PHI403 Lecture 01 — Metascience and Better Science

Opens the PHI403 course (sources/2023-anjum-rocca-phi403-causation-in-science) by asking: what are the non-empirical elements of science? Norms such as "more data is better", "predictive success counts", "the best theory has the greater explanatory power" are themselves normative, not empirical.

The lecture introduces concepts/philosophical-bias (Andersen, Anjum, Rocca 2019, eLife): implicit assumptions of an ontological, epistemological or normative nature that skew hypotheses, designs, and interpretations. Unlike cognitive biases, these cannot be removed — only made explicit.

Scientific methods embed commitments about causation: statistical correlation → Hume's regularity theory; comparative studies → difference-making (Lewis); experiments → Woodward's manipulability. EBM's explicit evidence hierarchy formalises this into a ranking.

The handout closes with the claim that philosophy is the critical friend of science: "philosophy-free science is simply not an option". Scientists need to know which assumptions they are buying into — this is the normative remit of the whole course.

Connections

Back-link: sources/2023-anjum-rocca-phi403-causation-in-science. Concepts: concepts/philosophical-bias · concepts/evidence-hierarchy · concepts/causation · concepts/regularity-theory-of-causation · concepts/interventionist-theory-of-causation.