Resolving the Raven Paradox: Simple Random Sampling, Stratified Random Sampling, and Inference to the Best Explanation

Download Edit this record How to cite View on PhilPapers
Abstract
We object to standard, simple random sampling resolutions of the raven paradox on the grounds that they relevantly diverge from scientific practice. In response, we develop a stratified random sampling model. It provides a better fit and apparently rehabilitates simple random sampling resolutions as legitimate idealizations of that practice. However, neither simple nor stratified models fare well with a second concern, the objection from potential bias. In response, we develop a third model on which we systematically check kinds of ways in which disconfirming cases—non-black ravens—might be caused. This provides a novel resolution of the paradox that handles both objections. Suggestively, this third approach resembles Inference to the Best Explanation (IBE) and relates confirmation of the generalization to confirmation of an associated law. We give it an objective Bayesian formalization and discuss the compatibility of Bayesianism and IBE.
PhilPapers/Archive ID
WARRTR-3
Upload history
Archival date: 2020-03-28
View other versions
Added to PP index
2020-03-28

Total views
51 ( #42,848 of 51,740 )

Recent downloads (6 months)
51 ( #10,771 of 51,740 )

How can I increase my downloads?

Downloads since first upload
This graph includes both downloads from PhilArchive and clicks on external links on PhilPapers.