From Here to There and Back: A Vocabulary for Processual Generalizations

• In Event: Rethinking Generalization

Abstract

Using in-depth case study methods such as Process Tracing, scholars are able to learn about the causal processes that link causes with outcomes within the studied case. While learning about how things worked in a particular case is important, many scholars want to draw broader lessons from a studied case that are relevant for other cases. Others might be interested in drawing on broader lessons as inspiration for what might be going on in particular cases.

Unfortunately, the existing literature focuses almost exclusively on the selection of cases as a strategy for generalization. While this is important, for all but the simplest causal phenomena, the process(es) linking a cause (or set of causes) and outcome together will be at least slightly different. How can we then claim the processes are similar in a typical case and other cases? To be able to claim similarity also requires that we simplify away the case-specific details of a process through theoretical abstraction. However, beyond the well-known intension/extension distinction, we lack a conceptual language for the abstraction of process theories. Additionally, the existing literature lacks a good methodological language for applying the ‘generalized’ lessons to understand other particular cases (‘here’ again).

The article first introduces a methodological language for abstraction at the processual level that involves raising/lower the level of granularity of a process theory (i.e. what links causes and outcomes together). Theoretical abstraction (or simplification) at the level of process can occur on a number of parameters: how many parts are included (i.e. the proportion of events that are matched in the process theory), the sequencing of the process, how specific are the actors and activities that compose each part, the temporal and spatial scope of the process, and the frequency and density of interactions between actors. Common to all of these is that we are dealing with functional equivalence. Taken as a whole, a highly simplified causal process theory would focus on a few key episodes. Further, in each of these episodes, very abstract terms might be used for the actors and the activities they perform. For instance, an entrepreneur might table a proposal, but in different cases what this looks like might be very different as long as it plays a functionally equivalent role.

The article then develops how to use processual abstractions to move from here to there and back again. Making processual generalizations from a case requires moving from ‘here’ to ‘there’ (target population) through the theoretical abstraction of a causal process from the studied case, as well as identification of the contextual conditions required for it to work as it did in the case (aka scope conditions). The second part of learning from cases involves moving from ‘there’ to ‘here’, where we apply the abstracted processual lessons to understand one or more unstudied cases. Moving from ‘there’ to ‘here’ involves what can be termed processual ‘concretization’, in which the ‘generalized’ (i.e. abstracted) processual theory is translated into a more detailed, case-specific processual theory.

Authors

• Derek Beach, University of Aarhus

• Gabriela Camacho, Humboldt Universität zu Berlin

• Sandrino Smeets, Radboud University Nijmegen