Free Will – A Critique of Wo lf ’ s Reason View Introduction Free will is intertwined with concepts which underpin a range of human activities – philosophical and non - philosophical alike. Free will is of concern to social and political theorists studying human freedom ; psychologists studying addiction ; legal theorists studying criminal liability At the centre of it all , it is moral responsibility itself which hangs in the balance. There are various accounts of the nature of free will. Some of these accounts are sceptical and entail that the existence of free will is a metaphysical ly impossib le Some suggest that the existence of free will requires the world to be a very certain way but, luckily for us, the world happens to be such a way. Other accounts hold that the existence of free will is common across the space of possible worlds. I am disposed to think that there are many possible worlds where free will exists I will not take a strong stand on whether free will actually exists. My intuition is that it does , b ut I leave it to psychologists and neuroscientists to settle. What I will take stand on is how high we ought to set the bar for ascriptions of free will My intellectual dues are paid to Susan Wolf , whose Freedom within Reason was enlightening. As far as praise is concerned, I think Wolf got it exactly right Where I depart from Wolf is on the topic of blame. This brings me to the thesis of this paper. I am going to defend the following claim: desert praise is compatible with determinism , but desert blame is metaphysically impossible. First on the agenda is an explication of the relation between free will and moral responsibility. Second is a discussion of Wolf’s Reason View and Frankfurt - type cases Third is a critique of some of Wolf’s commitments, which will help to bring my own position into focus. Fourth is an attempt to motivate my thesis, drawing out its implications and connections with other concepts. I will begin with moral responsibility. Moral Responsibility Moral responsibility is related to our practice of praising and blaming. When we praise an agent , we hold them morally responsible for some morally good action. When we blame an agent, we hold them morally responsible for some morally bad action. We can distinguish between a forward - looking account and a backward - looking account of moral responsibility. On a forward - looking account, we blame agents for morally bad actions because doing so allows us to secure various goods . These goods might involve relationship reconciliation, harm reduction, and the formation of moral character. On a backward - looking account, we blame agents for morally bad actions just because they deserve to be blamed. The consequences of the practice of holding the agent morally responsible are not important. In order to be morally responsible for an action, the agent has to fulfil certain conditions. What these conditions are is disputed , but it is natural to suppose that the agent has to fulfil an epistemic condition and a control condition. For instance, suppose that some agent performs action Φ , a morally bad action. However, our agent was not aware that Φ is morally bad ; perhaps Φ was morally bad due to some un intended consequence that our agent was not in a position to foresee. In some cases, this kind of ignorance can absolve the agent of moral responsibility. That is an illustration of the epistemic requirement This time, suppose that our age nt finds themselves in a situation where not performing Φ is morally bad. As much as our agent wants to perform Φ , they are unable to because they are tied to a chair. Although the agent d oes not perform Φ , they are absolved of moral responsibility since it was not in their control. Th at is an illustration of the control requirement I am going to be working with the following assumption: free will is whatever allows an agent to satisfy the control condition on moral responsibility ; a specifically backward - looking moral responsibility. This assumption may be mistaken. Even if it is, I hope that I am nevertheless saying something interesting about the control associated with moral responsibility, whether or not “free will” is the appropriate term for it. Free Will In Nicomachean Ethics , Aristotle claims the following: Where it is in our power to act, it is also in our power not to act, and where saying ‘No’ is in our power, so is saying ‘Yes’; so that if it is in our power to act when it would be noble, it will also be in our power not to act when it would be shameful, and if it is in our power not to act when it would be noble, it will also be in our power to act when it would be shameful. 1 Aristotle is here speaking about voluntariness . We must be careful not to equate this with our contemporary notion of free will. All the same, in this passage Aristotle has articulated something that many free will scholars take to be relevant to free will: the ability to do otherwise ; alternate possibilities; counterfactual power These concepts all point to a plausible condition on free will: ( C P ) An action is freely willed only if the agent could have refrained from performing the action 1 Aristotle, Aristotle: Nicomachean Ethics , ed. by Roger Crisp , p. 45 Aristotle puts forth a case where an agent is stuck on a ship. The agent has no oars and no control over the sails. The agent is at the mercy of the wind, which so happens to take them across the Mediterranean all the way to Egypt. Is the agent’s arrival in Egypt freely willed? It seems that the answer is “no”. A very natural way to e xplain this intuition is to assert our C P condition ; i f the agent is unable to prevent an event