Causal Determinism (CD) is the view that all of the laws of nature, together with all of the facts of the world at a certain time, imply one unique future for the world. CD has led to two major trains of thought concerning moral responsibility. Incompatiblists are those who think that if CD is true, then no one is morally responsible for their actions (i.e. CD is incompatible with moral responsibility). Compatiblists feel that one can still be morally responsible even if CD is true.
Many incompatiblists have based their views of incompatibilism on the intuitive and extremely prevalent notion that we need alternate possibilities in order to be morally responsible for our actions. Harry Frankfurt argued against this requirement of alternate possibilities. His argument suggests that any line of reasoning for incompatibilism that involves the idea that responsibility requires alternate possibilities will fail. Frankfurt argument for that responsibility does not require alternate possibilities has convinced many moral theorists, but not all.
Robert Kane has recently advanced a different incompatiblist condition of responsibility. The condition centers on agents having ultimate control over their actions. Interestingly, an implication of this condition (along with other relevant facts) is that alternate possibilities are required for moral responsibility. However, there seem to be some serious problems with Kane’s argument, or so I want to argue.
In order to fully understand Kane’s position, one must first understand an argument put forth by Peter van Inwagen which is designed to show that CD rules out alternate possibilities. The next step for many incompatiblists is to say that if van Inwagen’s argument is sound, then we are not morally responsible for our actions because alternate possibilities are needed for moral responsibility. After this skeptical argument has been established, Frankfurt’s criticism of it can be understood. Kane’s proposal can then be given and related back to alternate possibilities and Frankfurt.
Causal Determinism and Alternate Possibilities
Some have argued that if CD is true, then we lack alternate possibilities with respect to our actions. That is, if CD is true, then no one could do other that what she in fact does. In An Essay on Free Will, Peter van Inwagen argued for this view in the following way:
If determinism is true, then our acts are the consequences of the laws of nature and events in the remote past. But it is not up to us what went on before we were born, and neither is it up to us what the laws of nature are. Therefore, the consequences of these things (including our present acts) are not up to us (van Inwagen, p 16).
According to one plausible interpretation, Van Inwagen’s argument is as follows: We cannot prevent events in the past from occurring as we cannot change the past, nor can we prevent the laws of nature from obtaining as we have no power over the laws of nature. Furthermore, if we cannot prevent one event, and we cannot prevent that event from leading to (or causing, or implying, or entailing) a second event, then we cannot prevent the second event. Call this principle the Transfer Principle. Suppose, for example, that the following proposition, P, is true: The President of the United States goes to bed at 11 p.m. this Friday. If CD is true, then the laws of nature and the facts of the past entail P. As stated above, no one can prevent the facts of the past and no can prevent the laws of nature from obtaining. Furthermore, given the Transfer Principle, no one can prevent the laws of nature and the facts of the past from entailing P. Therefore, no one can prevent P from obtaining, not even the President. Van Inwagen argued that if CD is true, then analogous things would be true for all events, and therefore no one would have freedom to do otherwise.
This argument gives us a very strong reason to think that if CD is true, then we could never do other than what we in fact do. Some have claimed that unless we have alternate possibilities, we are not morally responsible for what we do. Harry Frankfurt captured this idea in a principle he called the "Principle of Alternate Possibilities" (PAP). PAP says "a person is morally responsible for what he has done only if he could have done otherwise" (Frankfurt p 829). The conclusion of van Inwagen’s argument, taken together with PAP, generates the conclusion that no one is ever morally responsible for what one does if CD is true. Many people take this to be a very troubling conclusion, one that has led some to question the veracity of PAP.
Frankfurt tried to show that PAP was false. The example that Frankfurt used to suggest the falsity of PAP has come to be known as a "Frankfurt-type case." The following is an example of a Frankfurt-type case: Imagine that while Bob is driving, he sees Paul starting to cross the road ahead of him. Bob decides that he does not want to slow down for Paul, and continues on at the same speed. Paul, who is unable to get out of the way of Bob’s car, is run over. As a result, Paul is killed instantly.
