In Causation and Free Will Carolina Sartorio offers an intricate argument for the claim ‘that actual sequences are sufficient grounds for freedom: they ground freedom, and nothing is required to ground it other than themselves, or their own grounds’ (p. 20). On this view, free will with respect to a choice or action is not explained in terms of the agent’s having alternative possibilities. Like Harry Frankfurt, Sartorio rejects the Principle of Alternate Possibilities (recast in terms of free will): Principle of Alternate Possibilities (PAP): a person does something freely only if he could have done otherwise. (Frankfurt 1969, p. 829) As is well known, Frankfurt opposes PAP on the grounds that it is vulnerable to counter-examples. One of Sartorio’s variants of these examples is as follows: Frankfurt’s case: Frank has reasons to harm Furt and makes the choice (C) on his own to shoot him. Unbeknownst to Frank, a neuroscientist has been monitoring his brain processes. The neuroscientist can predict the choices that Frank is about to make and can also manipulate Frank’s brain. Had the neuroscientist had any reason to believe that Frank wasn’t going to make choice C on his own, he would have intervened in such a way that Frank would still have made C. (pp. 13, 113) Frank acts on his own by making choice C, but he would have made that choice anyway, because the neuroscientist, who is monitoring his brain, would have intervened to cause Frank to make that choice had Frank not made it on his own. Sartorio agrees with Frankfurt’s claim that PAP cannot explain responsibility and free choice in these kinds of counter-examples. Sartorio’s whole enterprise is motivated by her assumption that in Frankfurt-style cases, cases in which the agent’s choice is unavoidable, the actual-sequence account is best positioned to explain what it is for an agent to act freely. On Sartorio’s view, free will with respect to choice or action X is a function of, and supervenes on, the actual causal (or quasi-causal) sequence that brought about X. She says emphatically that freedom is exclusively a function of the choice’s actual causal history. Basically, she contends that a free choice is grounded in a causal fact, namely, the fact that the agent’s judging that there were good reasons to make the choice causally resulted in the choice (p. 50). Sartorio does not offer a full-blown account of causation, but claims that causation is a ‘significant metaphysical relation’ (p. 49) that makes attributions of responsibility to agents intelligible. She also asserts that this relation has four key properties: (1) it admits of overdetermination (pp. 50-2); (2) it is difference-making (pp. 94-102); (3) it is extrinsic in that extrinsic factors (that is, factors external to the causal process) may affect it (pp. 75-6); and (4) it is intransitive (pp. 104-6). I will focus on Sartorio’s arguments for dismissing alternative-possibilities accounts of free will and defending an actual-sequence view. I will try to lay bare the connections between her account of free will and her conception of causation. In addition, I will differentiate three varieties of moral responsibility (choice-responsibility, action-responsibility and outcome-responsibility) and will briefly show the implications of keeping them apart for our understanding of our intuitions about free will. Sartorio says that she is not concerned with responsibility, but with the metaphysical precondition for moral responsibility, namely, freedom (pp. 7-8). In fact, it seems that if an agent is not in control of X, he cannot be morally blameworthy for X. The same point applies to being responsible for a particular outcome. Sartorio claims that an agent can be held responsible for an outcome Y just in case there is some X such that the agent is responsible for X, X causes Y, and the agent meets the relevant epistemic conditions (p. 81). Because free will as a component of responsibility is a morally-laden notion, intuitions about free will in cases where agents cannot do otherwise reflect moral intuitions. 1. The causal history account of Frankfurt’s case In Frankfurt’s case, intuitively Frank seems free to make choice C even though it appears that he could not have done otherwise. The fact that the scientist would have intervened had Frank not made C on his own does not exclude his acting freely. Frank controls his action despite his lacking alternative courses of action, and, according to Sartorio, what explains Frank’s control over his action despite the supposed violation of PAP is the causal history of his choice. An alternative-possibilities account would entail that Frank is not free with respect to his choice to harm Furt (nor with respect to his harming Furt), because this choice was unavoidable. Had he not chosen on his own to harm Furt, he would have made that choice anyway because of the neuroscientist’s intervention. According to Sartorio, Frank is free to choose