from occurring , then the event is out of the agent’s control and thus the agent is not morally responsible for the occurrence of the event C P may or may not be a necessary condition on free will. But is it sufficient? Imagine that an agent has a device inserted into their brain. This device is able to stimulate the agent’s brain in such a way as to make the agent decide to take any action. This device utilises quantum technology which allows it to make use of the indeterminism at the scale of sub - atomic particles. Suppose that some sub - atomic event in the device has a 50% chance of occurring. If the event occurs, the device will stimulate the agent’s brain so that the agent decides to perform Φ 1 If the event does not occur, the agent’s brain will be stimulated to perform Φ 2 instead. Whatever then happens, the agent could have done otherwise. Roll back the time and the agent might act differently. Whichever action comes about, is the action freely willed? It seems that the answer is once again “no”. A natural way to explain this is to point out that the source of the action is the device stimulating the agent’s brain and not the agent themself. Free will requires sourcehood . We can formulate another plausible condition on free will: ( SC ) An action is freely willed only if the agent is the source of the action. We now have two plausible conditions on free will. Our thought experiments so far suggest the following analysis of free will: ( FW ) An action is freely willed if and only if the agent could have refrained from performing the action and the agent is the source of the action. That is a very general analysis. A great deal hangs on how we analyse the constituent concepts of counterfactual power and sourcehood. I will now examine Wolf’s analysis of free will The Reason View The Reason View is an analysis of free will according to which the control required for moral responsibility can be identified with responsiveness to reasons . Here is how Wolf puts it: According to the Reason View [...] responsibility depends on the ability to act in accordance with the True and the Good. 2 2 Wolf, Susan, Freedom Within Reason , p. 79 In other words: [I] t is only the ability to do the right thing for the right reasons [...] that is required for responsibility. 3 An agent requires a normative ability to respond to moral reasons in order to be able to act freely. This seems like a very plausible analysis of the sourcehood condition. Consider a situation where a child is drowning. The agent, Jones, knows that the right thing to do is to try to save the child. Jones then decides to save the child because it is the right thing to do. Does Jones deserve praise? Surely, if anyone deserves praise, then it is someone who does the right thing for the right reasons. So let us answer “yes”. In that case Jones is responsible for their action, which means that Jones freely willed their action, which means that Jones is the source of their action. We have decided that doing the right thing for the right reasons is sufficient for the desert of praise. Is it necessary? Yes: [A]ccording to the Reason View, a person who does the right thing for the wrong reasons deserves no more praise than a person who doesn’t do the right thing at all. 4 And that seems to be right. We do not want to praise those who do the right thing merely because it happened to be in their interest to do the right thing. But we encounter a slight problem. Someone’s moral character may be so polished that they are psychologically determined to do the right thing. If an agent cannot refrain from doing the right thing, then, according to our FW analysis, the agent cannot be freely willing their action. And yet, the Reason View predicts that they are freely willing their act ions. There is only one way to resolve this contradiction; reject the Reason View, or reject the FW analysis. As it happens, we have good reasons to reject the CP condition in the form of Frankfurt - type Cases Frankfurt - type Cases Frankfurt - type cases are thought experiments which seem to involve an agent who (1) is unable to refrain from performing an action, and yet (2) is morally responsible for their action. I will paraphrase a Frankfurt - type case from John Fischer’s article. 5 This time, Jones finds himself at the voting booth. Black, a neurosurgeon, wants the Democrat candidate to win the election. Black secretly inserts a chip into the brain of Jones. The chip allows Black to monitor and control Jones’s brain activity. If the chip detects brain activity which suggests that Jone s is about to decide to vote for the Republican candidate, then the chip will stimulate Jones’s brain in such a way that Jones decides to vote for the Democrat candidate instead. If, instead, Jones decide s to vote for 3 Wolf, Susan, Freedom Within Reason , p. 81 4 Wolf, Susan, “Freedom within reason.” In Personal Autonomy, p. 272 5 Fischer, John, “Frankfurt - Type Examples and Semicompatibilism: New Work” (2011) in The Oxford Handbook of Free Will 2 nd edn , p. 244 the Democrat candidate, the chip will stay dormant. Jones decides to vote for the Democrat candidate, and the chip stays dormant. In this case, Jones cannot refrain from voting for the Democrat candidate. Black’s chip makes sure of that. And yet, it seems that Jones is morally responsible for voting for the Democrat candidate so long as Jones decided to vote for the Democrat candidat e without the chip’s input. Compare a slightly modified version of the drowning child case: Jones decides to save the child because that is the right thing to do. However, unbeknownst to Jones , Black is waiting round the corner with a gun. Had Jones