www.ttahko.net 12 January 2012
Outline 1. The idea of substance causation Overview of arguments for/against substance causation 2. All causation is substance causation Lowe s case for substance causation 3. Evidence of substance causation? No substance causation without substance dualism 2 / 23
1. The idea of substance causation Substance causation the core of agent-causal accounts can be considered a rival to event causation. Where event causation involves an event causing another, agent causation involves a substance bringing about some event. One classic question is whether agent causation reduces to event causation. If we take substance causation seriously, i.e. if agents are substances, then this does not seem to be the case. 3 / 23
On the face of it, agent causation is more mysterious than event causation. It may help to specify that agents are (simple) substances, but questions remain. On this view, agent causation is a form of causation similar to event causation, the difference being the involvement of a substance. There are difficulties though: Can we account for substances rather than their properties as constituents of the causation relation? 4 / 23
I will assume realism (non-reductivism) about causation). It s not necessary to take a stand on dispositional/relational views of causal powers here. However, the indeterministic aspects of substance causation should be clarified. Specifically, how should we understand the lack of determinate probabilities with regard to substance causation? Also, could there be a substance-causal property with a deterministic causal profile? 5 / 23
First, we are not told why the requirement of a determinate probability applies to nondeterministic event causation. If there can be a property that confers on a substance an indeterminate nondeterministic tendency to cause events of a certain general type, then why can there not be a property that confers the same thing on an event that is something's having that property? Second, we are not told why there cannot be a determinate probability in the case of substance causation. If there can be a property that confers on a substance a tendency to cause events of a certain type, then why cannot some such property confer on a substance a determinate nondeterministic causal tendency? (p. 195.) Randolph Clarke, Libertarian Accounts of Free Will. New York: Oxford University Press, 2003. www.ttahko.net 12 January 2012 6 / 23
One possible reply: It is not a requirement for substance causation that it be nondeterministic, and not a requirement for event causation that it must be deterministic. Rather, if we attempt to explain free will in terms of agent causation, then we need some restrictions for substance causation. Another such restriction is that the substance be a rational being, even though this is not a requirement for substance causation per se either. 7 / 23
The possibility of substance causation in general faces some objections as well: Events occur in time, whereas substances are timeless only events are valid causes. Addressed with an integrated account? Further, the way in which events and substances exist in space and time seems quite different. Can a substance affect the probability of a given event occurring? The roll-back argument Voluntary actions become chance events. 8 / 23
2. All causation is substance causation The idea is that only substances can be causes. Causation is uniform. This approach is familiar from Lowe. Has its roots in Aristotle. It may help to address some of the objections against substance causation. However, familiar cases of causation do appear to support event causation. 9 / 23
Lowe (2008: 145) makes an interesting case for the reduction of event causation to substance causation: Event e 1 caused event e 2 just in case there was some substance, s 2, and some manner of acting, F, and some manner of acting, G, such that e 1 consisted in s 1 s Fing and e 2 consisted in s 2 s Ging and s 1, by Fing, caused s 2 to G. The explosion of the bomb caused the collapse of the bridge. Substance causation also features in nature. 10 / 23
A magnetized piece of iron acts upon some nearby iron filings to make them move towards it: it attracts them. Some sulphuric acid acts upon some bits of copper to make them dissolve and transform their constituent atoms into copper ions: it oxidizes them. In all such cases, a substance brings about its effects by acting in some distinctive way. The same substance may act in different ways in different circumstances to bring about different effects. (p. 164.) E. J. Lowe, Personal Agency. Oxford: Oxford University Press, 2008. www.ttahko.net 12 January 2012 11 / 23
So, for Lowe, substance causation is ontologically fundamental. Only substances possess causal powers. This leaves him with the task of explaining probability in particular. Passive vs. active powers. E.g. radioactive decay is not a case of probabilistic causation. Rather, there is an active power in effect, and hence no stimulus is required. There is no causation here at all, just the already fixed probability of radioactive decay. 12 / 23
