating them. See [Prakken & Sergot 1996] and [Prakken & Sergot 1997].