to harm Furt, but he is unfree not to choose to do so. He is unfree not to choose C since the presence of the neuroscientist ensures that he will choose C even if after deliberating, Frank does not choose C. Sartorio claims that Frank is free with respect to C, not because of facts about some alternative, hypothetical causal sequence but because of the relevant properties of the causal sequence that actually caused C. Though Sartorio doesn’t want to provide an account of moral responsibility and says that her main concern is the freedom condition for responsibility (p. 35), it is clear, as I said at the outset, that our intuitions in Frankfurt-style cases are tainted with moral ideas. We think that Frank is morally responsible for shooting Furt because his choice is caused by his own assessment of his reasons, just as it would have been caused if the neuroscientist had not planned to intervene. The question is whether the neuroscientist’s back-up intervention is a sufficient justification for Frank’s wrongful choice. We think that it is not. Since we assume that Frank is morally responsible, we are inclined to believe that he is also free. In the alternative scenario, in which Frank’s choice is caused by the neuroscientist’s manipulation, we are certain that he is not morally responsible, and this motivates our conclusion that he is not free. Our beliefs about free will seem to be motivated by our moral intuitions, rather than the other way around. 2. Causation and responsibility On the counterfactual dependence account of causation, an event c causes an event e if and only if e would not have occurred in the absence of c. Sartorio claims that Frank’s assessment of his reasons caused his choice, but Frank’s choice doesn’t counterfactually depend on that assessment (p. 52). For the counterfactual dependence theorist, Frank’s deliberative processes did not cause the outcome (that is, Furt’s being shot). According to PAP, Frank would be free with respect to C only if he could have chosen otherwise. Frank is not responsible for the outcome because, if his assessment of his reasons had been different, he would have made C anyway, through the intervention of the neuroscientist. Since Frank is responsible because he has chosen to bring about the outcome on his own, we need a different account of causation to explain Frank’s freedom. We said that the availability of a back-up cause for his choice C (that is, the neuroscientist’s intervention) is not a ground for exemption from moral responsibility. Sartorio says that a cause is a difference-maker, but not in the counterfactual dependence sense. She holds that causes make a causal contribution to their effects that the absence of those causes does not make. In her words: Difference-Making: Causes make a difference to their effects in that the effects wouldn’t have been caused by the absence of their causes. (p. 94) Sartorio says that Frank’s assessment of his reasons to harm Furt did cause choice C, for the absence of Frank’s assessment of his reasons would not have caused C. Absences can be causes, according to Sartorio, but this specific absence would not make the same causal contribution to Frank’s choice as the corresponding positive event (that is, Frank’s deliberating on his reasons to harm Furt). If Frank’s assessment of reasons makes a causal contribution to C that is not made by the absence of that assessment, there is an actual causal sequence between Frank’s practical reasoning and his making C. His freedom to make C and his control of that choice is grounded in that causal history, understood in terms of Sartorio’s difference-making approach. For Sartorio, difference-making is a necessary condition of causation, but not a sufficient condition, because an event c may not cause an event e even if the absence of c does not cause e (p. 96). For instance, neither Frank’s wearing a hat nor the absence of Frank’s wearing a hat causes his choice. Why would the absence of Frank’s assessing his reasons not cause his choice in the intervention scenario? Sartorio claims that Frank’s judging that he has reasons to harm Furt caused his choice to shoot Furt because that judgment is not superfluous with respect to that choice (p. 95). In its absence, C would be caused by the neuroscientist’s intervention. If Frank decides not to shoot Furt, the absence of Frank’s assessment of his reasons to shoot Furt is causally efficacious with respect to the intervention of the neuroscientist, and that intervention in turn causes Frank to decide to shoot Furt. However, it’s not the absence of Frank’s judging that he has reasons to shoot Furt that causes his choosing to do so. In fact, causation is not a transitive relation, as Sartorio holds (pp. 104-6). So, Frank’s assessment of his reasons makes a causal contribution to his choice that the absence of that assessment does not make. It’s not clear that Sartorio’s difference-making view can explain Frankfurt’s