Chapter III: Critical Realism
3.3 Retroductive Inference
Retrductive inference involves moving from the level of observation and lived experience to postulate about the underlying structures and mechanisms that account for the phenomena involved (Mingers, 2003). Retroduction has been defined as ‘a mode of analysis in which events are studied with respect to what may have, must have, or could have caused them. In short, it means asking why events have happened in the way they did’
(Olsen and Morgan, 2004: 25).
Retroductive inference is similar to the abductive inference in philosophy (logic analysis). To describe the process behind the abductive inference, the American philosopher Charles S. Peirce worked out the concept of it. Peirce describes abductive inference on the one hand as a mode of inference with a defined logical form comparable to induction and deduction, and on the other hand as a more fundamental aspect of all perception, of all observation of reality. It involves what has been called redescription or recontextualization (Danermark, 2001: 89). Peirce used a method used by detectives in solving crimes, to exemplify the logic of abductive inference. The activities of various people, observations at the place of the crime and statements in interviews, are interpreted and gain a significance within the frame of an overall hypothesis about how the crime may have been committed. An experienced, insightful and creative detective is able to recontextualize what she knows about the crime (the clues) within the frame of different possible scenarios of how the crime could have been committed (ibid: 91).
By means of abductive inference we recontextualize and reinterpret something as something else, understanding it within the frame of a totally difference context. In this way we introduce new ideas of how individual phenomena are part of the structure and internal relations. We now focus on retroductive inference, which can be described as a mode of inference, by which we try to arrive at what is basically characteristic and constitutive of these structures (ibid: 96).
21
The general formation of abductive inference is as follows:
The surprising fact, C, is observed; 4
But, if A were true, C would be a matter of course, hence, there is reason to believe A is true.
We will later employ the above to form our narratives in Chapter five after the root causes are identified.
From the critical realist perspective, the best explanations are those that are identified as having the greatest explanatory power. Mainstream economics employs deductive or inductive inference to construct theories. Whereas, the deductive method requires researchers to operationalize their process, specifying what must to be done to measure a concept. That is, the researcher applying the deductive method begins with a theory and then derives one or more hypotheses from it for testing. Next, the researcher defines the variables in each hypothesis and the operations to be used to measure them in a specific, observable terms (Rubin, 2011). For example, some studies select explaining factors such as CPI, economic growth rate, and export value in a specific period and apply the deductive method to find out their correlation and regularities with the studied event.
This selection is subjective and the conclusion only suggests empirical regularities and statistical correlations of the data. Through the creation of operational variables, there is a tendency to measure and collect data only on what can actually be observed.
Deductive method bases its research process under numerous hypotheses and assumes those hypotheses will not change from start to the end. The contemporary critic on this, Nancy Cartwright, also a contemporary philosopher known for her contribution to causation, criticizes the covering law of explanation on the grounds that the alleged laws on which the explanation relies (according to the covering law) are not true, but rather generalizations that hold under special, typically ideal, conditions. These ceteris paribus
4 A is a hypothesis or an argument proposed by the author.
22
laws5, she claims, will not do the work required. Ceteris paribus generalizations, read literally without the ‘ceteris paribus’ modifier, are false. They are not only false, but held by us to be false; and there is no ground in the covering law picture for false laws to explain anything. On the other hand, with the modifier the ceteris paribus generalizations may be true, but they cover only those few cases where the conditions are right (Cartwright, 1983:
45).
On the other hand, the inductive method starts from data collection, after which the data are analyzed to see if any patterns emerge that suggest primary relationships between variables. Through a process of gathering substantial data and statistical analyses, it attempts to establish patterns, consistencies and meanings. The inductive method is problematic because we can never be certain that a recurring event will continue to occur.
Contrary to these limitations, however, the retroductive inference provides a more legitimate ground for critical realism. In an open system like the society we live in, it is problematic that we measure and collect data only on what can actually be observed, ruling out subjective and intangible evidence. It is also problematic that we take event regularities as the causation of an event because human beings are living in a transforming society, in that every individual behaves in accordance to the ever-changing roles they play. Unlike a moving object that is unanimated and follows laws of the nature, human beings think and react to external conditions. They may not behave in the same way when certain conditions emerge. The retroductive inference does not impose explanatory boundaries and hence, is suitable for social events based on critical realists.
Event regularities of mainstream economics have many problems; for example, is the knowledge appropriate enough to be universal? Also, can the logical answers created by mainstream economics be the fundamental causes of a complicated issue of salary stagnation that involved not only economic, but political and social issues. Critical Realism
5 Ceteris paribus means “holding other things constant.” This term is most widely used in economics and finance as a shorthand indication of the effect of one economic variable on another, keeping all other variables constant that could render an effect.
23
is a new research methodology in social science, in which the relationship of “if A, then B”
mentioned above is considered superficial, and researchers are encouraged to explore the underlying forces, the root causes, which generate the considered effect B. It contends that the society and individuals cannot be reduced to each other; casual mechanisms are stratified and structured, so to find out the causal relationship, we have to start with those strata and collect information and conclusions from empirical and deeper domains to answer fundamentally how an event happens. To this end, rather than deduction and induction, the retroductive inference is the appropriate method utilized by critical realism.
Despite the above, explanations derived from the retroductive inferences are always potentially open to revision. Unlike empiricism that considers the research results finalized, accepted theories in critical realism may be rejected in favor of the more convincing alternative, if the latter is better able to explain a phenomenon than the current one does (Sayer, 2002).
24