5.3 KiB
| title | section-id | sort | author | created | modified | language | description |
|---|---|---|---|---|---|---|---|
| Causation | metaphysics | 120 | Prof. James Okafor | 2026-02-07 09:00 | 2026-03-14 11:00 | en | Theories of causation from Hume to counterfactual and mechanistic accounts. |
Causation
Causation is one of the most fundamental concepts in our understanding of the world. We explain events by citing their causes; we hold people responsible for what they cause; science aims at causal understanding. Yet the analysis of causation has proved persistently difficult.
Hume’s Analysis
As discussed in the chapter on empiricism, Hume subjected causation to empiricist scrutiny and found no impression of necessary connection. His regularity theory holds that ‘A causes B’ means nothing more than:
- A and B are contiguous in space and time
- A precedes B
- All As are followed by Bs (constant conjunction)
This is a reductive analysis: causation is nothing over and above constant conjunction. There is no causal cement binding events together; there is only the pattern.
Problems with Regularity Theory
- Accidental regularities: Night follows day constantly, but day does not cause night. Constant conjunction alone does not distinguish causal from accidental regularities.
- Causal asymmetry: Barometers fall before storms, and there is constant conjunction between falling barometers and storms. But the barometer does not cause the storm; both have a common cause (air pressure changes).
- Absence of mechanism: The theory says nothing about how A brings about B. Causation seems to require some mechanism, not mere correlation.
Counterfactual Theory
David Lewis proposed a counterfactual theory: A causes B if and only if, had A not occurred, B would not have occurred. This is analysed in terms of possible worlds: A counterfactually depends on B if in the nearest possible worlds where A does not occur, B does not occur either.
Counterfactual analysis handles the barometer case: if the barometer had not fallen, the storm would still have occurred (the air pressure was the relevant factor). And it captures the temporal asymmetry: causes precede effects because the relevant counterfactuals run forward in time.
Problems with Counterfactual Theory
- Preemption: If A would cause E, but B would have caused E had A not occurred, and A fires first, then A causes E. But had A not occurred, E would still have occurred (via B). Counterfactual dependence fails.
- Overdetermination: Two independently sufficient causes simultaneously bring about an effect. Neither is necessary for the effect, so neither is a cause by the counterfactual criterion—but both intuitively are.
- Transitivity: Counterfactual causation is transitive in a way that generates unintuitive results. If A causes B and B causes C, then A causes C—even in cases where the causal chain passes through something that seems irrelevant.
Mechanistic Theories
Mechanistic theorists hold that causation consists in the transmission of something—energy, momentum, a conserved quantity—through a continuous process from cause to effect. Wesley Salmon’s process theory distinguished causal processes (which transmit marks) from pseudoprocesses (which do not).
The moving spot of light from a rotating projector is not a causal process: the spot at one position does not cause the spot at the next. The light beam itself is a causal process: energy is transmitted continuously.
Mechanistic theories handle many cases well but face difficulties with causation by absence (‘the failure to water the plants caused their death’) and distal causation.
Interventionist Theories
Judea Pearl and James Woodward have developed interventionist or manipulationist accounts. A is a cause of B if manipulating A (holding other things fixed) systematically changes B. This is close to the experimental intuition: causes are what we can intervene on to produce effects.
Interventionism is influential in the philosophy of science and in discussions of causation in social science and medicine. Critics ask whether it is circular: specifying what counts as an ‘intervention’ requires appealing to causal concepts.
Causation and Laws
Some philosophers ground causation in laws of nature: A causes B because there is a law connecting the type of event A is to the type of event B is. Laws, on this view, are not just regularities but something stronger—nomic necessities.
What makes something a law rather than an accidental regularity? The Ramsey-Lewis view identifies laws with the generalisations that appear in the best systematisation of the facts—the theory that best balances strength and simplicity. Necessitarian views hold that laws involve genuine necessity in nature, not just in our best theories.
Summary
- Hume’s regularity theory reduces causation to constant conjunction; it fails to distinguish causal from accidental regularities
- Counterfactual theories analyse causation as counterfactual dependence; they face problems from preemption and overdetermination
- Mechanistic theories require the transmission of a conserved quantity; they struggle with causation by absence
- Interventionist theories identify causes as what can be manipulated to produce effects; they are potentially circular
- The relationship between causation and laws of nature is itself disputed