case without altogether relying on the counterfactual dependence analysis and PAP. Consider the following assumption: Non-identity of Choices: For any pair of choices X and Y made by an agent A, if X has been caused by a certain deliberative process of A’s and Y has not been caused by that deliberative process, X is non-identical to Y, even if X and Y are choices to do the same action or to bring about the same outcome. If we accept this assumption, two consequences follow. First, Frank’s choosing to shoot Furt must counterfactually depend on his judging that there are reasons to shoot Furt. In effect, Frank’s deliberation causes his choice to shoot Furt only if that deliberation and the absence of that deliberation play different causal roles. Since the absence of Frank’s deliberation does not cause his choosing to shoot Furt when the neuroscientist intervenes, this manipulated choice is, according to the above assumption, non-identical to his choice to shoot Furt in the non-intervention scenario. Therefore, if Frank had not originally decided to shoot Furt on his own reasons, his (deliberative) choice to shoot Furt would not have been made. Second, our intuitions about whether Frank is acting freely do not violate PAP. In effect, Frank could have acted otherwise by avoiding the deliberative choice to shoot Furt, thus prompting the neuroscientist to cause him to (non-deliberatively) make C. Some authors have argued that Frank does have the ability to do otherwise in the original example. The so-called flicker-of-freedom defence of PAP maintains that Frank could have done otherwise by trying to make a different choice or by making a sign that he would choose to shoot Furt on his own. Sartorio’s reply is that such options are not sufficiently robust to ground Frank’s freedom (p. 14). However, making a non-identical choice is a robust alternative. Among other things, in the non-intervention scenario Frank’s deliberative choice is fully his own, but when his choice is caused by the neuroscientist’s intervention, it is, properly speaking, the neuroscientist’s choice. Frank is not responsible in this case; we think that the neuroscientist is responsible. This explanation of Frank’s control over his choice to shoot Furt is consistent with an alternative-possibilities view. 3. Is the alternative-possibilities view really false? Sartorio discusses some examples furnished by Peter van Inwagen which elicit the intuition that the agent in question is not free (and, therefore, not morally responsible) just because he cannot act otherwise (van Inwagen 1983, pp. 162-6). This supports the intuition (expressed in PAP) that free will requires the ability to do otherwise, which in turn presses the question whether it’s not possible to rescue the alternative-possibilities view. Consider the example of avoidable omission: Phones: I witness a man being robbed and beaten. I consider calling the police. I could easily pick up the phone and call them. But I decide against it, out of a combination of fear and laziness. Here I am clearly free in omitting to call the police. Contrast that case with one of unavoidable omission: No Phones: Everything is the same as in Phones except that, unbeknownst to me, I couldn’t have called the police. (The phone lines were down at the time.) Intuitively, I am not free in this case, because I could not have avoided the omission by calling the police. So, it seems that the possibility of my avoiding an omission is a requirement for the omission counting as an exercise of my free will. This reasoning might be extended to positive acts. First, consider the case of avoidable action: Not All Roads Lead to Rome: A man, Ryder, is riding a runaway horse, Dobbin. Ryder can’t get Dobbin to stop but he can steer him in different directions with the bridle. When they approach a certain crossroad, Ryder realizes that only one of the roads leads to Rome. Ryder hates Romans, so he steers Dobbin in that direction so that some Romans are hurt by the passage of the runaway horse. Ryder is surely free (and responsible) with respect to his hurting the Romans, because, among other things, he could have acted otherwise—he could have failed to steer Dobbin towards Rome. Compare this scenario with one in which Ryder’s act is unavoidable: All Roads Lead to Rome: Everything is the same as in Not All Roads Lead to Rome except that, unbeknownst to Ryder, all roads lead to Rome. (Ryder couldn’t have avoided harming the Romans.) These counterexamples show that the difference in our intuitions concerning my freedom in Phones and No Phones and Ryder’s freedom in Not All Roads Lead to Rome and All Roads Lead to Rome is not exclusively a function of the properties of the actual causal histories in each scenario, but also a function of properties of the parallel counterfactual histories. Sartorio defends an actual-sequence account by arguing that the difference in our intuitions concerning whether the agent