ABSTRACT
Though many authors recognize that mechanisms can involve stochasticity, there’s been little discussion about the nature of the stochastic relationships in mechanisms. I try to elucidate some issues that arise with stochastic mechanisms and propose new ways of thinking about them. I’ll focus mainly but not exclusively on what I call recurrent mechanisms—token mechanisms that operate in a similar manner at different times, or mechanism types that have many instances that operate in a similar manner. Scientists’ uses of “mechanism” seem to refer primarily to these mechanisms, which have also been the main focus of recent discussions of mechanistic explanation by philosophers.
