I develop process tracing (PT) for actor-centered and interpretivist studies in a way that allows agency and contingency. For the full chapter, see here.
In the social sciences, cases are generally unique: people’s interpretations, (mis)calculations, understandings, assessments, meaning-making, emotions, creativity, and spontaneity matter for outcomes and so does the socio-institutional context within which agents act.
But that cases and processes, when studied holistically, are unique does not mean we cannot learn broader lessons from them. The question is just: how can we? In this chapter, I suggest to treat mechanisms as akin to Weberian ideal types: abstract constructs that are adduced from multiple concrete, contextually embedded, and largely idiosyncratic instantiations.
In this conception, an instance or instantiation—the occurrence of a mechanism in empirical reality—is always rich, contextually embedded, and case and actor specific. The mechanism itself, then, is our (scholarly) abstraction. It is an analytical construct that defines, in abstract terms, how a given set-up or entity transfers motion in identical or closely similar ways over a variety of situations (partially adopted from McAdam, Tarrow, and Tilly, p.24).
So: whilst cases are always unique and actor and context specific, mechanisms, as ideal types or analytical constructs, are not! How this works in detail, and how actor-centered and interpretivist research may benefit from rethinking what mechanisms are and how they inform our understanding of processes and outcomes? See here.