SOPhiA 2021

Salzburgiense Concilium Omnibus Philosophis Analyticis

SOPhiA ToolsDE-pageEN-page

Programme - Talk

How to determine the optimal detailedness of probabilistic explanations: a defence of proportionality
(Philosophy of Science, )

Explanations in the special sciences like economics often refer to ``high-level'' features such as GDP. But are such high-level explanations better than lower-level ones providing more details, e.g., about individual consumption behaviour? What is the optimal detailedness at which to explain? In response, Woodward (2010) claims that a less detailed causal explanation is better if it is more proportional, meaning it cites a cause conveying enough and only relevant information about its explained effect.

Proportionality has been discussed primarily in deterministic contexts. Yet, the special sciences often rely on probabilistic causal relationships in their explanations. Can proportionality capture the optimal detailedness of probabilistic explanations? Kinney (2018) argues it cannot. In this talk, I defend proportionality as indicating the optimal detailedness of probabilistic explanations against Kinney's criticism.

First, I present Kinney's counterexample, purporting to illustrate that sometimes a more proportional probabilistic explanation is worse than a less proportional one. I show how this judgement stems from Kinney's criterion used to determine which probabilistic explanation is more proportional (an adaptation of Woodward (2010)).

Second, I argue that Kinney's criterion fails to determine the proportionality of probabilistic explanations. To do so, I revisit the concept of proportionality and argue that in perfectly proportional explanations there is a one-to-one mapping between the cause and the effect variable. In such a perfectly proportional explanation, I claim, a cause provides full information about its effect and vice versa. Based on these ideas, I argue for a necessary and sufficient condition for determining the proportionality of probabilistic explanations. Kinney's criterion does not satisfy this condition. Using a new criterion that I propose, I lastly show that the intuitively better explanation is the more proportional one in Kinney's example.

Chair: Daniela Schuster
Time: 18:50-19:20, 09 September 2021 (Thursday)
Location: SR 1.006

Ina Jäntgen 
(Erasmus University Rotterdam, Netherlands)

Ina Jäntgen is currently finishing her master's degree in philosophy and economics at the Erasmus Institute for Philosophy and Economics in Rotterdam. Before, she studied philosophy and economics at the University of Freiburg. In October 2021, she will start her PhD in Philosophy at the University of Cambridge. Her primary research interests lie in general philosophy of science and philosophy of the social sciences/ economics, especially in causation and scientific explanation. She also has interests in decision theory and formal epistemology.

Testability and Meaning deco