Temporal Horizons in the Study of Turkish Politics: Prevalence of Non-Causal Description and seemingly “Global Warming” Type of Causality

Abstract

In this article, I critically evaluate the causal and temporal dimension of social scientific studies focusing on Turkish politics. A very important and yet often neglected aspect of social scientific analysis involves the temporal dimension of causal processes. The temporal dimension of causal processes has direct consequences for operationalization and measurement, and hence it is an essential component of research design. Does the dependent variable (outcome) of interest unfold over the short term or the long term? Do the hypothesized independent variables (causes) unfold over the short term or the long term? Paul Pierson (2004) provided a classification of four types of causality based on the temporal dimension of causes and outcomes using metaphors of natural disasters: tornado, earthquake, meteorite, and global warming. Operationalization and measurement of long term causes and outcomes pose a major challenge, compounded by the challenges of periodization of causes and effects. Unfortunately, a large proportion of the studies of Turkish politics do not have a clearly discernible independent variable (cause) to begin with, and they are thus better characterized as works of "non-causal description." Moreover, many of the studies of Turkish politics tend to imply, but rarely state explicitly, a global warming type of causality (long term cause and long term outcome), which necessitates focusing even more intensively on such challenges of measurement and periodization. Yet the operationalization of the key (dependent and independent) variables is often missing even in articles published in reputable academic journals of Turkish politics and society. In the spirit of constructive criticism, I suggest a number of guidelines for research design in order to address the problems of causality and temporality discussed in this article, including awareness of multi-temporal equifinality.