decide d not to save the child, Black would have come out and threatened Jones to save the child or else lose their life. There’s no doubt that Jones would have saved the child if threatened. But it did not come to that, and Jones surely deserves praise for saving the child. I find this version of the drowning child case particularly interesting. Much of the literature is particularly focused on blame , so cases like this one have not been, as far as I can see, thoroughly explored . There are good reasons for this, but considering cases where one is seemingly forced to do the right thing stirs up some interesting intuitions Consider, moreover, what it would require for Jones to have an indeterministic kind of freedom which preserves their counterfactual power Consider Jones ’s deliberat ion about the situation: the right thing to do is to save the child; they ought to do the right thing; so, they ought to save the child . If Jones is determined by his reasons and deliberation, then they end up saving the child; the thing they ought, and want, to do It is not clear that adding an indeterministic counterfactual power into the mix is at all conducive to control. For Jones to decide not to save the child – despite their deliberations strongly pointing in the direction of saving the child – would require a sudden wave of irrationality which forces them to do something that they did not want to do . That does not seem conducive to freedom. Praise/Blame Asymmetry Wolf’s Reason View leads to an asymmetry. It claims that free will is the ability to do the right thing for the right reasons. An agent who does the right thing for the right reasons, but does not have counterfactual power, has that ability. That is not the case with agents who do not do the right thing for the right reasons. In order for an agent who does the wrong thing to have the ability to do the right thing for the right reasons, the agent has to have counterfactual power ; it must be true that they c ould have refrained from performing the wrong action , and performed the right action instead Wolf is committed to defending a compatibilist analysis of counterfactual power. Let us sum up the key ideas: (1) A n agent deserves praise /blame iff the agent has performed a morally right /wrong action and the agent is morally responsible for the action. (2) An agent is morally responsible for an action only if the agent has the ability to do the right thing for the right reasons. (3) If an agent does the right thing for the right reasons, then the agent has the ability to do the right thing for the right reasons. (4) If an agent does not do the right thing for the right reasons, then the agent has the ability to do the right thing for the right reasons iff the agent could have done the right thing for the right reasons (i.e., has counterfactual power) However, there is something peculiar about all of this Let us look at the situation from another angle. There are two conditions on desert praise: (i) the agent performs a morally right action, and (ii) the agent is morally responsible for that action. There are two conditions on moral responsibility: (i) control, and (ii) epistemic state. Hence, in a more fine - grained way, there are three conditions on desert praise: (1) The agent performs a morally right action. (2) The agent has control over the action. (3) The agent has the right epistemic state. According to Wolf’s account, an agent who does the right thing for the right reasons deserves praise. An agent who does the right thing for the wrong reasons, however, does not. This makes a lot of sense; but how do we explain this difference in desert? Either there is a difference in the moral value of the action, or there is a difference in the agent s’ responsibility for their action s . There is no difference in the moral value of the action, so the difference must be in the agent s’ responsibility for thei r action s . Now, how do we explain the difference in responsibility for action? According to our analysis of responsibility, either there is a difference in the control that the agents hold over their actions, or there is a difference in the agents’ epistemic state s It does not matter where we try to locate the fault, for, either way, we run into a problem. The Trolley Problem of Responsibility Let us utilise the internet ethicist’s favourite thought experiment: the trolley problem. Poor Jones is, once again, the guinea pig in the experiment. Consider first the following scenario s Scenario A: Jones stands next to a lever. In front of him is a set of diverging tracks. On the left track, a single person is tied down to the tracks. On the right track, five people are tied down to the tracks. There is a trolley barrelling towards the right track. Jones, a committed consequentialist, immediately realises that they ought to pull the lever so that the trolley switches tracks and runs over a single person instead of five. Jones’s commitment to doing the right thing pushes them into action, a nd they pull the lever. Suppose, for the purposes of the experiment, that some felicitous consequentialist theory is true across all scenarios Scenario B: Jones once again stands next to a lever ; the situation is the same as in scenario A. This version of Jones, however, is not such a committed consequentialist. Jones notices that the person tied to the left track is our beloved neurosurgeon, Black. As it happens, Jones has not forgiven Black for meddling with Jones’s agency during the recent election. Filled with a desire for vengeance, Jones pulls the lever. Scenario C: for the third time, Jones stands next to a lever. This time, however, the trolley is barrelling