Assuming that Bob is a rational and sane human being, it seems clear-cut that he is morally responsible for what he did. Now imagine that, unbeknownst to Bob, the road that he was travelling on was covered in ice. Due to the ice, even if Bob had tried to stop the car from hitting Paul, he would not have been able to and Paul would have been killed anyway. Let us further imagine that Bob does not try to stop the car. It seems that we would still want to hold Bob responsible for what he did even though he would have run Paul over regardless of whether or not he hit his brakes. If Bob had hit his brakes, then one might argue that he would not be morally responsible for what he did. However, the fact of the matter is that Bob did not hit his brakes. It seems, then, that Bob is morally responsible for what he did even though he could not have done otherwise.
As suggested above, if CD is true, then it seems that there would not be any alternate possibilities. That is, we could not do other than what we in fact do. However, the Frankfurt-type case seems to give very good reason to doubt the necessity of alternate possibilities for moral responsibility. In other words, it seems Frankfurt-type cases cast strong doubt on PAP. If Frankfurt is correct, and PAP is actually false, then much of the motivation for the view that CD is incompatible with moral responsibility is undercut. If incompatiblists are to maintain their ideals, then a new condition of moral responsibility will have to be advanced.
Kane, Ultimate Responsibility, and Alternate Possibilities
Robert Kane proposes a necessary condition of moral responsibility that he calls "Ultimate Responsibility" (UR). Kane feels that alternate possibilities alone are not sufficient to answer the question of whether or not CD is compatible with moral responsibility (Kane(a) p 130). He feels the notion of UR was what fueled many of the incompatiblist intuitions. According to Kane,
…to be ultimately responsible for an action, the agent must be responsible for anything that is a sufficient reason (condition, cause or motive) for the action’s occurring. If, for example, a choice issues from, and can be sufficiently explained by, an agent’s character and motives (together with background conditions), then to be ultimately responsible for the choice, the agent must be at least in part responsible by virtue of past voluntary choices or actions for having the character and motives he or she now has (Kane(a) p 130).
According to Kane, in order for us to be responsible for our actions, we must be at least partly responsible for the sufficient causes that move us to action. One way to understand the notion of UR is to think of it in terms of control. In order to be ultimately responsible for our actions, we must be in ultimate control of our actions. To have ultimate control over our actions means that we must have at least some control over the causes that are sufficient to move us to action, or must have had control over some action in the past that is now sufficient to move us to action. We need not have control over every action in order to be responsible. Suppose, for example, someone makes a decision and she is responsible for that decision. Assume that she had at least some control over the sufficient causes that moved her to action. Suppose further that making that decision is sufficient to cause the person to make some future decision. She would still be responsible for the future decision because she was responsible for the sufficient causes of that future decision.
This notion of control shows why Kane is an incompatiblist. If CD is true, then all of our actions are the result of (sufficiently caused by) events that have happened is the distant past and the laws of nature. There is no possible way that anybody could have control over these things, regardless of what sense of "control" is used. Therefore, no specific argument of control needs to be proposed in order to show that UR is an incompatiblist condition for moral responsibility.
Reflection on Kane’s condition of UR suggests a type of "regress" problem. We must first understand the nature of this regress problem in order to understand Kane’s notion of free choice. Once we understand this notion of free choice, we can fully understand the condition of UR that Kane has put forth.
When we perform what we consider to be free actions or decisions, we typically act in accordance with our thoughts, desires, intellect, knowledge, and many other mental aspects, which I will put under the umbrella term of "character." One major problem that many philosophers have had is to give an account of how we get the character we have. One intuitive response is that our character is the result of the prior decisions we make, and the subsequent actions we perform. When we choose a certain course of action, there are certain consequences that will result from that course of action. These consequences affect how we think and how we act. In other words, they affect our character.
Assume that we have a certain character, Z. Presumably, it was caused by the consequences of a certain course of action. Call the course of action and its consequences z. The picture we have so far is that Z was caused by z. (I will denote such relationships as z Æ Z.) Presumably, there would have to be some cause for z. As stated above, the reason we make the decisions and choose the courses of action we do is because of our character. Therefore, there must be some state of character that "existed" prior to Z and was distinct from Z. Call this state or aspect of character Y. Our picture so far is that Y Æ z Æ Z. What then, caused Y? Presumably, it was caused by some chosen course of action (and its consequences) that must have occurred prior to and is distinct from z. Call this course of action and its consequences y. At this point, we have y Æ Y Æ z Æ Z (Strawson, pp150-151).
Kane proposed that there is a way to stop this regression. According to Kane, at least some of our free mental actions are "efforts of will" which Kane calls "Self-Forming Actions" (SFAs). Such actions, according to Kane, stop the regress problem (Kane(b) p 107). Kane claims that free our decisions are nondeterministically caused. In order to understand nondeterministic causation, it is useful to compare it to deterministic causation. With deterministic causation, if one event, A, causes another event, B, then every time that A occurs B will also occur. One way to think of this (though not the only way) is to say that if A occurs, there is a 100 percent chance that B will occur. To say that A nondeterministically causes B, means that if A occurs, then B can occur, but it might not. In terms of probability, to say that A nondeterministically causes B is to say that A causes B and there is less than a 100 percent chance of B occurring if A occurs.
According to Kane, there are times in our lives when we have a genuine conflict over which course of action to take. We are "torn between competing visions of what we should do or become" (Kane(b) p 108). Consider an example of a businesswoman who, on her way to an extremely important meeting, witnesses a mugging in an alley. The woman is torn between different possible courses of actions. She could ignore what she saw and continue on to her business meeting, or she could help the victim in the alley. The woman would like to do both, but cannot. She must choose to do one or the other (Kane(c) p 68).
In the example, the woman wants both to go to the meeting and to help the victim. The fact that the woman wants to help the victim prevents her from simply going to the meeting, and the fact that she wants to go the meeting prevents her from simply helping the victim. Prior to choosing one course of action over the other, the final decision made is caused by an effort of will of the woman, such an effort is a mental event. This effort of will nondeterministically causes the decision the woman makes. That is, although the effort of will ended in the woman choosing a certain course of action, it could have ended with the woman choosing the other course of action. In other words, what choice will actually be made is "indeterminate." An effort of will of this sort is an example of an SFA.
Kane goes on to say that in order for a mental event such as a choice to be an SFA, it must satisfy the conditions of plural voluntariness, plural rationality, and plural control. These plurality conditions are a requirement for a choice to be an SFA. Choices are plural voluntary if they are both voluntary (not compulsive or coerced), and there is at least one alternative choice the agent could make which would also be voluntary (Kane(d) pp 109-110). A choice is plural rational if it is both rational, and there is at least one alternative choice the agent could make that would also be a rational choice (Kane(d) p 137). For an agent to be in plural control over her actions, the agent must be able to "bring about any one of the options…at will or voluntarily at the time. …it is to be able to do whatever you will…to do among a set of options, whenever you will to do it, for the reasons you will do to it, and in such a manner that neither your doing it nor willing to do it was coerced or compelled" (Kane(d) p 110).
By meeting these conditions, it seems that alternate possibilities are required for a free choice to be an SFA. If a choice is an SFA, then it must meet the above plurality conditions. However, those conditions, taken together, require that the agent who is acting must have been able to act other than the way she did. The conditions state that if a choice is an SFA, then there must have been some other choice the agent could have made which would have also been rational, voluntary, and such that the agent would have been in control of it. If this is true, then one must have been able to make some choice other than the one in fact made. Consequently, if one performs an SFA, then one must have had alternate possibilities.
It seems, then, that if Kane’s theory is true, then, contrary to Frankfurt, alternate possibilities are required for moral responsibility. Hence, it might be argued that Frankfurt-Type examples do not establish what they were designed to establish, namely, the falsity of PAP. This is pretty hard to accept given the intuitive pull of Frankfurt-Type cases. However, upon further examination, it seems that Kane’s own account of when a decision is free does not meet the UR condition. If this is true, then there are grounds to reject this account, and if it is rejected, it does not stand in the way of Frankfurt-Type cases undermining PAP.
Consider the first time a person is truly torn between multiple courses of action. Suppose Johnny is a little boy who has just been given two lollipops. Johnny really wants to keep both lollipops for himself, but he knows that his sister also wants one. Johnny knows, too, that his sister will be very sad if she does not get a lollipop and very happy if she does get one. Johnny both wants to keep the lollipops for himself and he wants to give the second lollipop to his sister. This is the first time in his life that Johnny has ever had such indecision over which course of action to take. After much deliberation, Johnny finally decides to keep the second lollipop for himself. This choice is the first time that any of Johnny’s choices have been an SFA.
On Kane’s theory, prior to performing that first SFA, Johnny has never been morally responsible for what he did. Presumably, all of Johnny’s actions up to this point have been caused by factors over which Johnny has no control (e.g. environment, upbringing, genetic make-up, etc.). According to Kane, Johnny now would be responsible because the decision he made was a result of an SFA. However, if we look at the events prior to Johnny’s first SFA, it seems that Johnny would not be responsible for his decision to keep the lollipop despite its being caused by an SFA.
Johnny was torn between two courses of action. The fact that Johnny wanted to follow both courses of action would be due to the same deterministic processes that prevented Johnny from being morally responsible prior to his first SFA. This is because so far in Johnny’s life he has not exhibited control over his actions, and therefore none of his prior choices have been SFAs. For his entire life up to this point, all of Johnny’s choices have been the result of his upbringing, environment, genetics, and so forth. There were no other factors that could have lead Johnny to action. As this effort of will was Johnny’s first self-forming choice, there was nothing else that could have made Johnny choose the courses of action that he did prior to the lollipop decision. However, Johnny is obviously not ultimately responsible for things such as his environment, upbringing, genetic make-up, and so forth because there is no way that Johnny could have had control over these things.
If Johnny was not responsible for wanting to keep the lollipop, and Johnny was not responsible for wanting to give the lollipop to his sister, then it seems that Johnny would not be responsible for whatever course of action he ends up taking. Kane could respond by saying that Johnny is responsible because he is responsible for the decision-making process (this being the only part of the equation not yet mentioned). However, Kane’s own account of free action seems to imply that Johnny’s first SFA was caused by some indeterminacy, something which Johnny could have no control over. A nondeterministic process such as this is analogous to flipping a coin and seeing which side comes up. One could argue that Johnny would still be responsible because he was responsible for flipping the coin. However, the fact that Johnny "flipped the nondeterministic coin" was also do to factors over which Johnny had no control. As stated above, Johnny is not responsible for having a character such that he would feel conflict in this given situation. To put it simply, Johnny is not responsible for wanting either of the two options, and he is not responsible for the nondeterministic process that actually causes him to act in one way or the other.
It appears that Johnny would not be ultimately responsible for his first SFA as he did not do it freely. If Johnny did not perform this SFA freely, there seems to be little reason to suppose that the decision the SFA caused was free either. It seems, then, that Kane’s own theory cannot meet the UR condition that Kane himself sets forth. Of course, this does not mean that the UR condition is false. It simply means that, given Kane’s theory, as it stands now, cannot meet the UR condition.
Support for Ultimate Responsibility
The purpose of this section is not to give a complete argument to support UR. Such an argument is beyond the scope of this paper. This section is simply designed to give some intuitive support for UR. In order to give this support, I will present a situation in which it is intuitive to say that the agent is not morally responsible for what he does despite the fact the he acts rationally, intelligently, and as a result of his character.
Imagine that Bill wants some diamonds. However, for obvious reasons Bill does not want to run the risk of getting caught while steeling these diamonds. Therefore, Bill kidnaps Paul and performs special neurosurgery on Paul. As a result of this surgery, Paul’s character is completely altered. Paul now has a desire to steal the diamonds and give them to Bill. However, Paul is not simply a mindless drone. He has reservations about acting on this desire. After all, Paul does not want to go to jail either. After much careful deliberation, Paul rationally decides that he should go ahead and steal the diamonds and give them to Bill.
