24.00: Problems of Philosophy Prof. Sally Haslanger November 7, 2001 Freewill III: Libertarianism I. Recap: Compatibilism v. Incompatibilism Hard determinism is the view that determinism is true and that because of this freedom is an illusion. The libertarian agrees with the hard determinist that freewill is incompatible with determinism, but disagrees about which claim should be rejected. The libertarian holds that we have free will, but this is only because determinism is false free acts occur and are undetermined. The hard determinist and the libertarian are both incompatibilists. Soft determinism, also known as compatibilism, maintains that determinism is true, but also that determinism is compatible with free will. So the compatibilist disagrees with both the libertarian and hard determinist about the compatibility of freewill and determinism. However, the compatibilist and the libertarian agree that we are (sometimes) free; the compatibilist and hard determinist both agree that determinism is true. Here is the picture: &nbs p; &n bsp; INC DET FREE Hard Determinist: &n bsp; Y & nbsp; Y & nbsp; N Soft Determinist: &n bsp; N & nbsp; Y & nbsp; Y Libertarian: &nbs p; Y & nbsp; N & nbsp; Y Compatibilism is appealing, but all the forms we've considered seem to have problems. So we should reconsider the possibility that the incompatibilist is right. And since we've already looked at the hard determinist brand of incompatibilism--that was Holbach--it is time to look at another form of incompatibilism, viz., Libertarianism: we have freewill, and determinism is false. Consider again the freewill dilemma (see also pp. 411-12 of the text): 1. If determinism is true, we can never do other than what we do; so we are not free. 2. If indeterminism is true, then some events--possibly some actions--are random; but if they are random, we are not their authors. So we are not free. 3. Either determinism or indeterminism is true. 4. Therefore, we never act freely. II. Libertarianism A. Chisholm and Agent Causation Of our readings, Chisholm and Kane are the representative libertarians. Both buy the basic incompatibilist argument: i) If x acts freely in performing action A, then x could have done otherwise. ii) If x's performing A is determined, then x could not have done otherwise. iii) So, if all acts are determined, then x never acts freely in performing action A. We saw last time that both (i) and (ii) are controversial. So rather than focusing on the determinism side of the freewill dilemma, let's consider how denying determinism might open the possibility of freewill. Recall that the problem with the idea that indeterminism solves the problem is that if my action is uncaused or random, then it seems no more free than if it is determined by prior events. Random events are as much out of my control as those that are caused by prior conditions. So denying determinism offers little help. However, consider determinism again: DET) Whatever happens is determined by prior events. Chisholm points out that we may deny determinism, i.e. maintain that: IND) Some things that happen are not determined by prior events, without maintaining that the exceptions are uncaused or random. We need only claim that some events are not caused by other events. He suggests [In defending freewill] we must not say that every event involved in the act is caused by some other event; and we must not say that the act is something that is not caused at all. The possibility that remains, therefore, is this: We should say that at least one of the events that are involved in the act is caused, not by any other events, but by something else instead. And this something else can only be the agent the man. (494) How is this supposed to work? Consider Aristotle's example: "...a staff moves a stone, and is moved by a hand, which is moved by a person." (Physics VII: 256a6-8). According to Chisholm, the moving of the staff and the hand are both events, and the hand-moving event causes the staff-moving event in the ordinary way (Chisholm refers to the causation between events as transeunt causation). Moreover, there is plausibly a brain event that causes the hand-moving event; again this is ordinary event causation. But, Chisholm maintains, if the action of moving the hand is free, then the brain event that causes the hand-moving event is not caused by another event, but is caused by the person or agent. (Chisholm calls such agent causation immanent causation.) I cause the neurons in my brain to fire in the way necessary to cause my hand to move, and there is no event which determines the firing of those neurons. So, according to Chisholm: I act freely in performing an action A iff (i) I, rather than some prior event, am the cause of A (or of some event that directly causes A), and (ii) I could have performed an action other than A. Incompatibility: If determinism is true I couldn't have performed an action other than A, so I wouldn't satisfy (ii). Freedom: I am the source of my actions; they are not determined by prior events. Objections: Many have rejected the idea of agent causation because it appears mysterious and not fully compatible with our conception of the physical universe. For example, as Chisholm himself points out, if the firing of the hand-moving neurons is not caused by a prior event, the agent must bring it about that the firing occurs without there being anything happening in the agent. As Chisholm explains, "...the agent himself cannot be said to have undergone any change or produced any other event (such as "an act of will" or the like) which brought [the action] about." (496). This seems to leave agent causation somewhat mysterious, e.g., what distinguishes a case in which there is an agent and a random neural firing, and an agent causing a neural firing? Chisholm answers this question by simply pointing out that there is something mysterious even about event causation: what distinguishes an event A followed by an event B, and an event A causing event B? [Can you