In maxheld83/pensieve: Tools for the Scientific Study of Operant Subjectivity

How, in spite of it all, can we understand one another? Both Q Methodology and the theory of Deliberative Democracy grapple with this, perhaps defining, challenge in our time of ideological strife. As outgrowths of American Pragmatism -- especially the proto-deliberative theorist and Stephenson influence John Dewey -- both enterprises also share conceptual roots. They both reject radical constructivism (e.g. Glasersfeld 1995), and presume that meaning can be "conscired" (Stephenson 1978), but that such mutual understanding is always contingent and provisional (Habermas 1984). Q's emphasis on "referential" subjectivity is also mirrored by the emphasis on authentic expression in deliberative practice.

This common outlook has been noted by some researchers, including especially Dryzek and Niemeyer (2007), who suggest Q as a method to measure the quality of deliberative fora. For deliberative theory, operationalisation remains a thorny problem: measures of deliberative quality must not be substantive, lest they collapse procedural prescription and substantive theory of justice, but entirely procedural standards risk becoming empty and formulaic. Conventional, often deductive survey measures are often too closed for necessarily open-ended deliberation, while purely qualitative approaches scale poorly and easily import researchers biases. As an abductive approach, with both qualitative and quantitative Q methodology appears to fit the bill, as Dryzek and Niemeyer have shown.

In this paper, we present alternative operationalizations for deliberative quality inspired by Dryzek and Niemeyer, but using alternative statistical procedures, including n-dimensional matrix decomposition, a rigorous within-subjects design and a conceptualisation of beliefs, values and preferences borrowed from game theory.

Using data gathered from over 80 participants on the topic on their desired and expected future work lives, as well as from a smaller citizen conference in Germany, we extract, interpret and publish multidimensional patterns of shared subjectivities. Results indicate that, using appropriate items, completing a Q-sort can in itself serve as a deliberative intervention, and that engaging resulting viewpoints may strengthen intellectual humility. Tentative (small-n) results from a rigorous within-subjects design for deliberative treatment effects also show some improvement in viewpoint structuration, as well as greater substantive consistence in line with hopes for deliberative democracy.

We also discuss how to draft and sample items as part of an operationalisation of deliberative discourse.

Q methodology, as perhaps any single intervention, is unlikely to be a panacea for our fractured public debate, but imperfect though it may be, it could just the next best thing to a thoughtful, personal conversation with a friend with whom you vigorously disagree.

maxheld83/pensieve documentation built on Jan. 21, 2020, 9:16 a.m.