Causation as Causal Influence

130859-Thumbnail Image.png
Description

This thesis proposes an extension of David Lewis's causal influence account of causation, providing a method to calculate the `degrees of causal influence.' By providing a quantitative approach to causal influence, I find that that the influence approach can assess

This thesis proposes an extension of David Lewis's causal influence account of causation, providing a method to calculate the `degrees of causal influence.' By providing a quantitative approach to causal influence, I find that that the influence approach can assess statements that involve causal redundancies, allowing the assessor to attribute primary causal responsibility to the contending cause with a higher net influence value. The causal influence calculation also addresses criticisms towards Lewis's influence account, namely those involving `inert zones' of influence, the use of the term `might,' trumping versus symmetric overdetermination, and Lewis's clause requiring stepwise influence. This thesis also compares the results of causal influence in multiple toy cases including Two Rocks, both the asymmetric and symmetric variants, demonstrating that causal influence overcomes many of the core issues in Lewis's initial counterfactual account of causation. Using the asymmetric Two Rocks variant, this thesis also provides a detailed example of how to use the calculation and a discussion of the calculation's limitations. The main drawbacks of the quantitative method for causal influence seems to be the effort that it requires and issues in finding measurable qualities to compare the similarity/difference between possible worlds. Using the Two Rocks case, however, the causal influence calculation reaches the same conclusions as what Lewis suggests. A primary remaining issue is applying the calculation to instances of causation by omission, however this seems to only be a problem in using the equations rather than a problem within the idea of causal influence itself. Also, there may still be issues in justifying comparative overall similarity. However, this is an issue that both the counterfactual and influence accounts face.

Date Created
2021-05
Agent