Fifth Narrative Science seminar: Uses of counterfactual narrative in political science and the parallels between philosophies of narrative and mechanism

Author: Dominic Berry
 
New year, new seminar speakers! This term we started with two papers that appeared extremely different on their surface, but which ended up having a great deal to say to one another. The first ‘Counterfactual Narrative in Political Science’ was given by Sharon Crasnow, and the second ‘Mechanism and Narrative’ was given by Phyllis Kirstin Illari. As ever we are very grateful to our speakers and the audience members. The following post offers a short summary of the talks and some of the discussion from the question period. It is not meant to offer a complete account of everything discussed. An audio recording of the session can be made available, please do contact us directly if you are interested.
 
Sharon Crasnow (Norco College)
Counterfactual Narrative in Political Science

Crasnow’s talk built on her existing interests in the uses of narrative knowledge in political science. The focus here though was on political scientists’ use of counterfactual reasoning, and in their case based work. These approaches are sometimes thought particularly useful when political scientists have limited access to data, exploring the space of possibilities by adoption of a counterfactual approach. Other times they are associated with a comparative approach between experimental groups and control groups. In contrast, Crasnow finds narrative and counterfactual reasoning operating far more broadly. For instance, being able to show that each given part of an overall causal story (often delivered as individual cases) can fit together and work more or less in unison, offers a narrative form of argumentation that makes essential epistemic contributions to the field.
 
As an illustrative and widely known example Crasnow made use of Frank P. Harvey’s controversial Explaining the Iraq War (2011). In a work that addresses the question of whether or not the US would have instigated the Iraq War without George W. Bush as President, we find a number of different uses and appeals to counterfactual argument. Sometimes this is to explore a space of possibilities, to see if evidence is available that would rule out potential causal accounts, other times, to identify ‘hinge’ events where clear alternative paths were available. The significance of individual cases in political science are largely built out of explicit or implicit assessments of their contingency, ramifications as hinge events, or as something necessary resulting from path dependence. These aspects of the work of political science therefore clearly evidence narrative knowledge.
 
This talk provoked a range of different questions. One participant wanted to know more about how an overall narrative that fits together different cases, relates to (or does not relate to) the cross-comparative work mentioned at the start of the talk. Is it the case, for instance, that providing an overall account is a way to intervene on such variance-based research? Crasnow suggested that the making of a good mechanism might direct you back to the comparative cases, telling you new things to look for. Others wanted to ask about the status of counterfactual reasoning in political science, hinting at the ways in which these approaches compete with historical analyses. Crasnow responded that though large counterfactual analyses such as Harvey’s are somewhat rare, the forms of reasoning she had discussed are widespread.
 
​Phyllis Kirstin Illari (UCL)
Mechanism and Narrative

---

​In this talk Illari reflected on her ongoing research into the philosophy of mechanism in light of the narrative science proposal. From the outset she highlighted that many of the things she was going to argue about what mechanisms are, and how they are evidenced, resonated with Crasnow’s talk, particularly regarding the fitting of things together. For instance, Illari’s preferred definition of mechanism, which identifies objects, the things they do, and how they are organised to bring effects about, was very similar to the previous talks’ emphasis on identifying hinge events and arguments about how they can be fitted together to produce particular outcomes.
 
Some of the particular work Illari focussed on included her work on evidenced based medicine, and more recent attention to ‘Infosec’ (information security). One idea from literary theory that struck Illari as being worthy of closer attention was that of Genette on points of view, and the distance between reader/viewer that different narrative arrangements can produce. When thinking about evaluation of different kinds of evidence for instance, it is essential to think about different kinds of audience, and the purposes to which the evidence is to be put. This concern is not only about how to make information and arguments legible, but also - in the case of Infosec - to know how to disguise knowledge or keep it secret (certain kinds of knowledge about how international computer networks actually work has a protected legal status).
 
In a recent co-authored article, focussed on how Infosec experts explain and evidence the primary mechanisms by which system attacks are carried out, she emphasised how this work requires assessments of what features of the overall network are stable enough to be worth considering (given how often technologies and infrastructures can change) and what can be openly debated (given legal restrictions). One way in which they can deal with these constraints is through the use of visualisations which provide an overall narrative, picking out key features, without necessarily going into details about causal connections. Indeed Illari and her collaborator, Jonathan Spring, were themselves compelled to translate some of the materials produced within Infosec into a diagram (created by Jonathan and which can be found in their article).
 
During the question period we focussed on some key elements. The first question focussed on the features of some of the Infosec diagrams, some of which seemed to be both ordered and spatially connected, while others were only ordered. Illari responded that these diagrams were often most useful for generating narratives that explain the ways in which an attack has been carried out. Others asked about the complexity of mechanisms in relation to the general and the particular: in the cases discussed, are practitioners moving from the particular to the general, or the particular to the particular? Illari responded that it is commonly both, and also admitted that though some philosophers like to find strong distinctions between claims in the general and in the particular, she finds it hard to separate the two. This kind of issue may well rest on the question a practitioner is pursuing.
 
Upcoming seminar
The next seminar will feature Ivan Flis on ‘Narrating an unfinished science: Scientific psychology in late-twentieth century textbooks’ and Adrian Currie on ‘History is Peculiar’. Please see our events page for details as to time and place.
 
And remember, get in touch if there is something you would like to know more about!

---

​