Event-based Theory
- Hume →
- Mill, Mackie (conditional)
- Reichenbach, Suppes (probabilistic)
- Lewis (counterfactual)
- Pearl, Woodward (interventional)
Basic claims
- Motivation: there is no power, disposition, or necessity…
- Against Aristotle’s notion of efficient cause (object/essence based)
- Hume’s criticism of necessity and power
- (Logical) positivists’ criticism of unobservables
- cf. A quote from Psilos
- Claims:
- Causality is a lawlike relationship between events
- [ball hitting a window] — [breaking of the window]
- [smoking] — [developing cancer]
- → Event & Law based
- What are events?
- Events, state of affairs, facts
Contentions:
- What counts as causal events?
- Granularity (Russell, see Kutach)
- Absence?
- Mental events?
- What are causal relationships/laws?
- Regularity theory: logical or probabilistic
- Counterfactual theory: counterfactual
- Bayes nets: functional
- Are causal relationship objective or subjective?
- Subjective: Hume, Kant(?), Pearl(?), etc…
- See Psillos
Mathematical Modeling by Causal Graphs
- Events: nodes (random variables)
- Causal relationships: arrows
- Questions
- How does it relate to probability? → Causal Markov Condition, Faithfulness Condition, etc.
- How to deal with intervention/counterfactual reasoning? → The graph surgery operation (do-calculus)
Refs