acts freely between each case and its counterpart is grounded in the actual causal histories rather than in the lack of an ability to do otherwise. Thus, there must be a causal difference between Phones and No Phones that is grounded in properties of the actual causal sequences. Sartorio claims that, whereas in Phones my failure to pick up the phone causes my failure to call the police, in No Phones my failure to try to call the police is causally inefficacious, because the lines are down at the time (p. 99). Sartorio appeals to the extrinsic nature of omissions to defend an actual-sequence account of No Phones: Extrinsicness: A causal relation between c and e may obtain, in part, owing to factors that are extrinsic to the causal process linking c and e. (p. 71) The fact that the lines are down is extrinsic in this sense. According to the difference-making approach, my failure to pick up the phone cannot make the same causal contribution to my failure to call the police as my picking up the phone. But, according to Sartorio, both my failing to pick up the phone and my picking up the phone are causally inefficacious with respect to my failure to call the police. She says that I can only fail to try to call the police (not fail to call the police): ‘In particular, my failing to try to call the police causally results in my failing to call the police in that scenario [Phones], whereas it doesn’t in No Phones’ (p. 75, italics added). However, the fact that my trying to call the police cannot result in my calling the police because the lines are down is the same as the fact that, if I fail to call the police, I cannot do otherwise (that is, call the police) because the lines are down. After all, the fact that the causal histories in Phones and No Phones are different depends on the fact that I am not able to call the police in No Phones because of an extrinsic factor. Therefore, it’s not clear that Sartorio’s treatment of No Phones does not rely on a PAP-account of free will with respect to omissions. Sartorio claims that extrinsicness holds with respect to both omissions and positive acts, but the claim is more plausible with respect to omissions. Whereas the extrinsic factor (ability to avoid) seems a conceptual requirement for omissions, it has a dubious place as regards positive acts. At most, it might seem a conceptual requirement for free acts, not for positive acts in general. In fact, the following proposition seems true: Omissions: An agent omits to do X only if he fails to do X and has the ability to do X. The parallel proposition with respect to positive acts is controversial: Acts: An agent performs X only if he does X and has the ability to do non-X (avoid doing X). Still, Sartorio argues, less persuasively, that the difference between Ryder’s steering Dobbin to Rome in Not All Roads (Lead to Rome) and his parallel act in All Roads (Lead to Rome) is not grounded in the fact that the former is avoidable whereas the latter is unavoidable. Again, the difference lies in the fact that in All Roads Ryder’s steering Dobbin does not cause the harm to the Romans, since an extrinsic factor (the fact that Dobbin will hurt the Romans anyway) pre-empts the causal connection. In All Roads, Sartorio claims, an extrinsic factor pre-empts the causal connection between Ryder’s act and the harm (pp. 73-4). Intuitively, Ryder is not free in hurting the Romans in All Roads, and yet his hatred of Romans causes his steering Dobbin (not only his trying to steer Dobbin) towards the road he chooses. Sartorio claims that Ryder is not free in All Roads because his act does not cause Dobbin to hurt the Romans. In fact, the absence of Ryder’s steering Dobbin doesn’t make a causal difference because the Romans are harmed anyway, and, therefore, Ryder’s steering Dobbin is not a cause of the harm according to the difference-making condition. However, the fact that Ryder’s failing to steer Dobbin does not prevent Dobbin from harming the Romans is the same as the fact that Ryder cannot avoid the harm. So, PAP does in fact apply to All Roads. 4. Varieties of moral responsibility Sartorio agrees that ‘our intuitions about control and responsibility themselves are not always precise’ (p. 141). However, she does not try to disentangle our background moral intuitions. I want to suggest that the differences in our metaphysical intuitions about Frankfurt’s case, No Phones and All Roads derive from the distinct kind of moral responsibility that we presuppose in assessing each example. I distinguish three kinds of moral responsibility: choice-responsibility, action-responsibility and outcome-responsibility. In the case of choice-responsibility we assess the agent’s intentions and choices. In the case of action-responsibility we evaluate the agent’s positive acts. Finally, in the case of outcome-responsibility we hold agents responsible for the outcomes they bring about. In Frankfurt’s example, we are asked to assess