In the example, our intuitions are such that we would probably not hold Paul morally responsible for what he did. There is some excusing factor for Paul. It is plausible to assume that this excusing factor is due to the fact that UR has not been met. Recall that to be ultimately responsible, the agent in question "…must be responsible for anything that is a sufficient reason (condition, cause, or motive) for the action’s occurring." In the above example, there are sufficient reasons "outside" of Paul that caused him to act the way he did.
Debate will (almost assuredly) arise over whether or not the UR condition can be satisfied. Most compatiblists would probably argue that the UR condition is far too strict and cannot be met. There are two responses to this criticism. First, just because there is, as of yet, no coherent account of freedom that could meet the UR condition, it does not follow that no account can be given. For example, if some plausible account of Agent Causation could be given, it would almost certainly fulfill the UR requirement. Second, even if no account of freedom could be given to meet the UR condition, this does not mean that UR is false. It simply means that we would not be able to assign praise or blame (moral responsibility) to people when they act. I grant that this is not an especially appealing option. People seem intuitively to want to be able to appraise agents (in terms of moral responsibility) for the acts they do. However, there also seem to be strong intuitions that assigning praise or blame incorrectly is somehow "wrong" or "unjust."
For example, assume that a house was burning down and Bob did nothing whatsoever to help. He did not try to save anyone, he did not try to put out the fire, he just stood there like a mindless lump. Assuming that Bob is responsible for his actions, and that we are "justified" in assigning praise or blame, it seems that we would not want to praise Bob for help to put out the fire when he did nothing to warrant that praise. If UR is correct and cannot be met, then any act would be analogous to this example. There simply would be no praise or blame that is merited. It seems to follow from our intuitions that we would not want to assign praise or blame in the first place (assuming, of course, that UR cannot be met).
Conclusion
Moral responsibility does not require our having alternate possibilities available. The Frankfurt-type case laid out in this paper strongly suggests this. However, Frankfurt-types cases do not address the view that control over the factors leading to action is necessary for moral responsibility. Not having such control seems to rule out Ultimate Responsibility. Lack of UR seems to be an intuitive and very plausible reason for saying that a person would not be responsible for her actions. If causal determinism is true, then it does not seem possible that anyone would have that control. By utilizing the principle of Ultimate Responsibility put forth by Kane, it seems that causal determinism and moral responsibility are incompatible.
To further back this claim, we relied on common sense examples in which we intuitively think the agent was not responsible. We intuitively would not want to assign any blame to an agent for doing something when she is not ultimately responsible for doing it. When someone pulls the trigger on a gun and kills someone, we do not blame the gun (or the bullet) for the murder. If someone were to build a sophisticated robot and then program it to rob a bank, we would not assign blame to the robot. One reason we do not blame the gun or the robot is because neither had any control over what they did. There were other factors over which neither the gun nor the robot had control that made them commit the crimes they did.
However, we have a tendency to think that we are not like the gun or the robot. We like to think that we have certain cognitive abilities that would differentiate us from both the gun and the robot. However, if casual determinism is true, then in many respects, we are like the gun and the robot. The gun did not have control over what it did. This is partly because there was a person (presumably) who was operating, or controlling, the gun. The gun was just "doing" the only thing it could do. Likewise, in a perfectly determinate universe, the very nature of the universe is controlling us and causing us to do what we do.
Kane’s own theory of freedom does not seem to meet the UR condition. His notion of Self-Forming Actions seems to be inherently flawed. Kane tries to use SFAs to evade the problem of regression concerning the origins of our act. However, Kane’s theory does not overcome the regress problem. Kane’s theory does not give and adequate account of how agents can gain responsibility-grounding control.
Kane also seems to rely upon alternate possibilities in order to maintain moral responsibility. It seems, however, that Kane need not make any such appeal. As long as the focus is on UR, then whether or not an agent has alternate possibilities when she acts seems irrelevant for moral responsibility.
References