Lowe thinks that the will is an active (noncausal), spontaneous power. This contrasts with classic agent causation. However, no special type of causation is required. Lowe vs. the roll-back argument: The will is a two-way, rational power. Contrast this with radioactive decay. Because of this, the roll-back does not entail that will is a mere chance event. This has strong Aristotelian connotations: there must be a reason for action. 13 / 23
3. Evidence of substance causation? If Lowe is right, substance causation is all around us in nature. But many think that event causation is a more plausible candidate here. In any case, how do we discover mental causation (needed for free will)? Maybe in terms of how things seem to us when we act? But how do we recognize these actions as featuring substance causation? 14 / 23
I wish to suggest that evidence of substance causation with regard to rational agents requires dualism. If rational agents are mental substances, then it is easy to see how we get evidence of substance causation. In Lowe s picture, substance causation itself is not the problematic aspect, but rather our understanding of rational agents as individual substances. The Aristotelian idea of rational agency is only feasible with mental substances! 15 / 23
The claim finds some support in Cover and Hawthorne (1996). They raise a problem for the combination of agent causation and physicalism. The core of their objection is that on a naturalistic view, human agents are composite objects rather than simple substances. Also, physicalism generally holds mental states to supervene on microphysical properties. A rational agent s free decision, then, supervenes on the agent s microphysical history up to the time of the decision. 16 / 23
Suppose that Jan freely makes a decision. Then, on the suggested view, the human animal that Jan is (and not, for example, that animal's brain) causes that decision. But Jan, like any human agent, could have been whittled down to a brain (kept alive in a vat). And if he had been, then, when he freely made the decision, presumably that brain (all that was left of Jan the human animal) would have caused the decision. Jan's brain in the vat (Jan) could have been intrinsically qualitatively identical to Jan's actual brain (encased in his skull) when he made his decision. The latter lacked an agent-causal power, but the former would have had that power. But fundamental causal powers, it is sometimes claimed, depend only on intrinsic properties. Hence there is a problem with a view that says that the human animal, and not that animal's brain, is the agent cause in a case of free action. (p. 211.) Randolph Clarke, Libertarian Accounts of Free Will. New York: Oxford University Press, 2003. www.ttahko.net 12 January 2012 17 / 23
Cover and Hawthorne conclude that agent causalists should take rational agents to be immaterial substances. Of course, this could also be seen to count against the possibility of substance causation. Clarke considers the key issue to be as follows: To uphold agent causation, we d have to take it as basic, i.e. not consisting of causation by lower level objects that the agent consists of. This is clearly in tension with physicalism. 18 / 23
Consider a further complication: One issue that was mentioned earlier concerned the timelessness of substances as opposed to events. This is supposed to count against substance causation in general. But, I think, less so if substance dualism is true. Provided that there is some support for substance dualism independently (which I m not arguing for!). As Timothy O Connor (2001) puts it, we require an ontology on which persons are enduring, ineliminable substances that are in some robust sense more than the sum of the constituents of their bodies. 19 / 23
The best (only?) way to understand persons as substances of the required type is to distinguish them from inanimate substances. Although complications remain about the role of emergence in this picture (maybe a strong form of property dualism can do the trick?). All this supports the conclusion that the Aristotelian idea of rational agency requires persons to be simple substances. 20 / 23
Lowe builds on this Aristotelian basis and, if he is correct, it cannot be upheld without substance dualism. It does not come as a surprise, then, that Lowe thinks his version of substance causation to require substance dualism. I do not wish to take a stand regarding dualism here. But it seems that a plausible case can be made for the incompatibility of substance causation and physicalism. 21 / 23
Two concluding remarks: 1. I take it that the most natural understanding of substance causation takes all causation to be substance causation. 2. To uphold free will, the very idea of substance causation requires adopting (some) version of substance dualism. 22 / 23
References: Clarke, R., Libertarian Accounts of Free Will. New York: Oxford University Press, 2003. Cover, J. and O'Leary-Hawthorne, J., "Free Agency and Materialism." In J. Jordan and D. Howard-Snyder, eds. Faith, Freedom, and Rationality. London: Rowman and Littlefield, 1996. Lowe, E. J. Personal Agency. Oxford: Oxford University Press, 2008. O Connor, T., Dualist and Agent-Causal Theories. In R. Kane, ed., Oxford Handbook on Free Will. Oxford: Oxford University Press, 2001, pp. 337-55. www.ttahko.net 12 January 2012 23 / 23