towards the left track, where a single person is tied down Jones spots Black tied down on the right track alongside four others. Unfortunately for Black, this is the vengeful version of Jones. Jones decides that the collateral damage is something that they are willing to live with. They pull the lever, diverting the trolley such that it runs over the five people, including Black, instead of the single person on the other track. Prima facie , this is what we might want to say about these scenarios: (A) Jones deserves praise. This requires that Jones did the right thing, and that Jones was responsible for that action ; and that seems to be right. (B) Jones does not deserve praise. This requires that Jones either did not do the right thing, or they were not responsible for their action. Jones did do the right thing, so Jones must have not been responsible for their action. (C) Jones deserves blame. This requires that Jones did the wrong thing, and that Jones was responsible for that action. What is clear is that there is an internal difference between Jones A and Jones B : Jones A was pushed into action by set A of reasons (the belief that pulling the lever is right, the desire to do the right thing, etc.), while Jones B was pushed into action by set B of reasons (the belief that pulling the lever will kill Black, the desire to kill Black, etc.). It is an open question whether we want to designate this as a difference in control or a difference in knowledge (or some combination of both) . What i s clear is that we must pin this down as the difference which explains Jones A’s responsibility (and thus desert praise) and Jones B’s lack of responsibility (and thus lack of desert praise). What is less clear is the difference between Jones B and Jones C. They seem to be pushed into action by the same set of reasons – the belief that pulling the lever will kill Black, the desire to kill Black, and so on. We settled that it is this set of reasons which explains why Jones B is not responsible for their action . How can the same set of reasons, and all the other psychological properties held by both Jones B and Jones C, make Jones B not responsible for their action and also make Jones C responsible for their action? There is an important difference between Jones B and Jones C ; the difference being the moral value of their actions. But that should make no difference to either the control possessed by each agent or their epistemic states ; the moral value is external to the agent ’s control over their action. If Jones C is responsible for their action, then so is Jones B. And if that is the case, then Jones B ought to deserve praise. But Jones B does not deserve praise. Compatibilism about Praise, Scepticism about Blame In order to deserve praise, the agent has to be pushed into action by the appropriate moral considerations. That is Wolf’s key insight, and I think this ought to be our starting point. Here is an intriguing question: where is this criterion to be subsumed under? Is it to be subsumed under control, such that if one is not pushed into action by the appropriate moral considerations then one is not in control of the action? Is it to be subsumed under the epistemic requirement, such that if one is not pushed into action b y the moral considerations then one is lacking some knowledge or understanding which absolves one of one’s responsibility for action? Is it to be subsumed under moral responsibility, though separately from the control and epistemic requirements? Notice that if we place this criterion under any of these three, then whenever one is not pushed into action by the right moral considerations one is not morally responsible for their action. The other option is to place this criterion outside the scope of moral responsibility altogether : we might say that an agent deserves praise/blame iff the agent has performed a morally right/wrong action, the agent is morally responsible for the action, and the agent’s reason for action was the right moral consideratio n. However , if we do that, we end up affirming that an agent can be responsible for a morally wrong action while no t deserv ing blame for the action. We might be able to get around that by positing that the re are conditions on desert praise different to the conditions on desert blame besides the moral value of the action. But it is not entirely clear how this can be a criterion separate from moral responsibility , since moral responsibility is meant to be whatever secures desert – something that we have already violated by suggesting that the reason for action is a condition separate from moral responsibility We have no choice, I think, but to take the right reason criterion seriously and accept the consequences of doing so. Let us formulate an analysis of desert praise: ( DP ) An agent deserves praise iff the agent performs a morally right action and the agent performs the action for the right reason. What we need is an analysis of what it means to perform an action for the right reason. I have already suggested that it involves being “pushed” into action by the right reason , but more can be said about what that entails. When an agent deliberates about whether or not to perform some action, various reasons may enter their mind. Sticking with Jones, suppose they are deliberating whether or not to save a drowning child. A relevant reason that may enter Jones ’s mind is that trying to save the child might be dangerou s; Jones may themselves end up drowning. This is only a relevant reason for Jones , however, if Jones already has the desire not to drown ; a desire that many of us have , of course Jones might also have other desires which conflict with the action of saving the drowning child which are much weaker , such a s a desire to not ruin one’s clothes. On the other hand, Jones , we hope, will also be influenced by various reasons which motivate Jones to save the child. One of those rea sons might be the glory t hat comes with being the hero. What we really want Jones to be motivated by, however, is the rightness of saving the child. That is what suffices for, and is required by, desert praise. But what does it take for Jones to be motivated by the rightness of saving the child? Jones needs to believe that saving the child is indeed the right thing to do. In the same vein, a thirst for water is not enough to motivate one to turn on the tap if one does not have the belie f that there is, or at least might be, water in the tap Moral reasons are, I think, a very special kind of reason s Moral reasons are always maximall y normative. When one has a desire for apples and a desire for strawberries, but can only eat one, which one it is rational to choose comes down to the particular agent’s strength of desire for each of those things. A moral reason, however, is always to be acted upon irrespective of one’s other reasons or desires , for a moral reason gives one unconditional motivation to act. In a sense, this is just what it means fo r a reason to be a moral reason Other kinds of reasons can be “defeated” Being thirsty gives one reason to seek out and drink some water , but if the water one finds is contaminated , then that defeats one’s reason to drink the water. A n agent can coherently think “I am thirsty, so I should some water, but not this water ”. It would not be coherent for an agent to think “I should , on moral grounds, save the child, unless I do not want to ”, or “unless it makes my clothes wet”. One qualification: it seems plausible that moral reasons can be defeated by other moral reasons: one should save the child unless, say, saving the child leads to the death s of ten more This will not invalidate my conclusion. Now, w henever someone performs a morally bad action, either (i) moral reason s are available to the agent, or (ii) it is not. Suppose that (i i ) , Jones does not see any moral reason to save the child. Perhaps they have been indoctrinated since a young age by a terrorist organisation, or perhaps they are a psychopath. Wolf herself is happy to grant that someone who is indoctrinated in such a way that they do not have the ability to respond to moral reasons is not morally responsible for their wrong actions. They do not have the ability to act in accordance with the True and the Good. Wolf’s strong intuition that such an individual is not morally responsible for their actions is what leads her to construct her theory as an alternative to mesh theories. Suppose, however, that Jones is a typical person, deliberating whether or not to save the child. The moral reason that Jones ought to save the child because it is the right thing to do just does not occur to the agent during their process of deliberation. Jones decides not to save the child Although like the terrorist Jones does not have the ability to respond to moral reasons, unlike t he terrorist Jones was not indoctrinated from birth Although no moral reason is available to Jones , perhaps it is Jones ’s fault that no such reason is available. They may be responsible for their own ignorance . Let us consider some possible scenarios where someone who is not indoctrinated from birth, or a psychopath, somehow cannot draw on basic moral reasons. Jones, under normal circumstances, very well knows that drowning children ought to be saved How did that not come up during their deliberation ? It seems to me that the only possibility is a sudd en bout of irrationality . This moment of irrationality might have been caused by the high - stress nature of the situation which impaire d Jones ’s deliberative faculties. Alternatively, it might be caused by some unusual flicker of i ndeterminism , or otherwise unusual neural activity, in Jones ’s brain which prevented that reason from surfacing up In any case, if Jones is the sort of agent who is usually responsive to moral reasons, but for some reason is not responsive to moral reasons in this instance, that can only put down to some kind of irrationality. It does not seem to me that irrationality is particularly conducive to control, and it does n ot seem to me that an agent can freely will themselves to be irrational in such a way. In these sorts of cases the agent seems to be absolved of moral responsibility. I have tried to show that whenever an agent is unable to draw on moral considerations during their process of deliberation, which is one state that an agent can be in when they perform a morally bad action, they cannot be held responsible for their action. For either they have become uncontrollably irrational, o r they were not given the opportunity to become responsive to moral reasons. The alternative is that (i) moral reasons do enter into the agent’s deliberation, and yet the agent decides to perform the morally bad action anyway. It is important to consider how thi s might come about. One possibility is that it is the wrong moral reasons which enter into the agent’s deliberation. It may be that the ag ent does the wrong thing while thinking that it is the right thing. Is this something that the agent can be responsible for? This is a massive issue that I cannot cover in detail. But I will make a few remarks. C onsider the trolley problem again, in its classic guise : ought the lever be pulled, as the consequentialist might suggest, or left alone, as the deontologist might suggest? Another crossroads: e ither both positions are reasonable, or one of them is not. If both positions are reasonable, then surely we cannot fault whoever happens to get it wrong. Perhaps had they had a slightly different intuition at some stage in their moral theorising, they would have ended up affirming the correct moral theory , and thus ended up performing the right action . As agents with limited cognitive faculties, this is not something that they could control. One may respond that it is difficult to imagine rationally coming to the conclusion that one ought not to save a drowning child. That leads us down the second path: some moral beliefs are not reasonable . One must have gone very wrong somewhere if they end up thinking that one ought not to save a drowning child. In that case, they must be irrational What about cases where one correctly identifies the right thing to do, where the right moral reason enters one’s process of deliberation, and yet the agent decides to do the wrong thing anyway? Wolf’s Reason theory, we must remember, is intended as an improvement on mesh theories , but not a complete replacement . I do not have the opportunity to discuss mesh theories in de tail, but the basic idea is that our wills are free when our desires form a harmonious mesh. Suppose that Jones desires to save the drowning child ; they desire it in the sense that they see it as valuable to do, on the basis of correct moral considerations. But suppose also that Jones desires not to save the drowning child for the reason that they are afraid of dying in the attempt As it happens, Jones has a higher - order desire for their desire to save the drowning child to win out They understand the weight of the moral reasons pushing them in that direction, and they really want to be the sort of person who does the right thing. Unfortunately, Jones ’s instinct for self - preservation is too strong, and despite their best efforts, the desire to keep their own life sa fe wins out. Jones is a slave to their own will , akin to an addict or a phobic. I would be prone to think that Jones does not deserve blame here. The alternative is that Jones desires to save the drowning child for various moral reasons and desires to not save the drowning child for other reasons, and also has a higher - order desire not to save the child T hey’re happy to be the sort of person who dismisses moral reasons However, m oral reasons are, by their nature, indefeasible The sort of person who wants to be the sort of person who dismisses moral reasons must be irrational, or else be unable to grasp the full normativity of moral reasons Conclusion I started this essay by outlining a few contested but common assumptions about the nature of free will and moral responsibility that I am taking for granted : that free will is a sort of control required for moral responsibility ; that the moral responsibility in question is the backward - looking, desert kind. I appealed to Frankfurt - type cases to claim that, dispute intuitions to the contrary, free will does not require counterfactual power. Finally, I assumed that Wolf struck gold with her identification of doing the right thing for the right reasons as a necessary and sufficient condition on desert praise. Keeping this core of Wolf’s theory as a constant, I tried to show on grounds of pure consistency that Wolf’s conclusions regarding blame are difficult to reconcile. It is difficult to explain, using Wolf’s theory, the difference in moral responsibility between an agent who does the right thing for the wrong reasons and an agent who does the wrong thing. I suggested that this is not a fault with the theory, but rather that it is a n uncomfortable consequence which we ought to embrace. I then tried to independently motivate this consequence – make the pillow easier to swallow – by pointing out that whenever an agent does the wrong thing, either m oral considerations are available to them, or they are not. I suggest that if they are not, then that is not something the agent is responsible for: either they have been indoctrinated, or , during their deliberative process, they were struck by an uncontrollable gust of irrationality. I suggest that if moral considerations are available to them, t hen either they are the right ones, or the wrong ones. If they are the wrong ones, then it is not something the agent is responsible for: it is to be attributed to the way we are cognitively limited as humans. If they are the right ones, then the agent’s action is not something the agent is responsible for: they were weak - willed, or else they did not really grasp the normative weight of those reasons or they were irrational. Whether the reasons are the wrong ones or the right ones, the agent is not to be he ld responsible. Whether any moral reasons are available to the agent or not, the agent is not to be held responsible. Bibliography Aristotle, Aristotle: Nicomachean Ethics , ed. by Roger Crisp, Cambridge Texts in the History of Philosophy, 2nd edn (Cambridge: Cambridge University Press, 2014) Fischer, John, “Frankfurt - Type Examples and Semicompatibilism: New Work” (2011) in The Oxford Handbook of Free Will 2 nd edn , ed. Robert Kane Frankfurt, Harry G. (1971). Freedom of the will and the concept of a person. Journal of Philosophy 68 (1):5 - 20. Vargas, Manuel, “ Instrumentalist Theories of M oral Responsibility ” in The Oxford Handbook of Moral Responsibility , eds. Nelkin, Dana Kay, and Derk Pereboom ( 2022) Oxford: Ox ford University Press Wolf, Susan, Freedom Within Reason (New York, NY, 1990 ) Wolf , Susan, “Freedom within reason.” (2005) In Personal Autonomy, ed. James Stacey Ta ylor, 258 - 74. New York: Cambridge University Press