Expressivity Landscape for Logics with Probabilistic Interventionist Counterfactuals
For researchers in logic and causality, this work offers foundational insights into the relative expressiveness of counterfactual logics, though it is incremental in extending known team semantics frameworks.
The paper provides complete characterizations of expressivity for logics combining probabilistic dependencies, conditioning, and interventionist counterfactuals within causal multiteam semantics. It shows that the strict tensor disjunction operator is necessary to capture all linear inequalities, establishing a strict hierarchy of expressive power.
Causal multiteam semantics is a framework where probabilistic dependencies arising from data and causation between variables can be formalized together and studied logically. We discover complete characterizations of expressivity for several logics that can express probabilistic statements, conditioning and interventionist counterfactuals. The results characterize the languages in terms of families of linear inequalities and closure conditions that define the corresponding classes of causal multiteams; we find that the strict tensor disjunction, an operator typical of team semantics but absent from the literature on causation, is needed to capture the full class of all linear inequalities. The characterizations yield a strict hierarchy of expressive power and some undefinability results.