Traditional metaphysicians have always viewed causation as a modal notion. They held that causes necessitate their effects. Hume, as an empiricist, attacked this idea because one cannot empirically observe these modal sequences. He argued that causation is constant conjunction or regularity of succession. Kant, a defender of the traditional approach, insists that causation is an a priori concept.
Hume’s Account of Causation
There is a view in the modal approach that there is a power, energy or special force that is the cause. Hume holds that all our ideas are observable from experience itself. Causality is this case is observing the first billiard ball striking a second, causing it to move. Basically we get impressions of temporal succession and spatial contiguity. We see one event succeeding another in a narrowly circumscribed region of space; not power, energy or necessitation.
The same thing applies to thoughts. We have a mental event that is followed by a physical event. I’m tired, but I remembered I need to do the dishes, so I get up and do them.
The Response to Hume
One type of critique is that Hume’s approach is too broad. Critics point out that there are noncausal patterns that regularly follow by events of another kind. A few simple example is,
The day is followed by the night.
A plant in London blows at a whistle at 9am for the start of the work day. Immediately after the whistle, workers in another plant in Manchester start work.
These examples show that the regularity of these events fit Hume’s approach, but these aren’t causal events. The night isn’t caused by the day. The plant in Manchester doesn’t start work at 9am because of a whistle going off at 9am in London, they just start the day at that time.
A second critique, is that there are difficulties for the regularity theory since it is presented by singular causal judgments (first ball strikes the second causing it to move). The problem here is that we have to say some individual episode causes some individual episode. There are many examples where there are many moving parts that don’t make it easy to identify any sort of causal event. For example, the assassination of Ferdinand caused the First World War, but it is unlikely that one is going to identify this with Hume’s approach.
The big shot that Human makes at the traditional metaphysicians is that power, energy and necessitation are not observable. Those sympathetic with the traditional approach will concede this point. Kant, for example, views that these things are just a priori or just innate/intuition.
The more popular approach is with Hume versus modal. These people do believe in the Hume approach, but believe that they need to supplement the analysis to fortify it against counter examples.
If one believes in the regularity approach they due owe a reply to the objections in the previous section. Neo-Humean’s view the second objection as the least serious. They view that our judgment is an assimilation of the case before us to some familiar pattern where we already have the regularity. For example, Ferdinand is assassinated and that leads to military tensions between two countries. This leads to other players picking sides, etc.
The other difficulty presented is seeing regularity of events that are not causal. John Stewart Mill tried to remedy by saying that causal regularity if unconditional; it holds no matter what - it does not hinge on conditions that need not obtain. In the case of the day/night case, fails this test because Mill claims that the sun could extinguish or the earth could cease rotating in the appropriate way.
Another strategy that has originated was proposed by the Logical Positivists. There view is what distinguishes a genuinely causal succession of events from merely accidental correlation is that the former has a status of a law of nature or is derivable from something that has that status.
Another approach is the notions of necessary condition and sufficient condition. J.L. Mackie is the most influential of this approach. Mackie thinks there is a background context that he calls a causal field. It’s basically the context in which we take our cause to operate. An example will help illustrate this better.
There is a house fire and the fire investigator determines that the cause of the fire is an electrical short circuit.
Mackie claims that we aren’t being told that the necessary condition of the house fire; they know there are a variety of factors that resulted in the house catching fire at the time it did. Nor is the electrical short circuit necessary to start the fire. There were other factors such as the flammable rags near by and the sprinklers being defective. The cause of the fire is neither a necessary nor sufficient condition. It is an indispensable component in a large bundle of factors - which were all present.
Mackie refers to this as INUS conditions, which are the first letters from the following sentence: What experts are calling the cause, then, is an insufficient, but necessary component in a bundle of factors that was unnecessary, but sufficient for the occurrence of the fire.
David Lewis also has a counter factual approach, where he views it as such: if c were not to occur, e would not occur. This fits into his possible worlds approach.