Frank’s choice-responsibility. I believe that the connection between Frank’s deliberations and his choice is a special causal relation. Reasons do not cause choices in the ordinary sense, but they do have a sort of internal relation to choices. The assumption about the non-identity of choices is a result of this special kind of causation between reasons and choices. Choices are not only caused, but also constituted, by reasons, and, hence, Frank does have the ability to make a different choice. Though outcomes caused by different deliberative processes may be identical, choices are defined by their reasons. Therefore, extrinsic factors cannot break the causal relation between reasons and choices, as they may break the causal relation between choices and outcomes. In choice-responsibility PAP applies through the assumption of the non-identity of choices. Therefore, the potential intervention of the neuroscientist cannot pre-empt the causal relation between Frank’s reasons and his choice in the non-intervention scenario. Outcome-responsibility does not encompass intentional, subjective elements and may certainly be affected by extrinsic factors. Thus, we may interpret No Phones and All Roads as only involving outcome-responsibility. In No Phones we may want to know whether I am responsible for the police not being called, and, if this is the question, the extrinsic factor that the lines are down is relevant. Intuitively, I am not responsible for that outcome because I don’t have the ability to avoid it. Similar reasoning applies to Ryder’s responsibility for the harm to the Romans in All Roads. Action-responsibility is an intermediate, hybrid notion. Both subjective and objective elements may be important depending on the emphasis. Our intuitions concerning action-responsibility are unstable precisely because they oscillate between the subjective and the objective interpretation of each case. In All Roads the question whether Ryder is responsible for steering Dobbin is a controversial one. We are inclined to believe that Ryder is not responsible for hurting the Romans when we emphasize the action’s outcomes, and we are inclined to think that Ryder is responsible for the act when we focus on his motive for harming the Romans. Though much more should be said, I can only suggest here that our ambivalence in such cases reflects the unstable nature of action-responsibility as a hybrid (subjective, objective) kind of moral responsibility. Unlike positive acts, omissions only allow choice-responsibility and outcome-responsibility, because they do not include a process causing the outcome. Unless motives and choices are clearly emphasised, when we are asked about responsibility for omissions we tend to understand the question in the language of outcome-responsibility. So, I am not responsible for the police not being called in No Phones, as that outcome was inevitable. However, our answer would change if we made it explicit that we want to establish my moral responsibility for choosing not to call the police (given that I am unaware that the lines are down). Intuitively, I am responsible for my choice, though this responsibility does not necessarily extend to the outcome. As I said, PAP bears on choice-responsibility via the assumption of the non-identity of choices. I conclude that Sartorio’s difference-making account of causation is unnecessary for explaining Frank’s responsibility for his choice. We can simply rely on some version of PAP and the assumption about the non-identity of choices. This assumption cannot be extrapolated to actions and outcomes. It seems that some avoidability principle is relevant for understanding all kinds of moral responsibility. When we hold an agent responsible for his choice, his performing a positive act, his failing to prevent an outcome or his bringing about an outcome, we assume that he could have avoided moral responsibility by doing something different. Sartorio discusses many other important topics in this book. For instance, she provides an excellent account of reasons-sensitivity that avoids difficulties for her actual-sequence account of free will. She points out that Frank’s choice in the non-intervention scenario reveals reasons-sensitivity because that choice is (actually) caused by both the presence and absence of reasons (pp. 124-6). Sartorio’s book offers a brilliant contribution to the literature on free will. It deserves serious study by philosophers and doctoral students. Legal theorists would also benefit from aquainting themselves with this sophisticated defence of compatibilism. Sartorio’s ideas will shape the terms of the debate about free will for a long time. References Frankfurt Harry G. 1969, ‘ Alternate Possibilities and Moral Responsibility’, The Journal of Philosophy , 66. van Inwagen Peter 1983, An Essay on Free Will ( Oxford, Clarendon Press). © Mind Association 2017
Mind – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera