Title:Monotone and additive Markov process duality

Abstract:This paper develops a systematic treatment of monotonicity-based dualities for Markov processes taking values in partially ordered sets. We show that every Markov process that takes values in a finite partially ordered set and whose generator can be represented in monotone maps has a pathwise dual, which in the special setting of attractive spin systems has been discovered earlier by Gray. This dual simplifies a lot in the special case that the space is a lattice and all monotone maps satisfy an additivity condition. This leads to a unified treatment of several well-known dualities, including Siegmunds dual for processes with a totally ordered state space, duality of additive spin systems, and a duality due to Krone for the two-stage contact process. It is well-known that additive spin systems can be constructed using a graphical representation involving open paths. We show that more generally, every additive Markov process can be formulated in terms of open paths on a suitably chosen underlying space. However, in order for the process and its dual to be representable on the same underlying space, one needs to assume that the state space is a distributive lattice. In the final section, we show how our results can be generalized from finite state spaces to interacting particle systems with finite local state spaces.