find any differences between the two sorts of cases? Are they equally mysterious?] But others find this response unconvincing. How, for example, does pointing to the agent as cause explain the event if there is nothing that the agent does to bring it about? B. Kane and Indeterministic Causation Robert Kane is sympathetic to Chisholm's libertarianism, but unsympathetic to agent causation. Kane begins by pointing out that freedom does not require each an every free action to be an exception to determinism, for actions that are determined by our freely formed characters are themselves free. For example, if I am a good person and always choose to do the right thing, my choices are free as long as my having developed a good character wasn't entirely determined, but was something I am responsible for. So for Kane, our focus should be on the potential freedom of "character building" acts or what he calls "self-forming actions" (SFA's, p. 505). Kane maintains that in moments of inner conflict when we are torn between two courses of action, the tension within us creates a neural indeterminacy. Faced with such a conflict, my brain could take either of two different paths, and prior states of the brain don't determine which path it will take. However, suppose I decide to take one course of action or another. If my brain proceeds down the path leading to that course of action, it can be said that I caused it to do so (so the resulting actions are not random ). Nonetheless, it is also the case that I wasn't determined to take that course because it was indeterminate that my brain would proceed in that way. Kane's argument depends crucially on the premise that: It may be the case that A causes B, even if A does not determine B. He offers a number of examples to support this premise: suppose a man slams his fist on a glass table in an argument. It is indeterminate whether the force of the blow will break the table. However, if he breaks the table it would be correct to say that he is responsible for breaking the table, and that he caused the table to break. So the slam of the fist caused the table to break even if it did not determine that the table would break. The analogy with agency would seem to work as follows. Let's use a famous case described by Jean Paul Sartre: a young man during WWII is torn between joining the resistance, or staying home with his ailing mother. In a case of such inner conflict it is indeterminate whether his brain will take path P (join) or path P* (stay home). Suppose he decides to join. If he does then pursue that choice and undertake to join, then his decision caused him to do so, even if it was indeterminate whether his brain would proceed down P. As a result, his actions are not random: they are caused but indeterminate. So, on Kane's view: I act freely iff my action is either a free self-forming action, or it is the product of a character that was formed by free self-forming actions. I act freely in performing a self-forming action A iff i) I could have performed an action other than A, i.e., my choosing A was not determined by prior events; ii) I decided to perform A; iii) My decision to perform A caused me to perform (or undertake to perform) A. Indeterminacy: If determinism is true I couldn't have chosen an action other than A, so I wouldn't satisfy (i). Freedom: My decision in favor of A caused me to perform (or undertake to perform) A. Kane's view seems to have an advantage over Chisholm's because it does not postulate agent causation: my decision (a datable event, presumably) was the cause of my actions (via neural pathways). Objections: ? How are we to understand the decision to perform A, e.g., the decision to join the resistance? By hypothesis, the decision was the cause of the (self-forming) action, but was the decision itself free? If in order to perform A freely, I have to freely decide to perform A, then does this generate an infinite regress? But if I didn't decide to perform A freely, then in what sense are my choice to join and my resulting actions free? ? It isn't exactly clear what Kane's view is. In particular, how does he understand the effect of my decision on the probabilities of the outcomes? If my action is free, must my decision have had an effect on the probabilities? How significant an effect? Note that in the case of the glass table, it would seem that the probability of the table's breaking at the moment in question is close to 0 if the man does not strike it, and is close to 1 if he strikes it with enough force; this significant change in the probabilities may be a basis for saying that the man caused the table to break. However, it would seem that on the picture Kane sets up, each neural pathway has (approximately) a .5 chance of dominating (it is their competition that supposedly gives rise to the indeterminacy). In such cases it is unclear that the agent's decision can be said to cause the result. (If a bubble has a 50/50 chance of bursting at a given moment, whether or not I touch it, and I touch it and it bursts, can I be said to cause it to burst?) ? What is the mechanism by which the agent's decision is having an impact on the probabilities of the different neural pathways and resulting courses of action? Is the agent's decision itself a neural event? Kane suggests, When neuroscientists described [the choice] in physio-chemical terms, all they would get are indeterministic chaotic processes with probabilistic outcomes. In short, if described from a physical perspective alone, free will looks like chance. But the physical description is not the only one to be considered. The indeterministic chaotic process is also, experientially considered, the agent's effort of will something agent is doing. (510) But it is surely difficult to understand how one and the same thing can at the same time be an indeterministic chaotic physical process and my effort of will. Where do we stand? Do you think that either Hard Determinism, Compatibilism, or Libertarianism offers a viable position?