Note: The following essay is a commentary on a conversation that took place in the American Journal of Sociology on the use of Verstehen (empathic understanding of human behaviour) between Watts (2014) and Turco et al (2017).
The goal of sociological study is to examine structural processes in a society consisting of human beings. Hence, a sociologist attempts to separate herself from the society in which she belongs and attempts to examine it “objectively.” In this procedure, it is inevitable that a sociologist ends up becoming the object of her own study.
The objectivity criterion in sociology assumes that a “self” can be separated from the “identity.” More specifically, this criterion assumes that a sociologist-self can be separated from other roles that the sociologist plays as a human. One senses that this is the impossible task that Watts (2014) wishes sociologists to achieve in order to have their work qualify as “scientific.” (p.324-5)
In describing the ‘frame problem’ of undertaking the exercise of verstehen, Watts says that “when we deploy our mental simulation apparatus to project ourselves into a particular situation… details are simply “filled in” by the unconscious mind, which draws on its assembled library of images, emotions, stereotypes, role models, and other stylized memorabilia.” As a result of this, “we treat the imagined details in exactly the same way as the specified details…and we calibrate our imagined reaction accordingly.” (p.328; emphasis is mine)
For Watts, the frame problem is not a “problem” in practice because “it is applied under conditions that are especially forgiving for prediction, namely concrete, immediate circumstances that are experienced repeatedly.” In addition to this, “the practice of theorizing about others using simulation is aided greatly by real-time feedback; thus, even when our mental simulations do make erroneous predictions, these errors can be corrected quickly.” (p.330-1) But in the process of theorizing, Watts notices that the further one goes from the context and the actors under study, the greater is the potential for error in making predictions. Yet, this shortcoming does not seem convincing in disqualifying the use of verstehen in teleological study of the society.
There is an underlying contradiction between Watts’ call for objectivity and his desire to not distance away from “a familiar actor and a familiar situation” (p.331). Attempts at objectivity require the distancing of the self from the minute specifications of the context and the actors, in order to be able to see the structural factors governing the social processes more clearly. For example, to be able to see suicide as a social fact, Emile Durkheim (2006) attempts to step away from the specificities of those individuals who commit suicide along with the context of each suicide, to look instead at the aggregate social process that can help understand variations in suicide rates across different religious groups.
As Turco and Zuckerman (2017) correctly observe, one finds in Watts’ position on verstehen an odd tension (p.1274). At times, Watts rejects the idea that empathetic explanation can have any teleological importance while at other times, Watts believes that it can lead towards a causal explanation.
Comparative and historical analysis, a method of data-gathering and interpretation, would not qualify as scientific under the criterion laid out by Watts. This is because, in addition to relying on the chronology of events to prove causality, it also uses narratives that show the causal link between two observations. In many cases, showing causality is not the goal of the research. Instead, comparative-historical research can be undertaken with the goal of understanding the correlation of different observations. If the offer of causal explanation acted as the sole entry-point for the study to be considered scientific, descriptive work that is prevalent not only in social but also natural sciences would lose its value. Hence, causal theory and scientific theory should not be collapsed into each other, a fallacy that Watts finds himself committing.
What Watts is calling for is a thick theory that makes strong claims for causality. For him, this can be achieved through the use of experimental methods and out-of-sample testing. (p.343) Yet, the thicker the theory, the less is the generalizability. Experimental methods, while having strong internal validity, have weak external validity. This is because, by the nature of its design, “experimental research tends to produce discrete findings that cannot simply be “added together” in any way” in order to shed light on macro processes (Mahoney and Thelen 2015, p.11). To trade-off generating thick theory (with strong causal claims but less generalizability) for thin theory relying on verstehen (and weak causal claims) for increased generalizability should not make the study become ‘less scientific.’ If the contrary were true, as Watts tries to argue in his text, then it is unclear where one would draw the line to qualify as scientific, given the spectrum of works that lie between the ideal types of purely empathetic and purely causal explanations.
To conclude, if the criterion dictating the entry of a socio-scientific work into “science” requires an escape of the human-subject from the human-self, it should push us towards questioning not the methods of our knowledge production but rather what is indeed meant by “science.”
Durkheim, Emile. 2006 . On suicide. London: Penguin.
Mahoney, James and Kathleen Thelen. 2015. Advances in Comparative-Historical analysis. Cambridge: Cambridge University Press.
Turco, Catherine J. and Ezra W. Zuckerman. 2017. “Verstehen for Sociology: Comment on Watts.” American Journal of Sociology 122(4):1272–91.
Watts, Duncan J. 2014. “Common Sense and Sociological Explanations.” American Journal of Sociology 120(2): 313-51