I’ll be giving a talk for the UC Santa Cruz Department of Linguistics on 19 November 2021:
The dark side of causation: inefficacious and shadowy influences
Bridget Copley, SFL (CNRS/Paris 8)
Formal semantics inherited a formulation of causation (from Lewis (1974) via Dowty (1979)) in which a causal statement entails the occurrence of the result. This formulation, however, runs up immediately against widespread linguistic phenomena (e.g. the imperfective paradox and non-culminating accomplishments) where , arguably, a causal statement should not entail the occurrence of the result. Popular solutions to this problem follow Dowty in introducing inertia worlds. An ontologically simpler solution, however, allows for a causal relation to be present without being “efficacious” (Copley 2005, Copley & Harley 2015). I present such a solution here, based on a causal model framework (Pearl, 2000*), and show how its use gives us a meaning for intentionally that accounts for the Knobe Effect (Knobe, 2003). In cases where an influence is inefficacious and no other influence in the model explains the actual result, a “shadowy” influence must be added to the model; I argue (following Copley & Kagan 2021) that such shadowy influences arise in Russian negated perfective sentences.
*Those who are not familiar with causal models might be interested to read chapter 1 of Pearl & Mackenzie 2018 (preprint online) for a gentle introduction.